MiFID e Version

MiFID 2.
0
CASTING NEW LIGHT ON EUROPES
CAPITAL MARKETS
REPORT OF THE ECMI-CEPS TASK FORCE

ON THE MIFID REVIEW
CHAIR PIERRE FRANCOTTE

Former Chief Executive Officer, Euroclear
General Manager, PLF International
RAPPORTEURS DIEGO VALIANTE

Research Fellow, ECMI & CEPS
KAREL LANNOO
Chief Executive Officer and Senior Research Fellow
CEPS & ECMI
CENTRE FOR EUROPEAN POLICY STUDIES

BRUSSELS
Acknowledgements
The rapporteurs gratefully acknowledge the contributions provided by all

participants in the Task Force, and in particular by the Chair, Pierre Francotte, who
went to extraordinary lengths to ensure that all members were fully heard. A list of
participants is provided at the end of this report. Mirzha J. de Manuel Aramenda
provided valuable research assistance and extensive operational support. Editing
and other tasks related to the administration of the Task Force were provided by
Anne Harrington, Margarita Minkova, Isabelle Tenaerts and Jackie West. All errors
should only be attributed to the rapporteurs.
Disclaimer
The views expressed in this report do not necessarily reflect the views and
positions of all members involved in the Task Force. The members do not
necessarily agree with all positions put forward and do not necessarily endorse
any reference to academic and independent studies. A sound and clear set of
principles has guided the drafting process, in order to preserve a neutral approach
to divergent views. All members were given ample opportunity to express their
views and if well-grounded they are reflected in the final text. Wherever
fundamental disagreement arose, the rapporteurs have made sure that all views
have been explained in a clear and fair manner. The content of the report, however,
can be only attributed to the rapporteurs. Data included in the text have been
generally considered as material and relevant.
ISBN 978-94-6138-081-4
Copyright 2011, ECMI and CEPS.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system
or transmitted in any form or by any means electronic, mechanical, photocopying, recording
or otherwise without the prior permission of the Centre for European Policy Studies and the
European Capital Markets Institute.
European Capital Markets Institute

And Centre for European Policy Studies
Place du Congrs 1, B-1000 Brussels
Tel: (32.2) 229.39.11 Fax: (32.2) 219.41.51
Website: www.ceps.eu
TABLE OF CONTENTS
Preface.......................................................................................................................i
Executive Summary ............................................................................................... 1
1. Introduction ..................................................................................................... 16
1.1 The MiFID review ..................................................................................... 16
2. The Markets in Financial Instruments Directive ..................................... 21
2.1 The legislative process ............................................................................ 21
2.2 The legal structure ................................................................................... 23
3. The Impact of the MiFID on European Markets: Some evidence
from cash equity markets.............................................................................. 28
3.1 Introduction .............................................................................................. 28
3.2 A more competitive environment ......................................................... 30
3.3 The introduction of new technologies .................................................. 36
3.4 Evolution in market microstructure and brokerage services ............ 39
3.5 A more comprehensive transparency regime...................................... 41
3.6 Greater investor protection .................................................................... 42
4. A Renewed Transparency Regime ............................................................ 44
4.1 The role of transparency in financial markets ..................................... 44
4.2 Pre-trade transparency for cash equities: Benefits and costs............. 49
4.2.1 Dark liquidity .................................................................................. 55
4.2.2 Pre-trade transparency waivers: where do we stand? .............. 59
4.2.3 A consolidated quote system ........................................................ 65
4.3 What pre-trade transparency for non-equity instruments?............... 66
4.4 Post-trade transparency regime: Fixing bugs and
extending scope ....................................................................................... 70
4.4.1 The trade reporting regime for shares ........................................ 75
4.4.2 Challenges with data consolidation............................................ 76
4.4.3 Bond markets ................................................................................. 85
4.4.4 Structured financial products ...................................................... 93
4.4.5 Over-the-counter derivatives ....................................................... 95
5. Reshaping Market Structure ......................................................................100
5.1 Introduction ............................................................................................100
5.2 Securities markets as network industry .............................................100
5.2.1 Markets Multi-sidedness and Securities Markets ...................102
5.3 Competition and fragmentation ..........................................................107
5.4 Classification and obligations of trading venues: A new role
for internalisation? .................................................................................117
5.4.1 Shedding light on over-the-counter equity trading ................121
5.5 Some aspects of market microstructure for organised trading .......134
5.5.1 Market setting: The role of trading mechanisms ....................135
5.5.2 The evolution of trading: Market efficiency and
financial stability ....................................................................................143
5.6 Access to market infrastructures for equities: Whats next? ............158
6. Provision of Investment Services ..............................................................164
6.1 The nature of investment services and investors ..............................165
6.2 Fiduciary duties .....................................................................................169
6.3 Organisational rules ..............................................................................174
6.3.1 Investment advice: Striking the right balance .........................181
6.4 Conduct of business rules .....................................................................185
6.4.1 Conflicts of interests under MiFID ............................................186
6.4.2 Client categorisation....................................................................189
6.4.3 Suitability and appropriateness tests ........................................192
6.4.4 Grasping the definition and economics of best execution
obligations ...............................................................................................198
7. A New Regime for Commodity Derivatives ...........................................205
8. Conclusions ...................................................................................................214
References ...........................................................................................................220
Annex I. List of Acronyms ...............................................................................238
Annex II. The objectives of the Task Force ..................................................240
Annex III. Market Infrastructure overview................................................. 242
Annex IV. Figures and Tables Illustrating Equity Markets ......................249
Annex V. List of Task Force Participants ......................................................259
List of Figures
Figure 1. Interaction with main EU regulatory actions ...................................19
Figure 2. MiFID timeline .....................................................................................20
Figure 3. MiFID original principles....................................................................24
Figure 4. Impact of the crisis on EEA equity turnover and volumes
(lit & dark) .............................................................................................................29
Figure 5. Registered trading venues in the European Economic Area (EEA)
and breakdown for equities ................................................................................31
Figure 6. Bid/ask average spread and implicit volatility on
EuroSTOXX, FTSE100 and CAC40 .....................................................................33
Figure 7. EEA equity markets 2010 pan-European market shares .............34
Figure 8. Most liquid EU shares by execution venue ......................................35
Figure 9. Pre-trade transparency and market quality .....................................51
Figure 10. Market depth and OTC activity .......................................................52
Figure 11. Current use of waivers ......................................................................62
Figure 12. Data vendors market shares (all asset classes) .............................79
Figure 13. Average quotes per minute ..............................................................84
Figure 14. Average daily number of IRS trades by currency .........................95
Figure 15. Average daily number of new IRS trades by maturity .................96
Figure 16. Primary securities markets .............................................................104
Figure 17. Secondary securities markets .........................................................105
Figure 18. Cost structures with (A) and without (B) size externalities .......109
Figure 19. Market fragmentation versus liquidity fragmentation ...............112
Figure 20. 2009 Exchanges revenues by activity
(main European markets) ..................................................................................116
Figure 21. Trading platforms and MiFID ........................................................118
Figure 22. Transparency obligations for venues under MiFID ....................124
Figure 23. Broker-dealer crossing system .......................................................125
Figure 24. Proposal for broker dealer crossing system (BCSs).....................127
Figure 25. OTC printed vs actual equity trading .......................................130
Figure 26. Dealer markets (trading models) ...................................................141
Figure 27. Average trade size on SETS ............................................................144
Figure 28. Average value of orders executed for most liquid
European shares .................................................................................................145
Figure 29. Aggregated S&P 500 market depth .............................................. 148
Figure 30. Shifting the regulatory centre of gravity. ..................................... 157
Figure 31. Equity landscape before and after MiFID and the
Code of Conduct ................................................................................................ 161
Figure 32. General clauses ................................................................................ 174
Figure 33. Exemptions for commodity instruments and firms ................... 206
List of Tables
Table 1. The MiFID review ................................................................................. 17
Table 2. Share of HFT by trading venue (shares of order books) ................. 37
Table 3. EEA trading venues volumes (thousands of shares;
by order book) ...................................................................................................... 41
Table 4. Dark liquidity in the EU (% of total EEA trading)*.......................... 56
Table 5. Dark liquidity in the EU (% of EEA RMs, MTFs and
BCNs trading)* ..................................................................................................... 56
Table 6. MTFs run/owned by exchanges ....................................................... 116
Table 7. Auction versus dealer markets ......................................................... 140
List of Boxes
Box 1. Defining dark in securities trading ......................................................31
Box 2. Defensive Strategies ..................................................................................54
Box 3. Transaction reporting: Current regime and CESRs proposals ..........72
Box 4. The US consolidated tape and quote ......................................................82
Box 5. Investing in bonds.....................................................................................87
Box 6. The TRACE experience ............................................................................92
Box 7. A case study of national stock exchanges:
Next steps after the demutualisation ...............................................................115
Box 8. The regime for systematic in|ternalisers .............................................120
Box 9. Organised trading for OTC derivatives: EU and US discussions ....127
Box 10. OTC equity transaction type standards .............................................131
Box 11. The role of issuers .................................................................................133
Box 12. Order types ............................................................................................137
Box 13. The US flash crash: what can be learned? .......................................146
Box 14. Regulatory shift: the growing role of macro regulators ................156
Box 15. A new regime for investment advice: the FSAs proposal ..............183
Box 16. MiFID and asset management at a crossroads..................................197
PREFACE
The Task Force aimed to shed new light on the main aspects of the Markets
in Financial Instruments Directive (MiFID), drawing on the vast practical
experience of the Task Force members, the timely input given by regulators
and the European Commission and the benefit of a substantial body of
academic research and analysis.
It was acknowledged at the outset of the Task Forces work that the
revision of MiFID was to be more than an ordinary assessment conducted a
few years after the introduction of a new directive, and that it was
important to take into account the lessons of the financial crisis in a context
of growing uncertainty about the future of Europe's capital markets. It was
believed that the review represented a significant opportunity to
strengthen the role of the internal market and to regain investors'
confidence, which had been badly damaged by the recent financial crisis.
As noted in the G20 context, MiFID is part of the process of strengthening
the effectiveness of regulation, especially in areas that had previously been
unregulated. Greater transparency, investor protection and market stability
are some of the main steps that will give financial markets sounder, safer
and more resilient foundations.
MiFID is one piece in a broader legislative and regulatory framework
regarding the financial markets and it is important to ensure that the
revised MiFID maintains appropriate coherence with other current and
forthcoming laws and regulations, whose interplay in the implementation
process should preserve legal certainty and a properly harmonised regime
for investment services.
The work of the Task Force and the final report, independently
drafted by the rapporteurs, come at a crucial moment for the MiFID review.
Through the Task Force, the industry has sought to make a contribution
that can play a major role in supporting the work of public authorities to
disentangle the effects of the crisis from the issues raised by separate
market failures or regulatory and supervisory gaps. The report is therefore
an attempt to make a thoughtful qualitative and quantitative analysis of the
|i
ii | PREFACE
major problems at stake and where feasible, to put forward proposals that
take into account market developments and reflect the latest findings in the
academic literature.
The Task Force was composed of market participants from practically
all relevant segments of the securities industry, whose interests and
sensitivities clearly did not always coincide. Therefore a consensus could
not realistically be reached on a number of important issues, but it did
create a unique opportunity for the Task Force to lay out the issues in a fair
and balanced way and to articulate the arguments of the different views of
these industry segments, providing legislators, regulators and supervisors
with a deeper insight into market views and the state-of-art literature. The
issues at stake are often arcane and highly technical, but the way in which
they are tackled by the public authorities in the coming months will have a
major impact not only on the functioning of the financial markets and on
the competitive position of the various players involved in the chain
between investors and issuers, but also, most importantly, on the well-
being of large and small investors and business units in Europe.
This is why the Task Force felt that it was more important to
articulate the issues and the different views that prevail than to purport to
submit elusive and vague consensus positions or high-level conclusions to
which each and all members of the Task Force could agree. The latter
approach would mask the real issues at stake and be an ineffective basis on
which public authorities could make their own policy decisions.
This objective guided the discussions in the five intense meetings
held in the second half of 2010. A key factor of success was going to be that
all segments of the industry would be able to express their views in an
open and constructive way, that the rationale for the various views would
be fully articulated and understood, and that reliable supporting data
would be provided to allow for a neutral and balanced assessment. It is to
the credit of all Task Force members that, notwithstanding the complexity
and sensitivity of the issues, they have successfully met this challenge. And
it is to the credit of the rapporteurs that they have produced a report that
sums up these issues in a lucid and balanced way. While this report will
not make the task of the public authorities seeking to define the future path
of MiFID altogether easy, it should at least allow it to be placed on a
stronger factual and conceptual basis. If one considers the complexity of the
task, this is a worthwhile outcome.
Pierre Francotte
Chairman of the Task Force
EXECUTIVE SUMMARY
The review of the Markets in Financial Instruments Directive (MiFID) is a

worthwhile but delicate exercise, coming at a time of great uncertainty for
the economic outlook and the impact of re-regulation in Europe. The
review is an opportunity to boost investor confidence and strengthen the
resilience, efficiency and transparency of financial markets and
instruments. Investor protection and market efficiency should remain the
guiding principles of the Directive, but it must also be compatible with the
growing demand for market safety and financial stability. This report is
therefore an opportunity to contribute to the debate led by the European
Commission and the new European Securities and Markets Authority
(ESMA) by providing an articulated and balanced position on MiFID.
While initially limited to strengthening conduct of business
requirements and reviewing mandatory areas, several factors now call for a
more extensive review of the Directive. New priorities have emerged as the
debate on the causes of the financial crisis has advanced and new
institutions have been established on both sides of the Atlantic to monitor
markets and macro-prudential risks. Also, the advantages of technological
innovation are seen differently today than they were before the crisis.
The functional approach of MiFID is set to become more prescriptive,
in line with other forthcoming legislation and the G20 objective to leave no
area of financial markets unregulated. By reducing the number of
exemptions and extending the scope of the Directive, the breadth of
implementing measures will expand, together with the role of ESMA and
national supervisors. In this context, there will be almost no scope for self-
regulation. Instead, the task will fall upon supervisors to keep rules up to
date with fast-moving market innovation.
The objective of the review is to improve market integrity, stability
and efficiency, as well as investor protection. Given the breadth of the
|1
2 | EXECUTIVE SUMMARY
changes foreseen, this Task Force has chosen to focus on three core areas:
1) Transparency
2) Market structure
3) Provision of investment services
While the review has identified some regulatory gaps, it mainly seeks
to improve the way in which the Directive is implemented and enforced by
national authorities. The European Commission and ESMA should
minimise the risk of adding layers of regulation where failures are the
result of inadequate supervision or enforcement. Clarifying intended
scopes of current regulation may help to create a more harmonised
framework of supervisory practices. Most importantly, the review should
clarify ambiguities in the legal text when the application of the law is
inconsistent.
Investors should benefit from the new MiFID. The revision of the
legal text should aim at solving legal and market divergences across
Europe, and make sure that the benefits of the new competitive
environment are spread along the value chain and passed on to final users,
retail and wholesale investors, as appropriate.
When it comes to assessing the effects of MiFID, it is difficult to
disentangle the impact of regulation from the effects of the recent financial
crisis. Nevertheless, the creation of a harmonised regulatory framework for
the provision of investment services in Europe is a paramount achievement
of the Directive.
MiFID has changed European capital markets in many ways. It has
brought greater competition between trading venues and between
investment firms both on trading costs and execution services. It has also
contributed to substantial investments in technology for trading and
platforms. The growth of dark trading venues, such as MTF (multilateral
trading facility) dark pools and broker-dealer crossing networks (BCNs),
has been another interesting consequence of the new environment. Finally,
the Directive has widened the scope of transparency requirements,
harmonised the framework of business conduct rules and improved the
protection for investors.
Still, there is scope to bring more clarity to some definitions and
further harmonise rules and supervisory practices. In other areas, such as
market quality (price formation) and integrity, the impact of the Directive is
not yet apparent, since evidence remains fairly controversial and
inconclusive.
MIFID 2.0: CASTING NEW LIGHT ON EUROPES CAPITAL MARKETS | 3
Transparency
Transparency plays a crucial role in the smooth functioning of financial
markets and the monitoring of systemic risk. It also ensures that the
process of price formation works well, through efficient price discovery
mechanisms. Regulatory action is needed in some respects, however, not
only to take stock of the recent financial crisis but also to assess the
experience gained since the transposition of the Directive.
Transparency is no panacea for market failures, however. Ill-defined
transparency requirements would harm efficiency in less-liquid markets
with no increase of investor protection or reduction of systemic risk.
Conversely, markets could become less-liquid and thus more volatile.
Hence, regulatory intervention should be proportional to the structure of
each market and take into account the dynamics through which orders find
their market-clearing price.
Retail and wholesale markets are intrinsically different due to their
divergent market structures. Yet, the design of transparency requirements
should not only look at the nature of investors, but also at the
characteristics of the market itself. The Commission may need to launch a
public consultation in order to request data that would constitute a
technical basis on which to better distinguish retail and wholesale actors for
each financial instrument.
Equity markets
Pre-trade transparency supports the smooth functioning of venues trading
mechanisms, as well as efficient price discovery and implementation of best
execution policies.
Under certain conditions, pre-trade transparency may impair market
liquidity. Hence, MiFID introduced waivers, which should be retained. A
move towards a more rule-based approach should be balanced with
flexible application and ongoing supervision in order to meet market
needs. However, conflicting views between members emerge when
discussing the breadth of these exemptions.
Turning to post-trade disclosure, the financial crisis called for a
further layer of market transparency. A new regime should include the
disclosure of aggregate data on capital markets to monitor systemic risk
and increase market integrity and efficiency. The extension of trade
reporting to both shares admitted to trading only on MTFs or organised
trading facilities and to equity-like instruments would be helpful, since all

these instruments serve similar purposes. However, for other financial
instruments, the mere extension of the rules for equities would most likely
generate inconsistencies, given their diverse nature.
In the post-MiFID environment, several aspects have contributed to
reducing the quality of data and hindering its consolidation. The MiFID
review should look at the standardisation of both data formats (code
identifiers, etc.) and flags to solve issues in some specific areas (e.g. OTC
trades). The relevance of trade flags stems from the support they offer to
liquidity discovery mechanisms across trading venues. Market initiatives
should reduce the number of trade flags, currently around 50, to fewer than
10 across Europe.
In this regard, ESMA should guide the currently industry-led
initiatives to improve standardisation and reduce inconsistencies.
However, either the Commission or ESMA should be able to impose
consistency if commercial initiatives do not lead to a satisfactory solution
within a reasonable timeframe.
The disclosure of data in real-time is fundamental for the efficiency of
price formation processes. New technologies can help reduce delays.
However, reducing the maximum allowed for trade reporting from three
minutes to one minute misses the point; regulators supervision should
make sure that firms are not taking more than real time for other reasons
than technical delays. In any case, the legal obligation is to report as close
to real time as possible and should be duly enforced. The industry is
working to make all market data that is not subject to delays freely
available after 15 minutes, in line with ESMAs recommendations.
Moreover, delays should be permitted in specific circumstances, with
appropriate calibration for trades done at the end of the day.
Despite the importance of a consolidated quotation system, priority
should be given to removing obstacles to the use of consolidated post-trade
data solutions. In particular, it is important to improve investor access to
both pre- and post-trade data. The challenge is to promote beneficial
competition for end investors and support best execution.
To achieve consolidation, formats would need to be standardised and
granularity increased, which would curb costs for users and increase
accessibility. Some market participants have already committed to reduce
costs for final users by unbundling fees for pre- and post-trade data; a step
in the right direction that data vendors and distributors would do well to
follow. When lower data collection costs are realised, they should be
passed-on to final users. Regulators and competition authorities should

draw attention to potentially unfair market practices and anti-competitive
market conditions that impede markets from offering data solutions at a
reasonable cost, rather than attempting to define when a cost is actually
reasonable.
The US experience with the unintended consequences of a
consolidated tape run by a public entity should suggest alternative
solutions. Consolidated tapes can be designed and offered by competing
data operators (so-called Approved Public Arrangements or APAs), once
the rules of the game have been clearly defined and duly enforced. These
tapes could cover not only shares but also other financial instruments
admitted to trading on regulated markets (RMs), MTFs or organised
trading facilities (OTFs), as long as a sound regime has been put in place.
Regulators should set the conditions to facilitate the consolidation
and timely delivery of data to investors in real time, fully unbundled
through APAs. Operators would have to meet strict requirements and be
responsible for the detection of multiple publications (misreporting or
double-counting). Only if the industry fails to meet these conditions, ESMA
should adopt the necessary arrangements to set a single consolidated tape
in the EU.
Confidential transaction reporting to regulators serves the integrity of
financial markets. Extending the scope of the reporting may reduce the risk
of manipulation on less-liquid markets and improve market integrity and
surveillance. Disclosure would then cover all financial instruments
admitted to trading on RMs, MTFs and OTFs, with no distinction as to
where and how they are actually traded. This move would need to be
coordinated with other EU initiatives, such as the review of the Market
Abuse Directive (MAD), and the harmonisation of current supervisory
practices across Europe. Information should be meaningful and not
overburden regulators supervisory activities.
Bond markets
A strong push towards pre-trade public disclosure for non-equity financial
instruments would require a rethink of the current market structure for
less-liquid asset classes, and consequently a shift in the intermediation
towards auction markets and from a mainly institutional demand to a more
retail and small professional one to ensure a constant and sufficient
demand over time. Pre-trade transparency is, therefore, strictly needed in
auction markets in order to stimulate investors willingness to bid, since

they are able to see other parties binding commitments in advance.
Some market participants agree that available data are sufficient and
that further pre-trade transparency is not really needed. Instead, at this
stage, they believe regulators should focus on improving post-trade
disclosure.
Other market participants, however, believe that an ad hoc pre-trade
transparency for bonds would be appropriate. This regime would apply
waivers, and would cover bonds listed on regulated markets or multilateral
trading facilities or organised trading facilities, whether order-driven,
quote-driven or inter-dealer platforms. Information should be more easily
available, to help price discovery and open the market to retail and small
professional investors.
A transparency regime for bond markets should provide meaningful
information to stimulate price discovery. The speed, breadth and depth of
information should be designed around dynamic liquidity measures.
Since there is not a single measure of liquidity readily available,
transparency requirements should be developed on an instrument-by-
instrument basis. This task should rather be left to secondary legislation,
such as Level 2 implementing measures or binding technical standards.
Liquidity is a dynamic aspect, which may take different forms
according to the characteristics of the financial instrument and the trading
mechanism. This reality should be taken into account which allows
exemptions or deferred publication in order to preserve an efficient and
sound price formation process. Dynamic measures of liquidity can be
designed around criteria such as frequency of trades, overall turnover or
prospective liquidity, product standardisation or transaction size. Finally,
since data are fragmented, data formats and flags may need to be further
standardised for the purpose of pre-trade transparency.
Derivatives and structured financial products

Derivatives and structured products are mainly traded on a bilateral basis.
These trades occur either purely over-the-counter or through inter-dealer
platforms under a request-for-quote (RFQ) model. A pre-trade transparency
regime, therefore, should be designed in a different way than for auction
markets. Liquidity in dealer markets, such as markets for bonds,
derivatives and structured products, is underpinned by quote-driven
auctions, inter-dealer platforms or bilateral negotiations. To function well,
bilateral markets (interdealer or purely bilateral) need less pre-trade
transparency than order-driven ones, such as equity markets. Executable

prices might thus not always be consistently available. The current market
structure, however, does not impede future market developments in the
years to come towards a different structure of intermediation and nature of
the demand.
Current MiFID rules on transparency were devised for equities, so
applying them bluntly to derivatives and structured products could
generate unintended consequences for the incentives for dealers to provide
liquidity by using their private information. They could potentially exit
illiquid products and confine their activities to those that are inherently
more liquid. Pre-trade transparency for complex non-equity products
would only work by first changing the way these products are traded and
the nature of the demand, which would need to be constant and sufficiently
high over time (as for auction markets).
A post-trade transparency regime for derivatives and structured
products should be more detailed and tailored to the nature of these
products. Where listed on RMs and/or MTFs, the regime could be
designed with the same methodology employed for bonds, but its
implementation should follow a phased approach.
Exemptions and due calibrations should be allowed in order to
preserve efficient price formation and guarantee the effective monitoring of
systemic risk. Calibrations should take into account the nature of these
markets and each financial instrument, rather than a division into broader
categories (e.g. by asset classes). A mere application of post-trade
transparency to a general list of instruments would definitely hamper
market liquidity. As for bond markets, measurements of liquidity should
be taken into account with due care to avoid adverse consequences in terms
of liquidity for wholesale participants.
The extension of trade reporting to non-equity markets can be
facilitated by current infrastructures, thereby reducing costs. It is critical,
however, that transparency requirements remain independent from the
eligibility criteria for trading and clearing. In effect, the requirements to
access a central counterparty clearing (CCP) are not based only on liquidity
itself but also on technical and legal aspects. Hence, the mere eligibility for
the clearing of an instrument is not an appropriate test for transparency
purposes, nor is it appropriate to assess the level of liquidity or the
frequency of trades.
Structured financial products (SFPs) and OTC derivatives are good

tools to free capital back to the real economy and better allocate risks and
resources. They help spread credit risk and transfer it to those who are
more able to bear it. However, the financial crisis has taught us that
spreading risk among counterparties through complex instruments does
not ultimately cancel it out.
Information about the underlying assets and net exposures should
always be publicly available in aggregate level to monitor systemic risk.
Regulators should have access to data via reporting, especially through
trade repositories. During times of financial distress, aggregated
information on net exposures would help contain herd behaviour set off by
market opacity.
Trade repositories have sufficient skills and capabilities to collect and
aggregate information about net exposures. Financial institutions would
need to disclose information on net exposures in a way that does not
compromise the confidentiality of sensitive information, but at the same
time provides meaningful information (more detailed) to regulators and to
the market. Trade repositories would then be able to aggregate this
information and offer a global picture.
Narrowing exemptions for commodity derivatives under MiFID may
have a substantial impact on the business of non-financial companies. Some
market participants advocate these changes as an important step towards a
more level playing field, with greater transparency and investor protection,
while others ask for further research on the consequences of narrowing
these exemptions, in terms of costs of hedging relevant exposures. The
need for consistency across several regulations in the commodity business
probably demands a more articulated answer.
Curbing speculation is a vague objective, given that the distinction
between hedging and speculative trading remains highly controversial.
Regulators, instead, should shed light on the risks of price manipulation
that arise from the accumulation of dominant net positions in future and
derivatives markets. Strengthening supervisory powers, especially through
position limits, could help to control price manipulation, in particular for
physical markets or markets for non-storable commodities (e.g. electricity),
but would not necessarily mitigate systemic risk. A more effective
alternative would be to use position management tools to impede net sizes
reaching a dominant position.
Market structure
Competition among trading venues on execution services and among
investment firms on the provision of other investment services have
generated positive effects, such as lower trading costs. Competition needs,
however, to be fair and based on a level playing field among MiFID official
trading venues. Markets benefit today from the interaction of various
groups of users and platforms.
Being a vital part of the network, competition appears where markets
are contestable, not only from a technological standpoint but also in terms
of fair market practices. This implies the need to lower barriers to entry and
exit for users and platforms (contestability), including sunk costs, as well as
to monitor market practices in a dynamic way.
This report also acknowledges the importance of ensuring a
harmonised approach in the application of MiFID requirements for official
trading venues. Regulated markets and multilateral trading facilities
should be subject to convergent legal obligations and supervisory oversight
across member states. This convergence already exists in some European
countries such as the UK.
On the classification of broker-dealer crossing networks (BCNs), the Task
Force discussed two different views.
On the one side, it is argued that some OTC trading escapes MiFID
rules on pre-trade transparency for trading platforms and systematic
internalisers, and does not provide sufficient post-trade information. It also
alleged that it escapes MTFs rules on access, discretion and surveillance. In
this view, some of the BCNs would perform the same function as RMs and
MTFs, while others may be more akin to systematic internalisers. BCNs
would operate as multilateral trading mechanisms by matching trades as
riskless counterparty. BCNs that do not meet the definition of OTC
trading would rather be classified either as multilateral trading platforms
or systematic internalisers (SIs). Otherwise, pre-trade transparency
obligations would be circumvented to the detriment of price discovery,
investor protection and market integrity.
Conversely, other market participants believe that advanced
brokerage, offered to wholesale counterparties through BCNs, meets the
MiFID definition of OTC trading. They note that trades may fall below the
standard market size only after splitting them to reduce market impact.
These child orders would then be internally matched, but mostly routed
to external trading venues, such as RMs and MTFs. The Directive, in their
view, would actually refer to parent orders, since fiduciary duties and
conduct of business rules apply to them as a whole. Finally, a Broker-dealer
Crossing System (BCS) should not be considered multilateral since the
dealers assume risks when providing best execution and other conduct of
business arrangements, which neutral platforms do not provide.
OTC equity trading plays an important role in financial markets, in
particular when it comes to the best execution of complex institutional
orders. Therefore, the review of MiFID should not ban these trading
activities. The review should rather clarify the criteria that define OTC
trades with proper flexibility, in particular what kinds of trades are subject
to OTC requirements under MiFID (e.g. child or parent orders). In
addition, it should foster the availability of data in order to allow the full
assessment and enforcement of best execution rules. Finally, it should
ensure that price formation processes and market quality are preserved
and, if possible, improved.
The data employed in the run-up to the MiFID review by the
Committee of European Securities Regulators (CESR) to ascertain the size
of OTC equity trading are not sufficiently accurate. Analysis based on this
data cannot be considered conclusive and has probably led to an
overestimation of the size of this market. More effort needs to be made to
assess market quality in Europe with accuracy and to clarify the actual size
of OTC equity trading, its origin and its impact on price formation
processes. There is a compelling need to improve the quality of market data
by reducing inconsistencies and increasing granularity through the use of
harmonised flags.
The G20 called for derivatives to be further standardised and traded
on electronic platforms, where appropriate. A greater push towards
standardisation and organised trading should balance the benefits of a
more transparent and orderly setting with the costs incurred by a
potentially lower availability of customised derivatives, which would mean
a greater possibility to leave some risk in the system not properly hedged.
These proposals raise challenges for regulators and market participants
with regard to investor choice, market liquidity and efficiency, and
potential overlaps with the MTF regime. The MiFID review needs to
address them in a way that is proportional and consistent with the
European Market Infrastructure Regulation (EMIR) and other relevant EU
directives.
Financial market infrastructures are networks that can operate in a

competitive environment. MiFID and competent authorities may need to
strengthen actions to keep barriers to entry and exit low, giving due
attention to economies of scale and scope, and other potential efficiencies.
The Code of Conduct improved price transparency. More remains to be
done, however, to solve existing commercial and technical challenges and
increase accessibility through unbundling and interoperability of current
infrastructures.
Ultimately, MiFID favours freedom of access by investment firms to
competing market infrastructures. In effect, while the original Directive
envisaged non-discriminatory access to competing infrastructures, the
transposition of this provision and its enforcement have been inconsistent
across EU member states, in particular in relation to the definition of the
legitimate commercial ground to deny access. The review of MiFID
should try to bridge these inconsistencies, together with other legislative
proposals such as the European Market Infrastructure Regulation (EMIR),
Securities Law Directive (SLD) and Central Securities Depositories
Regulation (CSDR).
Promoting structural changes in non-equity markets to give easier
access to retail investors may raise conflicting views. For some market
participants, the market for fixed income securities should remain
wholesale and dealer-driven. In their view even though a commendable
objective direct retail access to non-equity instruments may destabilise
these markets, as it may generate higher volatility with no liquidity
enhancements. These effects would ultimately heighten risks for retail
investors, given the increasing complexity of fixed income securities.
Instead, other stakeholders firmly support the opening of bond markets to
retail investors. They believe greater transparency would be a liquidity
driver for these markets. Under proper delays and exemptions, the
potentially negative impact of retail trading activities would be fairly
limited.
Issuers are important actors for innovation and a more efficient
allocation of resources in our economies. Some market participants believe
that issuers should be duly informed and agree where their shares are
traded in secondary markets. This proposal, however, does not find wide
support among both market participants and policy-makers, since in their
view it raises relevant issues of market efficiency, legal entitlement of the
right and conflicts of interest.
The systemic importance of modern capital markets highlights the

inner tensions between financial stability, market efficiency and
technological innovation. A well-functioning market must balance
efficiency and safety to avoid disequilibria. The role of technology in the
configuration of market infrastructure has become ever more important.
Markets are redesigning their infrastructures thanks to innovative
technology. In short, speed and volume will probably continue to grow, but
trading venues may have to deal with more frequent crises and outages.
Technological innovations and techniques have brought
revolutionary changes to trading. Among these changes, there are major
benefits, such as better order management and control of market impact; or
more efficient and faster feed of information into prices. Both these changes
generate gains in terms of lower spreads and better price discovery that
benefit participants throughout the value chain.
However, modern trading also presents a number of challenges, such
as an increase in fundamental market volatility, which in turn has brought
speed and volumes to critical levels. Advanced execution services like
direct-market or sponsored access have radically increased speed and
volumes for transactions, in an attempt to cope with increasing volatility.
Limits to infrastructure capacity nevertheless mean that higher speed and
volumes risk generating market disorder and financial instability.
To overcome these challenges, intermediaries and trading venues
need to strengthen their own monitoring. A coherent set of emergency
procedures in case of market disruptions should be designed in
consultation with market participants (e.g. circuit breakers). There are
several efficient monitoring systems already in place, which could serve as
model systems. Finally, trading rules should be harmonised across markets
to avoid instability arising from arbitrage.
Provision of investment services
Business conduct rules and organisational requirements play a crucial role
in the provision of investment services, by strengthening investor
protection and market integrity across Europe. However, a more
harmonised implementation of regulatory actions and supervisory
practices by member states should become a priority.
Best execution duties lie at the foundation of the fiduciary relationship
between service providers and clients. MiFID tries to grasp all factors
influencing the best execution of a financial transaction, and does it in a
very general manner. This situation, however, is not necessarily to the
detriment of final investors as long as execution policies are properly

implemented and data allows sufficient verifiability of execution.
Investors receive information about their execution policies that may
be sometimes incomplete. Even with most advanced execution metrics,
available data appear insufficient to make a thorough evaluation of
execution quality in a fragmented market.
Some market participants argue that best execution policies should be
applied dynamically and that simply complying with formal legal
requirements is not enough to provide authentic best execution. Other
participants however challenge the view that issues with best execution
come from execution policies that are MiFID-compliant. Both recognise that
those issues emerge from a consistent lack of data on execution quality
from trading platforms, which is under discussion in the debate on the new
transparency regime.
Although the crisis showed that some of the eligible counterparties
(ECPs) were not able to understand risk properly, the client categorisation
regime should not be subject to a major overhaul, other than those changes
currently proposed for local authorities. Some market participants suggest
that portfolio managers should be entitled to unilaterally require the
reclassification as retail clients. They argue that even though portfolio
managers are eligible counterparties under MiFID, they manage money on
behalf of professional and retail clients and they have the obligation to act
in their best interest. Others, however, advocate the inappropriateness of
this request since portfolio managers do not simply execute on behalf of
their clients, but they would directly gain (commissions on profits) from
rules that should theoretically only benefit retail investors.
Investment advice must always be suitable under MiFID. Some
market participants claim there should be a thorough review of the
mechanisms of incentives to make advice more independent. In
particular, they argue that investment advice should be based on a neutral
and independent system of remuneration, by only receiving fees from the
clients. No commission should be set by product providers.
Other market participants argue that obliging investors to pay for it
would increase the access costs to these services and dramatically reduce
the use of advisory services, with potential long-term costs for end
investors. They suggest keeping the advice fee embedded in the
commission, but disclosing whether the advisor is solely remunerated by
the client or also by a product provider.
All cost items and remuneration arrangements should be fully

disclosed before signing the contract. This would improve the ability of
investors to choose the service that best suits their own interest. Changes in
MiFID will need to be reconciled with other regulatory initiatives such as
Packaged Retail Investment Products, the Insurance Mediation Directive
review, and the Prospectus Directive.
The suitability test is crucial in the provision of investment advice
and portfolio management services. Views diverge on how to assess the
knowledge, financial situation and objectives of investors, and in particular,
on how deep the suitability test should look into investor habits and the
willingness to take risks. Some regret the lack of harmonisation in the
implementation of suitability requirements for discretionary portfolio
management. They would welcome action by ESMA to improve legal
certainty and reduce barriers to market investment products across the EU.
Other market participants, however, do not think any intervention is
needed in this regard and are satisfied with the current level of
harmonisation, which takes into account the natural market differences
among member states.
MiFID foresees a different regime for execution-only services based
on the product classification between complex and non-complex financial
instruments. Any change should take into account the fact that complexity
does not necessarily mean more risk. The objective should be to verify if the
product is in line with investors understanding of the ultimate risk that
they are finally going to bear. Regulation should not decide the level of risk
investors want to take.
Some market participants believe that certain UCITS (Undertakings
for Collective Investments in Transferable Securities) might have become
too complex to be easily understood by investors to skip the
appropriateness test, at least for retail clients. Others argue that
classification, in particular for UCITS, should remain as such, since a
change in classification could damage the UCITS brand outside the EU.
MiFID rules on conflicts of interest represent a first step in the
introduction of a common approach across Europe for the prevention,
identification, management and disclosure of such conflicts. Further
initiatives to strengthen the current regime and align supervisory practices
would enhance the treatment of these conflicts and benefit financial
markets. A harmonised set of sanctions should be combined with enough
flexibility for member states to adapt rules and procedures in line with
their national contexts.
Organisational requirements play a crucial role in ensuring business

continuity, market integrity and investor protection. A proper
implementation of the Directive should be ensured by harmonising
organisational requirements and supervisory practices across Europe, and
by removing ambiguities in the legal text. Ensuring consistency with other
upcoming legislation would then avoid inefficiencies. It would also
promote a uniform regime of investor protection and market integrity
within Europe, which would ultimately increase legal certainty and the
attractiveness of investment services.
1. INTRODUCTION
T
his report deals with the regulatory and economic challenges faced
by European regulators in the process of reviewing the Markets in
Financial Instruments Directive (MiFID). It does not attempt to cover
all issues and details raised by the MiFID review. Rather, it focuses on three
fundamental areas of the regulation of financial markets and instruments:
1) transparency, 2) market structure and 3) provision of investment
services. Within these three areas, the report examines the major trade-offs
faced by regulators, industry representatives and investors in the review
process. The summary of recommendations merges the findings of this
report with an objective and analytical contribution to the regulatory and
policy-making process
1.1 The MiFID review

The MiFID review is perceived by some as a way to address pending issues
of the original Directive, and by others as a way to reduce imbalances and
to level the playing field. Both views converge on the necessity to restore
investors confidence. As a result, the exercise will have a high political
visibility as one of the major responses to the recent financial crisis.
Revising a piece of legislation three years after its introduction is a
standard but worthwhile procedure. Yet, striking the right balance between
potentially conflicting objectives, such as financial stability and market
efficiency, is a delicate exercise. The review aims to further harmonise the
regulatory framework for investment firms and market infrastructures,
while the impact of the wider post-crisis re-regulation in these and other
areas remains unclear. It is a consequential piece of legislation, which
affects the competitive position of a number of players. It also comes at a
delicate moment for Europes capital markets, where it is necessary to
rebuild investors confidence and to strengthen the internal market. Due to
16 |
its level of detail, the review of MiFID is a very technical undertaking.

Therefore, European regulators need qualitative and quantitative data
through the support of private initiatives.
The Directive mandates a review of nine areas:
1) Extension of pre- and post-trade transparency requirements to
financial instruments other than shares (Art. 65, MiFID);
2) Removal of obstacles to the consolidation of data (Art. 65, MiFID);
3) Systematic internalisers provisions (Art. 65, MiFID);
4) Tied agents (Art. 65, MiFID);
5) Best execution (information on execution quality; Recital 76,
Implementing Directive);
6) Gold-plating;
7) Telephone recording (Arts 51.4 and 51.5, Implementing Directive);
8) Requirements for authorisation (Arts 31 and 32, MiFID); and
9) Commodity derivatives exemption (Art. 65, MiFID).
However, other issues will also be considered, including most of the
G20 commitments to increase the safety and stability of the financial
system. Amongst the most important categories are the following:
1) Transparency requirements (such as pre-trade transparency issues
and waivers, delays on publication time and a new transaction
reporting regime);
2) Regulatory boundaries between venues, e.g. levelling the playing
field between regulated markets (RMs) and multilateral trading
facilities (MTFs);
3) Legal treatment of broker-dealer crossing networks and other
platforms currently operated in the over-the-counter (OTC) space for
equity and non-equity financial instruments;
4) Issues of financial stability (rules for new trading technologies and
implications for financial stability);
5) Investor protection (e.g. selling practices, market access, client and
product categorisation) and
6) A special regime for SMEs.
See Table 1.
18 | INTRODUCTION
Table 1. The MiFID review

Mandatory areas Other areas
Data consolidation Pre-trade transparency waivers
Systematic internalisers regime Transparency for non-equity asset
classes
Tied agents Transaction reporting (with MAD)
Commodity derivatives exemption Trading venues classification and legal
treatment
Requirements for authorisation Disorderly markets
Gold-plating Selling practices (with PRIPs)
Best execution (data execution quality) Market access
Telephone recording Client categorisation
Product classification (with UCITS Dir.)
Special regime for SMEs
Conflicts of interest and inducements
Organisation requirements
Investment advice
Safeguards and asset segregation
Post-trade transparency delays
Source: European Commission (2010b).
The Committee of European Securities Regulations (CESR) has

grouped the issues of the review under three major areas: 1) transaction
reporting, 2) equity markets (structure and transparency) and 3) investor
protection and intermediaries (CESR, 2010a,b,c). Therefore, this legislative
initiative will have to interact with other important initiatives, such as the
review of the MAD (Market Abuse Directive),1 that cover diverse segments
of the investment services industry (see Figure 1).
1 Directive 2003/6/EC.
Figure 1. Interaction with main EU regulatory actions
UCITS
MAD PRIPs
MiFID
CRD EMIR

Source: Author
Some outstanding issues regarding MiFID, however, are not related
to regulatory gaps, but to the way in which the Directive was implemented
and enforced by national authorities. The European Commission and the
new European Securities and Markets Authority (ESMA)2 need to look into
the nature of these problems when considering changes to MiFID.
Regulators should minimise the risk of adding layers of regulation where
failures are rather the result of inadequate supervision or enforcement.
Clarifying the intended scope of current regulation may help to create a
more harmonised framework of supervisory actions. The revision will
affect not only the rules of the business conducted by those providing
intermediation as well as trading venue services, but also the interests of
investors (both institutional and retail) and issuers.
MiFID came into force in November 2007 (see Figure 2).3 As
cornerstone legislation, it had broad implications for many institutions
across sectors, including investment firms, trading venues and regulatory
authorities. On the one side, the Directive has unequivocally promoted the
2 Replacing CESR as of January 2011.

3 For details on the transposition of the Directive across Europe, see http://ec.europa.eu/
internal_market/securities/isd/mifid_implementation_en.htm.
20 | INTRODUCTION
creation of a new harmonised framework of rules across Europe that had

not existed before (e.g. conduct of business rules and a new pre- and post-
trade transparency regime). On the other side, this regulation has promoted
a competitive environment for the provision of investment services.
However, implementation and supervisory controls were not uniform
across member states, creating an uneven playing field among market
participants, which the review should solve. There needs to be a better
understanding of the impact of MiFID on European financial markets, as
the current economic and financial crisis may have distorted the real
impact of its implementation. The next sections provide some thoughts and
data for assessing the major effects of the Directive.
Finally, as shown in the timeline, the review must follow a strict set of
deadlines in order to be approved by 2011 and followed by implementing
measures by 2012.
Figure 2. MiFID timeline
Source: European Commission website and rapporteurs.

2. THE MARKETS IN FINANCIAL
INSTRUMENTS DIRECTIVE
2.1 The legislative process

The MiFID Directive and its implementing measures are one of the most
important components of the Financial Services Action Plan (FSAP). The
FSAP represents the follow-up of the 1998 European Councils Conclusions
and the Commissions Communication,4 condensed in the 2000 Lisbon
Agenda. The FSAP included 45 measures5 (originally 42; of which 29 were
directives) aimed at developing a single market for financial services, and
was renewed by the Europe 2020 agenda (European Commission, 2010a).
This plan represents one of the foremost attempts of the European
institutions to improve EU financial markets integration6 and to favour a
4 See European Council, Presidency Conclusions, 15-16 June 1998, Cardiff

(http://www.consilium.europa.eu/ueDocs/cms_Data/docs/pressData/en/ec/54315.pdf)
and European Commission, Financial Services: Building a Framework for Action, COM 625,
28 October 1998, p. 3 (www.europa.eu).
5Originally there were 42 measures, 29 of which were directives (plus the amendment to the
14th Company Law Directive, which was blocked). The Commission then added a
Communication on clearing and settlement (COM 312, 2004) and a Regulation on cross-
border payments (EC Regulation 2560/2001).
6 Financial integration of European markets has been on the policy agenda since the 1960s. It
was mentioned as early as 1966 in the Segr Report (see Segr, 1966). In May 2007,
moreover, the European Council reconfirmed the essential role of integrated financial
markets to strengthen the four freedoms of the Internal Market; see Council of the European
Union, Brussels European Council 8/9 March 2007: Presidency Conclusions, 7224/1/07
REV 1, 2 May 2007. A new effort to promote greater financial integration for a European
single market has been recently promoted by the Monti report (Monti, 2010), which put
economic and financial integration at the top of the EU agenda.
| 21
22 | THE MARKETS IN FINANCIAL INSTRUMENTS DIRECTIVE
better legislative process for innovative financial regulation. The FSAP

aimed at boosting cross-border transactions and financial integration; it
remains, however, hard to disentangle the effects of this set of measures
from the effects of the Economic and Monetary Union and more recently of
the financial crisis. Numerous new pieces of regulation such as MiFID,
MAD, the Prospectus Directive, Transparency Directive, International
Accounting Standards (IAS) Regulation, Takeover Bid Directive, Capital
Requirements Directive have provided European financial markets with a
common and unique framework of rules for capital markets.
The MiFID, the MAD, the Transparency Directive and the Prospectus
Directive have been designed and implemented according to a new
legislative process, called the Lamfalussy procedure.7 This new legislative
process entails three levels that represent a combination of regulatory
(levels 1 and 2), supervisory (level 3), and enforcement (level 4) actions. It
foresees extensive consultations with industry and expert committees (so-
called comitology), and strong cooperation between national authorities.
Level 1 directives are approved according to the regular co-decision
procedure.8
The recent financial crisis, however, calls for a deeper analysis of how
these measures and MiFID in particular performed vis--vis the
disruption in global financial markets. Nevertheless, the MiFID review
occurs in the context of discussions around a new set of directives that
make up the European strategy to tackle market failures, restore confidence
and promote the internal market and a more open market architecture to
favour cross-border transactions. However, economic, legal and social
barriers among member states, e.g. taxation, are slowing down the creation
of a fully integrated retail and wholesale pan-European financial market.
Under the overarching goal of ensuring a single market for financial
services, MiFID aimed at establishing a legal framework for integrated
securities and derivatives markets in Europe. The Directive attempted to
change Europes capital markets through liberalisation of investment
services and more investor protection. MiFID, in effect, was a sophisticated
7 Named after the Chairman of the Committee of Wise Men, Alexandre Lamfalussy, who
contributed to defining the terms of this procedure in the Final report of the Committee of Wise
Men on the Regulation of European Securities Markets, Brussels, 15 February 2001.
8Art. 294 (ex Art. 251, ECT) Treaty on the Functioning of the European Union (2010
Consolidated Version; hereinafter, TFEU), Official Journal of the European Union, C83/47, 30
March.
and far-reaching attempt of the Commission to provide the European

Community with valuable tools for meeting the challenges of the monetary
union and reaping the benefits of a single integrated market for financial
services. The revision of the legal text aims at solving legal and supervisory
divergences across Europe, which hamper market integration, addressing
unintended effects of the Directive, extending its coverage and making sure
that the benefits of the new competitive environment spread along the
value chain and are passed-on to final user, both retail and wholesale.
2.2 The legal structure

MiFID aimed at designing a coherent, risk-sensitive and comprehensive
regulatory framework for execution of financial instruments to ensure high
quality of execution and thus promote investor protection, and the
integrity, efficiency and orderly functioning of financial markets (Recital 5,
MiFID). By introducing harmonised rules for investment services and
securities trading, MiFID has opened up the market to competition. It also
provided a unique framework of investor protection rules, as well as
stricter rules to preserve market integrity and efficiency.
The Directive replaced the Investment Services Directive (ISD), which was
a first attempt to introduce a harmonised regime for investment services in
1993, based more importantly on mutual recognition supported by a
European passport. MiFID applies to all financial instruments, excluding
commodity derivatives dealt on own account (Art. 2 (i) and (k), MiFID) and
other instruments and institutions set in Art. 2. In particular, the purpose
of this Directive is to cover undertakings the regular occupation or business
of which is to provide investment services and/or perform investment activities
on a professional basis. Its scope should not therefore cover any person
with a different professional activity (Recital 7, MiFID). Moreover, the
Directive applies not only to MiFID-official trading venues and investment
firms providing investment services on a regular basis, but also to credit
institutions authorised under Directive 2000/12/EC (not including, for
instance, insurance and energy companies) and indirectly to Undertakings
for Collective Investments in Transferable Securities (UCITS) companies,
even though a formal exemption applies (Art. 2.1 (h), MiFID).9
As shown in Figure 3, MiFID aims at reaching two high-level
9 See Box 16.

principles set by the Financial Services Action Plan (FSAP): i) a single

market for wholesale financial services and ii) an integrated securities and
derivatives market. And this through the implementation of two general
goals: investor protection and market efficiency. In the aftermath of the
financial crisis, the need to ensure orderly markets and financial stability
has gradually become another important objective of the review and so too
of the Directive (see Figure 3).
Figure 3. MiFID original principles
Singlemarketfor
wholesalefinancial
services
Highlevel
principles
Integratedsecurities
andderivatives
market
European
passport
Investorprotection Orderlymarketsand Marketefficiency

financialstability Basic
principles
Organisational Conductofbusiness Marketaccessand

requirements rulesandmarket integrity
transparency
Source: Authors.
MiFID pursues these goals via a set of rules, covering four main
areas:
1) European passport (freedom of action across the EEA once received
the authorisation in one of the member states);
2) Organisational requirements (compliance officer, outsourcing,
internal control systems, record-keeping, conflicts of interests);
3) Conduct of business rules (clients classification, best execution, know-
your-customer rules, marketing, information to clients, handling
orders rules); and
4) Market structure and transparency (abolition of the concentration

rule,10 detailed transaction reporting, pre- and post-trade
transparency, other trading venue requirements).
The comprehensive regulatory regime for investment services and
securities trading as modified by MiFID had opened up capital markets
(in particular, cash equity markets) to competition and produced a new
harmonised framework for investor protection, as well as stricter rules to
preserve market integrity.
MiFID adopted a functional approach: it regulates the provision of
investment services irrespective of the institution carrying out the activity.11
The Directive applies to all market participants (in particular, credit
institutions, investment firms and market operators) acting within its
scope, i.e. the investment service activity. For instance, in order to increase
investor protection, investment advice an ancillary service under the ISD
has become a core service under MiFID, thereby obliging all market
players providing this service to be compliant with MiFID.
The economic downturn caused by the troubled financial situation,
however, has brought pressure to modify this approach towards a more
prescriptive one, which aims at ensuring that no areas of financial markets
will remain uncovered by the regulation (in line with other forthcoming
legislations, e.g. European Market Infrastructure Regulation). Reducing the
number of exemptions and extending the scope of the Directive will
increase the number of implementing measures to be defined by ESMA and
national authorities to make legislation effective. Markets will have almost
no space to self-regulate and ongoing supervision will need to keep this
broad set of rules up to date with fast-changing market innovation.
In order to foster competition, MiFID promoted two actions: the
abolition of concentration rules (for order execution services) and the
10Before MiFID, countries applied the concentration rule (Art. 14(3) of the 1993 Investment
Services Directive). This rule stated that member states had the right to concentrate all
orders on a regulated market. In practice, some member states exercised this right either by
restricting all trading or all but the largest orders to be executed on the primary exchanges.
Some member states did not exercise this right.
11A contrasting approach to regulation is the institutional one, which tends to regulate
market operators directly, by virtue of their status as institution or legal entity (as defined
by the legal text).
extension of the ISD European passport regime (for other investment

services). Firstly, by abolishing national concentration rules (pushing
trading on few regulated trading venues, i.e. stock exchanges, when
member states deemed it necessary for an efficient interaction of orders),
MiFID established a new regulatory structure in which all trading venues,
whether run by exchanges or other market participants, compete for trade
execution services on a level playing field. Secondly, the single
authorisation for investment firms and the right to access national
infrastructures on a non-discriminatory basis increased cross-border
competition between investment firms. In other words, competition under
MiFID has two dimensions: competition among trading venues (which
might be operated by market operators or investment firms, but must
follow the same rules when conducting the same activity) and competition
among investment firms when conducting intermediation business.
In parallel to opening up the market for trading venues and
investment firms, MiFID sought to harmonise and strengthen market
integrity and investor protection retail, in particular for the cross-border
provision of investment services, through the use of three sets of tools:
i. Conduct of business rules
The Directive regulates the relationship between retail clients and service
providers through: best execution duties (Arts 19.1 and 21, MiFID),
conflicts of interest rules (Arts 13.3 and 18, MiFID) and suitability and
appropriateness tests (so-called know-your-customer rules, Arts 19.4 and
19.5 MiFID). Other relevant requirements are related to: information to
clients,12 client agreements,13 orders handling14 and marketing rules.15
Finally, specific rules define exemptions for eligible counterparties (Art. 24,
MiFID) and help client classification.16 Overall, these regulatory
12 Art. 19.8, MiFID and Arts 40-43 Implementing Directive.

13 Art. 19.7, MiFID and Arts 14.3 and 39 Implementing Directive.
14 Art. 22, MiFID and Arts 47-49 Implementing Directive.
15 Art. 19.2, MiFID and Arts 24 and 27, Implementing Directive.
16 See also Implementing Directive, Arts 28 and 50. For non-complex instruments, clients can
opt out of these safeguards (appropriateness test), asking for execution-only service. For a
definition of non-complex instruments, see Art. 38, Implementing Directive, or Casey &
Lannoo (2009, p. 51). CESR is looking into the definition of complex vs non-complex
financial instruments. For instance, they might exclude from the group of non-complex
financial instruments all the instruments that embed a derivative and shares in non-UCITS.
UCITS may also be split into two groups consistent with their trading strategies, but the
proposal may come back at a later stage. See CESR (2010c, pp. 25-26).
requirements address typical issues in a fiduciary relationship between

service provider and client.
ii. Organisational requirements for investment firms and trading
venues
These consist of a range of broad obligations: a general duty to comply with
MiFID rules (Art. 13.2 MiFID); creation of internal mechanisms of control
(e.g. independent internal audit, IT, etc.; Arts 13.4-13.5, MiFID);
outsourcing (to be limited to non-core operational services; Art. 13.5,
MiFID); recordkeeping and safeguarding of clients financial assets (Arts 6-
8, MiFID). Other important requirements are: the appointment of a
compliance officer (objective and independent), rules on personal
transactions and effective procedures for risk assessment. In addition,
MiFID also includes specific requirements for trading venues, such as order
execution, access and surveillance.
iii. Transparency regime (pre- and post-trade; see also chapter 1)
While the first two sets of tools mainly apply to investment firms, the third
also applies to both investment firms and execution venues. Legal
requirements impose pre- and post-trade transparency for trades in shares,
while only an optional extension may be exercised by member states for
financial instruments other than shares. The MiFID transparency regime
points at multiple objectives, such as investor protection, market integrity,
market surveillance and market efficiency.
Post-trading infrastructures. Finally, MiFID recognises a general
principle that trading platforms need to provide users with free access to
post-trading infrastructures, for both clearing and settlement (Art. 34,
MiFID). The purpose of this rule is to preserve competition, avoid
bottlenecks along the value chain and avoid cross-subsidisation based on
anti-competitive practices. The Directive thus urges member states to push
regulated markets (RMs) to open up their infrastructure. It also recognises
that building of cross-border links between trading venues and post-trade
infrastructures will further favour the creation of a pan-European market
(Art. 35, MiFID). The right of investment firms to decide their clearing and
settlement infrastructure should apply whether or not the transaction has
been executed on a regulated market. In effect, MiFID asks RMs (Art. 34.2)
to make these links available, since they were the sole providers in the
provision of execution services when the Directive came into force.
3. THE IMPACT OF MIFID ON EUROPEAN
MARKETS: SOME EVIDENCE FROM
CASH EQUITY MARKETS
3.1 Introduction
MiFID is widely recognised as a major regulatory initiative of the European
Union in the financial services sector. It is an overhaul of the European
securities market, aimed at introducing a new, harmonised and competitive
design for the industry. Under the two high-level principles mentioned
above and diagrammed in Figure 3, the main objectives of the Directive
were market efficiency and investor protection.
It is difficult to disentangle the effects (positive or negative) of the
implementation from the impact of the recent financial crisis, which next
to the competitive pressure from newcomers has shrunk the business
volumes of the incumbent provider of execution services (see Figure 4).
Most notably, the reduction of frictional costs for trading and the following
increase in volumes may have been masked by the general cyclical
downturn brought about by the financial crisis. The crisis makes the overall
long-run effects of the Directive on the EU GDP difficult to grasp, but they
are estimated at 0.7% and 0.8% (London Economics, 2010).
However, the number of trades has been increasing, and volumes
seem to be moving to pre-crisis levels in 2010.
In terms of regulation, more needs to be done to fine-tune the current
legal texts with recent market developments and to avoid regulatory gaps.
Hitherto, MiFID has promoted a number of key changes:
1) Greater competition and market fragmentation,
2) Very significant investments in technology for trading and platforms,
3) Growth of dark pools and broker-dealer crossing systems (in line
28 |
MIFID 2.0: CASTING NEW LIGHT
I ON EUROPES CAPITAL MARK
KETS | 29
with maarket fragmenttation),

44) Wider sccope of transp
parency regimee and
5
5) Greater investor pro otection throu
ugh a harmon nised framew
work of
organisaational and con
nduct-of-busin
ness (CoB) rules.
F
Figure 4. Impacct of the crisis on EEA equity turnover
t and voolumes
(lit & dark)
14,562,761,413,782
-46.03%
7,860,089,362,2600
%
13.38%
1,6344,397,496,235 1,415,669,3556,321
838,628,584
+5.66%
8886,100,990
2008
2009
Turnover Volume
V Trad
des
Source: Th
homson Reuters, FESE.
F
The imp pact on price formation

f andd market integ grity remains highly
ccontroversial and with no definitive an nswer. Similarr changes in thet US
s
seem to have brought somee beneficial efffects too, in p particular in teerms of
l
lower implicitt and explicit trading
t costs17 and greater liiquidity (Angeel et al.,
2
2010).
1
17Explicit costs ofo trading are th
hose costs fully disclosed,
d such ass trading fees, which
w are
usually known before the investor enters into a deeal. Implicit costss of trading are th
u hose that
m not be fully
may y visible to invesstors before the transaction,
t such
h as opportunity costs of
e
executing an ord der on another platform
p that maay offer higher explicit costs, buut lower
i
implicit costs in terms of market immpact.
30 | THE IMPACT OF MIFID ON EUROPEAN MARKETS
3.2 A more competitive environment

Despite the original scepticism of the industry, the legislation has proven
successful in increasing market competition and efficiency, driving down
direct trading and post-trading costs, even though these benefits have not
been completely passed-on to end investors. As a result, competition has
been concentrating on prices (spreads between official trading venues have
increased) and execution services (provided by investment firms in
competition between them, brokers and some official MiFID trading
venues).18
MiFID therefore unleashed competition between:
Trading venues (by abolishing concentration rules and allowing
alternative trading venues), 19
Investment firms (by extending the coverage of the European
Passport to other services and promoting tied agents), 20 and
Market infrastructures (by providing the right of choice and access to
regulated markets and central counterparties, clearing and settlement
systems). 21
The CESR database has currently registered 243 trading systems, of
which 139 are MTFs, 92 RMs and 12 SIs. Concerning the European cash
equity market, there are 49 lit22 markets (27 MTFs and 45 Regulated
Markets) and dark markets23 (9 broker-dealer crossing networks and 23
MTF dark pools) as of June 2010 (Gomber & Pierron, 2010).
18 Official MiFID trading venues are: regulated markets (RMs), multilateral trading facilities
(MTFs) and systematic internalisers (SIs).

19 Arts 4.7, 4.14 and 4.15, MiFID.
20 Arts 23 and 31, MiFID.
21 Arts 33, 34, 35, MiFID.
22Lit markets are pre-trade displayed limit order book and quotes, in contrast to dark
books, which are limited order books that do not disclose book and quotes pre-trade. They
are only post-trade transparent.
23 See next Box for a definition.
Figure 5. Registered trading venues in the European Economic Area (EEA) and
breakdown for equities
Equity LitMarkets
MultilateralTradingFacilities
RegulatedMarkets
SystematicInternalisers 27
12 45
92 RMs MTFs
139 Equity DarkMarkets
23
DarkMTFs BrokerDealerCrossingNetworks
Sources: Authors from CESR and Gomber & Pierron (2010).
Box 1. Defining dark in securities trading

The term dark is often misperceived by the general public, in particular when it
comes to be associated with securities trading, and in particular equity trading.
Besides any possible pejorative meaning, dark trading is a legal form of trading,
foreseen by MiFID, that performs an important economic function in financial
markets. It consists of trading with no obligation of pre-trade disclosure of price
and volume. Specific conditions to benefit from the so-called waivers of pre-
trade transparency apply (see section 4.2.2 below). All other legal requirements
such as post-trade transparency, transaction reporting and market surveillance
rules (when trading is done on a MiFID trading venue) apply in the same way
as they do in pre-trade transparent (lit) markets. Dark venues are generally
accessible to non-retail investors who want to protect their trading interests from
market impact by not disclosing them before the transaction takes place. Market
impact may increase transaction costs and reduce incentives to trade. Other
reasons to use dark venues are explored in detail in section 4.2.
As the report illustrates, dark trading may be only damaging when it becomes
such a relevant part of the market that it de facto impedes the smooth functioning
of price discovery and formation processes on lit markets.
As used in this report, dark liquidity usually refers to all trading executed
in financial markets done with no pre-trade transparency, either under the MiFID
exemptions (waivers) or under the MiFID definition of over-the-counter trade.24
Dark trading venues or dark markets, however, refer to a sub-category of dark
liquidity, which includes dark pools and broker-dealer crossing systems (see
section 5.4). Dark pools are trading venues classified as MTFs (so-called dark
MTFs). Those venues refer to neutral trading platforms where trades find their
market-clearing price with no obligation of pre-trade disclosure of investors
trading interests.
On the one side, a more competitive environment produced positive

effects, as the reduction of direct trading (and post-trading) costs, with no
direct detrimental effects for price formation (CFA, 2009). Oxera
investigated trading and post-trading costs between 2006 and 2008 and
confirmed a relevant downward trend (Oxera, 2009). This trend was also
acknowledged by respondents to a 2009 CESR report (CESR, 2009a).
More controversial is the discussion on the reduction of implicit
trading costs (e.g. bid/ask spreads). In 2009, spreads have been higher than
pre-MiFID (London Economics, 2010), but moving down fast. It is hard to
disentangle the effects of the crisis (the stock indexes reached their bottom
level in February-March 2009) from the ones brought about by MiFID.
However, the latest data from 2010 seems to show a gradual decline of
spreads below pre-MiFID levels, in line with the stabilisation of financial
markets after the crisis (see Figure 6). Moreover, incumbents and
newcomers have been massively investing in infrastructure and new
technologies in order to reduce costs of execution and increase speed and
quality of execution services.
24As explained in section 5.4.1, OTC trades may not qualify as dark trade if they meet the
requirements set in Recital 53, MiFID.
Figure 6. Bid/ask average spread and implicit volatility on EuroSTOXX, FTSE100

and CAC40
Sources: Credit Suisse AES Analysis, 1 Nov 2007 to 30 Dec 2010; Credit Suisse (2011).
Source: NYSE Euronext (Response to CESR Consultation; CESR, 2010d).

On the other side, the growing number of trading venues has boosted
market fragmentation,25 thus increasing the search costs to source liquidity
and potentially also the market impact. Nevertheless, new technologies
seem to counterbalance this negative effect, delivering better order
management strategies with lower market impact. New technologies and
consolidated data solutions have drastically reduced the influence of depth
in the market of reference on trading decisions. In effect, estimates show
that the impact of market depth and volatility on implicit trading costs, and
therefore on trading decisions, is very low post-MiFID (London Economics,
2010). This result can be generally ascribed to the offer of multiple trading
venues to source liquidity on secondary markets for the same stock at a
reasonable price and cost.
Pan-European platforms. Fiercer competition, moreover, resulted in a
loss of market shares for national regulated markets and in the emergence
of new pan-European trading platforms (see Figure 7).
Figure 7. EEA equity markets 2010 pan-European market shares
Oslo Brs,
Turquoise, 2.868% 2.099% Other, 2.376%
Dark Pools, 2.06%
BATS Europe,
5.276%
LSE Group, London 13.293%
SIX Swiss 21.092% Milan 7.767%
Exchange, 6.082%
BME, 6.518%
Nasdaq OMX
NYSE Euronext,
Nordic, 6.679%
16.564%
Copenhagen 0.964% Amsterdam 4.292%

Helsinki 1.528% Brussels 0.793%
Oslo 0.081% Deutsche Brse, Chi-X Europe, Lisbon 0.428%
Stockholm 4.097% 11.825% 16.562% Paris 11.026%
Sources: BATS Europe, Thomson Reuters, FESE (Jan-Dec 2010; % total turnover, lit, auction,
and dark order books and hidden orders). For an overview of the main financial indicators
for equity and non-equity markets, please see Annex I.
25 See, more generally, chapter 1.

These new trading platforms are gradually increasing their market

shares to the detriment of the incumbents through the offer of pan-
European lit and dark order books for the secondary markets of most liquid
European shares.
Market shares. The pan-European competition is reflected in the
fragmentation of the most liquid shares in their national markets (see
Figure 8).
Figure 8. Most liquid EU shares by execution venue

Spain TheNetherlands France
100% 100% 100%
90% 90% 90%
80% 80% 80%
70% 70%
BME NYSEEuronext 70% NYSEEuronext
60% 60% Amsterdam 60% Paris
ChiX ChiX
50% 50% ChiX 50%
40% 40% 40%
BATSEurope BATSEurope
30% 30% 30%
20% 20% Turquoise 20% Turquoise
10% 10% 10%
0% 0% 0%
Telefnica Santander Royal Arcelor BNPParibas Total
DutchShell Mittal
Germany Italy UK
100% 100% 100%
90% 90% 90%
80% 80% 80%
70% DeutscheBrse 70% BorsaItaliana 70% LSE
60% ChiX 60% ChiX 60% ChiX
50% BATSEurope 50% 50%
BATSEurope BATSEurope
40% Turquoise 40% 40%
Turquoise Turquoise
30% Others 30% 30%
20% 20% 20% Others
10% 10% 10%
0% 0% 0%
Siemens Deutsche ENI Unicredit HSBC BP
Bank
Note: This graph includes the two most liquid shares (as ranked by CESR, by average daily
turnover) for the six biggest European equity markets.
Sources: CESR database and Fidessas Fragulator (% of volumes, YTD, 30 November 2010).
As shown below, market shares of incumbent-regulated markets

have consistently shrunk in the last two years. The most liquid shares in
their respective national markets have gradually been traded on new,
alternative pan-European platforms competing on execution and related
services, especially on cost of execution and access to services that allow
full exploitation of new trading technologies.
As a result of fierce competition, trading venues are struggling to

reach economies of scale, cutting costs and diversifying sources of
revenues. This may paradoxically induce the market in the medium to
long term to reconsolidate around a few operators offering pan-European
baskets of execution and related services for the most liquid shares in the
secondary markets. It is not clear what effect this final equilibrium will
have on European capital markets, since the benefits of network effects and
economies of scale, as a result of further integration, may need to be
balanced against a less competitive environment. However, competition is
usually measured through the level of market barriers to entry and exit,
rather than the number of competitors. Hence, the process should be
guided by principles of fair competition underpinning the MiFID text.
3.3 The introduction of new technologies

In the last decade, European equity markets have experienced evolutionary
changes in their infrastructure and trading methodologies. The
introduction of new technologies and greater competition has led trading
venues to cater for automated trading and investment firms to invest in
algorithmic trading and high-frequency trading (HFT) technologies.26
These developments stimulated new investments in market infrastructures
(in particular, new trading platforms). Speed and likelihood of execution
have become increasingly key aspects for professional investors, seeking to
reduce the market impact of their orders. As a result, new trading
methodologies and technologies have entered the scene and they seem to
be here to stay.27
High-frequency trading. Currently, more than a third of the
transactions executed on the UK FTSE 100 are made through HFT systems
(see Table 2) and much more through algorithmic trading. In the US, this
number was between 40% and 60% of the average daily volume in 2009
(Celent, 2009) and is set to be above 70% in 2010, as estimated in a recent
survey by Aite Group. Trading venues and broker-dealers have invested
massively in technology to accelerate and improve trade execution through
the use of sophisticated algorithms and to improve capacity. The NYSEs
26 HFT is a form of high-speed automated trading that it is often use by arbitrageurs and by
professional investors in order to improve their trading strategies, often supported by

complex algorithms.
27 For a more detailed analysis about the link between new technologies and current market
structure, see section 5.5.2.
speed of execution for small, immediately executable orders was 10.1

seconds in January 2005, compared to 0.7 of a second in October 2009 (SEC,
2010a, p. 6). NYSE Euronext Paris is going to move its servers to London to
reduce latency and to be closer to the main trading community in Europe,
while the LSE Group after the acquisition of the IT firm Millennium has
launched an infrastructure called Millennium Exchange, to which
Turquoise has recently migrated, while the main markets and third parties
markets such as Oslo Brs will migrate in 2011.
Table 2. Share of HFT by trading venue (shares of order books)

Trading venue High-Frequency Tradinga
Chi-X 40%
London Stock Exchange 32%
BME 25-30%b
NYSE Euronext 23%
Borsa Italiana 20%
Turquoise 19%b
Nasdaq OMX 13%b
a As a share of total trading value.
b As a share of total trading volumes.
Source: CESR responses to consultation paper on micro-structural issues in CESR (2010d).
HFT strategies. A report prepared by a Working Group chaired by

Pierre Fleuriot and commissioned by the French Minister for the Economy,
Industry and Employment has listed the main trading strategies that can be
pursued through HFT systems (Fleuriot, 2010):
Market-making (orders sent to capture the spread),
Arbitrage (instantaneous or statistical) and
Speculation (event-driven strategies to predict future price
movements).
This set of strategies falls under those typical functions served by
dealers and informed traders in financial markets (Harris, 2002). On the one
side, algorithmic trading (AT) and more specifically HFT systems generally
allow a faster and more efficient flow of information into prices, with
potential price discovery and gains in spreads (Hendershott, Jones &
Menkveld, 2008; Brogaard, 2010). AT also seems to continue after market

shocks and benefit from price reversal (Brogaard, 2010), and somehow
from asset volatility without necessarily adding up to a generally highly
volatile scenario. HFT is concentrated on most liquid stocks (Nasdaq OMX,
2010). On the other side, these systems could put trading platforms under
severe stress by increasing volumes and speed to critical levels. Its use is
expected to grow substantially in the coming years, as technologies become
more widely available to investors. Only the effects of their widespread
diffusion will clarify the real benefits and costs of technological
developments. However as with any other trading tool in financial
markets HFT systems may be used beyond their general purposes. In this
sense, some market participants may want to pursue more aggressive
trading strategies such as quote stuffing.28 Regulators should therefore
constantly supervise the actual use of these complex tools, set common
technical standards and define legal boundaries.
New trading technologies. AT and HFT systems, in fact, are often
erroneously seen as trading strategies, while they are only tools that allow
high-speed computation of orders and new and more efficient trading
strategies. However, AT and HFT have boosted new trading strategies,
such as statistical arbitrage, which have now become an essential aspect of
end investors trading decisions.29 Banks are continuously developing and
fine-tuning their order-routing systems to remain competitive. Quantitative
automated trading strategies, such as smart order routing through
algorithmic trading, have become mainstream in trade execution and
minimise the risks of sourcing liquidity to reduce market impact.
Quantitative strategies based on high-frequency trading tools are now a
more significant component of the market, and the success of this
technological progress has actually encouraged more entrants. In effect,
these sophisticated algorithms ought to slice big orders into small ones (so-
called, child orders) and spread their execution in a set time frame and
price range during the trading day. This substantially reduces risks of
28 Quote stuffing is a manipulative practice pursued through high-frequency trading
systems consisting of the introduction and instantaneous cancellation of a huge amount of

orders in order to slow down the competitors systems, gain processing time and deceive
the market.
29 Using data from listed firms, Hendershott & Riordan (2009) found that almost 50% of
liquidity on Deutsche Brse comes from algorithmic trading. On Nordic markets, 40% is
algorithmic trading and only 7% is pure HFT (Nasdaq OMX, 2010).
market impact. Overall, algorithmic trading seems to consume liquidity

when it is cheap, and supply liquidity when it is expensive, that is, when
spreads widen. This beneficial outcome does not seem to increase volatility
either (Hendershott & Riordan, 2009).
Market access. Moreover, trading platforms offer services to financial
institutions in order to install their servers close to their data mainframe
and reduce their latency. Other services to reduce latency, such as Direct
Market Access (DMA) and Sponsored Access (SA),30 allow entities to
directly or indirectly access a trading platform with different levels of
responsibility and controls over their actions (see section 5.5.2). These new
access and low-latency services are critical for certain advanced trading
strategies and seem to be a source of competitive advantage for trading
venues. In addition to them, other access services such as unfiltered
access31 may be important for high-frequency trading systems. However,
they bring new risks for the infrastructure, which need to predispose
adequate systems and controls to monitor and manage potential
misbehaviours or human errors (so-called fat fingers).
In conclusion, despite the indirect benefits of the availability of new
technologies in financial markets (lower trading costs), retail investors
remain only partially able to access technologies that facilitate their access
to multiple trading venues (e.g. lack of consolidated data solutions). In
effect, these technologies may benefit retail investors as long as they are
used to promote easier and cheaper access to pan-European markets.
Nevertheless, retail investors may have benefited from a modest
downward trend of trading costs, as a consequence of an increasingly
competitive environment.
3.4 Evolution in market microstructure and brokerage services
The introduction of new technologies and the new trading environment
brought about by MiFID32 have accelerated the evolutionary process in
market microstructure and brokerage services.
30 See Section 5.5.2.

31 Unfiltered access is a sort of naked sponsored access, in which the members client can
place an order on the trading platform that flows directly into the markets without a first
screening of broker-dealer systems.
32 Smith (2010) has found evidence that links latest developments in market microstructure
(e.g. increased use of HFT and automated trading) with implementation of RegNMS in the
US.
Concerning market microstructure, there are two relevant changes: 1)

the decline of dealer quote-driven markets in favour of limit order books
(dark and lit), even in the OTC space (Broker-dealer Crossing Systems,
BCSs); and 2) the growth of dark pools volumes.33
Limit order books. Firstly, the general request for more transparency in
trading instruments and infrastructures, and trading continuity plus the
evolution of technologies and trading systems have favoured the use of
limit order books. The use of smart order routing systems and other
sophisticated automated trading systems advocates the crucial role of
transparent limit order books (order-driven), which ensure more efficient
and less costly trading continuity than dealer markets as long as there is a
sufficient liquidity in the pool. MiFID has undoubtedly promoted a wider
regime of transparency and investments in new trading technologies.
However, on limit order books the market impact is potentially higher than
in quote-driven markets, since dealers have the knowledge of the order
before setting the price. Therefore, limit order books are usually
characterised by greater ability to absorb small orders than big ones
(Malinova & Park, 2008).
Secondly, the abovementioned change in market microstructure and
the increased fragmentation of liquidity have increased the requests from
market operators to deal with institutional orders on non-displayed limit
order books (or so-called, dark pools) in order to reduce market impact. In
effect, the public knowledge of an institutional order hitting the order book
may provoke front-running by parasitic traders aimed at using this
information to exploit the expected price movement at the expense of the
institutional orders, which may not find enough liquidity if prices move
away from the target (so-called exposure risk; Harris, 1997).
As a result, Table 3 suggests a growth of dark order book volumes in
the 3Q of 2010, in comparison with 2009. Specifically, the use of dark pools
of liquidity increased by more than three times last year, although it still
remains a small fraction of total volumes (1.72%). From three main dark
books in January 2009, this number was 11 at the beginning of 2010, and
new entrants (both MTFs and RMs) are increasingly offering similar
execution platforms.34 New systems that allow automated trading activity
also offer better reporting and clarity on volumes than before.
33 For a more detailed discussion, see section 5.5.

34 Section 4.2 will discuss the impact of dark order books on price formation and discovery.
Table 3. EEA trading venues volumes (thousands of shares; by order book)

2008 total 2009 total 2010 total %
% Tot. % Tot. 09/10
volume volume volume Tot.
Lit order 1,071,687,579 88.03 1,067,998,138 89.28 1,191,073,305 87.17 +11.52%
book*
Order
book - 143,871,002 11.82 116,440,019 9.73 146,209,746 10.7 +25.57%
auction
Dark 1,916,249 0.16 11,785,681 0.99 29,083,060 2.13 +146.77%
order
venues
Total 1,217,474,830 1,196,223,839 1,366,366,110 +14.22%
* Including hidden orders.

Source: Authors from Thomson Reuters.
Finally, brokerage service providers are experiencing a strong

consolidation phase as the proliferation of technologies and the need for
advanced execution services allow a more sophisticated offer through
centralised platforms. These platforms can thus exploit increasing
economies of scale and scope (network effects). Cheaper trading costs for
high-volume traders and institutional investors give an opportunity to
dealers to enter this market and combine advanced brokerage with execution
services, in order to minimise both market impact and risks, and to source
liquidity in a fragmented market. Broker-dealer crossing networks (BCNs)35
are the result of this evolutionary process, as they provide execution
services with best execution, on a discretionary basis.36
3.5 A more comprehensive transparency regime

MiFID recognises the role of transparency in financial markets as a means
35 BCNs represent linked groups of broker-dealer crossing systems (or BCSs) designed to
form one pool of liquidity and reach a critical mass to cross orders. Matching of orders
typically takes place with no pre-trade transparency and supported by best execution duties.
For the sake of simplicity, we will use these two terms (BCN and BCS) interchangeably in
the present text, as we are interested in the function they serve more than in the modality by
which they perform those services.
36 For a more detailed discussion of the economic and regulatory implications of BCNs,
please see section 5.4.

to enhance liquidity and increase investor protection.37 The Directive

extended the coverage and scope of the transparency regime to all equity
trading venues. However, the extension to non-equity markets was left to
the discretion of member states. Only two countries (Sweden and Italy)
decided to apply this transparency regime to fixed-income.
The scope of the transparency regime is fairly wide. Legal obligations
require the disclosure of pre-trade information (quotes) by the
abovementioned trading venues and post-trade information by investment
firms (directly or indirectly through a third party) wherever the shares
(admitted to trading on a regulated market) are actually traded. The post-
trade transparency regime applies not only to equity orders executed on
regulated markets, but also on alternative platforms and over-the-counter
crossing systems. Information, through different reports, is disclosed to the
market (trade reporting) and to regulators (transaction reporting)38 with
different requirements and granularity. Pre-trade transparency, however,
can be waived in some specific cases and in order to serve specific market
needs.
Despite the greater scope and the stricter requirements of the MiFID
transparency regime, the implementation has produced inconsistencies in
the quality of market data and the reporting methodology. Data for over-
the-counter transactions lack quality, including duplicate reporting, and
create uncertainty about its relevance in Europes capital markets.
3.6 Greater investor protection

Another important achievement of MiFID is in the area of investor
protection, through the harmonisation of a meaningful set of business
conduct rules and organisational requirements. Investor protection has
been levelled across Europe. Among the most important measures, the
introduction of conflict of interest rules, best execution and know-your-
customer rules (e.g. suitability and appropriateness tests), which specify
the general duty of the agent to act in the principals best interest (Art.
19.1, MiFID).
Main measures. The parallel introduction of know-your-customer
rules and best execution for all orders (and all financial instruments,39
37 See chapter 1.
38 See Arts 27-30, 44 and 45, MiFID.
39 But not all kinds of investors (e.g. eligible counterparties).
except spot foreign exchange products) has promoted a stronger

relationship with clients. This is true in particular between retail clients and
investment firms in the provision of specific services (such as order
execution services). These two rules have made the European and US
financial markets regulation closer than ever before. Exemptions for
professional investors (or eligible counterparties) or dilution effects
foreseen by the legal text allow them to contract around or opt-out of the
standard rules (e.g. best execution), in order to minimise transaction costs.
Moreover, the Directive has harmonised the framework of rules on
conflicts of interest, which previously fell under the principle of the home
country control.
Finally, concerns have been raised about the implementation of these
rules. Implementing measures are often adopted to reflect only the
minimum legal requirements without assessing the quality of execution
and the actual level of investor protection.
4. A RENEWED TRANSPARENCY REGIME
4.1 The role of transparency in financial markets

Transparency is the possibility by market participants to see the market
mechanisms, in particular information on market price of a security and its
determinants (order flow). It has gradually become a crucial aspect of
financial markets, in parallel with the evolution in market microstructure
and new technologies. In effect, with full information asymmetry there is
no price equilibrium and no exchange will take place (Madhavan, 1992),
due to strong adverse selection and moral hazard problems.40 Therefore,
the primary objective of transparency in its many dimensions is to
reduce information asymmetry and transaction costs in order to make the
exchange between buyers and sellers happen, thus reducing search costs
and promoting more efficient price formation. In other words, appropriate
transparency promotes price discovery in order to find the market-clearing
price (price formation) of a financial instrument.
40 Adverse selection is an informational problem that arises when products of a different
quality (e.g. lemons and good cars; junk and good bonds and so on) are sold at a single price
because of asymmetric information (inability of the buyer to understand the actual
quality/risk of the cars/financial product or borrower), so that too much of the low-quality
product and too little of the high-quality product are sold. The equilibrium will result in a
market price (due to this inability of the buyer to understand ex ante the quality of the
product) a bit higher than the real value of low-quality products and consistently lower than
the real value of high-quality products. Hence, the market equilibrium, in the mid-term, will
determine that only low-quality products are sold in the market. This fundamental issue can
freeze markets, thereby justifying mechanisms for quality signals, such as neutral third-
party informational tools (rating agencies, etc.) or regulatory actions to increase
transparency or simply pre-sale services. See, in general, Akerlof (1970) and Kraakman
(1986). Moral hazard is an informational problem created by the opportunistic behaviour of
the more informed party, who tries to exploit the informational advantage and the scarce
ability of the less informed party to monitor his/her activity. See, in general, Holmstrom
(1979) and Milgrom & Roberts (1992).
44 |
What information? The general problem typically is what kind of

information should be published and who should be entitled to receive it
(OHara, 1995). The implementation of a transparency regime is a complex
exercise. The kind of information that should be disclosed needs to be
carefully assessed, since full disclosure in some cases may have unintended
consequences. Therefore, the type of published information is strictly
related to the market microstructure (e.g. asset classes, trading
organisation) that underpins financial markets.41
The rapid increase of volumes in the last decade has boosted the use
of limit order books (auction markets), which have pushed down
transaction costs (explicit and implicit costs) and ensured continuous
trading as requested by new technologies (automated trading services).
Secondary markets, in particular for equity, have gradually moved from
more opaque quote-driven dealerships to auction markets (limit order
books). Limit order books need a certain degree of transparency. To ensure
that the order flow is transparent in a meaningful way, there are generally
two levels of data that may in some instances need to be disclosed: 1) pre-
trade data (quote) and 2) post-trade data (trade disclosure). Such data
finally need to be provided with sufficient depth, breadth and quality in
order to benefit markets.
To whom? The appropriate flow of information generally benefits
uninformed investors,42 as they increase their knowledge of the market.
However, as shown in the next sections, disclosing information may be
harmful for some types of uninformed traders and markets that are not
based on auctions. Therefore, the general goal to protect uninformed
traders who are not necessarily retail may need to be carefully balanced
against the risks to market efficiency coming from a fully transparent
market setting.
41 The main empirical literature on market transparency comes from research and empirical
studies on equities or financial instruments traded on organised trading platforms through

order-driven markets or quote-driven dealer markets. Studies in financial instruments
traded OTC are very limited since there is almost no publicly available information and
their market share was quite limited until the beginning of the new century. There is
currently no regime of trade reporting (public disclosure) and transaction reporting
(disclosure to regulators). For a discussion on market microstructure, see section 5.5.
42Uninformed investors are those who trade with no private information, but know that
prices will reflect the information of the informed traders (Dow & Gorton, 2006).
46 | A RENEWED TRANSPARENCY REGIME
Objectives. It is in the markets interest to have a certain level of

transparency. However, incentives may not be in line with public policies,43
so regulators may need to push market operators to disclose more
information. In effect, transparency requirements under MiFID (Arts 27-30,
44 and 45, MiFID) serve two public policy purposes:
i) Price efficiency
ii) Investor protection
Price efficiency. Public disclosure requirements (or trade reporting)44 to
the public should be proportional to the market mechanisms that shape
investor behaviour and through which securities are generally traded. For
sake of simplicity, this report splits the available market mechanisms in
open limit-order books (or order-driven; hereinafter, OLOB)45 and quote-
driven dealer markets,46 both ensuring continuity of trading.47 Both
systems currently support automated trading. In general, public disclosure
of the order flow improves the flow of information into prices and can
enhance liquidity48 because it lowers search costs (more efficient price
43 Transparency is something that everyone benefits from but very few would be willing to
provide information about their trades without explicit requirements. Therefore, the level of
transparency with no regulatory intervention may be sub-optimal. See also ESME (2009b).
44 The report distinguishes between trade reporting public disclosure pre-trade or post-
trade requirements from transaction reporting, which is a publication system for trade
reports to regulators.
45 OLOBs is a system in which submitted limit orders continuously receive immediate
execution against the book. Orders are submitted by public investors and dealers. It
operates as an auction because the price is formed through a continuous multilateral
matching of orders. A quote-driven continuous dealer market is a market in which investors
immediately execute their orders against the quotes provided by competing market-makers.
Bid/ask quotations typically depend on the size of orders. See, in particular, Madhavan
(1992).
46 Quote-driven markets are trading systems whose public quotes are the result of
competing quotations offered by market-makers and dealers for a certain amount of a

specific financial instrument. Main quote-driven markets are fixed income and derivatives
markets, whose trading activities can be done on exchange or over-the-counter.
47Trading may be also done through periodic auctions, batch or transparent (Pagano-Roll,
1996). In fact, periodic auctions that involve the submission of buy and sell limit orders
which are matched at a price/quantity set by the market or the dealer (batch auction) or all
crossed at a common price (transparent auction) are typically part of the opening call
auction, that is to form prices when there is a low level of liquidity.
48 Liquidity is a multi-faceted concept and it is difficult to find a definition that can perfectly
fit all dimensions of securities trading. A market is liquid when it is almost infinitely tight,
which is not infinitely deep, and [] resilient enough so that prices eventually tend to their
discovery). It also minimises adverse selection effects for uninformed

traders, with positive effects on opening spreads especially for dealer
markets (Pagano & Roll, 1996; Flood et al., 1999). OLOBs may benefit from
greater transparency, as enhances liquidity; reduces volatility; and
increases price efficiency (Giraud & DHondt, 2006). Prices are converging
faster to fundamental values in a more transparent market setting. For
instance, cash equity markets are so large and involve so many uninformed
investors that greater transparency requirements (as defined by MiFID)
seem beneficial, since they reduce volatility and improve market quality. In
thinner markets, with insufficient noise trading, volatility will increase
with market disclosure (Madhavan, 1996).
Competition. In addition, efficient information transmission between
markets may help to keep competition in the market between market-
makers and trading venues, and mitigates risks of fragmentation
(Madhavan, 1995). Artificially fragmented market settings may harm the
market with side effects on the cost of capital and the economy.49
However, wide trade disclosure principally for dealer markets (and
potentially for liquidity providers in OLOBs) may reduce transactional
efficiency in those markets, since disclosure of private information can
reduce market-makers incentive to compete on spreads (Bloomfield &
OHara, 1999). In a nutshell by increasing transparency an important
trade-off emerges. Regulators thus should look at striking the right balance
between overall benefits for uninformed traders and negative effects upon
informed traders (including dealers), who will not be able to benefit from
their informational advantage (market knowledge) if a wider transparency
regime applies. In some markets, the informational gap may be the only
incentive for dealers to deal with illiquid products, and thereby create a
market where none hitherto existed. A trade-off emerges in particular
when dealers play a more relevant role in the market, as in quote-driven
underlying value (Kyle, 1985, p. 1317). Three aspects emerge from this definition: i)
tightness (which is the possibility to turn over a position at the fastest speed technically
possible, when needed); ii) depth (which refers to the ability of the market to absorb
quantities without having a large effect on price; it is usually not constant over time in some
asset classes); and iii) resiliency (which is the speed at which prices return to their
fundamental value after a move due to regular trading or with more intensity to external
shocks).
49 See chapter 1.
markets for instance. Therefore, two factors must be considered when

imposing transparency requirements (Kumar et al., 2009):
1) The active role of informed traders in the market (e.g. a market
consisting of only institutional investors) and
2) The active commitment and participation of liquidity suppliers (e.g.
quote-driven illiquid markets).
The presence of these two factors may sensibly reduce the beneficial
effects of greater public disclosure, pre- and post-trade.
Moreover, the direct exposure of some types of orders (i.e. big size
orders) to OLOBs can generate market impact and thus push liquidity out
of the market (implicit transaction costs; Harris, 1997). Non-calibrated
transparency requirements may indeed harm liquidity, hence widening
spreads in quote-driven markets or reducing book depth50 in order-driven
markets and making markets thinner. In order to address these potential
trade-offs between transparency and market quality and efficiency, MiFID
has already set in the original text a list of exemptions from disclosure and
delayed publications in cash equity markets.
In non-equity markets,51 where trading mechanisms are mainly
facilitated by dealers, the transparency regime needs to be defined and
calibrated in a different way, since those markets have a fundamentally
different market microstructure from cash equity markets.
Investor protection. A transparent market may also generate
reputational capital, as an efficient transparency system can stimulate
surveillance by other traders and provide more control over insiders (less
moral hazard; Chowdhry & Nanda, 1991). It also gives more information to
investors in order to make more informed decisions for their investments,
as well as a more effective assessment on how intermediaries are
complying with their best execution policy. Moreover, greater disclosure
for investor protection purposes is also ensured through an ad hoc
transaction reporting system to regulators, which improves market
surveillance, market integrity and enforcement of current legislation. A
market subject to manipulation will infuse loss of confidence in the efficient
market functioning and raise problems of moral hazard.
50 Market depth is a measure of liquidity that represents the ability to absorb order flows
without large changes in prices (Glosten, 1994); it represents the volume required to
generate a market unit price change (Kyle, 1985).
51 This report will generally refer to non-equity markets as markets that are not primary or
secondary cash equity.

Conclusion # 1
Transparency plays a crucial role for the smooth functioning of financial markets and the
monitoring of systemic risk. It also ensures the correct functioning of the price formation
process through efficient price discovery mechanisms. However, regulatory actions in
certain areas are needed, not only as a result of the recent financial crisis but also from the
experience gained since the transposition of the Directive. Interventions should be
proportional to the market structure through which investors orders find the market-
clearing price. Disproportionate transparency requirements would harm market efficiency
in some illiquid markets with no increase of investor protection or reduction of systemic
risk, as the market would become less-liquid and more subject to manipulation and
eventually to market crashes.
4.2 Pre-trade transparency for cash equities:

Benefits and costs
In order to preserve the abovementioned objectives, MiFID designed a two-
level regime of transparency for cash equities, pre-trade and post-trade.
This structure seems to be well-designed, in particular for OLOBs, which
are continuous auction markets and inherently more transparent than other
forms of markets (Pagano & Roll, 1996).
Definition. Pre-trade transparency concerns the public disclosure of
quotes and trading interests. Appropriate and effective pre-trade
transparency requirements ought to serve investor protection and price
efficiency supporting price discovery (through real-time quote disclosure)
and best execution (through smart order routers). These transparency
requirements mainly apply to the official trading venues under MiFID
(RMs, MTFs and SIs; Arts 22.2, 27, 29 and 44, MiFID). In an integrated
market, more pre-trade transparency tends to even levels of information
between market participants. Overall as shown below evidence seems
to confirm that pre-trade transparency may be beneficial if the market is
large enough (as defined by Madhavan, 1996 and Baruch, 2005), i.e.
benefits for uninformed investors (liquidity demanders) are higher than
costs for informed ones (liquidity suppliers). Therefore, price discovery
seems still to be rewarding, since there is sufficient noise trading and space
for arbitrage (there is still private information that can be rewarded).
Benefits. The literature found conflicting evidence on the benefits of
pre-trade transparency, and the results seem to be different in relation to
the trading mechanisms that underpin the market structure.52 In OLOBs,

disclosure of quotes and order flow provide uninformed investors and the
entire markets with visibility of best prices and a part of the order flow
(usually first five best quotes) on real-time. If market players are able to
consolidate quotes across trading venues, the positive effects of these
requirements may offset the potential negative effects of liquidity
fragmentation (Glosten, 1994, Madhavan, 1995 and OHara & Ye, 2009).
Bohemer et al. (2005) found the opening-up of a limit order book to pre-
trade transparency to be beneficial. As a consequence, the new
transparency settings impacted on trading strategies: traders manage
exposure of orders with less market impact (higher cancellation rate and
shorter time-to-cancellation rate) and there are overall fewer specialists in
the market (informed traders). These changes improved market quality and
more specifically ameliorated:
i) Informational efficiency (smaller deviations of transaction prices from
efficient price) and
ii) Liquidity (increase of depth and decline in effective spreads).
The case study of Island in the US (Hendershott & Jones, 2005) shows
that reducing pre-trade transparency may have negative effects on the
trading venue. In effect, after Island decided to go dark, it lost half of the
market share, while effective and realized spreads went up drastically.
Other lit markets saw their market share increasing and spreads going
down, with unclear net overall effects. In 2003, Island redisplayed quotes,
but market share and spreads did not go back to the previous levels (at
least not in the short/medium term). Other evidence from Eom & Park
(2007) suggest that market quality increases with the level of pre-trade
transparency up to a point, after which the quality decreases, in line with
the characteristics of a concave function (see figure below). Traders adjust
their preferences about the real value of the share as quote disclosure
changes or as other traders behaviour changes. However, market quality
measured through six parameters53 would improve up to a certain level,
where an inflection point signals the reversal of the trend.
52 For an overview of literature on market transparency, see Sabatini & Tarola (2002).
53 Spread and relative spread; market depth (bid/ask spreads and limit orders size);
transitory volatility (from liquidity provision); market-to-limit order ratio; full information
trade cost (FITC); and 2 components of the trade execution costs, implied spread (adverse
selection and transitory cost). Eom & Park (2007, p. 323).
Figure 9. Pre-trade transparency and market quality
Market Quality
Pre-Trade transparency
Source: Author from EOM & Park (2007).
This pattern has also been suggested by recent studies on OTC equity
trading. There is no strong evidence of a positive relation between the size
of OTC trading and market depth (Gresse, 2010). However, Buti et al.
(2010b) find on a sample of US stocks that dark markets54 (non pre-trade
transparent venues) are not detrimental to market quality, although their
introduction pushed quoted spreads down until those venues reached a
certain market share.55 This intuition seems to be confirmed by an ongoing
study (Degryse, 2010), which tries to find hard evidence of a positive
relationship between market depth and size of over-the-counter trading.56
54 In the EU, the terms may only include MTFs or RMs applying pre-trade transparency
waivers. In this specific case, authors refer to all trading venues with no pre-trade
transparency (including broker-dealer matching systems).
55Authors identify this share at between 8% and 10%. However, they doubt the validity of
this threshold and warn that the causal relationship between dark pool and market quality
needs to be further investigated.
56Over-the-counter trading in these studies reflects a number that may be altered by double
counting and does not include all dark liquidity, as the report describes in the next sections
in more detail. However, it may include useful information on the impact of non pre-trade
transparent liquidity on market quality.
52 | A RENEWED TRANSPARENCY
R REG
GIME
Figu
ure 10. Market depth
d and OTC
C activity
Sourcce: Degryse et al. (2010).
A preview of o the results showed to our

o Task Forcce suggests that t
OTC C trading is beeneficial to maarket depth until it reaches a certain marrket
sharre (see Figure 10). The autho ors computed the depth of v visible markett for
diffeerent potentiaal price movem ments (10 to 60 bps). They y regressed daaily
meaasures of logarrithmic depth h with controlss (time dumm my variables, etc.)
e
and five OTC acttivity bucketss (<1%; 1-5%; 5-12%; 12-27%; >27%). As 57 A a
conssequence, the existence of nonn pre-trade transparent
t liq
quidity belowwa
certain amount seems to beneefit the whole market. This outcome appeears
to be
b the result off increased co ompetitive pressure by dark k venues not only
o
on handling
h largee trades but allso on price im
mprovements in comparison n to
thosse of other trrading venuess. With inter-m market compeetition, pre-trade
tran
nsparency in the main maarket allows free-riding aand creates new n
marrkets. Price diiscovery shou uld be inform mative and rew warding, so dark
d
pools can free-rid de on informaation about prrices due to n noise trading, i.e.
therre are still diffferent kinds off traders operating in this m
market (and most
m
of th
hem uninform med).
Costs. Howeever, as previo
ously stated, pre-trade
p transsparency may not
be always
a beneficcial. Madhavan
n et al. (2005) found
f evidencce on the Toro
onto
Stocck Exchange that openin ng up floor and CATS S (Computeriised
57Thhis exercise showed that the threshold of actual tra ading that seemss tolerable is betw ween
12% and 27%. Also in n this case, authoors (such as Buti et al., 2010), warrn that this thresh
hold
may not be accurate anda that more an nd better data sho
ould be made av vailable to the maarket
in orrder to further inv
vestigate these isssues.
Automated Trading System) markets pre-trade transparency to the public

(for members it was already available) may be harmful, in particular for
markets with specialists (floor trading). In effect, the results of this action
increased gaming risks around the opening price, spreads and volatility,
and lowered market depth (in particular for floor trading). The rationale
behind this case study relies on the assumption that limit orders can be
seen as an option at a specified price. Therefore, when limit orders are
forced to interact with market orders on the same pre-trade information
level, liquidity providers (informed traders) are less willing to provide free
options to market orders. This situation increases adverse selection effects.
As a consequence, liquidity decreases, spreads widen and price volatility
soars. However, levelling the amount of information available to traders
should also attract new investors to the table.
Exposure risk. Revealing specific trading interests to the public on the
limit order book may expose traders to unforeseen costs (exposure risk and
market impact; Madhavan, 1995; Harris, 1997, 2002). Order exposure may
be costly if the exposure (Harris, 1997):
1) Reveals traders motives,
2) Reveals the potential price impact of future trades or
3) Reveals valuable trading options.
Firstly, by revealing traders motives they will lose proprietary
information on the real value of the share, so reducing time for defensive
measures and exposing the trader to squeezers, who will be able to control
the order flow.
Secondly, orders exposure may attract parasitic traders,58 who profit
by anticipating large impact orders and taking liquidity on the order book
before the large order is executed (front-running).59 In this way, large
traders pay more for providing liquidity (they will buy at a higher price).
Thirdly, private information on exposure on the order book may lead
traders to set limit orders just above the limit price of the big trader, in
order to use it as stop loss if prices go down instead of up (they will sell to
the large trader). The large trader therefore gives a free valuable option to
58 Parasitic traders are a category of reactive traders (Harris, 1997), i.e. those who reply with
trading in response to trading opportunities given by other traders.

59 This practice is illegal if done by brokers on orders held on own account, in line with
Recital 19, Directive 2006/6/EC on market abuse (MAD).

other traders (with some risk for the last ones, though). In addition,
decreasing tick sizes may have led more orders to be hidden due to higher
exposure risk, since orders may be more easily subject to legal front-
running through smaller price unit changes.
In conclusion, pre-trade disclosure of information on highly illiquid
products may affect the price formation process of other related financial
instruments, or disclose private information that parties do not want to
disclose as part of private negotiations.
Conclusion # 2
Pre-trade transparency is fundamental for trading venues and their trading structure, as
well as for an efficient price discovery and a better implementation of best execution
policies. It is also important to exempt some trading interests (e.g. block trades) from pre-
trade transparency requirements, because as mentioned above this situation may
otherwise have a negative impact on the market. The breadth of these exemptions is under
debate and conflicting views emerge on several aspects, as explained in the following
sections.
Box 2. Defensive strategies

In order to manage exposure risks and aside from the illegality of some of these
practices, Harris (1997) categorised the potential defensive strategies that traders
can pursue to reduce adverse effects of exposing trading interests. These
strategies may be summarised as:
a. evasive (retaining information),
b. deceptive (confusing through the use of misleading information) and
c. offensive (counter-attacking with opposite orders).
Evasive strategies may use: i) multiple brokers, ii) anonymous/dark
trading, iii) counterparty selection strategies (by the trader or the venue where
the trader executes his/her orders), iv) slice and dice strategies (slicing orders),
v) internal matching systems, vi) indications of interests (IoIs),60 or vii) waiting
strategies.
Secondly, traders may disseminate inaccurate information through legal
mechanisms such as: i) trading in the opposite direction of the big order that
60 An indication of interest (IoI) is the name commonly used to refer to a message sent
between investment firms to convey information about available trading interest. IOIs are
also used by dark pools to attract order flow and to maximise trading opportunities by
enabling investors to find the contra-side of orders. The information provided in an IOI can
include the symbol of the security, the side (i.e. buy or sell) and volume/price of trading
interest (CESR, 2010b).
he/she tries to execute, ii) trading on a related market in order to divert market
attention, iii) introducing and cancelling orders in order to create uncertainty
around their trading interest or iv) introducing orders with lower size in order to
create different expectations and get better terms.
Finally, traders may reduce their exposure by adopting offensive strategies to
attack front-runners, by trading in the opposite direction and then removing the
interest.
4.2.1 Dark liquidity

Exposure risks and market depth (as above) may require the use of
liquidity that is not-pre-trade transparent (dark liquidity).61 In general, all
trades are important for price formation as they form liquidity, even if they
are not displayed. Dark liquidity below a still unclear market share may
bring some beneficial effects to the whole market. Recent findings on
hidden orders show that some opacity in the market improves liquidity in a
centralised limit order book with no fragmentation (Moinas, 2010). It can be
justified by the strong incentive for informed traders to sniff out where
hidden liquidity sits on the lit book (De Winne & DHondt, 2007). Hidden
orders stimulate informed traders to search for liquidity. In addition, if you
expose those orders, you may generate the effect of withdrawing liquidity
from the market, thereby making it more vulnerable to manipulation
because of its much lower thickness and depth (thinner). The presence of
hidden orders, therefore, stimulates greater activity for liquidity
demanders in order to find trading opportunities within the book. This is a
major difference between stand-alone dark pools, which do not have direct
liquidity links, and lit books (they do not hit the displayed limit order
book).
Purposes. Traders motivation to engage in dark trading have been
already implicitly or explicitly mentioned, but the IOSCO (2010c) has
recently listed the reasons in a consultation paper. They can be summarised
as:
1) avoiding private (or proprietary trading) information leakage,
2) minimising market impact costs (information leakage and/or
insufficient market depth),
3) ensuring better order management (lowering the risk of being
61 This term includes dark pools, hidden orders and BCNs.

picked-off),
4) avoiding also hidden orders as they may be sniffed out by
sophisticated software,
5) taking advantage of a price improvement and
6) minimising transaction costs.
The size of non-pre-trade transparent liquidity is constantly growing
across Europe (see Table 3) and currently represents 6.76% of total EEA
trading (off and on-order book; see Table 4) and 10.86% of EEA trading on
RMs, MTFs and broker-dealer crossing systems (BCSs).
Table 4. Dark liquidity in the EU (% of total EEA trading)*

2009 2010
Q1 Q2 Q3 Q4 Q1
Dark pools and hidden
orders** 4.71% 5.68% 5.6% 6.11% 5.26%^
Broker-dealer crossing
0.9% 0.9% 0.9% 1.4% 1.5%
systems
5.61% 6.58% 6.5% 7.51% 6.76%
* Trading under waivers, % of EEA total trading.
** Estimated market share as % of EEA total trading, assuming a constant average market
share of OTC trading of 38% in 2009 and Q1 2010 (CESR 2010b).
^ Does not include Poland.
Source: Authors calculation from CESR (2010b).
Table 5. Dark liquidity in the EU (% of EEA RMs, MTFs and BCNs trading)*
2009 2010
Q1 Q2 Q3 Q4 Q1
Dark pools and hidden

7.6% 9.2% 9.0% 9.8% 8.5%^
orders
1.43% 1.43% 1.43% 2.21% 2.36%
systems**
9.03% 10.63% 10.43% 12.01% 10.86%
* Trading under waivers, % of total EEA trading on RMs and MTFs.

** Estimated market share as % of total EEA trading on RMs, MTFs and BCNs, assuming a
constant average market share of OTC trading of 38% over time (CESR, 2010b).
^ Does not include Poland.
Source: Authors calculation from CESR (2010b).
Externalities. Different views emerge on where dark liquidity might

be concentrated. Some authors argue that stocks with high volumes, high
market depth and low volatility are mainly traded dark (Degryse et al.,
2008; Buti et al., 2010b). Others (Ready, 2009) suggest that dark liquidity
trading would be converging on stocks with high spreads, in order to find a
better deal on other venues. In effect, the idea is that although high liquid
shares may receive easy execution on dark pools liquid shares will receive
low execution costs on other venues as well, so the opportunity costs of
trading on lit pools are very small. There are few incentives to look to other
venues. With large spreads on lit books, instead, traders may search for
other venues giving price improvements (Cheuvreux, 2010). In the EU, this
should be even truer as traders get at least the mid-point (but also a volume
weighted average price, VWAP)62 from dark pools applying the reference
price waiver,63 while lit trading venues may offer very high bid/ask
spreads on that specific product. Ready (2009) also suggests that
institutional investors with big average positions and less turnover make
more use of dark liquidity, in line with the theoretical framework.
Dark liquidity may also generate negative externalities, in terms of
liquidity fragmentation and order migration. In effect, adding a dark pool
to a liquid market with OLOBs may lead orders to migrate towards this
new pool (cream skimming), which stimulates the arrival of new orders
(with liquidity-externality effects). The new environment seems to benefit
institutional investors, but it may decrease the price discovery ability of
retail ones, as they will not be able to see a part of the market (less efficient
price discovery, Buti et al., 2010a). Analysts find that overall the total
welfare increases, but the distributional effects remain. Moreover, the
abovementioned studies even though with limited evidence seem to
concur that an excessive volume of dark liquidity may generate negative
effects and market operators would not have incentives to reduce those
effects once they materialised. In effect, high network externalities and
effects on revenues may push players to act out the prisoners dilemma
(bad equilibrium). There may be no incentive to cooperate when a
damaging threshold is reached, even though the final outcome may not
benefit any of the pools in the market. Therefore, regulators may need to
62 See CESR (2010e, p. 3).

63 This venue uses prices formed on lit venues, mainly regulated markets.
monitor these liquidity pools to learn when their size begins to undermine
market quality (mainly price formation and discovery processes).
IoIs. In order to increase the market quality of the liquidity pool, dark
trading venues (dark MTFs/RMs) may adopt a non-binding indication of
interest (IoI) messaging system. This system is typically designed to attract
liquidity, by giving access to privileged private information on trading
interests, including size, price (or a targeted weighted average price), buy
or sell, and often the symbol of the security. IoIs are typically sent to some
traders on a discriminatory basis. Therefore, those investors may benefit
from the network of dark pools without competing for liquidity with other
traders.
Some market operators argue that this might be an elegant way to
circumvent the public pre-trade disclosure of quotes, discriminating
against potential investors who might be interested. At the same time,
disclosure of IoIs does not bind the counterparty to conclude the trade. On
the one hand, a non-binding IoI messaging system may increase legal
uncertainty by stimulating expectations (inefficient commitment) about the
stipulation of the contract (trade). On the other hand, MiFID recognises the
importance of non-discriminatory access to order information on trading
venues (RMs/MTFs) in addition to the overarching objective for organised
trading venues (as defined by CESR, 2010g) to comply with pre-trade
transparency requirements. Therefore, CESR (2010b) considers the
provision of private valuable order information on a discriminatory basis
as a violation of MiFID principles (unfair), in conformity with the SEC
policy (2009). However, the Committee did recognise the role of IoIs in
OTC trading if finalised to find selected counterparties to a large order
waiting for execution. This practice was already widely in use before the
implementation of MiFID.
OTC trading. Dark liquidity can sit on official trading venues, both
MiFID-compliant and over-the counter. MiFID recognises the importance
of OTC trading (Recital 5364) as part of the general freedom of investors
64It is not the intention of this Directive to require the application of pre-trade transparency
rules to transactions carried out on an OTC basis, the characteristics of which include that
they are ad-hoc and irregular and are carried out with wholesale counterparties and are part
of a business relationship which is itself characterised by dealings above standard market
size, and where the deals are carried out outside the systems usually used by the firm
concerned for its business as a systematic internaliser., Recital 53, MiFID. For more details
about OTC trading venues, see section 5.4.
(shareholders) and legal entities to trade shares privately if specific

circumstances are met. It also acknowledges that OTC trading can promote
the smooth operation of financial markets by providing in some instances
better investment services outside MiFID official venues. There are
conflicting views on the actual size of OTC trading (in terms of price
forming trading), which this report will consider in the next sections.65
CESR estimated OTC trading as 38% of trading in the EEA. However, as
CESR itself stated (2010b, p. 35), these data may be inflated by double
counting and misreporting; hence a lack of quality and accuracy urges
actions to reduce inconsistencies and increase granularity.
4.2.2 Pre-trade transparency waivers: Where do we stand?

Pre-trade transparency obligations are placed on both RMs and MTFs by
MiFID for shares listed on regulated markets (for Sis, a specific system of
publishing quotes applies).66 The Directive, however, acknowledges the
risk of market impact caused by transparency obligations for certain types
of orders and trading venues. In response, regulators have introduced a list
of exemptions in specific cases. In particular, MiFID recognises pre-trade
transparency waivers for market models and for orders of a certain type or
size. However, dark trading may deteriorate market quality if it reaches an
undefined overall size. Empirical evidence does not provide a clear answer
either to whether a threshold for dark liquidity and waivers would
capture the critical point or where this overall threshold should be placed,
whether by regulators or the market. Therefore, waivers of pre-trade
transparency should be constantly monitored and updated as soon as
market developments make it necessary. CESR (2010b) is thus proposing a
more rule-based approach, with ESMA defining binding technical
standards and ongoing supervision in line with its new powers as a new
European authority.67
65 For a more detailed discussion, see section 4.4.

66In particular, SIs should publish quotes for liquid shares if executed orders are below the
Standard Market Size (SMS); Art. 27, MiFID; Arts 21-25, Implementing Regulation.
67Art. 6 (1) (a), (2) (a), Art. 7 and 7e, adopted position at first reading by European
Parliament on 22 September 2010 (http://www.europarl.europa.eu/sides/getDoc.do?
pubRef=-//EP//TEXT+TA+P7-TA-2010-0339+0+DOC+XML+V0//EN#title3).
According to MiFID,68 there are four kind of waivers:

i) Reference price69
ii) Large in-scale orders (LISOs)70
iii) Negotiated trades71
iv) Order management facility.72
These waivers serve multiple functions:
i) Stimulating price competition (improvements) and containing
information leakage (reference price),
ii) Protecting orders from market impact (LISO) and
iii) Ensuring the smooth functioning of capital markets and execution
services (negotiated trades and order management facility).
Overall, waivers aim at reducing the exposure risks that may affect
market quality and integrity. Well-functioning financial markets allow the
efficient allocation of resources and thus the increase of total welfare for all
end investors.
Reference price. The reference price waiver applies to RMs and MTFs,
given the importance of alternative dark trading venues when it comes to
competition on trading costs and potential price improvements. Those
venues are typically MTFs that try to offer protection from information
leakage and market impact at competing prices, with no price discovery
(passive pricing system). European trading under reference price waivers
has been estimated around 25 billion in the Q1 of 2010, 5 times more than
For a detailed overview of the MiFID waivers (with examples), see CESR (2010e) and
68
Moloney (2008, pp. 826-830).

69 Art. 18.1 (a), Implementing Regulation. This waiver applies to trades that are crossed at a
price generated by another trading venue. In this case, the trading venue uses prices of more
liquid pools of liquidity in order to avoid the risk of not reaching the critical mass of
liquidity. See CESR (2010e, p. 3).
70 Art. 20, Implementing Regulation. The waiver applies to orders equal to or above a
minimum size specified in Table 2, Annex 2, Implementing Regulation. The calculation of
the normal market size should be made using the average daily turnover, which shall be
calculated as defined by Art. 33, Implementing Regulation. CESR (2010e, p. 17).
71 Art. 18.1 (b), Implementing Regulation. Specific conditions apply to negotiated trades that
avoid pre-trade market transparency as they are subject to different conditions than those
currently offered on public markets. CESR (2010e, p. 11).
72 Art. 18.2, Implementing Regulation. This waiver applies to orders held in an order
management facility (or Reserve) run by regulated markets or MTFs, which have the
potential to be introduced in the order book to be executed for instance against incoming
aggressive orders. See CESR (2010e, p. 14).
in the same period of 2009, but only accounting for 1% of total trading on
RMs and MTFs (see Figure 11). As mentioned above, the presence of
alternative dark venues can improve market quality and stimulates new
flows of liquidity with a beneficial impact on investors choice and
competitiveness of trading venues. This waiver is designed for venues
willing to offer lower trading costs and potentially price improvements in
relation to the referenced venue. Orders should be executed at mid-point
of the bid/ask spread of the primary market or at mid-point or best bid or
best ask of the European Best Bid and Offer (EBBO), as currently
established by CESR (2010e). However, market participants have
conflicting views on this issue. They are split between those who would set
a specific threshold (individual/aggregated) for the size of orders that use
this waiver or capping the volumes regarding trading under waivers), and
those who would leave the decision to the market on how many trading
venues should use this waiver with the obligation for these venues to
provide price improvements in exchange for the waiver.
Market views. The former group, in effect, implicitly proposes to set a
threshold (e.g. a percentage of Large-In-Scale Orders;73 see Fleuriot, 2010)
that de facto makes this waiver available only to a type of order (large ones)
disregarding the type of trading system (as also pointed out by some CESR
members; CESR, 2010b). In their view, MiFID foresaw waivers to limit the
market impact of large orders and price hence should be limited to the mid-
point as only at that point of the spread would it be justifiable for those
orders not to contribute to the price formation process. Against this view,
other market participants argued that this waiver was originally designed
to favour competition on explicit and implicit trading costs between dark
and lit books. In their view, therefore, these venues should be allowed to
match orders within the spread of primary markets or EBBO, with the
guarantee of a price improvement. It is important to ensure a price
improvement when the price comes from another system. A limitation
based on size of orders in their view would harm the market, as no
longer will small or child orders (of large parent orders74) be traded
dark, with no protection from information leakage and so exposure risks.
73As defined by Table 2, Annex II, MiFID Implementing Regulation; for the purpose of
determining the size of large-in-scale orders, the average daily turnover is defined by Art.
33, Implementing Regulation.
74 As part of slice and dice defensive trading strategies in order to reduce market impact.
In addition, this waiver currently captures only a small fraction of the

market (1%).
Figure 11. Current use of waivers

180 6.0%
160
5.0%
140
120 4.0%
100
3.0%
80
60 2.0%
40
1.0%
20
0 0.0%
Q1/08 Q2/08 Q3/08 Q4/08 Q1/09 Q2/09 Q3/09 Q4/09 Q1/10*
NegotiatedTrades(bn) LISO(bn) ReferencePrice(bn)

NegotiatedTrades(%) LISO(%) ReferencePrice(%)
Source: CESR (2010b).
LISO. The large-in-scale order waiver applies to orders that are bigger
than special thresholds set by MiFID for each defined liquidity band
(average daily order book turnover, ADT).75 It tries to soften the exposure
risk of large orders, which may generate market impact. It is largely used
by MTFs and RMs (for their hidden orders pools). As suggested by the
literature review above, hidden orders increase the quality of lit books as
they attract liquidity from large traders and informed traders, who try to
act more aggressively in order to capture this new liquidity flow. Therefore,
in order to ensure high market quality, the LISO waiver should work
properly and be able to capture fundamental liquidity in capital markets.
Market participants views are here again conflicting, and include
those that would reduce the threshold given that the average size of orders
75 See footnote 73.

is constantly dropping76 and those that fear inflating dark markets to the
benefit of a small part of the market that makes large use of this waiver,
while eventually deteriorating market quality. The former group claims the
importance of applying the waiver in a consistent and efficient manner. As
recently estimated in the UK market (LSE Group, 2010), the difference
between the standard market size (SMS)77 and the LISO thresholds is on
average 44 times larger (with high volatility within the basket of most
liquid stocks). In addition, it emerges from these calculations that by
reducing the LISO size on the FTSE 100 by 75% instead of the 25%
proposed by CESR, the capacity of the value of orders that may qualify for
the LISO waiver would reach almost 2% (circa 17% for the small and mid
cap index). Data show how the high variability of the LIS threshold
generally depends on the trading venue characteristics and size. This
would suggest that deciding the LIS size by liquidity bands may be correct
but the current thresholds set by each band may not reflect the changes in
the standard market size, which has been gradually going down in recent
years (in particular, with the introduction of MiFID). Therefore, this part of
the market is asking for a review of the LIS thresholds and a consistent
reduction in order to allow lit order books (in particular, those using
hidden orders facilities) to compete more efficiently especially with dark
MTFs.
The latter group, instead, believes that current thresholds are
sufficient and that a reduction would increase the size of non-displayed
trading and liquidity fragmentation, ultimately hampering the price
discovery process. Small investors claim that the reduction of the average
size of orders has been mainly caused by trading methodologies of
intermediaries, so a reduction of the LIS may only benefit them, with
unclear effects on end investors. In their view, the size of small end
investors orders has not decreased; therefore the way in which
intermediaries handle them should not affect in any case the way in which
end investors decide their investments.
76 See section 5.5.

77 Orders above the SMS are typically considered as orders that may deserve special
treatment due to their market size, also outside the official trading venues (OTC; Recital 53,
MiFID).
CESR also discussed the treatment of residual LIS orders (so-called

stubs), with no final answer but a majority of its members agreed that
stubs may need to become lit to avoid distorting price formation (CESR,
2010b). In their view, this will not have the same market impact. However,
other members of CESR and some market participants have highlighted
that the immediate publication of stubs may reveal the original size of the
large order or leak private information on the handling of these trades
(especially if the execution is not yet completed), leading to stubs not being
executed at all. If a stub is modified by the trader, market participants
generally agree that the LIS waiver should not apply.
Negotiated trades. The negotiated trade waiver is typically used for
transactions bilaterally negotiated with other parties, so not accessible to
other members on RMs and MTFs. A protection from pre-trade
transparency would here enable intermediaries to provide best execution
(CESR, 2010b) and give the opportunity to investors to exercise their
contractual power. Applying pre-trade transparency requirements to these
trades may deceive price discovery processes since economic terms of the
execution are the result of specific conditions that do not reflect current
market prices. Moreover, the waiver aims at avoiding some trades from
destabilising displayed continuous trading systems due to their systemic
importance. In Q1 2010, those trades were 3.7% of EEA trading on RMs and
MTFs, with a stable trend from 2008. From an economic and financial
standpoint, negotiated trades are very similar to OTC trades, as upstairs
trading.78 The main difference is that an OTC trade carries best execution
duties, since it is executed through a broker-dealer (execution on behalf of
the client), while negotiated trades do not benefit from best execution
duties since they are the result of face-to-face or back-to-back
transactions.
Market views. There is widespread agreement that this waiver should
be retained. It plays a crucial role in reducing systemic risk of OTC trades
that cannot be executed on a central trading mechanism. It is also used
when principal transactions (back-to-back) are not subject to current market
price but to a principal volume weighted average price (VWAP; CESR,
2010b). Some market participants believe that this waiver should be
extended also to RMs and MTFs that do not offer a displayed order book or
78 For instance, in the UK, MiFID gives the possibility to investment firms to report OTC
trades either as negotiated or as OTC.
continuous trading.
Turning onto the order management facility waiver, there is
widespread agreement that this waiver should be kept without major
changes. It allows better management of orders kept in reserves (e.g.
iceberg orders) in order to be used whenever the market would be under
stress due to several aggressive orders hitting the book at that moment. The
waiver therefore allows a more efficient execution service, attracting more
liquidity through brokers. CESR (2010b) clarified that differences between
RMs/MTFs and brokers applying this waiver should remain, as they
perform two different functions in financial markets.
Conclusion # 3
Under certain conditions, pre-trade transparency may impair market liquidity. Hence,
MiFID introduced waivers, which should be retained. A move towards a more rule-based
approach should be balanced with flexible application and ongoing supervision in order to
meet market needs. However, conflicting views between stakeholders emerge when
discussing the breadth of these exemptions. In effect, thresholds may need to be revised
regularly, in line with latest market developments. More specifically, regulators need to
carve out a new set of rules that promote the smooth functioning of capital markets and
meet investors needs with no adverse impact on market structure, which may ultimately
affect market liquidity, efficiency and investor confidence. Consistent and uniform
application across Europe should be ensured.
4.2.3 A consolidated quote system

The market entry of several trading venues in the last three years raised
concerns around the availability of pre-trade transparency data across
venues. The difficulty faced by investors, retail in particular, to access pre-
trade consolidated data solutions and a pan-European best bid and offer
(EBBO) may pose a potential obstacle to the creation of a competitive pan-
European market. The provision of consolidated quote solutions would
improve investors choice and increase competition between trading
venues on spreads, with the possibility for liquidity providers to compete
in a truly transparent environment. In effect, if investors are able to see the
best price across trading venues, they may be able to push intermediaries to
bridge links with these infrastructures offering better deals. Obstacles to
cheaper and more easily accessible consolidated quote solutions are in
general the same as currently experienced by the post-trade transparency
space (see next section). Once these barriers are removed, markets would
be able to offer new data consolidation services and meet investors needs.
Conclusion # 4
Despite the importance of a consolidated quotation system, priority should be given to the
removal of relevant impediments to widespread use of consolidated post-trade data
solutions. In particular, it is important to improve investors access to both pre- and post-
trade data solutions, in order to transform fragmentation into beneficial competition for end
investors. Data accessibility to regulated and OTC venues is key for data consolidation.
4.3 What pre-trade transparency for non-equity instruments?

Recital 46. Under this provision, MiFID, gives the faculty to member states
to broaden the transparency regime to financial instruments other than
shares. Only two member states have exercised this option and extended
this regime to bond markets (Italy and Sweden).
There are two main reasons why this option has not been widely
exercised: i) non-equity markets79 have a different market microstructure
from equity markets and trading mechanisms are designed to enhance
resilience in case of liquidity issues; ii) A general possibility to opt in may
not give enough incentives for member states to exercise the option,
thereby pushing them to behave strategically. In effect, the latter represents
a typical prisoners dilemma. Member states have no incentives to behave
cooperatively. If they deviate from their current application of MiFID, they
may put their markets at a disadvantage, in particular if they impose
stricter requirements for non-equity instruments.
Dealer markets. Non-equity markets, moreover, have different trading
mechanisms. Trades are typically dealt with a quote-driven dealer system
or bilateral negotiations with the support of an intermediary. Most
transactions are arranged via a dealer. Trading can be continuous or
periodic and dealers typically have an informational advantage over
investors as intermediaries. Accordingly, market-makers move first and
propose price schedules. Therefore, equilibrium may be found in dealer
markets (with one specialist or competing market-makers), since they are
more resilient and efficient as long as the market is not too big and the
information asymmetry not too large (Glosten, 1989; OHara, 1995). Dealer
79 CESR (2010f) has defined as non-equity financial instruments: a) corporate bonds; b)
structured finance products (Asset Backed Securities and Collateralised Debt Obligations);
c) Credit Default Swaps; d) Interest rate derivatives; e) Equity derivatives; f) Foreign
exchange derivatives; g) Commodity derivatives. Other non-equity financial instruments
that should be considered are sovereign and local authorities bonds, and equity-like
instruments.
markets are hence designed to deal with less-liquid markets, only if this
low liquidity condition is verified. Dealer markets, in general, can:
1) Provide stronger protection against exposure risks (Pagano & Roll,
1993),
2) Reduce information leakage,
3) Reduce costs for uninformed traders (if they have contractual power
due to the market illiquidity or competition in the dealer market) and
4) Better price if dealers can exploit private information and investors
can control their fair use (Madhavan, 1995).
There are other two specific reasons why most non-equity products,
even less complex ones, are currently traded in quote-driven markets, with
a strong dealership:
i) Structure of the demand and
ii) Structure of the intermediation.
Firstly, demand for such products comes mainly from institutional
investors and other dealers, since retail participation remains highly costly
in terms of cost and lack of sufficient knowledge. The nature of clients, and
the complexity, heterogeneity and large size of non-equity financial
instruments affect market structure and thus the prices of financial
instruments. In dealer markets, institutional investors can exploit their
contractual power and customise their transactions more than in a non-
discriminatory and open market setting (Biais & Green, 2007). Moreover,
dealers can have more control of their risk positions when they deal with
few market operators and exposures.
Secondly, the structure of the intermediation favours certain market
developments. In effect, the amount of capital committed to intermediation
by dealers may affect per se the choice of a specific level of transparency. If
dealers act as a principal in the transaction, on the one hand, there is more
chance to create a market for illiquid products by selling them direct to
investors, but on the other hand financial products will sit on dealers
balance sheets, with inventory positions that need some protection from
market impact through less public disclosure, in order to be still able to act
as intermediary. Then, if dealers only act as broker-agents, on the one side,
there will be less capital committed and thus less need for protection from
market impact and more public disclosure. But, on the other side, illiquid
products may not receive execution due to insufficient demand. The
complex nature of modern financial markets and economies requires
intermediaries to provide clients with more tailored and sophisticated
products and services, which may often be traded only in specific market
settings and with limited public disclosure. The use of electronic multi-
dealer platforms has surely increased competition between dealers, who
compete with executable quotes, and extended market accessibility to a
wider set of investors who qualify as members (typically only brokers or
institutional investors). A strong push towards pre-trade public disclosure
may require a rethinking of current market structure and intermediation of
less-liquid asset classes (such as some categories of bonds, OTC derivatives
and structured products), which may anyway lack price transparency since
they are typically based on proprietary valuation models (CESR, 2009).
Benefits and costs. Greater pre-trade transparency can indeed reduce
search costs and enhance price discovery processes as long as all relevant
investors can easily access these markets and dealers are not damaged by
information revelation. Flood et al. (1999) found that in a dealer market
opening spreads would be wider and trading volume lower, but price
discovery should be faster since traders would behave more aggressively in
order to find liquidity. Other authors therefore find a negative trade-off
between liquidity and price efficiency if transparency varies. Bloomfield
and OHara (1999) did not find any relevant impact of pre-trade
transparency requirements on dealer markets. Greater transparency, on the
one side, may reduce the informational advantage that dealers exploit in
exchange for the provision of liquidity and introduce uncertainty in
bilateral negotiations. The nature of the counterparty, for instance, is
usually included in the calculation of the price offered by dealers, which
cannot usually be estimated without an explicit request from the client. On
the other hand, however, the absence of pre-trade information may keep
the price discovery process for non-equity products very costly and
therefore inefficient, and it may promote a market setting with low
competition.
For all these reasons, a pre-trade transparency regime for non-equity
financial instruments should be designed in a different way from auction
order-driven markets or quote-driven dealer markets (duly adapted and
proportionate). In this regard, MiFID pre-trade transparency requirements
cannot be transposed as they now stand in order to prevent unintended
consequences on the incentives of dealers to provide liquidity through the
use of private information. In effect, pre-trade transparency may improve
search liquidity, since pre-trade information is asset-specific and helps to
reduce search costs and uncertainty around them, even in dealer markets
(Lagan et al., 2006). However, a pre-trade transparency regime could be
effective, depending on the number of potential counterparties that can
offer that tailored product and on the level of exposure of dealers

inventory positions to the market, since trade size is typically far higher
than the retail one. This leads dealers to offer executable quotes only on
electronic platforms where liquidity is selected and accessible to members
meeting certain requirements. On platforms where traded products are
commonly accessed by retail investors, it would be easier to implement
general pre-trade transparency requirements as long as the demand is very
high, which may not be the case for complex non-equity products. Current
market microstructure makes public disclosure of executable quotes
difficult and costly (Biais et al., 2006). Also, many non-equity products are
highly customised, so pre-trade transparency might be of little help. The
greater concern is liquidity, which is to find a dealer that can tailor a
transaction around the clients needs. If transparency requirements reduce
market makers returns, they may ultimately confine their activities to
those products that are inherently more liquid, exiting markets for less-
liquid ones. Pre-trade transparency for complex non-equity products could
only work by changing the way how these instruments are actually traded
towards open limit order books. To achieve this, the nature of the demand
needs to be constant and sufficiently high over time.
New proposals. The Commission (2010b, p. 28) has finally proposed
that investment firms willing to quote or receive a request for quote (RFQ,
probably run on organised trading facilities) would be requested to publish
price and volume available to the public, and eventually commit to it for
sizes below a certain threshold (retail size). The threshold will be specified
per asset class.
Conclusion # 5
A strong push towards more pre-trade public disclosure would require, in some
cases, a rethink of the current market structure for less-liquid asset classes, and a shift from
its mainly institutional demand to a more retail and smaller professional one. Conflicting
views in this area emerge around what should be the most efficient market structure for
these products. Liquidity in non-equity markets, such as markets for bonds, derivatives and
structured products, is currently handled through quote-driven auction markets, inter-
dealer platforms or bilateral negotiations led by dealers capital commitment.
For auction markets, whether led by dealers/market-makers (quote-driven) or
directly by demand (order-driven), pre-trade transparency is urgently needed. For inter-
dealer platforms (request-for-quotes model) or bilateral negotiations, where dealers commit
capital by being non-neutral counterparty, less pre-trade transparency than order-driven
ones (e.g. equity) is needed to function properly. Executable prices might thus not always
be consistently available. The alternative to a shift in market structure and in demand,

which may not necessarily occur, is to design a different transparency regime from the one
applied to equities. In general, pre-trade transparency may not be of any help to the market
if it does not serve price discovery of investors. For markets in which the demand is more
retail-driven, the need to have greater pre-trade transparency will be higher as it will
noticeably reduce search costs. However, an appropriate level of pre-trade transparency
may be beneficial for non-auction markets as well, as it may reduce investors search costs
and promote greater competition between dealers. Some market participants consider that
currently available pre-trade data is sufficient and legal requirements are not needed, since
non-equity markets do not necessarily function as equity markets and current transparency
have not limited participation to these markets.80 Others instead (and CESR, 2010f) believe,
however, that an ad hoc pre-trade transparency regime (with waivers) should apply for
non-equity products (in particular, bonds) listed on RMs or MTFs. Access to this
information should be made easier for retail investors.
Conflicting views also emerge as to how to design the transparency regime. Many
believe that preference should be given to the nature of the market rather than the nature of
investors (retail versus wholesale). For financial instruments mainly traded over-the-
counter on a bilateral basis, the introduction of any pre-trade transparency regime must
prioritise the avoidance of adverse liquidity consequences for involved counterparties and
systemic risk.
4.4 Post-trade transparency regime: Fixing bugs and extending

scope
Definition. Transparency may take on different forms and dimensions,
according to the mechanisms underpinning market microstructure and
investors needs. In conjunction with a pre-trade transparency regime,
MiFID established a regime of post-trade transparency, which applies to all
shares admitted to trading on a regulated market, wherever they are
actually traded.
Benefits. Post-trade transparency is particularly important to preserve
market integrity through the indirect control of market participants upon
competing investors (discouraging insider trading). The regime also
supports price formation processes since it reduces informational gaps and
investors search costs, with overall price efficiency gains, also for quote-
80 For instance, AFMEs www.investinginbondseurope.org provides pre-trade non-

executable price on more than 1,500 bonds (government, sub-government and corporate),
which may help end-investors find the best price for the financial product they are willing to
purchase. Regarding OTC derivatives, Markit publishes intraday indicative CDS prices for
instance (http://www.markit.com/cds/cds-page.html) and Bloomberg, Reuters, and
TradeWeb provide pre-trade information on interest rate derivatives.
driven markets (Flood et al., 1997; Bloomfield & OHara, 1999).

Accordingly, trade disclosure may reduce adverse selection and increase
the markets reputational capital, as the market will be less open to
manipulation (Chowdry & Nanda, 1991). Market participants may thus
develop better services and provide best execution at a lower transaction
cost, in particular
Costs. However, while transparency is certainly instrumental to
promoting liquidity in order-driven markets (e.g. cash equities) if
calibrated with delays for large transactions, quote-driven and dealership
markets may suffer from a strict trade disclosure regime (e.g. OTC
derivatives). Dealers may no longer benefit from their informational
advantage and may thus be subject to exposure risks81 and inventory
risks,82 since they need to manage large amounts of their own capital in
related instruments. Therefore, they may commit less capital to the market
and reduce their willingness to bear risk, thereby reducing liquidity
(Madhavan, 1995, 1996; Hendershott et al., 2007). Increased exposure risk
also means that uninformed investors may benefit from greater
transparency, with potential effects on market liquidity, however (Pagano
& Roll, 1996). Moreover, the disclosure of trade information (e.g. price)
under specific bilateral agreements may deceive the price formation
mechanisms and asset price valuation, as the price is the result of non-
replicable market conditions. Finally, Kovtunenko (2002) argues that with
full post-trade transparency, that is signalling the size of the uninformed
demand and observing other dealers quotes dealers can act strategically
to control the level of spreads as the size of uninformed demand varies,
and punish those who deviate from a collusive equilibrium (through a
price war).
Microstructure. Overall, transparency does not necessarily smooth
liquidity issues away, especially in quote-driven markets. Strict
transparency requirements may create a shift in market microstructure
towards a trading mechanism with less active participation of dealers and
without necessarily setting off a more active participation of other
investors. In effect, the literature has tried to shed light on the role of
81 As illustrated in Section 4.2.

82The risk coming from unforeseen price changes on assets held for market-making
purposes.
transparency in micro-structural processes, but without success. For

instance, there is no clear answer as to the cause of past changes in the
bond market microstructure; it is uncertain whether lower transparency
was the cause or the effect of that structural change (Biais & Green, 2007).
In coming years, aware of potential side effects, policy-makers will need to
decide whether they want stricter transparency requirements to promote
access by small professional and retail investors or more complex (not
necessarily riskier) financial instruments. There is a trade-off between
enhancing investor confidence and disclosure costs. Changes here may
modify the market structure, not always with beneficial effects.
Systemic risk. The recent financial crisis has added a further
dimension to transparency, which refers to monitoring and controlling
macro-prudential risks by competent authorities in specific markets for
financial stability purposes (e.g. OTC derivatives) through the use of
aggregate data on exposures in single or multiple financial instruments
(FSB, 2010). These data are increasingly available, typically through the
establishment of dedicated trade repositories per asset class. The need to
control systemic risk (and minimise spill-over effects) by macro regulators
has exercised some pressure to increase the overall transparency of
financial markets. This has occurred even in segments where transparency
was previously limited, due to a microstructure based on the role of dealers
and bilateral over-the-counter transactions.
Box 3. Transaction reporting: Current regime and CESRs proposals
Post-trade disclosure enables regulators to preserve market integrity and

surveillance across several trading venues by detecting transactions that may
breach market abuse rules, or represent other breaches of MiFID requirements.
MiFID introduced a regime of transaction reporting (Art. 25)83 and cooperation
between major financial authorities (Art. 58) that allows regulators to
confidentially access information concerning transactions executed in any
financial instrument admitted to trading on regulated markets. Both CESR and
the Commission proposed to extend the scope to also cover financial instruments
admitted to trading only on MTFs and organised trading facilities (OTFs;
Commission, 2010b, p. 46). The report can be sent to the competent authority by
the investment firms, by a third party on their behalf, by an approved trade-
reporting system or by the regulated market or MTF; and it shall be delivered by
83 In conjunction with Art. 9-16, Implementing Regulation.

the close of business the following day. The competent authority may request
additional requirements in exceptional circumstances.84 Data on transactions
should be stored for at least five years and ongoing reporting should meet
specific criteria, including details on names, dates, times, quantity, prices and
codes to identify parties and venues.85 Reports should be sent to the competent
authority no later than the following working day after the execution of the
transaction. Transaction refers to any agreement concluded with a counterparty
to buy or sell one or more financial instruments (EU COM, 2010b, p. 47).
CESR has currently proposed amendments to the transaction reporting
regime.86 First, the Committee has proposed to add a third trading capacity to
transaction reports (in addition to principal on own account and agent), the so-
called risk-less principal, which consists of the investment firm acting on its own
account and on behalf of the client (client facilitator). This scenario would
simplify the harmonisation of current transaction reporting standards across
Europe.87 The immediate implementation of this proposal raises some concerns
since automated reporting systems would need to be changed accordingly.
Secondly, CESR advanced a proposal to mandate the collection of client
identifiers by competent authorities, increase their standardisation, and include
client IDs when investment firms transmit orders for execution.88
On the latter, the Directive requires reporting of executed transactions,
which may not necessarily require the disclosure of the clients ID if the order is
executed by an investment firm other than the one that received the order
execution.89
The Committee then suggested extending transaction reporting to
members of RMs, MTFs and OTFs that are not authorised as investment firms
(e.g. market-makers) since they fall outside MiFID (assuming they do not provide
investment services). Trading activities of non-members also contribute to price
formation, and leaving them out of the sight of regulators may undermine the
84 Art. 4, Implementing Directive.

85 Art. 13 & Table 1 Annex I, Implementing Directive.
86 See, in general, CESR (2010a).
87 For instance, in some countries (such as the UK) the category of riskless principal is
already present in the transaction reporting regime.
88 CESR (2010a, pp. 9-16).
89 Following Art. 5, Implementing Regulation - for the purpose of the transaction regime
transaction represents the purchase securities financing transactions, the exercise of
options or of covered warrants, and primary market transactions (such as issuance,
allotment or subscription) in financial instruments falling within Article 4(1)(18)(a) and (b)
of Directive 2004/39/EC.
market monitoring function of national competent authorities. Alternatively,

transaction reporting obligations could be imposed on RMs, MTFs or OTFs which
have admitted these firms as a member, but these obligations would just shift the
reporting responsibility to entities that do not directly control this confidential
information. Reporting may thus be inaccurate for supervisory purposes.
In conclusion, the Commission (2010b, p. 49) to modify the reporting channels as
defined by MiFID, in line with the CESR advice. In particular, the Commission
will assess: 1) the viability of a consolidated European reporting mechanism to
which investment firms will report directly and competent authorities will have
direct access; 2) third party reporting firm should be approved as Approved
Reporting Mechanisms (ARM); 3) waiving reporting transactions if they have
already been reported to trade repositories or approved by a competent
authority, such as ARM. In line with the trade reporting system of consolidation,
the proposal would establish a European consolidator run most probably by
ESMA, which will work across asset classes. For trade reporting (public
disclosure), instead, the system of data consolidation will need to treat different
asset classes with separate tapes. Investors, therefore, will be able to subscribe to
the data solution that fits their needs at competitive costs.
Conclusion # 6
Overall, post-trade transparency serves various important functions in financial markets,
such as improving price formation and market integrity. In effect, trade disclosure may
stimulate self-designed market surveillance and increase visibility at the same time as
market liquidity, thus increasing investors confidence and promoting price discovery.
However as illustrated in the following sections there is a need for regulatory actions in
certain areas, not only as a result of the recent financial crisis but also of the experience
gained since the transposition of the Directive. Transparency is not panacea for market
failure, though, and interventions should be proportional to the market structure and
dynamics through which investors orders find their market-clearing price. Ill-defined
transparency requirements would harm market efficiency in less-liquid markets with no
increase of investor protection or reduction of systemic risk, as the market would become
less-liquid and more volatile.
Finally in order to reduce the risk of manipulation on less-liquid thinner markets
and improve market integrity and surveillance it would be appropriate to extend the scope
of the transaction reporting regime run by regulators to all financial instruments admitted
to trading on MiFID-official venues, with no distinction between instruments listed on
regulated markets, multilateral trading facilities or organised trading facilities.
Nevertheless, this work needs to be coordinated with other initiatives at EU level, such as
the review of the Market Abuse Directive, and with a thorough cost-benefit analysis. In
effect, the extension would help the harmonisation of transaction reporting regimes and
supervisory practices across Europe.
4.4.1 The trade reporting regime for shares

MiFID set out a trade disclosure regime (Arts 28-30 and 45, MiFID) that
covers all shares admitted to trading on RMs, irrespective of where those
shares are actually traded (in RMs themselves, MTFs, SIs or OTC).
Therefore, MiFID extended post-trade transparency requirements to shares
traded on new trading platforms and for the first time over-the-counter.
Some member states had no transparency requirements for OTC trades in
shares before the Directive came into force (e.g. Germany). In terms of
scope, while for equity-like instruments an extension of the current
regime for shares is broadly favoured, the picture is less clear for
instruments other than shares (CESR, 2010b).
Reports of completed trades should include information on trading
day, time, instrument identification, price, quantity and venue.90
Publication should be made as close to real time as possible, but in any
case within three minutes of the relevant transaction.91 Deferral to official
time publication can be granted for large transactions,92 as a waiver to pre-
trade transparency, in particular to afford liquidity providers enough time
to unwind inventory positions before investors become aware of them. In
this regard, CESR has proposed to raise the thresholds for delayed
publication and reduce the maximum granted delay to the end of the
trading day. The latter is currently set to three trading days after the trade
execution (Table 4, Annex II, Impl. Reg.).93
Conclusion # 7
The extension of public disclosure requirements (trade reporting) to equity-like instruments
and to shares admitted to trading only on MTFs or organised trading facilities would help
harmonise requirements among financial instruments that serve similar purposes. For
financial instruments other than shares and equity-like instruments, the mere extension
would most likely generate inconsistencies with the very nature of these financial
instruments (see following sections).
90 Art. 27, Annex I Table 1, Implementing Regulation.

91 Art. 29, Implementing Regulation. The three-minute buffer, however, should be only used
if it is not technically possible to publish this data in real time. See Recital 18, Implementing
regulation. CESR has recently proposed to reduce this buffer to 1 minute.
92 Art. 28 and Table 4 in Annex II, Implementing Regulation.
93 CESR (2010b), Consultation Paper...Equity Markets, op. cit., p. 18.
In terms of time limits for trade reporting, new technologies can help reduce delays.
The industry is working to make all market data that are not subject to delays freely
available after 15 minutes, in line with ESMAs recommendation. However, reducing the
maximum allowed delay from three minutes to one may prove immaterial since this delay
cannot be exploited by trading platforms in favour of their members. In addition, trades
may be affected by technical delays (latency issues) that ultimately affect all market
participants. In any case, the legal obligation is to report as close to real time as possible
and should be duly enforced. Delays should be allowed in specific circumstances, with
appropriate calibration for trades done at the end of the day. Current proposals should be
further evaluated since they may affect risk positions, with potential adverse consequences
on the market.94
4.4.2 Challenges with data consolidation

New trading venues have entered the market to offer execution services on
shares across Europe, increasing market fragmentation. MiFIDs strategy
for achieving a truly pan-European equity market lies in stimulating a
more competitive environment. However, this strategy currently strives to
deliver. Despite the industrys commitment to make consolidated data
solutions more accessible (both tape and quote)95 and provide data on a
non-discriminatory basis, data remain costly, in particular for retail and
small professional investors. In addition, quality and granularity of trade
reports (e.g. OTC) is sometimes insufficient to generate cost-effective
consolidation.
Pre and post-trade data on equity can be consolidated in two ways:96
1) Through direct access to market sources97 and
2) Through solutions offered by data vendors.98
94The LSE Group (2010) has calculated that for off-book securities admitted to trading in
their regulated market and traded under their rules reducing the deferred publication time
to the end of the day (with no exception) would mean that 27.2% of overall trade value in
that market (1.4% as number of trades) would be pushed to disclose positions by the end of
the day. This may create unforeseen consequences and push liquidity providers to reduce
the amount of capital committed to these markets.
95 A consolidated tape solution refers to the aggregation in one virtual basket of all post-
trade data of competing trading platforms in a specific market (in this case, the European
Economic Area, EEA). A consolidated quote solution should potentially offer pre-trade data
of all EEA trading platforms.
96 For an overview of the market for market data, please see ESME (2009a).
97 Market sources are regulated markets, MTFs and SIs that provide static pre- and post-
trade data of shares admitted to trading on their systems, or investment firms publishing
data with proprietary arrangements.
Market data sources offer real-time information for shares traded on

their systems. These data have been typically offered by incumbents (RMs)
with different levels of granularity99 at a cost to final users, while new
comers (MTFs) currently offer this service for free as part of their strategy
to increase market share. Fees vary according to the type of user and the
use they make of these data. There are currently three types of fees that are
flexibly applied by trading venues (ESME, 2009a):
i) License distribution fee,
ii) License non-display fee and
iii) Data fee.
The first type of fee is designed for data vendors that resell data with
special features or together with data on other asset classes. The second one
refers to the fee typically charged to investment banks and buy-side firms
that directly access the market feed for their own use, typically associated
with other services (co-location,100 etc.). Those fees are designed in
particular for algorithmic traders or investment firms that systematically
use high-frequency trading tools. Finally, data fees are those charged to
retail or small professional investors to access pre and/or post-trade data
directly from their computers. If not freely accessible in real-time, post-
trade data are typically offered for free with a delay of 15 minutes, in line
with CESRs recommendation (CESR, 2010b). Despite the fact that the
average fee charged by regulated markets has partially decreased,101 the
98 Data vendors are providers of data that typically collect a vast amount of information
from several market sources and rearrange them in order to make them more easily
accessible to retail and professional final users.
99 There are in general two levels of granularity: level 1 (which only includes the touch
price) and level 2 (which gives a view over the order book, typically the first five best and
bid offers).
100Co-location services allow members of the trading platform to install their desks close to
the central data storage of the platform in order to reduce latency and receive data at the
lowest time physically possible.
101The fee for the last trade price charged by major exchanges (95% of EEA lit books
current volumes, but not as % of listed shares) is currently 75 per month. The full cost to
access a consolidated tape and quote solution in the US is roughly 70 per month. However,
the US data solution does not only offer last trade prices, but complete pre- and post-trade
data (levels 1 and 2), including the Best and Bid Offer (BBO), and covers 100% of listed
shares. To have only the level 1 pre- and post-trade data service in Europe for all EEA
markets, the price is around 409 per month (Atradia, 2010).
full cost (including data vendors fees and IT costs, where applicable) of
access to pre- and post-trade data (full level 1 data only) remains fairly high
(CESR, 2010b), in particular for small professional and retail investors.
However, regulated markets freely distribute post-trade data after 15
minutes. Moreover, CESR has also recommended the full unbundling of
pre- and post-trade data transparency information, which exchanges have
partially implemented in recent months by splitting post-trade from pre-
trade data.102 Revenues from market data are an important item of the total
revenues of exchanges (see Box 7). A drastic cut of fees might not be
manageable for them in the short term, since the collection and disclosure
of data is a service that normally comes at cost for market sources.
The second way to access consolidated post- or pre-trade data is
through data vendors, which pay distribution fees to exchanges and resell
data in a consolidated fashion. Since the costs of accessing several trading
venues are high, end investors may opt for one-stop-shop solution, even
though it may not offer the same depth and quality as the direct market
feed. However, the costs of one-stop data vending solution seem to be high
as well. Unbundling of services in this market may substantially reduce
costs for end users, however. In effect, a right to access single offered
services on a non-discriminatory basis may reduce costs for non-
sophisticated end users and allow data vendors to generate significant
economies of scale and scope by aggregating data from several asset
classes. This change would stimulate further competition103 and potentially
drive overall costs down further over the long term. In effect, concentration
is rather high in the current market setting. Two players, Bloomberg and
Thomson Reuters, capture two-thirds of the demand taking into account
the whole market, including data from asset classes other than equity (see
Figure 12). The market share of these two players is higher if only equity
market data is considered. The whole market for equity data sales and
trading is roughly $4.45 billion, of which $1.8 billion in Europe (Burton-
Taylor, 2010). Also in this market, the offer is typically done through cross-
selling practices (bundling or tying),104 which allow partial market
102 See FESE press release (http://www.fese.be/en/?inc=news&id=141).

103Bundling increases the specificity of the product, reducing its homogeneity. Competitors
will ultimately find it more difficult to compete with a bundle of several data services, with
no possibility to compete with single data services.
104Tying is a widespread business practice that occurs when two or more products are sold
together in a package and at least one of these products is not sold separately. In this case,
segmentation. In addition, a lack of standardisation of data formats

between data vendors impedes interaction and does not allow users to
access separate services of data providers from the same IT platform.
Figure 12. Data vendors market shares (all asset classes)
Other
Morningstar 21%
1%
Thomson Reuters
SIX Telecure 39%
4%
Dow Jones
1%
S&P/CaplQ
1%
Moody's
Analytics IDC
2% FactSet 2%
2% Bloomberg
27%
Source: Atradia (2010).
The combined effects of unbundling data services and fees for

exchanges and data vendors may reduce overall costs for final users that
seek pan-European consolidated data solutions.
Quality of market data. Although MiFID harmonised transparency
requirements among member states, the quality of market data has
partially deteriorated in some areas, mainly due to the combined effect of
market fragmentation and incomplete specification and implementation of
reporting requirements (CESR, 2010b; see Box 10). Several aspects,
however, affect the quality of data and its eligibility for consolidation
the end user would be able to buy the tied product alone, but not the tying product without
the tied one. Bundling occurs when none of the components of the package is sold
separately, and components are offered in fixed proportions. It is the simultaneous sale of
two or more products as a package, with no possibility to purchase both or one of them
individually. For a more detailed analysis, see CEPS and Van Djik Consultants (2009),
Tying and other potentially unfair commercial practices in the retail financial service
sector, a study submitted to the European Commission, 24 November
(http://ec.europa.eu/internal_market/consultations/docs/2010/tying/report_en.pdf).
(ESME, 2009a):
1) Quotes should be executable and not indicative only,
2) Data should be delivered on real-time, with no delay (except where
waivers and calibration apply),
3) Data should refer to liquid markets and
4) Data formats and flags should be sufficiently standardised.
In particular, the use of diverse trade flags105 does not allow an easy
consolidation of data. The objective should be the reduction of flags to less
than 10 across all European trading venues. Moreover, on the one hand, the
insufficient granularity and monitoring of OTC reporting increases the risk
of misreporting and the uncertainty around quality of price formation and
of best execution (see section 5.4.1). On the other hand, trade reporting
obligations under MiFID create a serious issue of duplicative reporting (see
example in 0), which impedes an assessment of the actual size and shape of
the market.
APAs. In order to improve data accessibility and reliability, CESR
proposed to publish post-trade information through Approved Publication
Arrangements (APAs), which can be RMs, MTFs, organised trading
facilities (OTFs) or other operators (but not an investment firm itself). In
this way, trading venues would compete with data vendors for the
provision of market data to consolidators.106 In order to facilitate data
consolidation, APAs should define stricter requirements, approved and
monitored by competent authorities. However, it is fundamental that APAs
are designed around harmonised standards107 in order to reduce
105A flag is a code attached to the information on a trade that signals its status and/or the
venue where the trade has been executed. This report will refer to flags more simply as
codes for venue identification.
106NYSE Euronext has recently announced the intention to introduce a consolidated tape
with real-time post-trade data of all regulated markets, MTFs and OTC markets. See
http://www.euronext.com/news/press_release/press_release-1731-
FR.html?docid=929763.
107In order to reduce inconsistencies in trade and transaction reporting across asset classes,
free international open (i.e, non-proprietary) industry standards such as those developed
by the industry in line with the ISO 20022 standard (Universal financial industry message
scheme) may offer a high degree of reliability, transparency and standardisation to
ultimately enable data comparability and analysis by authorities with the possibility if
needed to adapt them to the characteristics of the market. Moreover, the introduction of
universal standards may reduce the risk of market segmentation and ensure a more
competitive and harmonised environment. CESR (2010g) has proposed the introduction of
ISO standards for post-trade publication fields.
differences in the definition and implementation of APAs across the EEA.

These requirements should minimise issues of double-counting and
misreporting. The Commission (2010b, pp. 31-32) proposed to introduce
criteria such as: high data security standards; access to data at a
reasonable cost and on a non-discriminatory basis; procedures to identify
erroneous trade reports; adequate resources and contingency
arrangements; and conflicts of interest procedures.
In addition, CESR (2010g) proposed two options:
i) To prescribe only standards that data sources (RMs, MTFs, APAs)
would need to use to disseminate post-trade data to end users and
ii) To prescribe not only those standards, but also a common message
protocol for the transfer of post-trade data by RMs, MTFs and APAs.
The second option carries higher costs for data sources, which may
need to be further assessed. In any case, full harmonisation would make
data consolidation easier, since there would be no requirement for data
consolidators to map the data they receive into their own standard (CESR,
2010g, p. 6).
EU Mandatory Consolidated Tape. Whether or not a new APA regime is
effectively implemented, CESR believes that a set of requirements should
be inserted in MiFID in order to provide easily accessible and less costly
consolidated market data solutions in the EEA. This proposal is called EU
Mandatory Consolidated Tape (MCT). Consolidated tapes should be
designed by the industry and disclose post-trade information on shares
admitted to trading on RMs or MTFs in the EEA, wherever the execution of
the trade takes place. All information should come from RMs, MTFs or
APA real-time and fully unbundled. Data will be then consolidated and
sold at a competitive price by a given operator, who will meet strict
requirements and be responsible for the detection of multiple publications
(double reporting or counting). Only if the industry does not deliver this
solution by itself, should ESMA adopt the necessary arrangements to set a
US-style consolidated tape in the EU, run by a non-profit entity. MiFID
should give enough powers to ESMA to act accordingly. The Commission
(2010b, p. 34) proposed that a European MCT could be structured in
different ways, under variable degrees of regulatory intervention. The
proposal favours the creation of a consolidated tape for all financial
instruments instead of only equities admitted to trading in the EEA,
wherever the execution takes place. It should be noted however that
building a tape for assets other than shares would be difficult to achieve in
the short term, since a consistent and harmonised regime of post-trade

transparency for non-equity financial instruments does not exist yet. MiFID
should ensure that data consolidators would be able to collect all tapes and
provide one access point to end investors. A market-led solution would
probably be the least costly and most efficient option (Option C).
Competition among data providers would be the best way to spur the
creation of consolidated data solutions outside equity markets. The
Commission is concerned about the cost of data, which it wants to keep
under reasonable bounds.
However, what constitutes reasonable pricing is difficult, if not
impossible, to define in practice. Instead, the Commission should aim at
ensuring competitive pricing. Only real competition between data
providers would keep costs at a reasonable level. The Commissions
attempt (2010b, p. 34) to impose a single no-profit or for-profit entity
(Options A and B) for the provision of the consolidated tape would not lead
to competitive pricing, but would only impose a risky price regulation.
Rather, it would ultimately define costs for end investors by splitting
revenues among trading platforms and APAs, which could create
inefficiencies and distort trading incentives (as showed by the US
experience, see Box 4 below). Furthermore, the Commissions proposal to
require APAs to be expressly authorised for the provision of consolidated
tapes would slow down the whole process and ultimately reduce
incentives to compete. In sum, achieving minimum consistency though
industry cooperation would probably be more efficient than strong
regulatory intervention. In this regard, ESMA should rather support
current industry-led initiatives to major impediments to cheaper and more
efficient solutions. However, either the Commission or ESMA should be
able to impose consistency if commercial initiatives do not lead to a
satisfactory solution in a reasonable time frame.
Box 4. The US consolidated tape and quote system

The US consolidated tape system was officially created in 1976 by the National
Market System in an effort to promote economic efficiency and allow brokers to
deliver best execution (Caglio & Mayhew, 2009). Three networks (A, B and C) -
respectively run by NYSE, AMEX and NASDAQ display real-time trade reports
(Consolidated Tape System) and market quotes (Consolidated Quotation System).
The published information refers to securities traded on all exchanges, regional
markets, electronic crossing networks (ECNs) and broker-dealer crossing networks
and collected through three different networks by dividing shares listed on
national and regional exchanges (Networks A and B) and NASDAQ National
Market and Small Cap (Network C). Networks A and B are governed by the
Consolidated Tape Association Plan (CTA) and the Consolidated Quotation Plan
(CQ), while network C is governed by the OTC/UTP Plan. These plans collect fees
charged for the access to the consolidated tape and quotation and distribute them
across all primary markets, in line with a defined allocation formula.
Prior to 2007, at least half of the revenues from Networks A and B were
distributed in proportion to the number of reported trades, while Network C
redistributed half of its revenues in relation to the number of trades and half as a
proportion of the volume of trades in those shares. The allocation formula boosted
a widespread rebate programme through which exchanges pushed their members
to split their orders even though there was no risk of market impact. In effect, this
system created strong incentives to generate volumes, to print trades.
The implementation of RegNMS108 in April 2007 modified this allocation
formula by splitting revenues as follows: 25% depending on the number of shares,
25% on the number of trades, and 50% on quote aggressiveness, i.e. frequently
displaying better prices and thereby helping to narrow quoted spreads, while
distinguishing manually displayed quotation systems. The introduction of this
formula was followed by the SEC requesting the exchanges to adopt a rule against
tape shredding.109 Both these policies have helped to partially increase the average
trade size (Caglio & Mayhew, 2009), which proves that the allocation formula does
influence current trading activities and raises a trade-off between control over data
consolidation and indirect incentives on trading. Moreover, since the introduction
of this formula, the average number of quotes displayed every minute has
constantly increased (see figure below). Soaring trading volumes and algorithmic
trading (Angel et. al., 2010) may not be the only explanation for this later effect.
Hence, the impact of this new allocation rule deserves further investigation.
In addition to these networks, in March 2007 the SEC established Trade
Reporting Facilities (TRF), which report directly to the consolidated tape any
trades executed on venues other than national and regional exchanges.
Securities and Exchange Commission (SEC), Release No. 34-51808; File No. S7-10-04,
108
April 2007, Rule 601 and 603.

Members of exchanges can break up their customers orders into smaller trades only for
109
compliance with best execution obligation (best price).

Figure 13. Average quotes per minute

600
500
400
300
200
100
0
2003 2004 2005 2006 2007 2008 2009
NASDAQlisted NYSElisted Russell2000 S&P500
Source: Angel et al. (2010).
Conclusion # 8
In the post-MiFID era, several aspects have reduced the quality of data and hindered
consolidation. In order to improve this situation, the MiFID review should look both at the
standardisation of data formats (code identifiers, etc.) and data granularity through flags.
The relevance of trade flags comes from the support they offer to liquidity discovery
mechanisms across trading venues. Market initiatives should reduce the number of trade
flags, currently around 50, to fewer than 10 across Europe. On the proliferation of data
formats, sources and vendors are working to reduce their number and bridge
inconsistencies. Achieving minimum consistency though industry cooperation would
probably be more efficient than strong regulatory intervention. In this regard, ESMA
should rather support current industry-led initiatives on standardisation. However, either
the Commission or ESMA should be able to impose consistency if commercial initiatives do
not lead to a satisfactory solution in a reasonable time frame.
Consolidated data solutions promote best execution and help to mitigate the
potentially negative effects of liquidity fragmentation. More accessible consolidated data
solutions should be delivered, in particular to retail and small professional investors.

Formats would need to be standardised and granularity increased, which would curb costs
for users and increase accessibility. Already some market participants have committed to
reduce costs for final users by unbundling fees for pre- and post-trade data; a step in the
right direction that data vendors and distributors should follow. When lower data collection
costs are realised, they should be passed on to final users. Regulators and competition
authorities should draw attention to potentially unfair market practices and
anticompetitive market conditions that impede markets from offering data solutions at a
reasonable cost, rather than attempting to define when a cost is actually reasonable.
The US experience with the unintended consequences of a consolidated tape run by a
public entity should suggest alternative solutions. Consolidated tapes can be designed and
offered by competing data operators (so-called Approved Public Arrangements or APAs),
once the rules of the game have been clearly defined and duly enforced. These tapes could
cover not only shares but also other financial instruments admitted to trading on RMs,
MTFs or organised trading facilities, as long as a sound regime has been put in place.
Regulators should set the conditions to facilitate the consolidation and timely
delivery of data to investors in real time and fully unbundled through APAs. Operators
would have to meet strict requirements and be responsible for the detection of multiple
publications (misreporting or double-counting). Only if the industry does not deliver with
these conditions, should ESMA adopt the necessary arrangements to set a single
consolidated tape in the EU.
4.4.3 Bond markets

Public trade disclosure of financial instruments other than shares has been
typically left to the market, in particular given the different market
structures that shape trading mechanisms for corporate and sovereign
bonds. In recent years, certain initiatives have been taken to offer public
access to partial pre-trade and post-trade data, such as non-executable
prices available through AFME110 and Xtraker.111 Despite initial scepticism
about post-trade transparency for bonds (CESR, 2008; EU COM, 2008),
problems posed by liquidity during the financial crisis for instance the
evaluation of illiquid products have led regulators to claim a post-trade
transparency regime, even though no other significant market failures have
been detected (CESR, 2009b). Moreover, the growing interest of retail
investors in these markets (IOSCO, 2004) has drawn the regulators
attention to allowing their needs to coexist with those of wholesale and
110 See: www.investinginbondseurope.org

111 See: http://BondMarketPrices.com
institutional investors.
Bond markets (corporate, sovereign and other public bonds)112 are
typically quote-driven markets or OTC bilateral negotiations, in which
intermediation plays a critical role. On the buy-side, the demand for these
financial instruments is overwhelmingly institutional, with very limited
retail participation (CESR, 2008). The vast majority of bonds are listed on
regulated markets, and, if traded there, they are typically pre- and post-
trade transparent, like shares. However, this situation refers to a very
limited number of securities; only those where there is enough liquidity to
allow trading on a regulated market. Typically, opacity prevails as
intermediaries are rewarded by their informational advantage in less-liquid
products. Portes et al. (2006) listed the characteristics of bond markets,
which explains their typical low liquidity. In effect, bonds (Biais & Declerk,
2007):
1) Attract long-term investors (hold and buy strategies),
2) Are difficult to short sell (at least in the same market) and so to
manage inventory risks,
3) Have few differences between each other (predictable returns) and
4) Are less concentrated than stocks (each issuer have several bond
issuances outstanding).
In addition, transaction costs for bonds usually decrease with high
ratings, short maturity and size (Edwards et al., 2007). Investing in short
maturity and bigger sizes may only provide incentives for a part of the
demand, which may ultimately affect liquidity in the market.
These characteristics have boosted institutional demand, which has
reciprocally influenced market microstructure. For instance, bond markets
in the US were typically order-driven markets with strong retail
participation in the 1920s. In a few years without relevant changes in the
role of bonds in financing the economy the weight of institutions reached
a turning point in which small trades were led towards a market
equilibrium designed for large investors (Biais & Green, 2007). Where
dealers commit capital on illiquid products, they will only do this in
exchange of less disclosure, since this is the equilibrium that rewards their
112This report considers sovereign bonds as part of the wider public bonds category, which
has been defined by CESR (2010f), p. 12. In particular, it includes bonds issued by
governments, governmental authorities and national/international organisations financed
by governments.
informational advantage (Madhavan, 1995). The use of big amounts of

capital to maintain and develop trading activities implies high fixed costs
(especially, opportunity costs), which requires a proper system of
incentives for the firms committing those resources.
Proposals to increase transparency in dealer markets usually generate
conflicting views in terms of their benefits and costs. Legislative proposals
will need to strike the right balance of requirements concerning public
disclosure. On the one hand, transparency is a public good and its optimal
level may not emerge spontaneously in the market. On the other hand,
opacity is not a market failure in itself and, to a certain extent, should be
well calibrated, especially when it comes to dealer markets.
A transparent market setting may improve price formation and
investor protection by reducing information asymmetries and thus adverse
selection and moral hazard. A more transparent (post-trade) market setting
can promote:
i) A more efficient price formation process by reducing search costs for
end investors (Bloomfield & OHara, 1999; Edwards et al., 2007);
ii) Greater competition among dealers and market-makers on risk-
sharing (Naik et al., 1999) and
iii) A fairer environment as long as information is publicly available to
all to increase their knowledge about the market (level playing field;
Casey, 2006).
Box 5. Investing in bonds

Bonds are usually associated with the idea of a safe investment since they are
meant to provide stable payoffs (fixed interest/income) until maturity. In effect,
bonds are products that can be considered safer than more aggressive
instruments (such as equity and some derivatives), in particular when investors
buy them at issuance and hold the product until maturity. This strategy is mostly
followed through plain vanilla bonds, which typically give no options for the
issuer and can be designed as straight (zero coupon, bullet; interest is paid at
maturity) or coupon (interest is paid in each specified period), linked or not to the
inflation rate. In this case, risks would be limited to:
Default or restructuring risks (the probability that the institution will default
or restructure its own debt);
Inflation rate risks (the probability that the inflation rate is higher at
maturity than at the time of purchase, which reduces the cash flow value);
Exchange rate risk (when payments are made in another currency, the
probability that the exchange rate makes the value of the cash flow lower in
comparison to the original investment); and
Exogenous risks (the probability that external unforeseeable events would
delay payments of interests, principal, or lead to a debt restructuring).
However, where investors dispose of a bond before maturity the level of risk
dramatically increases. In this scenario, the final outcome of the transaction
(specifically, the bond price on the secondary market) can be affected by these
additional factors (Fabozzi, 2007):
Interest rate risk (the probability that the bondholder will bear opportunity
costs and expressed in the decline of the bond price when the official
market interest rate increases);
Floating rate risk (the probability that the interest rate paid to the
bondholder would be lower than the prevailing fixed interest rate paid for
bullet bonds);
Yield curve risk (the probability of a shift in the long-term yield curve due to
changes of the general market conditions);
Reinvestment risk (for bonds that periodically pay interest and principal,
which may need to be reinvested at least with the same return);
Downgrade risk (as mentioned above, the probability that the rating on the
issuers credit risk worsens, making the bond price decline);
Credit spread risk (the probability that the premium over a default-free
benchmark typically 10y US Treasury or 10y German Bund can increase
due to movements in related markets, as credit default swaps);
Liquidity risk (the probability that the market does not have enough
liquidity to price bonds at their theoretical level);
Volatility risk (the probability that the yield would follow an unexpected
and highly volatile pattern);
Call and prepayment risk (where the bond in exchange for a higher return
includes a provision that allows the issuer to call back the bond before the
maturity date by repurchasing it at market price or an pre-established one);
Sovereign risk (where the issuer is a government or government-funded
institution, the probability that the government exercises the sovereign
powers to unilaterally decide to default or restructure its debt).
As a result, investing in bonds may involve a high level of complexity,
which ultimately requires a carefully designed transparency regime. Ill-defined
transparency requirements may have unintended consequences for market
liquidity and investors participation in the bond market.
A more competitive environment due to more willingness to take

on risk, knowing the position of other dealers (risk-sharing) may reduce

informational advantages and ultimately lower transaction costs and
bid/ask spreads (Goldstein et al., 2007). Lower search costs may stimulate
efficient pricing, greater use of electronic multi-dealer platforms to diffuse
quotes, and more interconnection between current venues. This may push
inter-dealer platforms to expand their reach towards small professional
investors or small intermediaries that struggle to receive best execution
with the current regulatory gap, including retail investors (Ferrarini, 2009).
Moreover, bond prices are often used to calculate default probabilities and
as a benchmark for the valuation of illiquid bonds or other illiquid
instruments/assets (e.g. pricing matrix). Efficient pricing mechanisms
engender positive liquidity externalities (Amihud et al., 1997;
Bessembinder et al., 2006; Cici et al., 2008), which spread benefits to the
whole market. Finally, aggregate disclosure of bond holdings may increase
systemic liquidity113 (Lagan et al., 2006), reducing uncertainty around
exposure in times of uncertainty.
In terms of costs, in a multi-stage transaction process, Naik et al.
(1999) have showed that big inventory positions may suffer price revision
risk, despite greater transparency reducing information asymmetries
between dealers and hence improving inventory risk-sharing. In effect, in
following periods, prices will reflect private information fully, minimising
the informational advantage and reducing the capital committed to a sub-
optimal level. Therefore, the commitment of dealers may be affected in the
first place. In effect, it would become more costly to manage inventories
given the potential opportunistic behaviours of a typical prisoners
dilemma.114 Opportunistic behaviours may be also pursued through
parallel exploitative actions in linked markets (such as swap or CDS).
Excessive transparency can also damage liquidity, as informed traders will
act knowing their competitors positions and vice versa. Strict transparency
113Systemic liquidity is the market liquidity in stressed times, when liquidity is typically
driven by the homogeneous herd behaviours of investors (e.g. panic or euphoria). It is
specifically related to market participants behaviour. It is the opposite of search liquidity,
which is the liquidity in normal times when liquidity is typically driven by search costs for
end investors. See, Lagan et al. (2006).
114The prisoners dilemma is a situation in which parties behave opportunistically in order
to maximise their outcome. By acting uncooperatively, parties will end up in a worse
position than if they had coordinated their actions.
requirements may give them enough information to be able to enforce the

artificial price of the oligopolistic market equilibrium (Kovtunenko, 2003).
Disproportionate information about securities payoffs may reduce
informational advantages for informed investors with respect to dealers,
thus pushing part of this liquidity out of the market.
Trade disclosure. Finally, greater trade disclosure may also have an
unclear impact on systemic liquidity by increasing homogeneity of
valuation and risk management tools on the one hand, which may be put
under critical pressure when market conditions in the linked market
worsen (cascade effects), and, on the other hand improving asset valuation
tools.
Due to the reasons mentioned above, speed, breadth (granularity)
and depth of information should be designed around dynamic liquidity
measures, which may apply to some related asset classes (such as liquid
structured products and CDSs). Post-trade information that is useful for
price discovery purposes115 is generally information about the
characteristics of the bond (price, volume and identifiers), frequency of
trading, credit ratings at the issue, and whether bonds are listed and/or
traded off-exchange (flags). Data should be consolidated across trading
platforms wherever possible (IOSCO, 2004). To preserve systemic liquidity
and to monitor systemic risk, some aggregate information about notional
value and current holdings should be provided to regulators, even on a
confidential basis.
Transparency requirements. The Commission confirmed that, while
transparency requirements will be differentiated by asset class, the new
regime will cover all bonds with a prospectus or admitted to trading on a
RM or MTF (even if traded OTC; EU COM, 2010b, p. 27). The extension
does not cover bonds that may be traded only OTC, as a result of a private
placement. CESR (2010f), proposed the introduction of a post-trade
transparency regime for corporate and public bonds, which are defined as
transferable debt securities [...] with a maturity of at least 12 months.
However, both CESR and the Commission only took into account the size
of transactions as a valid liquidity measure to set thresholds and delays for
post-trade transparency. Liquidity measures typically encompass a wide
array of measures (not only size) that may need to be considered
115The report does not consider information requested for transaction reporting purposes at
this stage, which is covered by current MiFID text and in the earlier section of the report.
altogether. CESR did not consider the frequency of trades to set the
threshold but the possibility to delay publication by 15 minutes in
exceptional circumstances can eventually help where products are
infrequently traded. To give an example, on Xtrakter,116 almost 75% of the
43,000 corporate bond issuances are traded mainly wholesale and less than
10 times a month (orders can be up to 15% of all the bond issuance). In
October 2010, the highest average daily number of trades for a single bond
was 129 times. Moreover, while 80% of government bonds are traded
electronically (50-60% of volumes), only roughly 30% of corporate bonds is
traded on electronic platforms. This setting may drastically affect the level
of data that can be obtained in the 15 minutes buffer defined by CESR and
the Commission.
However, bonds infrequently traded may need some protection from
full or partial post-trade transparency, especially when traded in big
wholesale sizes. In effect, full disclosure of these trades may stimulate
actions against the holder in other related markets (as CDS market), in
order to exploit the private information. Nevertheless, this issue may be
tackled with a specific scheme of delays (as done for equities). This scheme
could also take into account the frequency of trades. Delays calibrated by
the initial issuance, the frequency of transactions and their size may be
indispensable for these markets to function well. It would also play a key
role in winding up and unwinding big inventory positions.
Finally, it is fundamental to make sure that standards of publication
(e.g. identifiers, condition codes, transaction definition, etc) are clearly
defined and implemented across trading venues. This would eventually
stimulate data consolidation, which may be difficult to implement where
publicly available quotes are not executable. In any case, even partial
coverage would increase the attractiveness of these markets and,
116As mentioned above, Xtrakter publishes some information about bonds, available on
www.bondmarketprices.com. Xtrakter runs an approved reporting mechanism to the FSA in
London, which has processed 578 million transactions in 2009. It offers bid/offer quotations
and transaction reporting (through TRAX) for equities, fixed income securities (a vast
majority) and OTC derivatives. Xtrakter produces bond liquidity data on a monthly basis
that includes 28,000 to 35,000 securities per month; its database contains nearly 400,000
documents. Published data is only on fixed income securities. Overall, Xtrakter already
captures at least one leg of 60-70% of all fixed income market (at least one leg transacted in
London).
potentially, overall liquidity.
Box 6. The TRACE experience

With the introduction of a Trade Reporting And Compliance Engine (TRACE), on
1st July 2002, the US officially introduced a trade reporting system for corporate
bonds, a big-bang for US corporate bond markets which covers agency debt from
March 2010 and asset-backed securities since 2011. This system reports post-trade
information within 15 minutes of transactions below $5 million at par value. It
consolidates the time of execution, price, yield and volume for 100% of OTC
transactions in over 43,000 securities, representing 99% of the corporate bond and
agency debt market activity.
Despite concerns of the industry due to the loose protection of inventory
positions and large trades, regulators have implemented this solution in a
relatively short timeframe for the entire market. On the one side, it is extremely
difficult to measure any unintended long-term effects of TRACE on liquidity and
market innovation. In particular, it is unclear whether lower costs are not offset at
the expense of promoting sub-optimal market equilibrium in the long term, by
impairing dealers capital commitment. However, some academic studies have
provided empirical evidence on the effects of TRACE. The system has:
Reduced trade execution costs (Bessembinder et al., 2006; Edwards et al.,
2007),
Reduced spreads on more liquid bonds (Goldstein et al., 2007) and
Improved corporate bond valuation (Cici et al., 2008).
These studies suggest that the impact of TRACE has been very positive on
trading costs and corporate bond valuation, while less so on liquidity over the long
term.
Finally, the nature of the US system where dealers operate rather as
broker-agents, with low capital commitment makes the implementation of a
regime such as TRACE easier than in Europe. In effect, in Europe dealers typically
act as principals to the transaction, competing with other dealers to buy bonds that
may sit on their balance sheets afterwards. Bonds are then sold to investors by
offering executable prices on a platform or bilaterally. In this respect, in order to
keep liquidity in the market, the introduction of a transparency regime should be
gradual, to preserve large trades and relevant inventory positions in the market.
Moreover, European and US bond markets also have other differences (Biais &
Declerk, 2007):
European markets are more competitive, since national banks compete with
dealers to offer bonds to wholesale and retail investors.
Frequency of trades is higher in Europe than the US.
Effective spreads post-TRACE are not lower than European ones.
This situation does not imply that TRACE has not generated positive effects in the
US and it does not preclude a similar regime in Europe from having a positive
impact on European markets as well. A more competitive market setting plays a

crucial role in reducing or increasing these differences.
Conclusion # 9
A transparency regime for bond markets should provide meaningful information to
stimulate price discovery. The speed, breadth, and depth of information should be designed
around dynamic liquidity measures. Since there is not a single measure of liquidity
readily available, transparency requirements should be developed on an instrument-by-
instrument basis. This task should rather be left to secondary legislation, such as Level 2
implementing measures or binding technical standards.
Liquidity is a dynamic aspect, which may take different forms according to the
characteristics of the financial instrument and the trading mechanism. This reality should
be taken into account where allowing exemptions or deferred publication in order to
preserve an efficient and sound price formation process. Dynamic measures of liquidity can
be designed around aspects such as frequency of trades, overall turnover or prospective
liquidity, product standardisation, or transaction size. Finally, since data is fragmented,
data formats and flags may need to be further standardised for the purpose of pre-trade
transparency.
Promoting structural market changes in non-equity markets to give easier access to
retail investors may raise conflicting issues. For some market participants, the market for
fixed income securities should remain wholesale dealer-driven. In their view even though
a commendable objective direct retail access to non-equity instruments may destabilise
dealer-driven markets, as it may generate higher volatility with no liquidity enhancements.
These effects would ultimately heighten risks for retail investors, given the
increasing complexity of fixed income securities. Other stakeholders, however, firmly
support the opening of bond markets to retail investors. Greater transparency may be a
liquidity driver for these markets. Under proper delays and exemptions, the impact of retail
trading activities would be fairly limited.
4.4.4 Structured financial products

Structured financial products (SFPs) are securities whose cash flow
payments come from a pool of assets. Securities are typically issued by an
independent legal entity, so-called special purpose vehicle (SPV). Credit
enhancements, which are different levels of credit seniority, give them the
possibility to reach a wide range of investors (retail and institutional). SFPs
can be divided into three major categories (CESR, 2008):
i) Asset-backed securities (ABSs),
ii) Collateralised debt obligations (CDOs) and
iii) Asset-backed commercial paper (ABCPs).
ABSs are securities issued and financed by the sale of specific assets.
In effect, there are several sub-categories, named after the underlying asset.
Some of them are: residential mortgage backed securities (RMBSs);

commercial mortgage backed securities (CMBSs); and other securities
backed by other assets/receivables (as loans, auto loans, etc.).
CDOs are securities typically collateralised with a portfolio of other
fixed income securities. The income generated by these securities actually
feeds the cash flows generated by CDOs over time.
ABCPs are commercial notes collateralised by other assets that have a
short maturity (180 days maximum).
SFPs often are very illiquid, in particular during times of financial
distress. In effect, the size of the SFPs market was drastically reduced as a
result of the crisis. Banks heavily involved in securitisation and SFPs
issuances suffered important losses and, in extreme cases, defaulted. The
crisis highlighted relevant market failures, such as distorted system of
incentives and lack of reliable information in the market. While the demand
for these products gradually recovers, 90% of new issuances are still done
only for repo purposes (with no private placement). It is highly unlikely
that the market will revert to 2006-7 levels, at least in the short term. As
fixed income securities, SFPs have many similarities with bonds and in
theory are designed to reproduce a similar payoff structure. However, the
financial design of SFPs is often more complicated and is in general not
directly linked to the credit risk of a legal entity, but to a pool of assets
whose information may not be publicly available or not meaningful for
investment purposes. Limited information about SFPs can be accessed by a
limited number of investors. This information suffers from a lack of
completeness and dispersion in memoranda (a sort of prospectus), rating
agency reports, issuers presentations, and pool reports. The complexity of
these documents often makes it difficult to extract meaningful information
for investment purposes (IOSCO, 2009b).
CESR (2010f) and the Commission (2010b, p. 27) have finally opined
that the transparency regime for structured products should cover all asset-
backed securities (ABS) and collateralised debt obligations (CDOs) for
which a prospectus has been published (i.e. including all ABS and CDOs
admitted to trading on EEA RMs) or which are admitted to trading on an
MTF. Due to the perceived illiquidity of these markets, CESR suggested a
phased approach for a post-trade transparency regime. Any extension of a
post-transparency regime to non-equity markets needs to be properly
calibrated. The Commissions consultation report on MiFID also takes stock
of the CESR advice (European Commission, 2010b, p. 28), which will apply
the approach to other non-equity products too. A crucial role will be played
MIFID 2.0: CASTING NEW LIGHT
I ON EUROPES CAPITAL MARK
KETS | 95
b trade repossitories (see neext section).

by
4
4.4.5 Over-thhe-counter derivvatives
OTC Derivativves. Derivativees are financial instrumentss whose valuee (price
O
o the contracct) is derived
of d from the va alue of an un nderlying asseet (e.g.
e
equity, bond, or commoditty) or market variable (e.g.. interest rate,, credit
r
risk, exchange rate, or sto ock index) (V Valiante, 20100, p. 1). Alm most all
t
transactions (aaround 84%) are executed
11
17 d OTC, i.e. bilaaterally negotiiated or
t
through inter--dealer platfo
orms that disp play to their m members execcutable
q
quotes collecteed from a gro oup of dealerss with a limiteed role for reg gulated
m
markets and official
o alternaative trading platforms.
p 118 Th
hese instrumeents are
r
regularly useed for hedgin ng, funding, speculation and arbitrag ge. The
f
frequency of trades
t is very low. The figu ure below show ws the daily number
n
o trades per currency and maturity thatt are executed
of d in the OTC interest
i
r
rate swap (IR RS) market, wh hich represennts almost 75% % of the globaal OTC
d
derivatives maarkets (measured in notiona al value).119
F
Figure 14. Averrage daily numbber of IRS tradees by currency
1
117Updated figurre from BIS, by comparing notio onal amount of aaverage daily turnover of
OTC and exchange-traded derivatives;
O d seee http://www.b bis.org/publ/rpffx10.pdf?
n
noframes=1, and http://www.biss.org/statistics/ottcder/dt1920a.pd
df.
1
118For a more detailed
d discussio
on on the naturee of these produ
ucts and marketss, see in
general Valiante (2010).
g
1
119The notional value
v (or simply notional)
consistts of the face valu
ue of the OTC derivatives
ccontracts.
96 | A RENEWED TRANSPARENCY
R REG
GIME
Sourcce: TriOptima.
Figu
ure 15. Averagee daily number of
o new IRS trad
des by maturity
Sourcce: TriOptima.
Credit default swaps. Ano other importan nt class of OTTC derivatives is
creddit default swwaps (CDSs). These
T contractts allow a cou
unterparty to get
full protection ov ver an exposurre held by thee subscriber w with a third paarty.
The protection seller
s periodiccally receives a premium in exchange for
paying the diffeerence betweeen the face value v of the obligation (e.g.
(
secuurities) covered d by the CDS contract and thet current maarket value, wh hen
a crredit event occcurs. Alternattively, the insu
urer can pay thhe face value and
a
takee the obligatio on in exchang ge. The CDS is triggered when a speccific
creddit event occurs. In addition to a defau ult, a CDS con ntract can be also
a
trigggered by other events, such as a debt restructuring g or moratoriium
(ISD
DA, 2003).
Credit deriv
vatives have played
p a crucia
al role by provviding a sourcee of
priccing for some corporate bon nds. They also o offer protectiion for inventtory
posiitions in bond d markets sincce CDS markeets allow goin ng long on bon nds
withh no need to provide high h cash paymen nts as collaterral. Furthermoore,
bonnds covered by CDS con ntracts are regularly
r acccepted for reepo
tran
nsactions, whicch give a stro ong incentive to deal with iilliquid produ ucts.
Receent evidences (Shim & Zhu,, 2010) supporrt this statemeent and show that t
CDS S trading actu ually has boossted liquidity in Asian bon nd markets. This
T
self--reinforcing reelationship beetween CDS and bond markets has fosteered
creddit derivatives markets in n the last deecade. Despitee quick grow wth,
howwever, CDS sin ngle-name con ntracts are still very illiquid, with over 99%
% of
trades occurring fewer than 10 times a day from December 2009 to June
2010 (DTCC, CDS Data Warehouse). CDS index contracts, however, have a
more liquid market but only 20% of trades are actually traded more than 10
times a day and 10% more than 50 times a day from March 2010 to
September 2010 according to DTCC CDS Data Warehouse. CDS indexed
contracts are also highly standardised (ISDA, 2003) and electronically
traded. Aggregated data are periodically published by DTCC Deriv/Serv,
although no data are provided on net exposures of systemically important
financial institutions.
Trade disclosure. A post-trade transparency regime for SFPs and OTC
derivatives may be designed around requirements similar to the ones
adopted for bonds. However, since transactions are mainly OTC bilateral
negotiations with very low frequency and high size, information may be
not only meaningless but eventually harmful, especially if the financial
instrument acts as a benchmark for evaluating similar products. Therefore,
exemptions and delays for these financial instruments should be on
average wider than for bonds and primarily based on the size and
frequency of trades. Nevertheless, SFPs and OTC derivatives typically lack
price transparency, given the use of proprietary valuation models to price
complex products (CESR, 2009b). In this case, aggregate price indexes
should be publicly available in real time, whenever possible. In effect, given
the low number of transactions in these products, a continuous flow of
information may not necessarily be available to support the pricing of these
products by competing binding indication of interests from end investors
(as OLOBs). Therefore, proprietary pricing models become indispensable
tools.
Aggregate data. Securitisation and OTC derivatives may be formidable
tools to free capital back to the real economy and promote better allocation
of risk and resources. They contribute to transfer credit risk and spread it in
the markets to those who are more able to bear it. However, the experience
of the financial crisis has taught us that spreading risk among many
counterparties through complex financial instruments does not ultimately
cancel it. Information about the underlying assets and net exposures should
always be publicly available in order to monitor systemic risk. Competent
regulators should have access to data via reporting, especially through trade
repositories. During stress time, aggregated information on net exposures
could help contain herd behaviours set off by market opacity. It is thought
that this form of transparency can improve systemic liquidity (Lagana et
al., 2006). Investment firms would have to provide this information, which
would subsequently be aggregated by trade repositories (TRs), with
sufficient skills and capabilities to collect and aggregate this information
(European Commission, 2010b, p. 13). However, aggregated information on
net exposures might not be readily available; hence financial institutions
could explore the possibility to modify their data management system in
order to collect information on net exposures in a way that does not
compromise the confidentiality of sensitive information. Trade repositories
would then be able to aggregate this information and offer a global picture.
Trade repositories would therefore play a double function: a) disclosing
aggregate data on net exposures (trade reporting); and b) providing specific
data on transactions to regulators on a confidential basis (transaction
reporting). The transparency regime should cover all derivatives centrally
cleared and those reported to trade repositories (Art. 6.1, EMIR).
Conclusion # 10
A post-trade transparency regime for derivatives and structured products should be more
detailed and tailored to the nature of these products. Where listed on RMs and/or MTFs,
the regime could be designed with the same methodology employed for bonds, but its
implementation should follow a phased approach.
Exemptions and due calibrations should be allowed in order to preserve efficient
price formation and guarantee the effective monitoring of systemic risk. Calibrations should
take into account the nature of these markets and of each financial instrument, rather than
a division into broader categories (e.g. by asset classes). A mere application of post-trade
transparency to a general list of instruments would definitely hamper market liquidity. As
for bond markets, measurements of liquidity should be taken into account with due care to
avoid adverse consequences in terms of liquidity for wholesale participants.
The extension of trade and transaction reporting to non-equity markets can be
facilitated by current infrastructures, thereby reducing costs.
Finally, harmonising the scope of eligibility for clearing purposes with the legal
duty to be pre- and post-trade transparent would leave big gaps in the actual
implementation of the transparency regime. In addition, it may also distort trading
activities and redirect them towards non-centrally cleared products, since on top of the
costs of centralised clearing firms would bear the risk coming from pre- and post-trade
transparency with no assessment of size and frequency of transactions. Extension of post-
trade and transaction reporting to non-equity markets can be facilitated by current
infrastructures (reducing costs) but it is critical that transparency requirements remain
independent from trading and clearing eligibility requirements. In effect, the requirements
to access a central counterparty clearing (CCP) are not only based on liquidity itself but
also on the standardisation of technical and legal aspects, and therefore the mere eligibility
for clearing of an instrument should not be considered to be a proper test of liquidity and
frequency of trading for transparency purposes.
For instance, as shown above, CDS contracts are quite standardised and over 75%
of trades are confirmed on the same day. Dealers expect as much as 90% of these contracts
to be cleared on CCPs.120 While expectation about eligibility to central clearing are high,
this choice does not necessarily consider market liquidity, which remains still very low with
less than 20% of CDS contracts traded more than 10 times a day and less than once for the
vast majority of them. Potential delays and issues affecting the process of centralisation of
clearing on CCPs may ultimately affect the implementation of disclosure requirements in
these markets.
120 See http://www.finextra.com/news/fullstory.aspx?newsitemid=20458.

5. RESHAPING MARKET STRUCTURE
5.1 Introduction
The structure of financial markets changed dramatically over the last
decade. New trading technologies and growing volumes have generated
important network externalities, which have ultimately reshaped the
original market design based on natural monopolies. These developments,
together with innovative policy decisions, have also favoured changes in
three areas: competition and market fragmentation micro-structure and
infrastructure. In this framework, MiFID has managed to provide a
comprehensive regulatory framework that boosted the aforementioned
developments and minimised potential unintended consequences for
market quality and investor protection. This report does not address the
issue of what market structure works best. For this question, there is, in
principle, no final answer. The best market structure is the one in which a
market-clearing price can always be found (OHara, 1995, p. 269).
5.2 Securities markets as network industry

Definition. A network is a market setting in which the output is the result of
the interaction of several inputs (so-called nodes) that generate
externalities when combined in a certain way. Three aspects characterise a
network industry (Economides, 1993): complementarity, compatibility and
coordination.
The output of a network infrastructure is typically a composite good,
which mixes and matches inputs that are complementary.121 For instance,
a phone call is the result of a network that links the caller with the receiver.
121A and B are two complementary products when an increase in the price of A also reduces
the quantity of B.
| 100
The link between the caller and the adapter/switch is complementary to

the link between the adapter or switch and the receiver, which at the end
compose a phone call. For the formation of the composite output of the
network, the relationship among the nodes, their complementarity and
their compatibility, is essential. In this sense, complementarity only exists if
the inputs composing the final good are compatible, i.e. both follow the
same standards and appear coordinated in time and space.
Network externalities. The network output can generate two kinds of
externalities: i) direct and ii) indirect. An externality is direct, when any
additional customer/node increases the size of the whole network by 2*n
potential new goods. The overall number of complementary goods is n*(n-
1). This externality is also called production externality and is generated by
the economies of scope produced by an additional node. An externality is
indirect, when an extra customer does not necessarily represent 2*n
potential goods, but only increases the size of the network and the
possibility of finding a good matching. It is also called consumption
externality or size externality, i.e. the value of a unit of the good increases
with the expected number of units sold (Economides, 1996). This kind of
externality relies upon the presence of economies of scale and represents
the inverse of the classic law of the demand since the value of the good
increases with the number of units sold (or produced).
Network settings. Networks can have two basic market settings or a
combined one: one-way, two-way and mixed.
A one-way network (e.g. TV broadcasting) is a network where the
output is function of n consumers and m developers/suppliers of products.
Producers develop the product and make it available through the grid.
This generates indirect network externalities: more users will increase the
number of developers and vice-versa. However, there is no strict
complementarity between consumer and developer. One extra customer
does not necessarily add 2*n potential complementary goods, but it will
indirectly increase the overall value of the network since it will be more
likely that an exchange will take place at its best value (consumption or size
externality).
A two-way network is a network where n consumers interact with
n1 other consumers over the switch, that is the product that allow
consumers to access different suppliers. This network setting generates
direct network externalities: one additional consumer equals n new
combined products or 2*n complementary nodes (due to its different
102 | RESHAPING MARKET STRUCTURE
nature). As mentioned above, a market for local or international phone calls

is a typical two-way network setting. The externality effect of having one
more customer directly affects the utility function of the other
consumers/investors (direct or production externality).
Finally, the two basic settings may interact according to the level of
complementarity between groups of users/platforms/goods. Interactions
allow for combinations of goods or consumers/investors whose
willingness to trade generate both one-way and two-way network
externalities. This network setting will occur more frequently in multi-
sided markets, whereby the interaction of types of users/goods/platforms
takes place.
5.2.1 Markets Multi-sidedness and Securities Markets

Multi-sided markets. Networks can be classified according to their pricing
structure, i.e. the possibility that platforms can charge different prices to
different groups of users. The market would be considered as two-sided
or more generally multi-sided when the pricing structure of respectively
one or multiple platforms is based on:
i) The interaction of groups of users-only (Rochet & Tirole, 2003) in
which n receivers interact with m senders through a platform; often
the platform subsidizes the entrance of n receivers on the oneside
with the interest of m to get in contact with n on the other side;122 and
ii) The interaction of both groups of users and platforms (Armstrong,
2006) n users can interact with m other users; the value of the network
is the possibility that you can interact through different platforms, so-
called multi-homing123).
122The videogame industry (e.g. Microsoft Xbox 360, Sony Playstation, etc.) epitomises a
good example of a two-sided market, in which the value of the network will increase by
stimulating the interaction between users of the platform (gamers) and developers of game
software. The platform will therefore charge a more modest fee to developers (typically free
access or sometimes even paying them) in order to attract more users and generate size
externalities. The balance between the two fees (price distortions) will change as the
platform gains more users and increases its value or from the potential intensity of network
externalities that attracting a group of user may generate.
123B2B exchanges (e.g. Alibaba.com, etc.) are platforms that link business buyers and sellers.
Sellers are charged for posting their products, while buyers typically access those platforms
for free. The possibility to offer the content in multi-homing i.e. on multiple platforms
increases the value of the overall network as it increases the potential reach of sellers
proposals, whether or not both sides of users are both multi-home. Users wish to reap the
Both interactions may generate direct or indirect externalities as

characteristic of one-way or two-way networks.
Securities markets can be considered as networks, whose output is the
exchange of the financial instrument at the best available terms (a
combination of complementary goods, i.e. investors bid and offer). The
structure of markets for financial instruments is a combination of different
network settings that allow the interaction of liquidity demanders and
suppliers through buying and selling securities. Through this interaction of
bids and offers, resources will have more chances to find their best
allocation in the economy. In effect, transaction costs would be prohibitive
in a world with no exchange platforms; investors or issuers would have to
bear the full cost of finding the counterparty who best values their
securities, which would ultimately lead to resources being inefficiently
allocated. Markets can reduce transaction costs (by facilitating transactions
on agreeable terms) and allow more efficient allocation of resources, but
not all market designs reduce costs to the same extent (Coase, 1960;
Demsetz, 1968).
Primary markets. Securities markets (in particular for shares) can be
divided into primary and secondary markets.124 Although they both match
liquidity demand and supply, their network settings may be different since
they might be based on the interaction of groups of users, on multiple
platforms or both. In primary markets, securities are sold by issuers for the
first time, as a combination of issuers offer and a set of buyers bids. These
markets may be extremely volatile before the security is sold at the market-
clearing price. Volatility occurs at the time where the security is proposed
to potential buyers at a theoretical value, which will change according to
their willingness to buy (the number of offers).
benefits of network externalities in an environment of non-interconnected platforms (Rochet

& Tirole, 2006).
124 The report discusses the interaction of investors interests in trading platforms. Little
attention is actually given to the network settings of the post-trade industry, which typically
take a different shape from the one of trading platforms.
Figure 16. Primary securities markets
Source: Authors.
As shown in the figure above, primary markets gather investors and

issuers in the same place. It is a single-home market, since issuers list their
shares in the market of their choice, which typically coincides with the
main national regulated market for both legal (official listing) and
economic reasons (most liquid primary markets). In this case, the national
regulated market will be the only place where buyers can purchase those
shares when the issue begins. Cross-border listing is possible but generally
occurs at different terms. The pricing structure of primary markets is two-
sided; issuers are willing to pay for accessing the platform (listing fees),
while members of the market get access to those executable quotes at more
favourable terms (or within the membership access fee they actually pays)
when the security is issued. The platform subsidizes the interaction of n
investors with the interest of m issuers to offer them their securities (Rochet
& Tirole, 2003). For listing of SMEs which typically represent a less
mature market with lower liquidity (so size externalities) issuers may get
more favourable terms to access the trading platform to place their
instruments.
Figure 17. Secondary securities markets

Investorsoffers
Membership
/AccessFee Intermediaries
DMA/SA
Trading Trading
PlatformB PlatformE Trading
Trading PlatformH
PlatformA
Execution
Fee Trading Trading
PlatformC PlatformG
Trading Trading Trading
PlatformD PlatformF PlatformI
DMA/SA
Membership Intermediaries
/AccessFee
Investorsbids
Source: Authors own elaboration.
Secondary markets allow the interaction between investors bids and

offers (the complementary goods) in order to achieve the exchange of a
financial instrument (liquidity). Investors can access the platform through a
switch (intermediaries, e.g. brokers or dealers) or directly through Direct
Market Access (DMA) if the firm has enough expertise, but under the
supervision of an intermediary.
Two pricing structures are usually charged in this network industry
setting. Firstly, both bids and offers pay a fee for every executed order,
which is generally the same for bids and offers. Commissions typically vary
according to the number and size of orders, thereby promoting more
volumes and revenues, to increase the value of the entire network
(Economides, 1993) and reach the critical mass through size externalities
(Pagano, 1989). The nature and pricing structure of the network at this
stage is thus one-sided with indirect size externalities (as a one-way
network). In addition, access fees are passed on to final investors by
intermediaries (spreads) if the investor does not directly access the market.
Two-sided markets. Secondly, trading platforms generally charge a
membership fee to grant access to the venue (see Figure 17). For investors
who do not meet the admission criteria, fees are embedded in the
intermediaries executable bid/ask spread (e.g. retail investors). The same

occurs for investors who prefer to access markets through intermediaries
(brokers/dealers) because they deal with several trading venues and find it
inconvenient to pay multiple membership fees. In effect, there are several
categories of traders/investors, who may need to access multiple platforms
to execute their orders to get best execution (Harris, 2002). Markets with
automated continuous auction systems (called broker-dealer crossing
networks) have a pricing structure that is typically two-sided (non-
neutrality of access fees125). This pricing also exists in markets where
dealers commit own capital or provide crossing services with or without
fiduciary duties (or quote-driven dealer markets). Fees in these two
markets are designed to stimulate the interaction between uninformed and
informed investors, thus generating strong network externalities (of the
indirect type).
In effect, four conditions help us to define secondary securities
markets as two-sided:
1) Both groups of users (bids and offers) gain from accessing the
platform.
2) Too high transaction costs would not allow private bilateral
agreements on how to deal with externalities (Coase, 1960).
3) Users pay access fees.
4) Access fees are non-neutral.
Despite strong size externalities (economies of scale), securities
markets here two-sided markets have shown the ability to support
competing trading venues (multi-homing; Armstrong, 2006) where at least
one group of users needs to access multiple platforms, as a reflection of the
diverse needs of market participants beyond simple price factors.
Advanced technologies (smart order routing systems, etc.) have provided
investors with tools to exploit direct network externalities (economies of
scope) coming from the interaction of competing (trading) platforms, in
particular if complementary (Van Cayseele & Reynaerts, 2010).126 The value
125The non-neutrality of access fees is the primary condition of two-sided markets (Rochet
& Tirole, 2004). This means that the overall volume of transactions depends on the allocation
of fees between groups of users. The platform can influence the overall number of
transactions by modifying the membership for a group of users, keeping the other fees
constant. In effect, members acting in different capacities may get (in relative terms) a lower
fee (as they bring uninformed and low-risk investors to the market) than a firm requesting
membership in its own capacity only.
126 The role of economies of scale and scope in the post-trading infrastructure have been
of the network derives not only from its size externality and its ability to
reach a critical mass of liquidity as originally perceived by markets and
policy-makers. It also comes from the interaction of competing trading
venues with different market designs to meet multi-faceted investors
needs. In effect, today an investor can pursue different strategies (e.g. price
improvements and speed) by executing simultaneously two different
orders on multiple platforms.
Conclusion # 11
Todays markets benefit from the interaction of groups of users and platforms, as well as
competition/interaction between platforms (including dealers networks). As a vital part of
the network, competition strives if markets are contestable, not only from a technological
standpoint (sunk costs), but also in terms of fair market practices and an evenly applied
market regulation (dynamic view). This implies the need to abate barriers to entry and exit
and monitor adopted market practices. Competition authorities should play an increasingly
important role in this regard.
5.3 Competition and fragmentation

Concentration rules. The structure of financial markets has remarkably
evolved in the last decade. Most notably, equity markets have experienced
unprecedented changes that have redefined the interaction between
investors and trading platforms. The tension between reaching critical
mass (with its self-reinforcing network effects) and meeting investors
need for diversified execution services (as a result of their multi-form
discussed at length, in particular for settlement services. Some recent studies (Schmiedel,
2002; Schmiedel et al., 2006; Van Cayseele & Wuyts, 2006) suggest that there are potentially
strong economies of scale in the EU post-trading landscape, which may lead to further
alliances and mergers between incumbents. Divergent views arise around the policy
implications that these authors suggest. On the one hand, some authors (Schmiedel et al.,
2006) argue that to fully exploit the economies of scale of network infrastructures, mergers
and further integration in the post-trade industry should be promoted. On these theoretical
rationales, for instance, the ECB decided to go ahead with Target2Securities (T2S), which is a
project to create a pan-European platform for securities settlement accounts in central bank
money. On the other side, others (Van Cayseele & Wuyts, 2007; Van Cayseele & Reynaerts,
2010) noted that economies of scale are limited and are typically exhausted before the pan-
European scale is achieved. In addition, bridges and multi-homing services suggest that
economies of scope may be significant. Hence, they suggest that regulators should carry on
further investigation before unbundling services or creating monopolies in any part of the
value chain.
nature; Harris 2002) has generated a conflict between consolidation and

fragmentation. Prior to the introduction of MiFID, stock exchanges
benefited from the application of concentration rules, which forced
intermediaries and investors to carry out any transactions in listed shares
on their main national market. The aim was to preserve price formation
and market integrity by concentrating trading activities in one consolidated
limit order book, even though the primary economic objective and nature
of a trading network is the reduction of transaction costs to increase
volumes. Infrastructures for order execution (exchanges) have been widely
considered as a natural monopoly until the widespread diffusion of new
technologies and internet. This conception was based on two premises,
firstly, that amassing liquidity in one place would generate positive
externalities and, secondly, that liquidity would gravitate around the most
liquid market (Pagano, 1989). However, technology has transformed
securities markets into networks where users and platforms interact.
Interaction reduces transaction costs (marginal costs127) close to zero by
generating positive externalities, given economies of scale and scope. In
effect, the rationale of a natural monopoly lies in the basic condition that
only one firm can produce the output (here the exchange of financial
instruments) at the lowest marginal costs and at the minimum efficient
scale (Varian, 2006). Under this assumption, costs would be declining with
a growing number of users (size externalities, see Figure 18 A below) as
opposed to an ordinary cost structure (Figure 18 B). Regulation must set
prices equal to marginal cost (that is the price of a perfectly competitive
market). Otherwise, the monopolist would produce or generate a sub-
optimal volume of trading. However, defining the price of perfect
competition (price equal to marginal costs) notably is not a perfect science.
127 The marginal cost is the cost of producing one additional unit of the final output.
Figure 18. Cost structures with (A) and without (B) size externalities
A B
Price Price
MC
ATC
ATC
MC
Quantity Quantity
Note: MC = Marginal Costs) and ATC = Average Total Costs.

Source: Authors.
If at least one market participant could replicate the cost structure of

the monopolist (A), the market should no longer be considered a natural
monopoly (it would not be naturally designed for one market participant
since a monopolistic equilibrium would be sub-optimal in terms of prices
and volumes). As shown below, many studies in the last two decades
suggest that competition between trading venues is possible and somehow
beneficial, thereby discarding the presence of a natural monopoly in the
market for securities (for a review, see Lee, 2002, amongst others, and
Economides, 1996).
In effect, transaction costs (implicit and explicit) have plummeted
across European and US markets after the abolition of the concentration
rule in both jurisdictions, which has contributed to fragment the trading
landscape (Lannoo & Valiante, 2010). In Europe, the abolition de facto took
place in 2007-08 when member states fully transposed MiFID rules in their
national legal systems, while in the US Rule 390 had been already abolished
by the New York Stock Exchange (NYSE) in December 1999.
Contestability and fragmentation. The contestability128 of financial
markets has been increasing since the 1980s, when automation of trading
128A perfectly contestable market is a market in which barriers to entry and exit are low
enough to give newcomers enough incentives to enter the market and compete with the
incumbents (Baumol, 1982).
(and computers) has gradually become widespread, allowing the

dematerialisation of securities and new forms of trading for investors as
well as higher speed and capacity for trading venues. Technology made the
dematerialisation of securities possible, together with new forms of trading
for investors and higher speed and capacity for trading venues.
Automation of trading, in effect, has drastically reduced entry barriers in
terms of sunk (fixed) costs since technologies have become commoditised
and easy to customise (Domowitz & Steil, 1999). However, only with the
commercial diffusion of the internet in the mid-1990s, securities markets
became widely contestable, given the abating of geographical barriers.
Floor trading specialists have been replaced by sophisticated algorithms
executed from only a few locations around the globe, as they are more able
to deal with market impact (Domowitz, 1990). Nowadays, retail investors
can potentially access every market in the world from their home PC.
Hence, in the last decade new demand for securities has flowed into
financial markets and volumes have been constantly increasing across the
planet (see regional and global indicators in Annex I). The number of
uninformed investors has grown and become less rare than it used to be.
Overall, soaring volumes have promoted important changes in the
structure of the market, and as a result open limit order books (OLOBs)
have become a stable institution (Glosten, 1994).
Increased contestability means more possibilities for newcomers to
enter the market for order execution and matching services. Market
fragmentation to be distinguished from liquidity fragmentation129
became a recurring aspect across European and US markets after the
abolition of concentration rules. Incumbent exchanges have been
demutualised by becoming for-profit firms (for a review, Levin, 2003) and
put in competition with alternative trading platforms (so-called
Multilateral Trading Facilities130), internalisers and crossing networks run
by brokers/dealers (see next section).
Benefits and costs. Fragmentation has positive and negative
implications, due to two potential effects:
129Market fragmentation concerns the co-existence of several trading platforms in the same
marketplace. This does not necessarily mean that this market setting favours the
fragmentation of liquidity by reducing interconnection and thus hampering price discovery
processes and market quality. This report uses the terms fragmentation and market
fragmentation interchangeably.
130 Art. 4, MiFID.
1) Order creation
2) Order diversion.
The interaction between these forces determine the impact of co-
existing trading systems and the optimal degree of transparency, in
particular for a dealer market setting with low probabilities of trade
execution and experiencing the potential entry of a competing venue
(Degryse et al., 2009).
On the one hand, fragmentation fosters competition between trading
venues on access and execution fees (or bid/ask spreads), as well as
competition on execution services between brokers/dealers. Competing
trading venues also provide competing execution and matching services
(e.g. dark order books, crossing systems, internalisation, etc.). As long as
investors can access consolidated data (through mandatory disclosure) and
use technologies to split up orders (Madhavan, 1995), trading venues may
be able to offer diversified services to meet multi-faceted investors needs
and to produce strong economies of scope (interaction of many users and
liquidity pools and between pools). Those needs are very diverse (Harris,
2002) but can be perhaps summarised as:
Handling big size orders (hedging, inventory, etc.),
Investing in information (arbitrageurs, informed traders, etc.),
Closing a position as fast as technically possible (close to real-time)
and
Minimising costs.
For all these reasons, the market will most likely not coalesce in a
single pool of liquidity. The provision of new services and competitive fees
may then increase volumes even further (order creation). Several empirical
studies have shown so far that competition/fragmentation has produced
beneficial effects on market structure by reducing spreads, fees and
attracting new liquidity in some markets (Glosten, 1994; McInish & Wood,
1996; Battalio et al., 1997; Boehmer, 2005; Foucault & Menkveld, 2008; for a
review, see OHara & Ye, 2009; Riordan et al., 2009; Angel et al., 2010;
Gresse, 2010).
Cream-skimming. On the other hand, market fragmentation may have
some drawbacks for liquidity as it reduces beneficial liquidity externalities
(economies of scale). Orders will be diverted on other venues and not
consolidated in a single order book, with implications for the likelihood of
execution and potential unintended consequences on liquidity (Mendelson,
1987). Furthermore, orders will stop competing on the same order book
with potential cream-skimming effect. In effect, conflict between order

creation and diversion may push intermediaries and trading venues to
select part of the demand (typically the uninformed one, since less risky),
leaving remaining order flows to deal with informed investors. The cream-
skimming activities of intermediaries (outside the electronic order books)
and some venues can spread orders in the market with strong adverse
selection problems for markets that suffer from discrimination (in
particular, if it is a quote-driven dealer market; Hagerty & McDonald, 1996;
Easley et. al., 1996; Battalio, 1997; Gajewsky & Gresse, 2007). The impact of
these potential distorting effects will depend on the overall size of the
market, if the market is big enough the effects of cream-skimming activities
will be fairly limited for liquid shares, while effects on less-liquid markets
may need to be assessed case-by-case. Finally, fragmented markets have to
deal with fragmented surveillance systems and rules. Authorities need to
make sure that organisational and surveillance requirements are fully and
uniformly applied across all trading venues. Ongoing supervision should
then make sure that practices are sufficiently implemented.
Figure 19. Market fragmentation versus liquidity fragmentation
Market Fragmentation Liquidity Fragmentation

Costs
Benefits
(total surplus) C E
X
b B
No-return point
Surplus
Y
a
Technological (mid-90s/early 2000)
and regulatory breakthroughs
(Abolition Rule 390 - MiFID)
n (few separate platforms) n (several interconnected

platforms)
N (users and platforms)
Source: Authors.
Market vs liquidity fragmentation. The graph above explores the market

structure dynamics. B and C are respectively functions of benefits and costs
for all market participants (e.g. investors). Their value varies as a function
of the number of users and competing platforms. Two potential equilibria
may be found (Y and X).
The benefit function B is a concave function that represents the
evolution of benefits for all market participants associated with an increase
of users and platforms. The first equilibrium Y refers to the market setting
before new technologies and the abolition of concentration rules unleashed
platform competition. Before reaching Y and immediately afterwards, the
cost function is very steep given big economies of scale (due to a few
disconnected platforms that have reached the critical volumes in Y).
Similarly, benefits increase fast, as on the one side investors can access a
trading platform to match other interests and avoid expensive bilateral
negotiations and on the other side trading platforms can gradually
reach their profitability and benefit from a combination of economies of
scale and scope from Y onwards.
In effect, the development of advanced technologies (e.g. automated
trading and internet) have lowered sunk and access costs, thus the thesis
that national exchanges were natural monopolies with limited economies of
scale was gradually dismissed. This situation led regulators to abolish
concentration rules, which had protected that market equilibrium (Y).
The combination of regulatory and technological breakthroughs
(more recently in Europe than in the US) has promoted market
fragmentation and competition between platforms. As a result, costs for
setting up trading platforms (sunk costs) and trading costs for final
investors have further gone down. Beneficial effects due to a greater
number of users (economies of scale) and multiple platforms (economies of
scope) have increased benefits for final investors who can access better-
tailored execution services and for trading platforms which benefit from
a potentially higher range of investors and services to compete on. This
situation has created a highly competitive landscape around costs of
trading and offered diversified bundled execution services (e.g. direct
electronic access services)131 in order to gain a bigger part of the surplus.
131Competition between replicable bundles can be fierce and somehow beneficial.

Conflicting views arise when a bundle is deemed to create non-replicable portfolio effects,
As a result of these competitive forces among trading platforms and

investors on the order flow (informed versus uninformed), market
fragmentation/competition may generate costs, such as market impact for
larger size orders and costs for routing orders across several trading
venues. Therefore, once passed the equilibrium Y and after reducing costs,
greater benefits await market participants due to the combined effects of
economies of scale and scope. Costs (e.g. costs of interconnection) and
lower benefits (such as reduced profits for trading platforms) for market
participants will push the two functions towards another equilibrium (X).
This point is at a higher level than Y, where the market absorbs the entire
surplus, regardless of its distribution. The stability of this equilibrium is
unknown, since it is up to the market (and regulators) to understand that a
greater number of competing platforms may cause greater costs than
benefits, even if the number of users would increase. Hence, the strategic
behaviour of market competitors which look for an opportunity to seize a
bigger part of that surplus may result in liquidity fragmentation and
strong negative externalities. Empirical studies are not conclusive about the
optimal market capacity for trading platforms that financial markets can
currently bear. Economides (2008) argues that perfect competition (with
zero marginal costs) would provide a network that is smaller than the
socially optimum. Hence, the second equilibrium (X) implies a non-
perfectly contestable market.132
weakening the concept of efficiency defence; see Case COMP/M.2220, General

Electric/Honeywell, Commission decision of 3 July 2001 and Evans & Salinger (2002).
132 A perfectly contestable market is a market in which barriers to entry and exit are absent
(Baumol et al., 1982). The threat of potential hit-and-run strategies keeps markets highly
competitive. Besides other critical assumptions, an important one behind this theory is the
absence of sunk costs. In the case of financial markets, those network infrastructures suffer
fairly high sunk costs that are going down thanks to technologies. Accordingly, even though
the market becomes more and more contestable due to shrinking sunk costs (thanks to new
technologies), these costs will not be zero; therefore competitors may need to benefit from a
quasi-perfectly competitive and contestable market in order to have enough incentives to set
up a competing network infrastructure.
Box 7. A case study of national stock exchanges: Next steps after the
demutualisation
In the last two decades, national stock exchanges have undergone evolutionary
changes in their governance and business models. Three major trends happened
to be part of these changes:
1) Demutualisation
2) Consolidation
3) Diversification
The demutualisation of European national stock exchanges started in the
early 1990s and had been ongoing for more than a decade (for a review, see
Levin, 2003). Nowadays, the once nationalised and user-governed stock
exchanges are for-profit entities facing fierce competition from new trading
platforms, but they still keep part of their institutional role in financial markets
(Lee, 1998). For instance, they still provide primary markets (listings) and junior
markets for small and medium enterprises (SMEs).
National stock exchanges have been leading the process of consolidation,
since the process of demutualisation started, and in reaction to the prospects of
abolishing the concentration rule and opening up execution services to
competition. Exchanges have been working to increase their economies of scale
in order to compensate the drop in trading revenues. This situation brought a
period of intense talks around potential mergers and some important ones
eventually took place. For instance, the merger in 2000 between Paris,
Amsterdam and Brussels (Euronext), followed then by Lisbon (2003). The
merger between the London Stock Exchange and Borsa Italiana was part of this
process too. In addition, a final act (for the moment) was the merger between the
NYSE and Euronext in 2007 that created the worlds biggest exchange.
Currently, exchanges are important players that are redesigning their
business models to keep pace with changes in market structure. In effect,
consolidation is also driven by the importance of acquiring sufficient know-how
and economies of scale to develop their business model in other parts of the value
chain or to reinforce their original vertical (silo) models (see Figure 20.).
Diversification is the word that better identifies current market
developments in the business of exchanges. These firms have massively invested
in strategic mergers and new trading platforms, in particular by successfully
acquiring new MTFs (such as Turquoise, Smartpool and NYFIX Euro-
Millennium) or running their own (see Table 6) in order to build know-how on
new trading mechanisms (e.g. automated dark order books or HFT). Investing in
technologies to increase speed and capacity of infrastructures is key for their
future. For instance, NYSE Euronext recently launched its new universal trading
platform (for all markets) and LSE Group will soon launch the new Millennium
platform for the main UK markets (after the start of Turquoise with this system).
Figure 20. 2009 Exchanges revenues by activity (main European markets)
SIXGroup
NYSEEuronext
DeutscheBrse
NASDAQOMX
LSEGroup
BMEGroup
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Information Trading Listing SettlementandCustody IT Other
Source: FESE.
Table 6. MTFs run/owned by exchanges

2008 2009 2010
16 19 26
Source: CESR MiFID Database (accessed on July 1st).

Business diversification has also been pursued by developing trade
execution services for assets other than equities. For instance, for Deutsche Brse
(through Eurex) and NYSE Euronext (after the acquisition of LIFFE in 2001), such
services are already profitable. The proposed merger between NYSE Euronext
and Deutsche Brse is aimed at giving a boost to diversification133.
133 For more details about the merger proposal, see http://www.nyse.com/
press/1297768048707.html.
Furthermore, regulatory changes have effectively promoted greater use of

organised trading platforms for non-equity products, together with advanced
data services to keep up with competing post-trading services and market
fragmentation (by consolidating data across asset classes). Exchanges will have to
give more emphasis to their expertise in these areas in order to compete with
incumbent dealer markets and other actors.134
5.4 Classification and obligations of trading venues: A new role

for internalisation?
Trading platforms bring together buyers and sellers interests in financial
instruments to promote best allocation of capital in financial markets.
Trading platforms can be classified as either neutral or non-neutral.
A neutral trading platform allows matching and public
dissemination of executable bids and offers (or asks), which are typically
consolidated in an open limit order book (displaying best bid/ask offers)
that may or may not be pre-trade transparent. Binding manifestations of
interest are sent to the platforms by investors, typically through their
intermediaries or dealers, who can also trade on their own capacity. Access
to the platform is defined on a non-discretionary basis and the company
running the platform does not trade on own account or act on behalf of
investors either as a principal, agent or riskless counterparty.135 This is the
main characteristic of so-called multilateral trading mechanisms. MTFs
contribute to price discovery by publishing meaningful pre-trade
information (executable quotes), which allows traders to execute their
orders almost in real-time. Furthermore, liquidity providers may commit
capital to support continuous trading.136 An undiscriminating platform
has three main characteristics (Madhavan, 1992; Glosten, 1994):
i) It will breakdown with high asymmetry of information (adverse
selection and moral hazard).
ii) It lowers spreads for small quantities (cheaper and more accessible
for retail and small professional investors; fairer markets).
134 For an interesting view on the role of exchanges, see Harris (2010).
135 Recital (6), MiFID. By contrast, Dealing on own account means trading against
proprietary capital resulting in the conclusion of transactions in one or more financial
instruments; Art. 4, MiFID.
136 For a review of potential trading mechanisms, see next section.
iii) It invites competition between trading platforms as lower execution

fees will attract more investors.
Figure 21. Trading platforms and MiFID
Trading Platforms
Neutral Non-Neutral
Price discovery Preferencing (best execution)

Multilateral Bilateral
Non-discretionary rules Discretionary rules
RMs MTFs SIs OTC (Recital 53)
MiFID Trading Venues
Source: Authors.
A non-neutral trading platform brings together buyers and sellers
orders that are already on its own books (own account) or are part of
dealings carried out by wholesale counterparties (e.g. asset management
companies). Therefore, execution services are provided as part of activities
to handle in-house orders (internalisation) or to cross them on internal
crossing networks (over-the-counter, OTC). Both activities can be
considered preferencing activities (Lee, 2002) developed by banks and
brokers/dealers in order to promote best execution outside MiFID official
venues (RMs, MTFs and SIs), as well as to increase competition on pricing
with main markets. These forms of trading present a trade-off between the
risks of market fragmentation and cream skimming on the one side, and
better prices and lower dealing costs (in particular, for the institutional
demand) on the other. A discriminating platform (Glosten, 1994), finally,
makes markets:
1) Less likely to breakdown (it selects liquidity) and
2) Unresponsive to competitive reactions (liquidity is typically locked-
in by switching costs and competition may not be profitable as
liquidity is limited.137
However, technologies have brought down switching costs and introduced new trading
137
methodologies. This is actually increasing competitive reactions also on these platforms for
MiFID designed a set of official trading venues operating as trading

systems for shares admitted to trading on RMs. On the one hand, there are
two major categories of neutral and non-discretionary trading venues
providing order execution services under MiFID: regulated markets (RMs;
Arts. 4.14 and 42), and multilateral trading facilities (MTFs; Art. 4.15). The
Commission is consulting on the proposal to introduce a new category of
trading venue for OTC derivatives, the so-called organised trading facility
(OTF, see Box 9), which represents the evolution of current inter-dealer
platforms currently classified as MTFs. The definition propounded by the
Commission does not clarify yet if the platform will be a discretionary or
non-discretionary trading system.
As trading venues, RMs and MTFs perform order execution services
with the same organised trading functionality (Recital 6, MiFID).
However, RMs have more stringent requirements for admission to listing
and trading, higher disclosure costs for issuers and obligations to monitor
potentially abusive practices and preserve orderly trading.138 Specifically,
securities regulation has always recognised an important role of RMs for
investor protection; this may have affected the way in which MTFs have
been sometimes treated by national authorities.139 CESR (2010b), in this
respect, has formally proposed to level these rules, which will oblige MTFs:
i) To have the same arrangements as RMs for conflicts of interest in
their operations,
ii) To be equipped with adequate mechanisms to manage the risks to
which they are exposed and
iii) To set arrangements for the sound management of technical
operations (e.g. risks of system disruptions).
The third requirement will need to be extended also to OTFs
(European Commission, 2010b, p. 19).
Finally, in order to trade a share admitted to trading in another
regulated market, regulated markets have to wait 18 months from the
original admission, plus the publication of a summary note to the
prospectus by the issuer, while MTFs run by investment firms are not
the provision of execution services.

138These other requirements are mainly coming from Art. 43 MiFID, the Transparency
Directive (2004/109/EC) and the Prospectus Directive (2003/71/EC).
139 See Moloney (2008, pp. 797-813).
requested to follow this procedure. As explicitly stated by MiFID (Recital

6), these players operate with the same business model and thus should be
subject to the same set of rules.
Systematic internaliser. On the other hand, the Directive also
established a new trading venue, the so-called systematic internaliser (SI).
This venue executes trades on own account on an organised, frequent and
systematic basis (Art. 4.1.7, MiFID), outside RMs and MTFs. It applies non-
discretionary rules in handling trades and pre-trade transparency for client
orders on liquid shares (for illiquid ones, only on request) below the
standard market size (SMS).140 However, besides publication issues, pre-
trade transparency is not meaningful since quotes are not executable
because execution happens at the discretion of the SI. The contribution to
price discovery is therefore minimal.
Conclusion # 12
Competition between trading venues on prices and execution services and between
investment firms on the provision of other investment services has generated some positive
effects, in terms of promoting investments in trading platforms and reducing trading costs
for shares. Competition needs, however, to be fair and based on a level playing field between
MiFID-official trading venues and their structural characteristics.
In this respect, this report acknowledges the importance of ensuring a harmonised
approach across national supervisory authorities in the application of MiFID requirements
for official trading venues. This may require a further alignment of the remaining
differences in the legal obligations (and supervisory practices) applied to regulated markets
(RMs) and multilateral trading facilities (MTFs). This alignment of legal obligations
already exists in some European countries (e.g. the UK).
Box 8. The regime for systematic internalisers

The low number of firms currently registered as systematic internalisers (only 12)
has suggested that the current regime may not be a great success. Along with
CESRs advice (2010b), the SI regime does contain some ambiguities in its text. In
particular, Art. 21.1 (Implementing Regulation) sets the conditions to be
considered where trading venues execute trades on an organised, frequent and
systematic basis on own account (bilateral trading mechanism). These
conditions to be considered all together are:
140Art. 27, MiFID; Arts. 21-25, Implementing Regulation. The Standard Market Size (SMS) is
defined in Annex II, Table 3 of the Implementing Regulation, which changes in relation to
the average value of transactions.
1) The application of non-discretionary rules and procedures and the activity

has a material commercial role;
2) The development of activities through personnel or automated technical
systems, with no consideration if they are also used for other purposes;
and
3) The continuity and regularity of the activity.
Condition 1, in particular, has raised doubts about the term non-
discretionary having the same meaning as for RMs and MTFs. CESR (2010b)
suggests keeping this reference but clarifying that it was originally intended only
to avoid non-systematic trading (including this clause in their commercial
policies).Dealing on own account is by definition a discretionary activity, so the
requirement cannot refer to the trading system itself. Furthermore, by assessing
whether the activity is a significant source of revenue, competent authorities
should be able to determine the materiality of the commercial role covered by the
firm. CESR suggests setting a threshold for the activity to be considered
significant. By contrast, the Commission (2010b, p. 17) proposed to replace the
more vague condition of material commercial relevance with a quantitative
threshold above which, if the firm also meets the other two conditions, it will be
required to register as SI to the competent authority.
Condition 3, finally, does not explicitly clarify what is the meaning of a
continuous and regular activity, besides what common sense would suggest.
In addition, some pre-trade information (quotes) displayed by SIs is not
meaningful since it is often one-sided and in a very low size (even one share).
Therefore, de facto the market does not see liquidity and activities on the venue.
In order to keep the level playing field even for retail traders in RMs and MTFs,
CESR suggested to fix a minimum quote size (10% of SMS of any liquid share in
which the firm is SI). Quotes then need to be two-sided and periodic trading data
reports should be published more regularly (e.g. monthly; CESR, 2010b). Finally,
SIs are not properly flagged in the post-trade disclosure, which has prompted
CESR to propose a flag for trades executed by internalisers (ISO 9362). CESRs
proposal on quotes, flagging and trade reports has been also endorsed by the
Commissions recent consultation report.
5.4.1 Shedding light on over-the-counter equity trading

OTC trading. Besides internalisation, another form of preferencing activity
has been historically offered by brokers, and known as internal crossing.
Part of the liquidity in the market even for liquid equity markets has
always followed a different way of execution, generally called upstairs
trading (Madhavan, 2000). Those services provided on a discretionary
basis have been evolving with time and are currently offered through
sophisticated internal crossing engines. These engines utilise complex
algorithms to achieve best execution, by either matching internally or
executing on MiFID official venues (RMs and MTFs). These internal
crossing systems are also more commonly defined as broker-dealer
crossing systems (BCSs, or broker-dealer crossing networks BCNs if several
BCSs interact in one or more pools of liquidity). In addition, crossing
activities are also subject to fiduciary duties and conduct of business rules
(Arts. 19, 20, 21, MiFID; OTC) as well as arrangements to disclose conflicts
of interests. While post-trade transparency obligations apply to them, they
are not affected by pre-trade transparency or organisational requirements
(CESR, 2010b).
MiFID captures internal matching activities offered by
brokers/dealers through the definition of transactions carried out on an
OTC basis. Three conditions should be met for trades to fall under the
definition of OTC equity trade (Recital 53, MiFID):
i) Transactions must be ad hoc and irregular;
ii) Dealings take place with a wholesale counterparty; and
iii) Dealings must be above the standard market size (SMS).
The interpretation around the implementation of these conditions is
currently different across the industry. Market views consistently diverge
around the business nature of BCNs, which for some perform the same
business as RMs and MTFs and therefore should be classified as such.
Market views. One side of the market believes that there is part of OTC
trading that should be traded on RMs/MTFs and SIs and are therefore
subject to strict pre-trade transparency requirements and more effective
post-trade reporting, while they acknowledge the importance of OTC
trading in the MiFID text. Most will agree that all trading happening in
Europe, including the one taking place under the OTC label, should be
properly classified in order to preserve market quality and prevent any
significant part of the market from escaping pre-trade transparency
without any economic justification (see below) and other MTFs rules on
access, discretion and surveillance. This part of the market argues that OTC
performs the same function as RMs and MTFs, and if not properly
classified, could increase liquidity fragmentation and weaken investor
protection.
Moving from the assumption that the OTC trade reports used were
meaningful and mostly accurate, a recent study tries to shed more light on
OTC trading (Gomber & Pierron, 2010). Using current data on OTC trades,
the authors show that about 39% of OTC trades are below retail size, 48%
are below standard market size, 87% are below large-in-scale and 73% are
too small to have a market impact according to their proprietary
methodology. According to them, this should be enough to clarify that
these networks deal with trades below the standard market size, even
though orders are partially sent to regulated markets and MTFs. In
addition, crossing systems are designed as multilateral trading
mechanisms that match trades, acting as riskless counterparties. Hence,
87% of trades could potentially escape the size-related pre-trade
transparency waiver imposed on RM and MTF trades (Gomber & Pierron,
2010). These findings imply that the MiFID conditions to be considered
OTC trading are not met and therefore there should be a proper
classification for these trades under current MiFID-official trading venues
(see below).
This first position concludes that, even if OTC trades would be
considered bilateral, 52% of them seem to be escaping MiFID rules of pre-
trade transparency for SIs. In any case, they see BCSs as multilateral
trading systems, that are systems managed by firms not operating on own
account or taking any risk in the transaction (recital 44, MiFID). As a result
multilateral trading systems should be classified as RMs and MTFs.
A countering view contests the abovementioned interpretation of
recital 53. In their view, broker-dealer crossing systems (BCSs) are fully
compliant with MiFID, which recognised their role in financial markets.
However, those platforms should be better classified in the new MiFID text.
124 | RESHAPING MARKET
A STRUCTUREE
Figu
ure 22. Transpaarency obligation
ns for venues under MiFID
mechanism
Execution
Multilateeral trading B
Bilateral trading
Classification
Regulated Multilateraal Systeematic Reciital 53

market training facility intern
naliser OTC
C basis
Pre-trade Pre-trade Pre--trade

transparency transparenccy transp
parency
with venue with venuee (Anonnymous
S
SI)
Transparency
requirements
identification identificatio
on
Post-trade
Post-trade
P transparency y
Post-trade
trransparency (Anonymous SI S
transparency
w venue
with or OTC)
with venue
id
dentification identification
Limit orderr
display rulee
Sourcce: Gomber & Pieerron (2010, p. 11)).
According to t this secondd position (seee graph below w), the advannced
brokkerage servicees offered to cllients (wholesaale counterparrties) through the
use of BCNs meeet the MiFID conditions
c for trades carried
d out on an OTC
O
basiis, since they are:
a
Dealings with
w wholesalle counterparrties (e.g. assset managem ment
companies),,
Dealings (pparent orders)) above the sta
andard markett size and
Dealings irrregularly transsacted with brokers/dealerss.
Figure 23. Broker-dealer crossing system
Broker/Dealer Clients
External liquidity pools
(AM, etc)
Parent
Orders
Broker/Dealer
Split and
Front Desk Diced
(algorithmic
trading)
Internal
Open Markets
Crossing System
Child
Orders
External
Lit Markets Dark Markets Crossing
Systems
BCN
Source: Authors.
They note that trades may be below the standard market size only
after the broker/dealer has accepted to execute the trade and split the
parent order into child orders, eventually routing them to several trading
venues to minimise market impact (including internal crossing systems, as
BCSs or BCNs, but also RMs and MTFs). They argue that the Directive
refers to parent orders since fiduciary duties and business conduct rules
apply to the parent order as a whole and not separately to each child
order. Nevertheless, in their view, most trades are sent to regulated
markets and MTFs for execution, while usually less than 1/3 is internally
crossed. They argue that advanced brokerage services need to interact with
internal crossing systems, which allow a smoother handling and execution

of big orders. The general rationale for trading OTC is best execution of
trades outside official markets. Moreover besides the exemption of recital
53 this position also argues that BCSs cannot be considered multilateral
since they do not act as riskless counterparties since the riskless position
is part of the execution they carry on behalf of their clients. Since MiFID
obliges them to provide best execution and apply other conduct of business
arrangements, they should not be considered acting as riskless
counterparties. In addition, recital 44 says that RMs and MTFs should be
managed with non-discretionary rules, which is not the case for BCSs. Most
of all, they claim that internalisation is gradually becoming an
indispensable complementary execution service to those offered by official
markets, rather than a competing service as originally feared.
Finally, some players claim that imposing a trading threshold may
represent a threat for the provision of complex and tailored brokerage
services, as it needs to be accurately and constantly updated. Clarifications
of the defining criteria of OTC trades under MiFID and availability of data
for full assessment and enforcement of best execution, and other conduct of
business rules are needed. Once these aspects are defined, the burdens of
the obligations would set a natural threshold to the development of these
trading systems with no threat to price formation processes in the open
markets, since it would become highly costly to carry them out on a broad
scale.
Commissions proposal. The European Commission (2010b, p. 11) has
proposed a new sub-regime for BCSs, which will de facto assign a new
classification to these venues under MiFID organised trading facilities (see
next box). This classification will work by setting two boundaries: i) if third
parties will enter orders in the crossing system, the BCS would transform
the system in an MTF; ii) if the broker/dealer executes internally against its
own capital, the system would change in a SI.
This proposal tries to accommodate conflicting views and it may
represent a radical change for the definition of SI and MTF. For instance, if
third parties enter orders in the system, it does not necessarily mean that
the system is based on non-discretionary rules (such as an MTF). Therefore,
this proposal will de facto modify incentives to provide discretionary
execution services (e.g. upstairs trading) under MiFID, with unclear
broader implications.
Being a sub-category of OTF, the broker/dealer running the BCS may
also need to add an identifier for post-trade transparency requirements and
report the daily number, value and volume of transactions. It will also add
an identifier to transaction reports, to show when the transaction is
executed on the system.
Figure 24. Proposal for broker dealer crossing system (BCSs)
systems
become
Systematic Multilateral trading

internaliser facility
- Matching against - Third parties enter

own capital orders in the system
Source: European Commission (2010b).

Box 9. Organised trading for OTC derivatives: EU and US discussions
Following the G20 commitments,141 the European Commission has proposed that
where appropriate, trading in standardised142 OTC derivatives143 [will] move to
exchanges or electronic trading platforms (EU COM, 2010b, p.12). Besides the
fact that some trading on fixed income products (bonds and structured products)
and OTC derivatives takes already place through platforms registered as MTFs
across Europe, the Commission taking up the positions of the European
Parliament (Langen, 2010) and CESR (2010g) has proposed a new category of
venue, the so-called organised trading facilities (OTFs; EU COM, 2010b, p. 13).
141 In particular, G 20, Declaration on strengthening the financial system, London Summit,
2 April 2009 (http://www.londonsummit.gov.uk/resources/en/PDF/annex-strengthening
-fin-sysm) and G 20, Leaders Statement: The Pittsburgh Summit, Pittsburgh Summit, 24-
25 September 2009.
142 Standardisation is a multifaceted concept. It refers to specific technical processes,
economic and legal terms of a financial product that allow a straight-through processing
(STP) of a derivative transaction (Valiante, 2010, p. 11). A product is standardised when the
interaction between those aspects does allow an STP of the transaction. The use of electronic
means is only one aspect of standardisation. Other relevant aspects are uniform contractual
agreements and the use of plain vanilla terms, where possible.
143 For a definition, please see Section 4.4.5.
The Commission suggests that OTFs should meet the following conditions:
1) Non-discriminatory144 multilateral access;
2) Support pre- and post-trade transparency obligations;
3) Report transactions to trade repositories; and
4) Have a dedicated facility for execution of trades.
For the consultation document, ESMA would set requirements to
determine when a derivative is sufficiently liquid to be traded exclusively on
OTFs or other organised venues. The decision could be based on liquidity
measures such as frequency of trades and average size of transactions, as well as
additional criteria such as the degree of investors participation. Ideally,
requirements may need to coordinate with rules proposed for swap execution
facilities (SEFs) in the US. Since the market for OTC derivatives is global,
uncoordinated responses could create regulatory arbitrages.
Furthermore, rule-making should take into account that non-equity
products (in particular, OTC derivatives) usually have a different market
structure than equity markets (limit order books), with dealers posting executable
quotes at investors request (Request For Quotes model; RFQ) or through
bilateral transactions. The RFQ model is required by the nature of the transaction
that is essentially bilateral, and by the nature of the demand that is mostly
institutional and require a certain degree of customisation. A greater push
towards standardisation and organised trading should balance benefits of a more
transparent and orderly setting with costs entailed by lower availability of
customised derivatives and so greater possibility to leave in the system some risk
not properly hedged (Valiante, 2010, p. 45).145
Trading in bonds and structured products is already partially done on
organised trading venues like MTFs, but transparency obligations do not
necessarily apply to them, due to the exemption in MiFID (Recital 46). In effect,
there is a risk that the requirements for OTFs may overlap with those for MTFs,
as they provide a execution service. The border between being qualified as a MTF
rather than an OTF should be clearly spelled out.
In order to promote more organised and transparent trading for non-
equity asset classes, the Commodity Futures Trading Commission (CFTC) has
proposed146 that those systems should have robust risk controls and should
display either indicative or executable quotes from members. In order to avoid
144 The Commission does not clarify if non-discriminatory should have the same meaning
as non-discretionary, as currently defined by MiFID for RMs and MTFs (recital 44).
145 CESR (2010g) believes that trading of standardised derivative products on organised
trading venues is to be encouraged by regulators, even though not mandated at this stage.
146 Commodity Futures Trading Commission (CFTC), Core Principles and Other
Requirements for Swap Execution Facilities, (www.cftc.gov).

unintended consequences for market structure, trading venues should be able to

deploy both RFQ and electronic order book trading systems. In effect for US
regulators this solution would allow a gradual shift from currently bilateral
trading to SEFs by enhancing the opportunity for dealers to compete on
executable quotes whenever possible, without necessarily undermining the
management of their huge inventory positions (by publishing an indicative quote
or falling under a large-in-scale waiver). The US proposal, at the moment, does
not clarify whether these venues will have non-discriminatory access and pre-
trade public transparency obligations, such as EU OTFs. The RFQ model under
SEFs should have at least five market participants from which other participants
can simultaneously request quotes. The order book model, if chosen by market
participants, should develop a platform in which all of them can post bids and
offers for other participants and decide with which of the displayed bids and
offers transact.
Finally, in Europe, some trading platforms registered as MTFs have already
started to bring together, through organised trading on order books, executable
quotes and interests of dealers and other market participants that have chosen to
stream their quotes on these platforms. Public pre-trade disclosure is limited to
non-executable quotes, such as the mid-price.
OTC trade disclosure. Besides specific positions on market structure

and how equity trading on an OTC basis currently occurs, there is a
consensus that the quality of OTC data needs to be effectively improved,
reducing inconsistencies (duplicative reporting) and increasing granularity
(in particular, through the use of specific flags). Post-trade transparency
requirements for OTC trading have been introduced by MiFID for the first
time in many countries. According to CESR (2010b), up to 38% of EEA
trades occur on an OTC basis, but it also recognises serious risks of
misreporting and double-counting. These data are collected by Thomson
Reuters from market data sources that collect themselves data as defined
by MiFID. At such levels, OTC trading may raise doubts about its beneficial
role in European capital markets. Some argue that this size despite being
historically high for Europe contradicts the MiFID concept of OTC being
an exceptional type of execution. Separately, since all OTC trading is by
definition pre-trade dark, if this figure is correct, such level of non- pre-
trade transparent trade may undermine market quality and price formation
mechanisms.
However, there are two important caveats to the interpretation of this
data. First of all, given the poor consistency and quality of OTC data, the
publicly reported OTC may not be the same as what is actually traded
OTC. Secondly, and even more importantly from the perspective of either
of the two positions outlined above, CESRs data only include the market
share of OTC as reported through Markit BOAT or national exchanges. In
effect, these data do not permit a trade-by-trade analysis or break down the
overall 38% figure without assuming any double-counting or misreporting
(see Box 10). These issues contribute to increasing uncertainty and
opaqueness in the market. Without such an analysis, it is impossible to
draw from the market share of the OTC alone any definitive conclusions
about whether the OTC classification is being correct and whether, as a
consequence, there is sufficient transparency and market quality in the
market.
OTC breakdown. In this respect, a recent study (Nomura, 2010) reports
an earlier investigation made by the UK Financial Services Authority (FSA)
that estimates the share of reported OTC trades affected by double
counting.
Figure 25. OTC printed vs actual equity trading
Source: FSA (from Nomura, 2010).
Since some trading venues are not officially recognised by MiFID and
divergences exist between the regulatory and economic definition of trade,
what the Directive considers as trade to be made transparent does not
necessarily match the definition of trades that contribute to price formation.

Therefore, there is a difference between what is printed as defined by
MiFID and what really contributes to price formation (actual trading).
As suggested by the figure above, in the UK market, roughly 35-40%
of OTC trading represents give-ups from brokers,147 plus other categories
of trades that may not be considered as real liquidity, either price forming
or informing. In addition, CESR (2010g) investigated other cases of double-
counting or misreporting (see box below).
Box 10. OTC equity transaction type standards

In order to improve the quality of OTC trade reporting, CESR (2010g) has
proposed the introduction of transaction type standards.148 These standards
should increase granularity and help at the same time to breakdown OTC trades
into data that should be reconsolidated as either price forming or informing data,
while avoiding double counting or misreporting.
Hence, CESR proposed a flag for each type of transaction below:
i) Benchmark trades (when price is calculated over time, as result of different
variables, e.g. VWAP (volume-weighted average price);
ii) Agency cross trades;149
iii) Give-up/in trades;
iv) Ex/cum dividend trades; and
v) Technical trades.150
In order to avoid double reporting due to the complex application of Art.
27.4 (implementing regulation), CESR also clarifies that it should generally be the
executing broker that makes the report. If it is unclear who the executing
broker is, the selling investment firm provides the report.
147 Give-up trades are orders executed by a broker (A) on behalf of another one (B), who is
officially executing it on behalf of a client. Since the order should be made by the broker B
on behalf of the client, broker A should give up the trade (once executed) to broker B. In this
way, the trade is printed when broker A buy shares on behalf of broker B and when the
trades are given up and bought/sold by broker B in order to justify his/her execution to the
client.
148 To amend Arts. 3 and 27(1)(b), Implementing Regulation.
149When a broker brings together clients orders to buy and sell, both transactions
conducted as one transaction.
150 Technical trades is a generic category that includes non-addressable liquidity or
exchange of shares [] determined by factors other than the current market valuation of
the share (CESR, 2010g, p. 11). This category should include back-to-back transactions to
ensure the proper functioning of the financial system.
Therefore, the Committee has highlighted three potential sources of

double-counted or misreported trades:
i) Riskless principal (the investment firm buying and selling on own account
simultaneously to make the report);
ii) Agency cross (the investment firm crossing the two trades should make
the report); and
iii) Single/Multiple OTC transaction on behalf of a client (only the selling
investment firm or the trading platform where the trade has been sent
should make the trade report, if not agreed otherwise).
An example of double reporting is when a client orders to buy 10 million
shares at a guaranteed 2 hour VWAP. The investment firm decides then to split it
into child orders of 6 million shares on the main RM, 3 million on a MTF and 1
million on another MTF. The investment firm will adjust the average price to the
actual VWAP for client booking. The uncertainty surrounding who should report
in this case obliges the investment firm under MiFID I to report as OTC the
client booking of 10 million (client-leg). However, the market also sees 6 million +
3 million + 1 million as traded shares (market leg). Therefore, trade reporting will
show 10 million as execution on MiFID-official trading venues and 10 million as
OTC, even though the real trading volume was 50% of what has been reported
and none of it over-the-counter.
The CESR proposal clarifies that only the market-leg should be reported either by
the selling investment firm or by the RMs or MTFs (if orders are sent to the
market). In the abovementioned example, all 10 million shares would be trade
reported by the RM and the two MTFs. In the case of BCNs, the trades that are
matched in the internal crossing system would be trade reported by the executing
(crossing) investment firm, while other trades sent to the market would be
reported by the trading venue where orders have been sent for execution.
Conclusion # 13
It is within the scope of this report to clarify the different views around the classification of
trading venues and OTC trading (BCNs). However, this report does not challenge the role
that OTC equity trading plays in financial markets, but rather acknowledges it, in
particular where it comes to guaranteeing effective best execution for complex institutional
orders. Therefore, the review of MiFID should not ban these trading activities but rather
recognise their relevance by properly classifying them.
This report also acknowledges the importance of ensuring a harmonised approach
across national supervisory authorities in the application of MiFID requirements to official
trading venues. This may require a further alignment of the legal provisions and
supervisory practices applied to regulated markets (RMs) and multilateral trading facilities
(MTFs). This alignment already does exist in some European countries, such as the UK
where the remaining differences have been levelled.
Data employed in the run-up to the MiFID review by the Committee of European
Securities Regulators (CESR) to ascertain the size of OTC equity trading are inadequate.
Analysis based on this data cannot be considered conclusive and has probably led to
overestimates of the size of this market. More effort needs to be made to accurately assess
market quality in Europe and clarify the actual size of OTC equity trading, its origin and
its impact on price formation processes. There is a compelling need to improve the quality of
market data by reducing inconsistencies and increasing granularity through the use of
harmonised flags.
Box 11. The role of issuers
Issuers play a crucial function in financial markets, as they promote innovation

and a more efficient allocation of resources in our economies. Through the access
to capital markets, they favour a better allocation of resources by bringing
together those who need them with those who have a surplus of capital and wish
to allocate it in the best way possible. Issuers represent the direct link between
financial markets and the high street, in particular for small and medium
enterprises (SMEs) that have limited possibilities to access alternative funding
channels. After a single or multiple listing, it actually matters where financial
instruments are traded in secondary markets (Amihud & Mendelson, 1996; Di
Noia, 2001). On the one side, fragmentation may discourage issuers participation
if they do not keep control over the value of their shares over time. Some believe
that a right for the issuer to decide to go public and where securities should be
traded in secondary markets should apply. Such a proposal would entail the
right of issuers to, at least, be duly informed and asked for acquiescence in
relation to their issued securities in order to be traded in venues different to
that/those of issuance. The proposal would not be claiming for the ban of multi-
market trading, but rather the need for issuer to be informed and in agreement to
such fact. On the other side, assuming interconnection between trading venues,
multimarket trading may allow more efficient pricing, lower costs and greater
investor protection (those who provide liquidity to issuers).
In this sense, competing flows of information may also generate prices more akin
to the firms fundamental value. It is highly controversial that once the issuer has
placed financial instruments (mainly shares) and collected capitals at relatively
low cost, the management of the company (or the shareholders majority) should
be able to influence the way these instruments are traded in secondary markets,
even though they are not the companys property anymore. Conflicts of interests
between issuer and trading venues or their members may impose additional
costs. A proposal that would allow issuers to decide in which secondary markets
their shares should traded, however, has been historically challenged by both
market and policy-makers (e.g. the SEC151).
It is also relevant, in particular, to ensure that the interconnection among venues
is effective and does not permit the formation of inferior prices or market abuses
due to potential arbitrages with main markets or proper supervision cannot occur
(extreme case). Finally, the possibility to trade instruments freely across trading
venues is a crucial aspect for market liquidity; it broadens both the appetite of
investors and competition between trading venues, although venues may
ultimately suffer from any competitive constraint set by the issuer. For SMEs,
finally, the Commission (2010b, p. 20) proposed a special regime in order to
facilitate a cheaper access to capital, as well as to enhance inter-linkages between
markets.152 A set of requirements would apply to firms with market capitalisation
below 35% of the average market capitalisation. In this respect, the regime
proposed does not seem necessarily to rely on a set of lighter requirements but
rather on adapting current rules applying to trading venues and their members to
SMEs size and nature. The framework of rules will be defined in the framework
Directive, which will leave space to member states to tighten or loosen the regime
in line with the specific market conditions. This set of principles may boost a race
among member states to increase the level of standards, as long as the market
recognises these aspects as a further incentive to provide capital to SMEs.
Otherwise, as it seems to be in this case, the incentive for member states would be
to lower regulatory barriers in order to further reduce costs of access for SMEs
and end investors, as long as appropriate protection for both parties will still be
in place.
5.5 Some aspects of market microstructure for organised trading

The microstructure of financial markets has been deeply researched in the
last decades. However, technological developments bring rapid and
sweeping changes to the structure of markets, which may need further
investigation. This section will look at the following topics:
i) Market settings (e.g. auction versus dealer markets);
ii) Trading tools (high-frequency trading, algorithmic trading, direct
151The Unlisted Trading Privileges Act of 1994 made official that no approval from the SEC
was needed to trade securities (not only shares) in other markets than the market where
they were issued. See, Pub. L. No. 103-389, 108 Stat. 4081 (1994) (codified as amended at 15
U.S.C. 78l (f) (1994)). The US Congress has always expressed its favour towards multimarket
trading and that an issuer does not have the right to veto exchange trading of its
securities; Release Discussing Exchanges' and NASD's Proposed Rule Changes, Exchange
Act Release No. 34-22,026, 50 Fed. Reg. 20,310, 23,313-14 (1985).
152 For a more accurate proposal, see Demarigny (2010).
electronic access); and

iii) Other micro-structural issues (e.g. decreasing average size of orders).
As mentioned above (see footnote 48), liquidity in financial
instruments depends on depth, resiliency and breadth of the market. To
allow an efficient interaction between these variables, the market
microstructure needs to be designed according to several factors, such as
financial stability, the role of financial regulation, efficiency, market
integrity and investor protection.
5.5.1 Market setting: The role of trading mechanisms

The market structure of capital markets defines the set of rules and
mechanisms that lead the interaction among different trading interests and
platforms and the incorporation of information into prices, by minimising
frictional costs (implicit and explicit trading costs). These rules may involve
price and quantity discovery (Francioni et al., 2008). Market microstructure
is generally influenced by two factors: i) the nature of liquidity suppliers
and demanders (traders motives); and ii) external factors, such as
technological developments. In particular, the automation of trading has
produced radical changes in order routing, information dissemination and
trade execution (Domowitz, 1996), which use complex mathematical
algorithms. As explained below, technological change is a two-way
process; technology has influenced market structure, which in turn has
stimulated investments in technology. Similarly, the level of market
transparency comes as a result of the design of market microstructure, but
can also ultimately stimulate changes in the structure itself.
The key aspect of a trading mechanism is transforming the latent
demands of investors into realized transactions. This transformation is
based on price discovery, the process of finding market-clearing prices
(Madhavan, 1992, p. 608), which is to a great extent influenced by the level
of market transparency. In general, two types of trading mechanisms let
markets decide transactions clearing prices:
i) Auction markets
ii) Dealer markets
The former puts investors in competition (e.g. limit orders; see
Boehmer et al., 2005) to find the best market-clearing price (downstream
competition). The latter stimulates upstream competition between dealers
to offer the best bid/ask quotes (e.g. RFQ models). Both trading
mechanisms can be continuous or periodic. A continuous trading
mechanism entails the uninterrupted submission and matching of bids and

offers, which can be submitted by investors (auction markets) or
dealers/market-makers (dealer markets). A periodic trading mechanism
brings together those binding interests (typically collected over time) at
some point in time.
Auction markets are widespread in financial markets. There are two
types of auction markets:
i) Batch (call) auctions
ii) Continuous auctions
Batch auctions collect over time bids/offers from investors and
execute these orders at a particular point in time at a single price (to
maximise volumes, call auction) or at slightly different simultaneous price.
Call auction are frequently used in financial markets, especially in the
opening of markets when liquidity is low and the market needs to find a
market-clearing price to start its session.
Continuous auction markets are the most diffused form of trading,
which allows amassing many transactions in one place and generating
economies of scale that abate (implicit) transaction costs (Economides &
Schwartz, 1995). The use of continuous auction markets and their
transparent setting has been increasing in the last decades, as the flows of
cross-border financial transactions have soared. The global interconnection
of financial markets has been made possible thanks to the development of
automated forms of trading and to new communication technologies (e.g.
the internet), which have brought together liquidity demand and supply
across markets, especially for liquid financial instruments.
According to this background, open limit-order book (OLOB) so-
called order-driven market have become a widespread form of auction
market, since it permits continuous trading in which investors public
reservation prices are submitted simultaneously and consolidated in a
single order book, which shows the best bids and offers at different price
levels. This market design has some relevant characteristics:
i) The need for high volumes to allow continuous trading (efficient
system for small trades; Baruch, 2005);
ii) A meaningful competition between orders (information reward153 as
153Investors must receive proper reward from their investment in information (which can
assume different forms; on fundamentals, rumours, etc.). The order flow, therefore, should
contain private and public information. Traders who act on the belief that they own private
a result of information asymmetry) and a greater flow of information

into prices; and
iii) A sufficient level of transparency (allowing investors to see the order
flow and make their investment decisions accordingly).
Auction markets are inherently more transparent than other market
settings, since the price formation mechanism needs transparency to foster
competition between investors. The resulting outcome is more information-
efficient prices, which may ultimately reduce volatility (Baruch, 2005).
Efficient prices also support the evaluation of assets and hence better risk
management processes.
Furthermore, alternative ways can promote competition between
investors and raise the incentives to submit bids/offers, keeping volumes
high. In particular, the flexibility of the execution system and thus the
possibility to enter different limit order types allow investors to compete
with each other on the basis of their information (see next box). In effect,
traders will submit orders on the basis of the information they consider
private and will behave strategically154 when it comes to i) liquidity and ii)
asset value. They submit orders on the basis of information such as
potential arbitrages, information on the fundamental value of the asset or
the need to divest own liquidity.
Box 12. Order types

Limit order books make possible the interaction between several types of orders.
Besides the modality and conditions for the execution, orders always include a
binding interest to buy or sell a financial instrument when certain conditions are
met. The set of conditions applicable to orders that a trading platform can adopt
on its system is potentially unlimited.
Therefore, below are summarized the most-used types of orders:
Unconditional limit155 (when a trader enters an order to trade a number of
shares at a certain price that is not yet verified);
Market (when a trader enters an order to trade a number of shares at the
currently displayed price or at the closest value);
information which in the end is already public and embedded into prices are called
noise traders.
154 They will base the decision on their beliefs about other traders future choices.
155 The report generally uses limit orders and unconditional limit orders interchangeably.
Marketable (an order at the limit or better price);

Tick sensitive (an order to sell or buy at an uptick or a downtick);
Stop/Take profit (market or limit-sell or buy order for an open position,
respectively long and short position, triggered when the price moves
through the stop price in order to contain losses or take profits from the
position held);
Market-not-held (an order to leave the decision of execution to the broker);
At-auction (orders only allowed for the call auction at the end or beginning
of the day; if not executed, these orders are typically cancelled);
Immediate-or-cancel, IOC (a market order that requests the immediate
execution of the order; if part of the order is not executed at that market
price, it should be cancelled);
Fill-or-kill, FOK (a slightly different IOC order since in this case the order
should be cancelled in full if it cannot be filled at that immediate price);
Market to limit (another variation of IOC and FOK; if the order is not
executed in full at the current immediate price, the rest of it will be
cancelled and re-entered as limit order with price equal to the price of the
executed part of the initial order);
Trailing stop (a sell order that sets a stop price at a fixed number of ticks
below the market price, which will follow the market price where above
the executed price; if the price is below the execution price the stop price
will be the floor that will be triggered if the price reaches this value);
Market-if-touched (a buy or sell market order triggered when the price
reaches a certain level; a combination of a market and limit order);
Box Top (a market order that is automatically changed in a limit if the
order is not executed at market price).
These orders are all market or limit orders that may or may not apply
detailed conditions, in relation to the strategy that the trader wants to achieve
(limit risk, speed of execution, price improvements, etc). Further, these orders can
be pre-trade displayed or hidden when waivers apply and combined together to
pursue trading strategies. Finally, all orders can be seen as orders with a limit
price subject to certain conditions. For instance, a market order is a limit order in
which the limit price is the market price at that specific instant.
Sources: Harris, 2002; Comerton-Forde & Rydge, 2004 (www.interactivebrokers.com).
Order flow competition. To meet investors needs, several types of

orders interact on the order book and get executed in a sequential order.
The order flow competition between traders through orders has been
partially investigated. Many types of limit and market orders may create
several equilibria between orders that have not been researched enough.
For instance, Foucault (1999) suggests that the volatility of the asset is the
main determinant of the equilibrium between market and limit orders. In
effect, when the volatility of the asset increases there is more possibility for
limit orders to be picked-off by new information and suffer adverse

selection (winners curse). Market orders will be more costly in terms of
price impact and spread because there will be more competition between
investors in their use. This situation also explains why trading at the end of
the day is dominated by market orders, as the probability of execution of
limit orders decreases.
As suggested by Parlour (2010), in a world in which information
moves faster across assets classes due to technological developments and
financial interconnection, the order flow is intrinsically more informative
and the costs of adverse selection due to fast-moving information
encourages the use of market orders, which may be seen as limit orders
with price set at the market value in the instant of the observation of the
order flow. In addition, the increasing need to look at shorter time horizons
to fill the gap in returns caused by the dismal growth of developed
economies has further increased volatility of asset values, which are
affecting the microstructure of financial markets. Financial institutions are
massively investing in routing technologies and speed of execution, since
investors need to observe the order flow in real-time to use market orders.
In effect, since volatility has created a strong disequilibrium between
market and limit orders, the only way to win competition between market
orders and improve the liquidity of the order book is to increase speed and
volumes. More speed and higher volume ultimately reduce the portion of
the price on which non-retail investors can compete on and create potential
risks in terms of capacity for trading platforms. In the past, immediacy of
execution was possible through specialists acting on the trading floor.
Today, the only way to see the order flow and reduce market impact by
investing accordingly is through investments in technologies such as smart
order routers and direct access with high-frequency trading tools. Retail
investors were and will always be slower than professional investors.
However, the typically small size of their limit orders protects them from
market impact and the reduction of tick sizes (see also next section).
Finally, order-driven auction markets are designed to handle small trades
at the lowest transaction costs but only as long as the market for that
financial instrument is sufficiently liquid.
Table 7. Auction versus dealer markets

Auction Markets Dealer Markets
PROs PROs
- High transparency - Handling of illiquid products
- Price information efficiency and block sizes
- Order flow competition - Resilience (to information
- Low transaction costs asymmetries)
(especially for small trades) - Execution certainty
CONs CONs
- Market impact - Costs
- Market breakdown - Opacity (price formation)
(information asymmetries) - Accessibility (if OTC)
- Operational risks (capacity) - Low competitive pressures
Source: Authors.
Dealer markets. Another form of trading, often complementary to

auction markets (dealer markets can be designed as multilateral auction
systems), is the dealer market. In this market setting, trades are typically
dealt with quote-driven dealer markets or bilateral negotiations with the
support of an intermediary that submits executable bid/ask quotes. The
submission of quotes can be: i) bilateral or ii) multilateral.
Bilateral models are usual in bilateral OTC transactions, which are in
general used in illiquid markets where financial instruments may need
further customisation. The RFQ model (see Box 9) also represents a bilateral
dealer market since executable quotes are only available on request by that
specific counterparty through the use of an inter-dealer platform. RFQ
markets have been developed for less-liquid products such as corporate
bonds or structured products, while OTC bilateral contracts are still the
main negotiation mechanism for many types of derivatives.
Dealer markets can be designed as (multilateral) auction model,
through a quote-driven market or a facilitator. In the former, dealers
submit executable quotes on a continuous basis or periodically (batch
auction system). A dealer competes with other dealers/market-makers on
the size of the offered bid/ask spread. A quote-driven market is a market
model used for liquid markets such as stocks or futures (e.g. NASDAQ).
Given free entry into market-making business, quote-driven systems are
equivalent to continuous auction mechanisms (Madhavan, 1992). In the
latter model, a facilitator (e.g., inter-dealer platform) brings together
different buying and selling interests as riskless counterparty (multilateral
trading system) but it does not interpose itself to counterparties and the
transaction between the buyers and the seller is based on a bilateral RFQ
model. Therefore, quotes are only indicative and may vary in a pre-defined
range. The trading platform remains a neutral (and multilateral, as Recital
6, MiFID) trading system if non-discretionary rules to submit quotes by
dealers (inter-dealer platform) and orders by clients and dealers (dealer to
client) apply.
Figure 26. Dealer markets (trading models)
PureRFQmodel
Bilateral(OTC)
Private
negotiation
Continuous
DealerMarkets
auctions
Quotedriven
auctionmarkets
Batchauctions
Multilateral
Interdealer
Facilitator
(RFQmodel)
Dealertoclient
Source: Authors.
Benefits and costs. On the one hand, dealer markets may be beneficial
for some asset classes. In particular, these markets can ensure order
execution with minimal exposure risk (market impact; Pagano & Rell,
1993). This effect is achieved through a system of competing dealers that
exploit their informational advantage; namely, the possibility to see the
order size and price before it is executed, which does not necessarily
happen in order-driven markets (Malinova and Park, 2008). This
informational advantage comes as a reward for their investments into
capital commitments and advanced technologies. The use of capital on own
account also ensures that investors can trade highly-customised and
illiquid financial instruments to better meet their risk profiles, by

minimising information leakage. Some investors may also get better terms
if they can exploit a contractual power (institutional investors) or bring a
kind of liquidity that maximises the possibility of dealers to benefit from
their market knowledge. The informational advantage of market-makers
(due to their investments in market knowledge) and their intensive capital
commitment in the development of bilateral or multilateral platforms,
make dealer markets potentially more resilient to information risks than
order-driven markets. Dealers also provide markets with innovative
ventures and financing, as well as tools to manage particular risks that
would be prohibitively costly in other market settings.
On the other hand, the extreme flexibility of these market settings is
usually offset by fairly high costs to deal with financial products. For
instance, corporate bonds should be financial instruments with lower
trading costs than the related stock of the company as they are less risky
(Biais & Green, 2007). However, this is usually not the case because the
market for corporate bonds is often very illiquid (as explained above) and a
market with dealers that commit capital seems so far to be the only
structure able to deal with this constant illiquidity. In addition, the pricing
mechanism is not as transparent as in OLOBs since it is not based on the
reservation price of investors, but on the proprietary valuation models of
dealers. Bid/ask spreads in dealer markets are the outcome of these models
and can only be influenced by the potential threat of competing market-
makers/dealers, which may offer more aggressive bid/ask spreads in
order to erode rivals market shares and increase volumes. In effect,
another potential drawback of dealer markets is the structurally high
concentration levels of the dealer industry (capital is limited!), which may
ultimately determine monopolistic or oligopolistic settings. A model with
monopolistic/oligopolistic dealers may be undesirable not only in terms
of output and prices which are typically fixed above marginal costs but
also because the assignment of stocks between dealers may be arbitrary
and lead to a sub-optimal distribution (Stoll, 1978).
Finally, accessibility for retail investors to some bilateral dealer
markets may be difficult without the support of intermediaries, who
typically pass the higher costs on end investors. It may even impossible (as
for OTC derivatives instruments) because dealers may decide to deal only
with some specific counterparties due to counterparty risk and product
complexity.
The ability to price discriminate thanks to their information

advantage makes this market setting more suitable for discriminating
platforms (see section 5.3). Accordingly, market-makers move first and
propose price schedules. Hence, if the market is less-liquid and the
information asymmetry fairly large, some authors suggest that dealer
markets (with one specialist or competing market-makers) may be both
more resilient and efficient, thereby making it easier to reach equilibrium
(Glosten, 1989; OHara, 1995). Dealer markets are therefore designed to
deal with less-liquid markets, as long as this low-liquidity condition is
always verified.
Hybrid markets. Capital markets have developed different market
microstructures, dominated by OLOBs or dealers. In particular, equity
markets regularly combine auction markets (OLOBs) with the presence of
liquidity providers (market-makers/dealers) that offer liquidity to the
market to ensure the continuity of trading. In this way, markets try to
combine the benefits of both market designs, ensuring a high degree of
transparency (with due exemptions and delays), market integrity, and
investor protection (for uninformed traders). This market setting combines
the efficient pricing mechanisms of auction markets, and greater protection
for retail investors, with stronger resilience to market shocks (information
asymmetry). Stronger resilience comes from the capital committed by
liquidity providers in order to ensure continuity of trading and to always
maintain available executable quotes.
5.5.2 The evolution of trading: Market efficiency and financial stability

Automation and the internet. The automation of trading and the introduction
of the internet have fostered quick developments in capital markets in the
last two decades, in particular for equity markets, pushing volumes up and
making markets in general more accessible. The possibility for investors to
more easily compete amongst each other on a global basis and across asset
classes has led them to invest even more in advanced technologies for
trading with a rapid increase of volumes. In particular, current markets are
experiencing faster and greater flows of information into prices. This
situation has made the order flow more informative than ever before, but
private information embeds many complex variables (such as risk
exposures, gross positions, macroeconomic outlook, and regulatory
boundaries) that have increased the complexity of trading strategies and
ultimately asset volatility (as rationale behind the financial systems bias
towards boom and bust cycles; Group of Thirty, 2010).

Easier and cheaper market accessibility has brought a new wave of
investors and raised volumes, which remains at higher levels than at the
beginning of the 21st century, despite the recent financial crisis. This
situation has inevitably promoted a shift in the structure of financial
markets. Equity markets, in particular, have gradually moved from more
opaque quote-driven dealer markets to OLOBs run by advanced trading
platforms able to process millions of trades in the blink of an eye.
Algorithmic trading. As a consequence, complex flows of information
(e.g. multimarket trading, risk exposure, etc) need trading software that can
collect and process a wide range of variables in order to reach an
investment and trading decision, as well as route orders across multiple
venues (Biais, 2010). The introduction of algorithmic trading technologies
aims at replacing the human trader with machines that overcome cognitive
limits of human beings and process complex information quicker and
completely. Further, algorithmic trading reduces search costs and market
impact by splitting big size orders and automatically executing them on
several venues (split & dice strategy; see figures below).
Figure 27. Average trade size on SETS

70,000
60,000
50,000
40,000
30,000
20,000
10,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Q1
2010
AverageTradeSize()
Source: LSE Group (2010).

Figure 28. Average value of orders executed for most liquid European shares
80,000
70,000
60,000 BP
HSBC
50,000 Unicredit
ENI
Siemens
40,000
DeutscheBank
TOTAL
30,000
BNPParibas
ArcelorMittal
20,000 RoyalDutchShell
Telefonica
10,000 Santander
0
Aug07
Aug08
Aug09
Oct07
Dec07
Feb08
Oct08
Dec08
Feb09
Oct09
Dec09
Feb10
Jun07
Apr08
Jun08
Apr09
Jun09
Apr10
Source: CESR MiFID Database.
The figures above show the trend in the average size of orders
respectively for the London Stock Exchange SETS and for some of the most
liquid EU shares. The decline should reflect a substantial increase in the use
of split and dice strategies, which are based on trading algorithms that
route orders across markets after the big parent order has been split into
small child orders. The average size of retail orders has not followed the
same pattern. Nevertheless, the figure above suggests that MiFID has also
contributed to this decline by perhaps increasing competition in execution
services, thereby pushing investment firms to invest in new technologies
and diffuse them at a faster pace. In effect, besides the decline in the
average value of executed orders from 2007, the second figure also shows
that the difference between average values of orders among liquid shares is
drastically reducing to a tiny range (except for Telefonica and Santander).
This may be the result of a wider use of new trading technologies, which
typically treat orders and split them in a similar way. On average the
market seems to deal only with retail orders. Hence, this situation should
create more liquidity for small retail trades and less liquidity for
institutional orders that do not use split and dice strategies (explaining in
part the growth of dark pools of liquidity; see Section 4.2.1).
Algorithmic trading (AT) gives more speed and strategic thinking to
the investment decision. This trading activity, however, may also generate
costs. Firstly, it raises the asymmetric information between fast and slow
traders, which may discourage slow traders from investing and create
disequilibrium, thus volatility. No evidence that AT raises volatility has
been found so far (Hendershott & Riordan, 2009). Secondly, the increased
number of messages and potential human errors in defining the most
appropriate algorithm during the trading day may generate operational
risks and ultimately cause market crashes (see next Box), with potential
spiral effects due to the interconnection between orders on the order flow.
This situation would irremediably harm the investor confidence that keeps
global financial markets together (in particular OLOBs).
AT can be:
1) Directional (market trend) or market neutral (HFT),
2) Short or long,
3) Fully or partially hedged and
4) Short-term or overnight.
Algorithmic trading solutions are typically embedded in proprietary
software made in-house or outsourced to external IT companies.
Finally, investing in algorithmic trading presents high fixed costs,
mitigated by relevant economies of scale. Network effects and the risk for
slow traders to be easily picked off bring market participants to run to get
the best algorithmic trading technologies. However, on the one hand,
overinvestment may not be socially useful but, on the other, trading
platforms may need to have harmonised monitoring practices to avoid
downside spiral effects (Biais, 2010).
Box 13. The US flash crash: What can be learned?

On 6 May 2010 between 2.40pm and 3.00pm US stock markets experienced a
previously unseen phenomenon. Within a few minutes, the S&P 500 index
plummeted by almost 100 points (-8.2%) and the Dow Jones Industrial Average by
over 1,000 points (-9.2%). Indexes then recovered their previous levels almost 20
minutes later, at the end of the so-called flash crash. Some stocks, such as Procter
& Gamble, dropped nearly 40% before rebounding. The whole market swung
within a 10% range down in 10 minutes due to a series of events.
The SEC report (2010b) split the crash into two major events: i) a liquidity crisis in
the E-Mini (and SPY) future contracts156 market; and ii) a liquidity crisis in the
main stock index market (S&P 500). The general market trend on that day was
already affected by downward pressures from the morning due to the European
debt crisis, with high volatility (S&P volatility index was above 22%) and an
overall sentiment of uncertainty concerning the global economic outlook. The e-
Mini future contract market was very thin (the traded value was around $2.65
billion, 55% less than the early morning).
The market crash propagated from a liquidity crisis in the e-Mini future
contracts. In effect, against this background of low liquidity, an investment fund
trader submitted an aggressive selling algorithmic programme for a big position in
e-Mini contracts (75,000, for a value of $4.1 billion 157), which was a hedge for an
existing position in equity. Only two other sell programs of that size were executed
in the previous 12 months in that market. In addition, those other sell programs
took 5 hours to complete the execution, while in this case the same number of
contracts was executed in 20 minutes. The strong sell pressure, according to SEC
findings, was initially absorbed by high-frequency traders (HFTs), which support
previous evidence on the supply of liquidity offered by algo-HFT when the market
is volatile and fairly thin (Hendershott & Riordan, 2009; Chaboud et al., 2009),
together with fundamental buyers and cross-market arbitrages, which transferred
the selling pressure to equity markets. However, HFTs do not hold big positions
for a long time so they started to sell aggressively in e-Minis, causing an increase of
HFT trading, increasing volumes (HFT generally trade high volumes but do not
hold more than 4,000-5,000 contracts). The artificially high volumes made the
initial sell algorithm feed the market with more sell orders. The trader had set this
algorithm at 9% of total volumes in the previous minute, without considering price
and time, which ended up generating 75,000 contracts that could not be absorbed
at once. The human error of the trader was the assumption that higher volumes
meant higher liquidity. HFTs began to sell and buy contracts among each other,
while the market was going down and the buy-side demand dropped to less than
1% of the mornings levels. In short, the market actually dried up because of the
uninformed use of advanced trading technologies (see graph below).
156 The E-Mini is a stock index instrument traded in electronic future and equity markets. It
is a derivative product designed to replicate the S&P 500 Index. The number of outstanding
contracts is not fixed at any given time. This type of product was introduced by CME in the
1997 and trades exclusively on the CME Globex electronic trading platform 24h a day (SEC,
2010b). The first liquidity crisis also concerns the S&P 500 SPDR exchange-traded fund
(SPY), which replicates the S&P 500 index as well. For simplicity, the report only mentions
the e-Mini market since both markets were affected by the same kind of events.
157 The value of contracts traded was down from the morning to $2.65 billion.
Cross-market arbitrageurs transferred the downward pressure to equity

markets, which fell of 3% like the e-Mini market in 3 minutes. Demand collapsed
and the buy-side only held 1,050 e-Mini contracts and equity markets lost another
2%, bringing the total to 5%. At that moment only 35,000 contracts of the sell
algorithm were executed and the market was stopped for a few seconds by CME,
due to the awkward trading activities and in an attempt to reduce selling
pressures. When the market resumed, in effect, it started to recover but the net
imbalance between buyers and sellers was of over 30,000 contracts (for sellers). The
sell algorithm then completed its sell programme at 2.51pm due to the restored
liquidity attracted by the quick recovery, with high losses overall for the
investment fund since no price and time conditions were added to the algorithm.
Figure 29. Aggregated S&P 500 market depth
Source: SEC (2010b).

In the equity market, besides the initial fall of the index, the fall and pause of
the e-Mini market sparked fears that there was a more fundamental reason than
just human error behind the price decline. Therefore, some market-makers and
liquidity providers widened spreads or withdrew from the market, causing selling
pressures to continue their action and buying interests to go even lower. As a
result, prices started to behave irrationally since markets started to match stop-
market orders with stub quotes originally set by market-makers at levels away
from current market prices in order to fulfil continuity of two-sided quotes
obligations, even when a market-maker had withdrawn. As soon as the market
participants verified the integrity of their data and trading systems, sell-side and
buy-side interests returned and the market quickly recovered its previous value,
but with big losses for some market participants. As the SEC reports, after the
crash, the exchanges and the Financial Industry Regulatory Authority (FINRA)
agreed to cancel orders that were clearly a result of errors in the trading rules.
To sum up, the crash was the combination of human error with the use of stop-
market orders by main US exchanges, which pushed prices to behave irrationally,
so generating heavy losses not only for the traders that run the algorithm but also
for markets. The SEC also concluded that pausing trading systems was beneficial
and that there should be clear and sound procedures to break erroneous trades to
ensure certainty and strengthen investors confidence. In the aftermath of the flash
crash, the authority introduced circuit breakers, which are procedures to halt
trading across markets when prices swing in the previous 5 minutes of a value
higher than 10%. According to these rules, systems should be paused for 5
minutes. Finally, it is thought unlikely that the same phenomenon would occur in
Europe, given that, on this side of the Atlantic, there is a more limited use of stop-
market orders and much less-liquid ETFs markets (low interconnection). However,
it would be beneficial for Europe to also adopt a set of rules to interrupt price
volatility in extreme circumstances, taking the solution in the US as a benchmark.
High-frequency trading. As mentioned above, the general increase of

asset volatility has changed the equilibrium of order flow competition (for
OLOBs), beyond any classification by asset class. Greater competition for
market orders between investors has led market participants to invest
hugely in IT services and high-frequency trading (HFT) technologies,
which allow investors to see the order flow and execute orders in almost
real time, i.e. in microseconds (and soon nanoseconds). HFT is gradually
becoming indispensable for algorithmic trading strategies but it has also
supported the emergence of other strategies. In particular, as anticipated in
Section 3.3, HFT may pursue the following three major strategies (AFM,
2010; Chi-X, 2010):
1) Market-making
2) Arbitrage
3) Informed trading (or speculation)
Firstly, HFT may supply liquidity to markets through systems able to
analyse the order flow at the fastest speed technically possible, reducing
risk-sharing on inventory positions. As suggested in the box above, HFT
may absorb temporary imbalances between buy and sell orders. HFT has
the important function of temporarily absorbing market shocks. However,
HFT positions are usually market-neutral (non-directional), hedged and
continuously updated (high cancellation rate), and closed-out by the end of
the day (AFM, 2010). Market-making means that the role of HFT is
temporary, so in the event of extreme imbalances, markets would need to
restore fundamental demand to avoid market collapse (for instance, by
halting trading across markets for a short timeframe).
Secondly, HFT allows cross-markets and arbitrage across asset classes

to benefit from price anomalies. Moreover, it also permits so-called
statistical arbitrage, which uses historical datasets to find repeating
patterns in the price trends that can be profitably exploited.
Thirdly, HFT may use low-latency execution to pursue strategies
based on investments in information. Three examples are: trading
according to indicators that try to predict future performance; placing and
cancelling orders to find hidden liquidity; or screening algorithms to find
failures and profit from them.
Both algorithmic and HFT are widely diffused and more often used
as complementary tools to achieve specific trading strategies. These
technologies are essential for institutional investors and brokers/dealers,
and, as a result, new trading strategies are becoming widespread (see
Section 3.3). These new strategies are often deemed to increase volatility
but no empirical evidence has been found so far that supports this view
(Hendershott & Riordan, 2009; Chaboud et al., 2009; Broogard, 2010).
The implementation of trading strategies through algorithms and
HFT needs trading tools that can operate with the lowest technically
possible latency (proprietary and roundtrip).158 To reach the highest speed,
and therefore the lowest latency, several factors need to be considered.
Among the most important ones are the following (AFM, 2010):
1) Complexity of the trading algorithm (e.g. smart order routing),
2) IT infrastructures (computing power),
3) Efficient connection (network latency in terms of capacity, speed and
stability of data),
4) Physical distance from the data source (trading platform) and
5) Access services to the platform.
Investment firms and trading platforms are investing in IT
infrastructures and special services to reduce latency (respectively,
158Latency is the time/delay that occurs for a package of data to be sent/received from one
point to another. It is the technical delay caused by the inability of the physical
network/infrastructure to act/react in real time. Latency can be split in round trip and
proprietary (AFM, 2010). Roundtrip latency is the technical delay caused by the structure of
the trading platform, which may take a minimum amount of time to accept, execute and
confirm orders, besides time taken by the security check done by the firewall. Proprietary
latency is the technical delay caused by the hardware, software, IT infrastructure, and access
arrangements of brokers/dealers, or investment firms or investors (direct access) that
transmit orders to the platform for execution.
proprietary and roundtrip latency), such as co-location services,159

connection services (e.g. optical fibres, connection capacity) and direct
electronic access (DEA) services. Several European equity markets (e.g.
NYSE Euronext, Oslo Brs) are moving their trading systems close to the
major European financial centre (London) to offer similar advanced
execution services and higher speed of execution (by reducing physical
distance) through proprietary trading systems, often provided by third
parties. Some investment firms are physically relocating in a mid-position
among several platforms with the support of latest high-speed connections.
This solution or multiple co-locations can ultimately reduce the risks of
liquidity fragmentation and improve arbitrages across trading venues.
Direct electronic access. Amongst advanced trading services, direct
electronic access (DEA) arrangements can potentially reduce latency, costs,
and trade errors (Celent, 2008); but they may have multi-faceted
implications. These services, in particular, may have beneficial effects for
clients (low entry costs and administrative hurdles of membership),
sponsoring firms (attracting HFT volumes, coupled with volume discounts
of trading venues), and trading platforms (attraction of new liquidity).
DEA arrangements allow firms that are not registered intermediaries (or
investment firms under MiFID rules160) to have direct access to the trading
venue through: i) using an intermediarys member status/contract; or ii)
becoming direct member or direct access (IOSCO, 2010b).
In the first case, a non-member will be able to directly access the
trading venue to execute orders by using:
1) The intermediarys infrastructure (Direct Market Access, DMA) or
2) Its own infrastructure, but under intermediarys control and rules
(Sponsored Access or SA).
Sponsored access, in particular, lowers the economic barriers to enter
a platform. It allows direct access to markets for more participants, with
159Co-location services have de facto substituted the role played by the floor trading in the
past. Instead of bringing together specialists and investment companies traders on the floor
to get information and execute orders faster, todays main equity markets put together in
the same room, close to the central data server the servers of trading specialists, brokers,
dealers, market-makers, and investment firms.
160To which apply also rules on regulatory capital, Capital Requirements Directive,
Directive 2006/49/EC (Art. 12, MiFID).
their own technology, without intermediaries. In both DMA and SA, the
intermediary is responsible for the customers conduct161 and transactions
are flagged with the ID code of the intermediary. Moreover, the
intermediary monitors customers activities (and eventually reports
anomalies to authorities; FSA, 2008) and typically applies specific
requirements to get DEA to the trading platform. In addition, trading
venues may unilaterally decide to restrict access only to certain type of
customer. DMA permits the intermediary to apply pre-trade filters and
post-trade controls, since orders are directed to markets through
intermediaries infrastructure. The proportion of trading value done
through Direct Market Access is expected to grow up to 15% by 2011
(Celent, 2008).
For SA services, intermediaries typically receive a drop copy of the
order at the same time when it is sent to the market (AFM, 2010). This
situation makes it difficult to apply pre-trade controls, even though it
allows faster execution because only pre-trade filters set by the trading
venue are actually applied. Pre-trade filters are particularly important to
detect erroneous trades or human errors (fat fingers), rather than for
detecting market abuses. Ex ante and ex post controls are applied to SA
services by the trading venue or by third parties on behalf of the
intermediary as well. SA is a typical service for a small number of highly
sophisticated intermediaries and their clients. IOSCO (2010b) suggests
applying:
1) Specific requirements to DEA customers, such as limits to
transactions notional value,
2) Pre-trade controls and
3) Post-trade controls.
In the case of Direct Access (DA), firms that are not investment firms
under MiFID may be granted membership to the trading platform. In such
cases, they would apply both their own internal controls and the controls of
the platform by executing orders directly on the trading venue, using their
own ID code. As a consequence, in order to avoid credit risk, members
should apply sound financial requirements and become clearing members
of the CCP for the purpose of balancing their trading activities with
161 Even though IOSCO (2009a and 2010b) recognises that it may be difficult in certain
jurisdictions to act against a non-member for violation of market rules or it may be difficult
to show an intermediarys lack of supervision.
potentially high notional values (IOSCO, 2010b). The European Parliament

(Swinburne, 2010) has recommended a clamp-down with an explicit ban
on DEA naked arrangements, which grant market access without filters
to intermediaries and trading venues (unfiltered market access). This
Report does not find evidence that such agreements have been put in place
by any European trading venue.
Overall, DEA arrangements create specific challenges:
1) Risks of market manipulation and insider dealing if only traditional
controls apply, for instance, by offering much smaller and not
harmonised tick sizes that can make supervision more complicated,
especially if HFT is involved;
2) Risks of creating an unlevel playing field (unfairness) by allocating
limited physical space on a discriminatory basis, as for co-location
services, or because of the geographical position;
3) Risks that the limited capacity of trading venues to receive/send
messages can restrict access in a discriminatory way; in this regard,
the adoption of incremental fees for messaging may be a less invasive
way to keep control of the platforms capacity constraints;
4) Risks that human errors or erroneous trades may happen more
frequently, with devastating effects on investors confidence; and
5) Risks that members may not be sufficiently capitalised in case they
would be held responsible for their actions (credit risk; in particular
with DA).
In order to face these challenges, the Commission (2010b) is
consulting market participants on the possibility to amend MiFID as
follows:
1) Require HFT over a specific threshold to be authorised as investment
firms,
2) Impose new general organisational requirements and non-
discriminatory clauses,
3) Set a minimum tick size,
4) Impose an obligation on HTFs trading a significant number of shares
to keep providing liquidity on an ongoing basis (as for market-
makers) and
5) Require to keep (limit) orders on the order book for a limited amount
of time before cancellation or, alternatively, requiring firms to keep
their cancellation over execution ratio over a specified level.
On points 1 and 4, in particular, it should be noted that HFT is

difficult to define (CESR, 2010b, p. 40)162 and if defined it would not be
easily identifiable on the order book. It seems that the Commission refers to
HFT where it pursues market-making strategies, since large positions may
boost the credit risk of HFT firms. Notwithstanding that credit risk is a
potential issue for all traders accessing financial markets with insufficient
financial resources, it is rather the way in which traders access financial
markets (in which HFT can be more easily identified) that needs to be
further scrutinised. IOSCO (2010b) suggested setting minimum standards
for DEA customers, which include appropriate financial resources and
appropriate procedures to assess customers knowledge and technical
proficiency, or setting position limit filters that avoid exceeding credit
limits and gaining exposure to unacceptable risks (p. 22). Hence, since
HFT uses DEA arrangements (trading or not on own account), by
regulating the three ways how traders access markets directly with no
direct intermediaries involvement (but partial responsibility) would be
an easier-to-handle and more functional approach (for instance, by
requiring of the DA that their members are investment firms and/or
clearing members).
Moreover, point 4 may raise problems of incentives for HFTs to
provide liquidity, if they are obliged to do so on an ongoing basis. As
illustrated above, HFT supplies and consumes liquidity in different
circumstances, and since they are not allowed to offer spreads, it would
reduce their incentives to trade high volumes if market-making obligations
are imposed upon them.
On point 5, as mentioned, incremental fees based on sent/received
messages may be a less invasive way and allow trading venues to fine-tune
fees with the actual capacity of their trading platform. An organisational
requirement to deal with capacity constraints may stimulate these changes.
Disorderly markets. The abovementioned risks (in particular,
operational and credit risks) may cause disorderly markets and create
financial instability if no appropriate regulation and supervision has been
put in place. The Commission (2010b) suggests requiring trading venues to
set appropriate harmonised procedures to halt trading to mitigate the risk
162It is worth noting how the regulatory approach is changing in comparison to the current
MiFID text, in which regulation imposes obligations on market participants as a result of
being the provider of a specific service rather than being identified with a specific status
(such as, high-frequency trader).
of errors (e.g. circuit breakers; Swinburne, 2010). Harmonised regulatory

obligations to ensure fair and orderly trading and to monitor members
compliance with rules (e.g. adequate risk management systems; FSA,
2008) may need to be applied across markets, since otherwise it would
create great instability (order imbalances). Some market participants also
argue that those measures should be reserved for critical situations, while
solutions that allow trading continuity should be by far preferred. CESR
has been requested by the level 1 mandate of the new MiFID to define
binding technical standards to ensure fair and orderly markets, including
DEA arrangements (CESR, 2010b).
In case markets experience outages/crashes, the FSA (2010b) has
suggested fairly sound requirements, as follows:
1) Maintaining fair and orderly markets (by suspending trading
activities if the orderly functioning is under threat and requiring
sound arrangements to deal with technical operations);
2) Informing members/participants of trading conditions (e.g. by
disseminating non-misleading information) and setting order
management procedures (e.g. by cancelling orders already in the
system in case of outages or cancelling transactions that come as
result of erroneous matching due to price anomalies in market
crashes);
3) Restarting trading with transparent and non-discretionary rules and
procedures to avoid affecting fair and orderly trading on related
markets (members should be clearly informed about the status of
their outstanding orders, sound and transparent rules should apply);
4) Setting contingency arrangements for trading venues using the
reference price of the market where trading has been suspended (also
for asset classes other than equity markets). Extending these
arrangements to investment firms in order to ensure continuity in
meeting best execution obligations in the absence of a reference price
(it may be desirable for them to have connectivity to multiple
venues).
When the failure does not only pertain to a specific problem of a
venue, under such market circumstances, any eventual decision by the
competent authority or corresponding body to halt trading in any given
security would need to be enforced among all trading venues. In addition,
trading venues may need to periodically assess their trading capacity and
test their systems accordingly (stress tests). Other important controls
should be falling on trading venues: approval of participants; sufficient

capacity and resilience to cope with high volumes; real-time market
supervision; market volatility controls (trading suspensions, etc.);
erroneous order cancellation (trade bust policies); and market abuse
monitoring/surveillance.
Box 14. Regulatory shift: the growing role of macro regulators163

The link between the international regulatory activities around issues of financial
stability has brought significant implications for the revision of the MiFID and the
work of securities regulators. Regulatory activities face three major challenges:
i) Global economic vulnerability affects sovereign tolerance for volatility in
financial markets;
ii) Disagreement on economic policies will increase the need for regulation that
delivers financial stability (macro-prudential regulation); and
iii) A possible trade-off between new trading technologies and market stability.
Securities regulators are not directly represented at G20 level and are more
frequently compelled to comply with decisions made somewhere else. In effect, the
centre of regulation has shifted from national technical issues to G20 level (with
IMF playing a major role; see graph below). In terms of enforcement, however, the
centre of gravity is still at national level, even though Europe is trying to increase
the enforcement powers of the new European authorities. These different
approaches may create some clashes between new financial regulation (which
represents the current priority at global and national level) and its concrete
application at national level. Benefits and costs of macro-prudential financial
regulation may need to be further investigated.
For instance, exogenous factors brought about by the global regulatory
framework are influencing the discussion around the functions performed by some
market operators (such as HFTs and market-makers), i.e. the impact of HFT on
volatility and liquidity, or the role of market-makers in concentrating risks and
interconnections. New regulatory tools coming from the top are actually at
centre of the discussion. For instance, concerning HFT, the introduction of circuit
breakers or stricter financial requirements. For market-makers and financial
institutions, exposed to certain macro risks in general, the introduction of trade
repositories, margin/capital requirements and liquidity requirements are
proposed.
163This section builds upon Barbara Matthewss presentation at the Task Force meeting on
15 October 2010 (Matthews, 2010).
Figure 30. Shifting the regulatory centre of gravity.

Before 2008 After 2008
National G20?
Informal IMF?
Intl
Informal
EU Intl
EU +
national
Source: Matthews (2010).

Moreover, several aspects may be deemed to impact volatility and so push
regulators to rein in new areas of financial markets, with a potential impact on
MiFID. These variables are:
1) Exchange rates (and FX);
2) Anaemic growth;
3) Persistent unemployment;
4) Latent or lurking inflation;
5) Deflation;
6) Budget deficits; and
7) Unsustainable sovereign debt burdens.
If new technologies are perceived as a determinant of volatility, they will be
subject to stricter regulation. Regulators attention is gradually turning towards
trading networks and the implications of regulatory actions on such network
infrastructures as financial markets.
In conclusion, another important aspect is data access, in terms of best execution
and regulatory/supervisory actions (price, cost, speed, counterparty exposures,
etc). Regulation may need to define proper mechanisms to launch early warnings
and to identify systemic risks. Indirect effects of actions in parallel segments of
financial markets due to their interconnectedness can increase the level of
uncertainty in the market and bring tough regulatory responses.
Conclusion # 14
The systemic importance of modern capital markets highlights the inner tensions among
financial stability, market efficiency and technological innovation. A well-functioning
market must balance efficiency and safety, since inefficiencies will sooner rather than later
raise problems of safety and vice versa. Technology plays a central role in the configuration
of market infrastructure. In short, speed and volumes will likely continue to grow but
trading venues may have to deal with more frequent crises and outages.
Overall, technological innovation, combined with new trading techniques, has brought
revolutionary changes to trading platforms. Among these changes, there are major benefits,
such as better order management and control of market impact; or more efficient and faster
feed of information into prices, which generate diffused gains in terms of lower spreads and
better price discovery.
However, modern trading also presents a number of challenges, such as an increase
in fundamental market volatility, which in turn has brought speed and volumes to critical
levels. Advanced execution services like direct-market or sponsored access have radically
increased speed and volumes for transactions, in an attempt to cope with increasing
volatility. Yet, limits to infrastructure capacity mean that higher speed and volumes risk
generating market disorder and financial instability. To overcome these challenges,
intermediaries and trading venues need to strengthen their own monitoring. A coherent set
of emergency procedures in case of market disruptions should be designed in consultation
with market participants (e.g. circuit breakers). There are a several efficient monitoring
systems already in place, which could serve as model systems. Finally, trading rules should
be harmonised across markets to avoid instability arising from arbitrage.
5.6 Access to market infrastructures for equities: Whats next?

Market access. Competition among network infrastructures needs a certain
degree of market contestability in order to bring costs down for final users
and avoid oligopolistic settings. Current development of competition has
led to a dramatic cost reduction in trading and post-trading costs, even
without interoperability. Greater and cheaper market accessibility is a key
objective of MiFID, to be achieved through competition and greater
contestability. MiFID grants to investors/intermediaries non-discretionary
market access to trading venues (recital 6), to investment firms freedom of
access to regulated markets in other than their country of origin (Art. 33),
and most notably it gives freedom to investment firms to decide the
infrastructure for clearing and settlement (Art. 34). In addition, member
states need to ensure incumbent infrastructures grant access to newcomers
unless a legitimate commercial ground makes the link unavailable. In this
regard, the definition of legitimate seems to leave some space for
ambiguity. Incumbents typically allege one of the following three reasons
to refuse access:
i) Efficiency and cost reduction,
ii) Not meeting specific requirements or standards set by the
infrastructure itself or
iii) Technical difficulties.
Efficiency and costs reduction has been frequently dismissed by the

Commission, which wishes to avoid high market concentrations, even if it
comes at cost (see GE/Honeywell)164. As to technical reasons, these are
usually assessed case-by-case by looking at available technologies and
standards. With regard to the accessibility requirements set by the
infrastructure, these should be clearly defined and applied on a non-
discriminatory basis (see Code of Conduct, 2006, 25-33)165.
Market architecture. The opening up of competition in the upstream
market (trading) has also generated some beneficial effects in terms of
market accessibility for end investors. However, in the post-trading
segment accessibility remains somehow limited, also affecting the overall
costs of trading. Looking only at costs however, does not reflect the
redistributive effects that a different market structure can generate. As a
result, it is unclear how current markets can balance efficiency and
competition with market integrity and stability. Excluding any ideal world
of perfect competition, there are three possible scenarios:
1) Open architecture
2) Closed architecture and
3) Semi-open (or semi-closed) architecture.
The first scenario would require unleashing competition at every
stage of the post-trading architecture, imposing painful unbundling
obligations (disregarding potential economies of scope) and
interoperability arrangements. The second scenario entails the creation of a
closed architecture, whereby market players and regulators work for
further integration disregarding potential economies of scope. This market
design tends to maximise economies of scale only and ignores beneficial
portfolio effects and interoperability in favour of quasi-monopolies or
monopolies at various stages of the post-trading value chain. Finally, there
is an intermediate formula that does not seek to impose a specific market
design but rather to interpret market trends and reap the benefits of scale
and scope economies. For instance, this approach does not judge bundling
164 See Case No. COMP/M2220 General Electric/Honeywell, Regulation (EEC) No. 4064/89
Merger Procedure, Article 8(3), (http://ec.europa.eu/comm/competition/mergers/

cases/decisions/m2220_en.pdf).
165See Code of Conduct for Clearing and Settlement, (http://ec.europa.eu/internal_
market/financial-markets/docs/code/code_en.pdf).
or mergers as harmful as long as they generate important economies of

scope (efficiency defence), are replicable (together with the infrastructure),
and do not create a big concentration in the market. In effect, according to
this approach, the post-trading infrastructure should be evaluated under
the essential facility doctrine,166 i.e. the infrastructure should not be forced
to open the access to its infrastructure as long as it does not meet all the
following requirements:
1) The facility controlled by one market player is the only way to access
the relevant market.
2) The facility is not replicable at a reasonable cost.
3) The monopolist refuses access to the facility or offers it at unbearable
costs or under unfair conditions.
4) The absence of a valid economic justification.
On the other hand, this third approach would minimise barriers to
entry and exit by promoting interoperability agreements between
incumbent infrastructures and newcomers (in line with Art. 34, MiFID and
forthcoming European Market Infrastructure Regulation), and minimising
the presence of monopolies at any stage of the post-trading architecture, if
no natural monopoly conditions are verified.
Code of Conduct. To favour a more open post-trading architecture and
to promote a full application of the principle originally set in MiFID (Art.
34), in November 2006, the European Commission endorsed the Code of
Conduct for Clearing and Settlement, which was signed by the Federation
of European Securities Exchanges (FESE), the European Association of
Clearing Houses (EACH), and the European Central Securities Depositories
Association (ECSDA). The Code represents the last self-regulatory effort by
the industry to meet high-level objectives, before the financial crisis
promoted a more interventionist approach through regulation. In effect, the
Code was a valuable explorative work to achieve three goals:
1) Transparency of prices and services,
166 The essential facility doctrine is a theory that was originally elaborated in the US
antitrust case law; see, U.S. v. Terminal Railroad Association, 244 U.S. 383, 1912. An extension
of this doctrine was elaborated in Europe too in several cases, in particular, European
Commission, Magill TV guide/ITP, BBC and RTE, in GUCE 1989, L 78/43; Court of Justice,
Oscar Bronner GmbH & Co. KG vs Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG,
Mediaprint Zeitungsvertriebsgesellschaft mbH & Co. KG and Mediaprint Anzeigengesellschaft mbH
& Co. KG, C-7/97, p. I-7791. Amongst the most important literature, see in general Areeda
(1990).
2) Access and interoperability and

3) Unbundling of services and accounting separation.
After three years, the Commission (EU COM, 2009b) made an
assessment of the progress achieved so far, which highlighted how the
Code of Conduct had positively contributed to price transparency.
However, conflicting evidence was found with regard to access and
interoperability and unbundling of services and accounting separation
(Lannoo and Valiante, 2009). In effect, despite the Code promoting basic
principles (such as right of standard access and reciprocity) and some
additional competition between infrastructures, commercial and technical
barriers to cross-border access to infrastructures remain fairly high (EU
COM, 2009b), in addition to other legal, economic, fiscal and social barriers
that still impede the full development of a European post-trading
architecture in particular for settlement services (Giovannini Group, 2001,
2003; CESAME, 2008). More competitive clearing services, however, have
been effectively delivered in some markets, including silo market models
(e.g. Swiss equity market), even though the situation seems to have stalled
for interoperability agreements in other markets (see figure below).
Figure 31. Equity landscape before and after MiFID and the Code of Conduct
Before...
LSE NYSE SWX DB BME Nasdaq

Trading
BIT Euronext OMX
2
Clearing
SIS Eurex
CC&G LCH
xclear
LinkUpMarkets
NCSD VPSS
Settlement
SIS ISD EECSD

M.Tit. Euroclear Clearstream Iberclear
Sega
LTCSD LVCSD
...and after the Code
NYSE NYX LSE SIX BB Nasdaq Nasdaq

Trading
Euronext Arca Turquoise BIT Group Equiduct Chix OMX2 OMX BATS DB BME
Europe
3 2 3 2 2 2 2 3 2 2 2
Codeofconduct INTEROPERABILITYAGREEMENTS
EuroCCP EMCF
Clearing
Eurex
LCH1 SISxclear
Citi
LinkUpMarkets
Settlement
Euroclear
SE/FI
VPSS
M. SIS ISD EECSD

Euroclear Clearstream Iberclear
Tit. Sega
LTCSD LVCSD
EuropeanCSDs
Source: FESE.
Note (from the presentation): In green the interoperability agreements in place and in yellow
those on hold due to review by the regulators or to discussions between the parties. For a
complete overview of the status of the requested links between post-trading infrastructures,
As shown above, after the Code came into force, many requests for
interoperability and access were sent to incumbent infrastructures but only
two were actually put in place. In addition, MiFID also stimulated the entry
of new competing infrastructures in the clearing space.
Despite conflicting findings, costs of trading and clearing services
went down by 60% between 2007 and 2009, while costs of settlement
services seem to have gone down only in some markets, also due to the
competitive pressures coming from the upstream trading market thanks to
the introduction of MiFID (Oxera, 2009; EU COM, 2009b, p. 3). The
fragmentation of European financial markets infrastructure, in particular
for equities, is still led by geographical aspects (i.e. national markets; see
figures above).
In the aftermath of to the financial crisis, risk reduction seems to
prevail over cost reduction, which was a main objective of the Code. As a
result, interoperability agreements are being more deeply scrutinised for
potential threats to infrastructure integrity, while important guidelines are
still in the implementation phase (Sections 3.5.3 and 3.8.3, FESE et al., 2007).
In particular, the crisis drew attention to potential risks, such as
competitive pressures that could lead to a race-to-the-bottom in terms of

risk management practices. Moreover, asymmetry in collateralisation
agreements should be the object of further examination and sound
internationally-agreed standards.
The whole post-trading sector will be eventually reshuffled by an
array of new European regulations. It is essential that EU co-legislators
ensure strong coherence among these legal texts by giving to each piece of
legislation a specific role, minimising conflicts and promoting legal
certainty.
Conclusion # 15
Academic literature has consistently suggested that market infrastructures, particularly for
equity markets, are network structures that should operate in a competitive environment.
MiFID should therefore take action to keep barriers of entry and exit low, but with due
attention to economies of scope and potential efficiencies. The Code of Conduct succeeded in
improving price transparency and stimulated interoperability. More remains to be done,
however, to solve existent commercial and technical challenges in terms of access,
interoperability and unbundling. Ultimately, MiFID will favour freedom of access by
investment firms to competing market infrastructures. In effect, while the original MiFID
Directive (Article 34) envisaged a level playing field in terms of non-discriminatory access
to competing infrastructures, the transposition of this provision into national law and its
enforcement has been inconsistent across EU member states. As such, greater efforts need
to be made to ensure consistency in the enforcement of the regulatory framework. For this
purpose, not only the revision of MiFID will play a role, but also concurrent legislations
such as European Market Infrastructure Regulation (EMIR), Securities Law Directive
(SLD) and Central Securities Depository Regulation (CSDR).
6. PROVISION OF INVESTMENT SERVICES
MiFID represents an important attempt to design a harmonised framework

of rules for the provision of investment service in the European Economic
Area (EEA). It adopts a functional approach by regulating market
participants according to the actual investment service167 they provide.
Within its scope, the Directive therefore applies to all market participants,
with some important exemptions.168
In parallel with a more competitive environment for trading venues,
MiFID sought to harmonise and strengthen the level of investor protection
and market integrity through two sets of rules:
1) Organisational requirements (ORs) and
2) Conduct of Business (CoB) rules.
The former aims at strengthening market integrity (i.e., fighting
market abuses and attempts to manipulate markets) and improve
resilience. The latter targets greater investor protection and market
efficiency. These rules consist of fairly detailed provisions, which should be
read in conjunction with the general obligation for investment firms to act
honestly, fairly and professionally in the best interest of clients (Art. 19.1,
MiFID), and the rules on trade and transaction reporting (see Section 1).
167The definition of investment services and ancillary services can be found in Annex I, Sec.
A and B, MiFID. The latter are typically provided in combination with the former. However,
no authorisation to operate as MiFID investment firm should be granted if only ancillary
services are provided (Art. 6.1, MiFID). Ancillary services can only be provided in
combination with investment services to benefit from the European passport. The definition
of investment service under MiFID should be read in combination with the exemptions set
by Art. 2.1.
168See, in particular Art. 2.1, MiFID. Amongst others, MiFID most notably set exemptions
for insurance companies, energy firms, and UCITS and pension funds (Art. 2.1(a)(h)(i)(k)).
164 |
The overarching objectives of investor protection, market integrity, and

market resilience also operate as high-level principles (informing the
legislation and serving as guidelines for its interpretation). For instance, in
order to increase investor protection, investment advice, (an ancillary
service under the ISD), became a core investment service under MiFID,
subject to the authorisation of the competent financial authority. This
means that even market players providing only this service169 can compete
at European level and need to comply with MiFID.
6.1 The nature of investment services and investors

Nature of investment services. The exchange of financial instruments and
more generally the search for the best allocation of resources involves a
wide array of (financial) services, which typically fall under MiFID, either
as core investment or ancillary services. The nature of services
sometimes raises economic and regulatory issues, particularly whenever
the quality of the service is observable only after its use or is not verifiable
at all. The former situation is an example of experience goods, and the
latter is a case of credence goods.170 Investment services, besides those
activities that are not offered to third parties (e.g. dealing on own
account), are made up of aspects from both experience and credence
goods.171 Most notably, investment services may be affected by:
1) Strong information asymmetries between providers and customers
(including limited financial education),
2) High switching costs due to sunk investments (lock-in effects),
169Generic advices are not included in the definition, see Recital 81, Implementing
Directive.
170 Products and services can be classified in three categories: search goods; experience
goods; and credence goods. A search good consists of a product or service for which it is
possible to assess the quality before the purchase. Search elements include those attributes
of the relationship that are easily detected and understood by customers. An experience
good, however, is a product or service for which the buyer can evaluate the quality only
after the purchase and its use. Finally, a credence good is a product or service whose value
and quality cannot be assessed even after its use, as features cannot be easily compared with
other products or services. See Nelson (1970).
171For a general discussion of investor protection and competition policy issues in financial
services, see Renda & Valiante (2010).
166 | PROVISION OF INVESTMENT SERVICES
3) Cognitive biases in decision-making and judgment that affect

investors and
4) Difficulties for the customer to shop around.
Asymmetric information. Firstly, two particular aspects generate
asymmetric information between investment firms and their clients: i)
limited financial education and rationality; and ii) private information. The
former is related to the type investor. In effect, retail investors have
typically less knowledge of investment products and financial markets than
professional and institutional ones, who usually own the necessary
resources to reduce any informational gap vis--vis investment firms and
can leverage their contractual power. However, all investors can behave
irrationally and increase their informational gap (see below). The latter
cause of asymmetric information comes from the investment firm that
finally owns private information as a result of its investments for the
provision of investment services. For instance, a dealer knows better than
its clients how order execution works, precisely because order execution is
a core part of its business.
Switching costs. Asymmetric information between providers and
customers may spark other market effects, which may include transaction-
specific investments (or sunk costs).172 For instance, for a retail investor, the
relationship between the provider and the customer is often based on trust
given the overwhelming difference in knowledge and access to
information. This situation generates sunk and switching costs, since
when a customer changes provider it loses the transaction-specific
cost/investment, and faces an additional, equally sunk cost to familiarise
with the new provider. This effect is particularly strong for investment
products, where the fiduciary relationship with the service provider may
be a fundamental transaction-specific asset, in particular for retail investors
(see Llewellyn, 1995; next sections). Transaction-specific costs/investments
may amount to switching costs, which lock in consumers and ultimately
give the investment firm the opportunity to unilaterally increase its prices
over its rivals since customers would de facto be locked in. Under these
circumstances, competing investment firms would have to offer discounts
that offset switching costs in order to attract rivals customers. For retail
172Sunk costs are irreversible costs paid once to produce or to consume a specific service or
product. They represent a barrier to entry for producers and a barrier to switch for
consumers. Sunk costs linked to investment services are usually higher for retail investors,
who cannot benefit from scale economies.
investors, in particular, the following costs impede switching providers

(Klemperer, 1995; OFT, 2008; Renda & Valiante, 2010):
Transaction costs, e.g. documentation, time, fees, search costs, other
information costs, learning costs, etc.;
Exit costs, due to, e.g. loyalty programmes, etc.;
Uncertainty costs, since the quality or suitability of a product can only
be observable after purchase (experience attributes), or can never be
fully observed by the customer (credence attributes) and
Psychological costs, mainly in the case of goods with a high significant
credence attributes (mutual trust), such as investment advice.
In particular, experience and credence attributes represent a
transaction-specific investment for customers that make switching very
difficult (so-called status quo bias). Retail investors may potentially suffer all
the above costs, while wholesale investors typically encounter only
transaction costs and an opportunity costs when they switch provider. An
example of opportunity cost may be the risk of suffering market impact in
order execution, which can ultimately restrain institutional investors from
changing provider.
Another aspect that may influence the provision of investment
services, for retail investors in particular, are so-called cognitive biases.
These are investor behaviours that violate the assumption of rationality
and standard economic principles. In particular, investors (Posner, 1998;
Jolls et al., 2000):
Do not necessarily maximise their utility in every circumstance,
Do not often have stable preferences (e.g. sunk costs and initial
allocation of entitlements may dominate in their decisions) and
Do not have the ability to process or accumulate an optimal amount of
the information available in the market.
The violation of rational assumptions lies on three cognitive limits
(Jolls et al., 2000): i) bounded rationality, ii) bounded willpower and iii)
bounded self-interest. 173
173For an overall analysis of potential cognitive biases resulting from these cognitive limits,
see Renda & Valiante (2010), pp. 51-55. For instance, the overconfidence bias means that
people tend to overestimate (to be overconfident about) the probability of an outcome if an
example of the event has recently occurred (linked to the prospect theory and the precedent
behaviour). Therefore, consumers are generally overconfident in their abilities and in their
Bounded rationality (Simon, 1957) refers to limits faced by human

beings in terms of accessible information, mental capacity and available
time. It should be distinguished from rational ignorance.174 In addition,
individuals are often path-dependent in their choices, especially when
they cannot fully appraise the value of something they already possess
(Korobkin & Ulen, 2000; Sunstein, 2000; Jolls, 2007). As to bounded
willpower, it leads people to act in conflict with their long-term interests,
even though they anticipate pervasive effects in so doing (e.g. smokers).
Finally, bounded self-interest may push people to care about treating
others fairly because they want to be treated in the same way. Agents will
act nicer or nastier depending on how the other party treats them. All
these cognitive limits also occur with the provision of services not related
to investment. However, the experience and credence attributes present in
investment services can magnify the effects of cognitive limits on judgment
and decision-making processes. As result, MiFID has laid out a regulatory
scheme that offers greater protection to final investors with particular
attention to retail investors, who suffer more from experience and credence
attributes.
Moreover, due to the abovementioned costs, only an insufficient
number of marginal consumers is actually able to shop around, thereby
limiting the flow of information that helps to correct market failures. In
effect, by shopping around, informed consumers generate a positive
externality on uninformed ones (as long as they are able to see them) that
may have too little incentive to acquire information (Hynes & Posner 2001).
MiFID Requirements. MiFID brought together all these concerns and
introduced a more paternalistic approach to consumer protection. For
instance, MiFID intervenes on firm-client relationships by setting
obligations such as the good faith clause (Art. 19.1) or best execution
duties (Art. 19 and 21, MiFID). In effect, these duties try to palliate two
phenomena present in firm-client relationships:
1) A lack of effective monitoring (moral hazard) and
2) A lack of ability to assess the quality of the service (adverse selection).
future fortunes. For example, many people invest, believing that they can beat the stock
market, or they underestimate the risk that illness or unemployment may cause difficulty in
repaying a loan (p. 54).
It represents the situation in which the investor is not able to or finds it inconvenient to
174
put effort in understanding a clear set of information in a specific amount of time.

The former has broad implications for how services are offered and
disclosed to clients. MiFID organisational requirements play a primary role
in containing adverse effects by promoting a more effective monitoring and
discouraging moral hazard. Adverse selection, in particular, may drive
good quality services out of business with consequences for the stability of
the market. In this regard, MiFID business conduct rules ensure that
investors are aware and able to understand the risks they take and, ideally,
evaluate prices against actual product quality.
The introduction of stricter pre-trade and post-trade transparency has
also increased the flow of pre-contractual and post-contractual information,
which contributes to reduce adverse selection and moral hazard for end
investors (besides any test of suitability and appropriateness foreseen by
MiFID). It follows that MiFID organisational, conduct of business and
transparency rules have been more consistently applied, with positive
effects also on market structure and its long-term efficiency.
6.2 Fiduciary duties

Contract incompleteness. A contract between two parties for the provision of
a product or service entails a complex web of aspects, which cannot be fully
anticipated. Contracts are therefore incomplete by their very own nature,
no matter the sophistication or evenness of parties contractual power.
The degree of completeness of a contract depends mainly on four
factors (Hermalin et al., 2007, p. 11):
i) Specification costs
ii) Monitoring costs
iii) Enforcement costs
iv) Need for pre-contractual commitment
Firstly, high specification costs make the drawing up of the perfect
contract impossible for several reasons. Most notably, a contract cannot
usually cover all potential contingencies in the provision of a good or
service, especially when deferred in time, as it is impossible to predict all
future events. This universal cost affects every contractual provision of
services or goods.
Secondly, costs of monitoring that the transaction has been executed
may be comparably prohibitive (or even be not observable or verifiable at
all).175 This situation hinders the possibility of devising contracts effectively.

Thirdly, contracts may not be enforced because costs can be too high
in absolute or relative terms (in comparison to the potential benefits).
Enforcement is strictly related to monitoring costs. If a party cannot verify
the execution of a contract, a competent authority or judge may not be
capable of doing so either. In this case, in addition to default rules,176 the
provision of experience and credence goods or services is combined with
strong remedies (e.g. punitive damages) and fiduciary duties (general
clauses). In particular, when verifiability is low, strong remedies may
reduce the agents pre-contractual commitment to act in its own interest.
Last but not least, in order to enter into a contract, parties need to
trust the other partys commitment; otherwise, they may not take on the
risk of going into the agreement. This is particularly true in transactions
where reputational aspects are crucial to the success of the deal, such as
investment advice.
Investment services, as mentioned above, involve an agency
relationship with high specification and monitoring costs, due to their
experience and credence attributes. Moreover, even though investment
services are frequently observable, the scarce ability to verify their
execution (low verifiability) often makes enforcement too costly.
This complexity can have differing degrees of intensity, in relation to
the type of transaction cost or information asymmetry. For instance, retail
investors may feel an investment service is too complex for their level of
knowledge and wholesale investors may fear the economic implications of
complexity on other aspects of their business. For all these reasons, in
addition to greater competition, regulators have pushed for the recognition
of fiduciary duties in the provision of investment services. The concept of
fiduciary duty originates from the Common Law duty of the agent (e.g.
Observability and verifiability are indispensable to define a contractual contingency.

175
However, they may come at high cost, since they include tasks such as investigation,
measurement, documentation and monitoring (Hermalin et al., 2007).
176Default rules are those rules that apply to a legally binding agreement if not decided by
parties otherwise. They are effective in reducing transaction costs setting the ground for a
contractual agreement and they may encourage parties that enjoy an informational
advantage to reveal their type. Default rules have the role to balance the contractual power
between structurally strong and weak counterparties in terms of private information.
However, default rules need to be carefully drafted as they may increase transaction costs if
parties need too often to contract around them because they are or ineffective or too costly.
broker) to be loyal to the principal (e.g. investor). Fiduciary duties are

typically coupled with disgorgement, that is, the restitution of any gain
obtained by the agent as a result of its actions against the principals
interests (Easterbrook & Fischel, 1993). In this way, legislation has
gradually overcome its formalistic approach to the law of contract (the
sanctity of the contract) based on the fiction of arms-length transactions
and equally powerful counterparties. By recognising that parties do not
frequently have the same bargaining power, legislation has gradually
afforded more protection to the weaker party of the contract. In effect, the
party with in theory more contractual power can hardly impose its own
contractual conditions today.
Financial regulation has followed the same trend and increased
protection for final investors, moving from an original duty to read to a
duty to disclose, and now to a duty to behave. In Civil Law countries,177
a preference to deal with fiduciary relationships through bilateral
negotiations (contracts) has often given a crucial role to national legislation
in dealing with these issues. Therefore, by applying Common Law-based
principles such as fiduciary duties, the European legislator has decided to
introduce, in a pre-existing legal system mainly based on Civil Law
principles, concepts that need the underpinning of strong national and
supra-national enforcement mechanisms, which have been conspicuously
absent at European level (Pistor & Xu, 2002, pp. 32-33).
More generally, fiduciary duties apply to situations in which there
are:
1) Information asymmetries (moral hazard and adverse selection)
2) Transaction costs
3) Cognitive biases and
4) Market frictions (limited competition).
Also the literature on marketing portrays trust, satisfaction and
average perceived cost (including switching costs) as key determinants of
177Civil Law countries are those countries in which law is primarily codified or designed
with other written regulations, while the role of the judges is to verify their application with
low space for interpretation. By contrast, Common Law countries are those where law is
developed through court decisions. However, since the last century these two systems have
been getting closer. For instance, a typically common law country such as the US has
recently approved an enormous package of financial reforms (Dodd-Frank bill) that will
need to be implemented primarily through detailed regulation.
customer loyalty (amongst others, Graf et al., 2008).

Additional cognitive biases may affect retail investors behaviours
and make the provision of investment services even less observable and
verifiable, and so de facto even more akin to fiduciary duties. Wholesale
investors, however, typically have the contractual power to modify contract
terms if they are not devised properly, while for retail investors costs are
typically too high. Hence, many fiduciary duties have been formulated
differently for retail and professional investors.
The breadth of these duties is very important, though. On the one
hand, strict fiduciary duties may have two unintended consequences. They
may preclude investors ability to contract around specific terms; or make
pre-contractual commitment too high, generating clients overreliance in
the positive outcome of the transaction. As a consequence, a protective
approach would end up damaging precisely those investors the regulation
seeks to protect. On the other hand, a loose definition can make
enforcement too costly and expose the beneficiary/principal to two types
of agents wrongdoing (Cooter & Freedman, 1991):
i) The fiduciary/agent may misappropriate the principals asset or
some of its value (moral hazard); and
ii) The fiduciary/agent may neglect the management of the asset
(negligence or violation of the duty of care).
In the US, the fiduciary duty is recognised in two kinds of situations:
when the contract explicitly recognises a fiduciary relationship (e.g.
principal-agent); and when specific circumstances surrounding the
transaction and the relationship occur. Trust, confidence and influence
are typically those circumstances that denote the existence of a fiduciary
duty (Frankel, 1983, p. 829). In effect, a fiduciary duty should not be
considered as a moral belief, but legal ground should be found in the
potential sources of failure for the contractual relationship (Easterbrook &
Fischel, 1993). A fiduciary relation is a contractual one, characterised by
unusually high costs of specification and monitoring.
Due to high specification and monitoring costs, MiFID and more
generally EU regulation have gradually recognised fiduciary aspects to the
agency relationship between the provider of financial services and its
customers (in particular, for retail investors). The use of general clauses in
the legal text, such as good faith or fairness, has afforded judges much
needed flexibility, especially in Civil Law countries,178 to formally recognise

fiduciary duties and to devise them in line with the nature of the
counterparty and the relationship. MiFID, in particular, has introduced a
general clause that explains the fiduciary duty between providers and
clients by asking the provider to act in clients interest (Art. 19.1). However,
general clauses should not be seen as vague statements, but have a clear
contractual meaning. For instance, the concept of fairness can be seen as a
deviation from a benchmark transaction that shifts the balance in favour of
the more entitled party (Jolls et al., 2000). There are two general clauses
upon which organisational requirements and conduct-of-business rules are
designed (see figure below):
1) The general requirement to act honestly, fairly and professionally in
accordance with the best interests of its clients (Art. 19.1, MiFID);
and
2) The general requirement to make sure that all information,
including marketing communications, addressed by the investment
firm to clients or potential clients shall be fair, clear and not
misleading (Art. 19.2, MiFID).
The enforcement of fiduciary duties has been typically left to
competent authorities through information coming mostly in via trade and
transaction reporting, plus other substantial supervisory powers (Art. 50,
MiFID) and the possibility to file civil litigations. However, no harmonised
system of sanctions is in place so far in Europe. The Commission (2010b) is
looking into the possibility of granting further powers of investigation to
competent authorities. It is also examining the feasibility of harmonising
the regime for remedies, such as sanctions based on the losses generated by
the infringement, and not on the disgorgement (restitution of profits).
178Civil Law countries are those countries were law is primarily codified or designed with
written regulation, while the role of the judges is to verify their application with low space
for interpretation. By contrast, Common Law countries are those where law takes shape
through court decisions. However, since the last century these two systems are getting
closer. For instance, a typically common law country as the US has recently approved an
enormous package of financial reforms (Dodd-Frank bill) that will need to be implemented
primarily through detailed regulation.
Figure 32. General clauses
Investment Firm - Client
Organisational Requirements
Conduct-of-business rules
Best interests of clients Fair, clear, not misleading

(art. 19.1) information (art.19.2)
General Clauses
Conflicts of interest Information to

Best execution
Suitability & clients
Appropriateness
Source: Authors.
Finally, competition between investment firms may improve the

quality of the provision of investment services without making it necessary
to tighten legislation or monitoring. Competition may generate pre-
contractual incentives to behave in the best interest of investors, but with
unclear effects on the implementation of rules already in place. On the one
hand, in particular, if firms do not act cooperatively, competition may
reduce monitoring costs and promote greater compliance with fiduciary
duties. On the other hand, however, if firms do cooperate, competition may
result in a collusive equilibrium based on a shared soft interpretation of
fiduciary duties across firms, if they are not soundly defined.
6.3 Organisational rules

MiFID organisational requirements (ORs) aim at promoting integrity and
orderly functioning of financial markets (Recital 5), as well as integrity,
competence and soundness among investment firms and entities that
operate regulated markets or MTFs (Recital 2, Implementing Directive).
For instance, ORs improve monitoring and compliance with conflicts of
interest rules and prevention of misselling practices by requiring
disclosure, recordkeeping and other material obligations. The Directive

(Art. 13)179 sets a wide array of organisational requirements, which are
needed in order to gain authorisation but can be slightly aligned with the
logic of the firms relevant markets. This list of requirements has had a
major impact on the internal controls of investment firms and must be
applied also by UCITS management companies if they perform portfolio
management and advisory services.180 Moreover, ORs apply not only to
investment firms but also to official MiFID trading venues (RMs and MTFs,
Art. 14, 26, 39 and 43). However, diverse interpretations by member states
of the proportionate approach set by MiFID (Art. 13.4) have promoted a
fragmented implementation of these requirements across trading venues,
de facto setting the scene for different rules being applied to similar
activities. The review of MiFID should look at the possibility to finally align
material obligations.
Besides a general clause to comply with MiFID rules and procedures
(Art. 13.2, MiFID), the Directive set strict requirements to improve internal
mechanisms of:
1) Compliance
2) Risk management
3) Order handling
4) Transparency to clients
Compliance. Firstly, the Directive set organisational requirements to
ensure the creation of adequate policies and compliance with MiFID rules
through:
A general duty to establish adequate policy and procedures and
comply with MiFID rules (Art. 13.2 MiFID),181
The appointment of a compliance officer (objective and
independent),182
Obligations for tied agents (Art. 23, MiFID) and
Recordkeeping (13.6, 25.2, MiFID).183
179 Plus Art. 5-25, Implementing Directive and Art. 7-8, Implementing Regulation.
180 See Casey & Lannoo (2009), pp. 141-143, footnote 3.
183 Art. 13.6, MiFID; Art. 16, 17, 51, Implementing Directive; and Art. 7-8, Implementing
Most notably, the investment firm is required to appoint a

compliance officer that should be able to independently ensure compliance,
with enough powers and resources to solicit the firm to correct any policies
or procedures that do not comply with MiFID obligations. The Directive
also requires firms to keep records of all services and transactions
(including transaction reports) taking into account proportionality with
regard to the type of business and the range of investment services and
activities performed. Investment firms then must retain all required records
for a period of 5 years or longer if requested. Records that refer to
obligations and rights for the investment firm and its clients shall be
retained for a period that is at least equal to the duration of the relationship
with the client (Art. 51.1, Impl. Dir.). Records shall be finally stored in data
management systems and be easily accessible for future reference by
competent authorities (Art. 51.2, Impl. Dir.). CESR (2010c, p.10) has
remarked the importance of record keeping obligations in four areas:
i) Reception and transmission of orders
ii) Execution of orders on behalf of the client
iii) Dealing on own account
iv) Portfolio management
For some group of consumers, the investment firm should keep
records of the investment advice as well. However, CESR noted that, since
investment advice services are usually offered face-to-face, setting
recordkeeping obligations would be costly and very difficult to implement.
Other ways to improve quality of investment advice should be taken into
consideration (see Section 6.3.1).
Risk management. Secondly, ORs define rules concerning the firms
risk management system. In particular, the Directive sets:
A system of internal controls (e.g. independent internal audit, IT, etc.
(Art. 13.4-13.5,),184
The possibility to outsource non-core operational services (Art.
13.5),185
A client order handling system (Art. 22.1 and 22.2)186 and
Regulation.
184 Art. 5, 7, and 8 Implementing Directive.
185 Art. 13-14, Implementing Directive.
186 Art. 47, 48 and 49, Implementing Directive. Art. 31-32, Implementing Regulation.
The safeguard and administration of clients financial assets (Art.

13.7, 13.8).187
This framework of rules aims at improving internal risk management
and ensuring continuity in the provision of investment services.
Furthermore, MiFID attempts to improve the control of counterparty risk,
as well as economic, financial and operational risks. In particular, the
Directive forbids excessive risk-taking actions by using clients assets to
deal on own account if the client has not explicitly expressed his/her
consent. These rules have not prevented the recent financial crisis, but they
presumably played a crucial role in limiting the negative effects of cases in
which the broker/dealer also offered custodian services and made use of
clients assets to deal on own account or did not segregate assets with third
parties, putting additional credit risk on assets that should have been
protected from bankruptcy procedures (e.g. the Lehman Brothers case).
However, asset segregation with third parties is currently foreseen by
MiFID only for clients that are not credit institutions. Clients can eventually
oppose the decision to segregate the assets but only with regard to a
specific entity, however.
In addition, MiFID (Recital 27) exempts investment firms from
applying client asset protection rules when full ownership of funds and
financial instruments has been transferred to cover any client obligation.
The Commission is looking into this matter and may decide to remove this
possibility (at least for retail clients assets), which was a main issue in a
recent bankruptcy procedure (Lehman Brothers; EU COM, 2010b, p. 70).188
Investment firms would also be required to improve disclosure on asset
segregation and diversify the placement of client funds using different
third parties.
Transparency to clients. Finally, MiFID has set organisational
requirements, which amounts to a new discipline for information
disclosure to clients, with the particular role of containing conflicts of
interest.
More specifically, the Directive defines:
187 Art. 19, 20, 42, 43, 48, and 49, Implementing Directive.
188The Commission proposes to give an option to member states to exclude title transfer
collateral arrangements in case of professional clients and eligible counterparties.
Rules on personal transactions,189

Conflict of interests rules and inducements disclosure (Art. 13.3, 18.1,
and 18.2)190 and
Information to clients rules (Art. 19.2, 19.3 and 19.8).191
Disclosure to clients should protect clients interests and information
should be fair, clear and not misleading (Art. 19.2). Most notably, MiFID
organisational requirements introduced new rules to prevent investment
firms from acting in conflict with their clients interests by identifying,
managing and, if they cannot be eliminated, disclosing these conflicts to
prospective clients before approaching investors to offer a transaction. In
order to comply with the general clause on acting honestly, fairly and
professionally in accordance with the best interest of a client (Art. 19.1),
the investment firm should disclose fees, commissions and non-monetary
benefits (inducements) that are neither directly generated by the
investment service provided to clients nor do they help to increase the
quality of the service (Recital 39, Impl. Dir.). This rule aims at limiting
undue influence by interest groups on the provision of services to clients.
Furthermore, the legal text should be consistent with the forthcoming
legislative proposal on Packaged Retail Investment Products (PRIPs),192
which will define the terms of pre-contractual information. The
Commission also proposed to increase the level of information to be
disclosed to clients when dealing with complex financial instruments (EU
COM 2010b, p. 58). Information about risks/gains ex ante and ongoing
quarterly reports on the instruments and their underlying should be
disclosed to clients.
New proposals. The European Commission (2010b) is reviewing some
of the organisational requirements discussed above. In particular,
concerning inducements, it is currently foreseen that if the amount of
inducements cannot be clearly determined, the method of calculation
should be disclosed (Art. 26 (b)(i), Impl. Dir.; CESR, 2010i). As MiFID does
not define the requirements to be disclosed in the form of a methodology,
in line with the Commissions view, this has created difficulties to quantify
or provide a tool to quantify the level of inducements. In this respect, the
189 Art. 11-12, Implementing Directive.

190 Art. 13.3 and Art. 18, MiFID; and Art. 12, 21, 22, 23, and 26 Implementing Directive.
191 Art. 24 and 27, Implementing Directive.
192 European Commission Communication, COM(2009) 204 final.
Commission has proposed to introduce:

1) Ex-post reporting obligations,
2) More details and templates for disclosure to clients,
3) Other requirements to be assessed by supervisors to define if
inducements enhance the quality of the service,
4) A ban on inducements provided by a third party in case of portfolio
management services and
5) A ban on inducements provided by a third party for intermediary
providing independent advice.
Inducements may not only distort incentives to provide an
investment service in the best interest of clients, but they may also be used
to give compensation to brokers that may potentially distort incentives.
The Commission (2010b, p. 65) also proposes to clarify and modify
rules on activities concerning dealing on own account.193 Most
importantly, orders executed using matched principal (so-called back to
back transactions194) would be considered as dealing on own account, but
they would not fall under the Capital Adequacy Directive (Dir.
2006/49/EC). They would not fall either under MiFID, unless the
transaction was tagged as market-making or was executed outside
organised MiFID venues on an organised, frequent and systematic basis
(Art. 2.1(d)).
Moreover, the Commission (2010b, pp. 66-72) proposes to strengthen
authorisation and organisational requirements in general. In particular,
members of the board of directors would pass a fit and proper test, which
would evaluate their time commitment and other aspects relevant for the
exercise of their functions. The Commission also proposes to promote the
inclusion in the board of non-executive directors with supervisory
functions.
193Dealing on own account means trading against proprietary capital resulting in the
conclusion of transactions in one or more financial instruments, Art. 4.1(6), MiFID.
194 Back-to-back transactions consist of a chain of securities transactions among multiple
counterparties (typically investment firms) involving the purchase and sale of a security, for
settlement on a single date. It may refer, for instance, to the case in which the investment
firm C buys a security from investment firm A and, at the same time, it sells the same
security to investor B during the same day. These are transactions perfectly matched with
very limited risk for the investment firm C.
Sales policies. Additional requirements would need to be taken into

consideration for the launch of products, operations and services. The legal
text would be modified (EU COM, 2010b, p. 68) to include these
requirements, which would need to be applied before the financial
instrument is offered to clients (sale policy). In particular, investment firms
should:
i) Run a compatibility test with clients needs.
ii) Strengthen duties of compliance with MiFID rules and risk
management of the product.
iii) Subject products to stress test as appropriate (Art. 5, Impl. Dir.) and
periodically review distribution and performance of products and
services.
iv) Ensure that staff receive appropriate training for new products.
v) Ensure that the board has more effective control on these aspects.
In this way, MiFID introduced a further layer of organisational
requirements to be applied before approaching the client. Sales policies are
generally designed to increase the guarantee and the protection of
investors, rather than to improve pre-sale services or increase the
possibility of investors to shop around. In this sense, increasing the
protection of investors does not necessarily mean improving their
awareness of the risk they take.
Telephone recording. In addition, the proposal entails the removal of
member states discretion upon the use of recording for communications
between the investment firm and the client. Recording would become
mandatory while still preserving individuals privacy and respecting the
rules on data protection. Member states may extend the obligation to
services other than the receipt and transmission of orders and execution of
orders and transactions concluded when dealing on own account in all
financial instruments (e.g. portfolio management; EU COM, p. 75).
Records would be stored for at least 3 years. This type of organisational
requirement would be a strong deterrent against market abuses and
promote greater compliance with MiFID rules, but it would come at a cost
for market participants and competent authorities. (CESR, 2010c).
However, it would promote greater enforcement across Europe, in
particular if supervisory arrangements between authorities on their use are
effectively harmonised.
Conclusion # 16
Organisational requirements play a crucial role in ensuring business continuity, market
integrity and investor protection. A proper implementation of the Directive should be
ensured by harmonising ORs and supervisory practices across Europe and removing
ambiguities in the legal text. Moreover, strengthening consistency with other upcoming
regulations would avoid inefficiencies and promote a uniform regime of investor protection
and market integrity within Europe, which would increase legal certainty and the
attractiveness of investment services.
6.3.1 Investment advice: Striking the right balance

The advice. Investment advice is topic subject to heated debate in Europe,
in particular after the scandals that shook financial markets a few years ago
(e.g. Parmalat) and more recently as a result of the financial crisis, both of
which have dramatically impaired investors confidence. The nature of the
investment advice service lies on a trust-based and usually long-term
relationship, which emerges from a situation of strong informational
asymmetry between the provider and the client (mostly retail investors).
This relationship has wider implications and it usually qualifies as a
transaction-specific investment (TSI), as illustrated in the seminal
contribution by Williamson (1975). In effect, customers face sunk
investments/costs that typically deter them from switching providers
(changing provider frequently means losing the previous TSI and facing an
additional one in order to become familiar with the new provider).
The information and behavioural biases involved can lead to rather
high perceived switching costs, difficulties in evaluating alternatives, and
consequently low price transparency and customer mobility.195 This leads
customers to rely on proxies to establish the value of a given investment
option the proxy, in this case, being the advice received by their service
provider. For this reason, so-called relationship banking is brought into
the sphere of consumer protection. In more detail, the provider-customer
195In this respect, the retail financial services sector can be said to differ noticeably from
other economic sectors. For example, in the telecommunications or energy sectors
consumers are less likely to suffer from an information asymmetry, provided that they have
visibility of the quality of service and relative price of the offer. At the same time, the
elements of trust, transaction specific investments and bounded rationality are much
less important in these fields. Finally, especially in the telecommunications sector,
consumers are more likely to shop around for better offers.
relationship can be described by two dimensions: i) the depth of the

relationship, stemming from the off-contract relationship (typical in some
brokerage services), and ii) its thickness, defined as the information
conveyed to the bank through the multiple financial contracts and services.
In particular, the thickness of the provider-customer relationship offers
customers a range of potential advantages, but may also give the informed
provider the opportunity to capture rents by engaging in hold-up
behaviour and false representation of both market and contractual
conditions. In many cases, this relationship is strengthened by personal
feelings with the person who represents the provider.196
Independence. Investment advice is usually provided by investment
firms that also offer other investment services. This situation, next to the
abovementioned aspects, highlight the importance of proper advice for
the efficient allocation of capitals, which needs to overcome the risk of
conflicts of interest. The investors quest for independent advice has
drawn attention to improving the quality of this service. Divergent views
emerge around the specific way investment advice could be boosted to
help restore investors confidence. On the one hand, those who believe that
greater disclosure and stricter organisational requirements can enhance the
quality and attractiveness of investment advice. In particular, in a report
sponsored by EFAMA (2010), asset managers suggested implementing,
amongst the others, four actions at point of sale:
1) Disclosure of the nature of the distributors services,
2) Full and detailed disclosure of all cost items and the principles of
remuneration arrangements,
3) Introduction of a standardised advisor certificate and
4) Granting investors a right of withdrawal.
Disclosure should include the origin of the income of the adviser,
whether it comes only from the fees charged for the advice or depends on
the products sold and in which way. This side of the market challenges the
idea that imposing an independent charging structure, with no alternatives,
will increase the costs of advice and ultimately discourage investors from
using, with strong direct and indirect repercussions on the efficient
allocation of resources in the market. All information should be disclosed to
investors before signing the contract, in order to allow them to shop
196For a more comprehensive analysis of the interaction between switching costs and
cognitive biases, see Renda & Valiante (2010).
around. Disclosure would increase investors confidence in the advice

service, thereby promoting its use. Moreover, the advice service needs to
meet the suitability test in MIFID (see next section), which helps to improve
its quality and design the service in the clients best interest.
In the area of pre-contractual information changes in MiFID will need
to be reconciled with other regulatory initiatives at European level (such as
PRIPs, IMD review, and Prospectus Directive).
On the other hand, some market participants (in particular, retail
investors) claim that besides the importance of improving disclosure and
the role of the suitability test there should be a thorough review of the
incentives which impede advice to be independent. In particular, they
argue that, in order to be considered independent, the investment advice
should be based on an independent system of remuneration, by receiving
only fees for the provision of the service. No commission should be set by
product providers (see box below). The strict independence of advisers
could be achieved by severing the links between fees and products.
However, this severance should be handled with care by allowing
alternative forms of advice to be ultimately offered to clients.
Box 15. A new regime for investment advice: the FSAs proposal
The UK Financial Services Authority (FSA, 2010a) proposed the introduction of a
new system of adviser charging for retail investment products. If investment
advice is to be disclosed as independent, it would need to have its own charges
either upfront or taken from the investment returns. Furthermore, charges should
not be supplemented with other commissions given by product providers, even in
form of non-monetary benefits (inducements). The purpose is to ensure the
independence of advice for retail clients and avoid the rules being circumvented
through the use of soft commissions. Fees should be contracted and fully
disclosed upfront (even potentially ongoing charges)197. In addition, only advisers
who provide a fair and independent analysis on a wide range of retail investment
products would be considered independent under the FSA proposal. The adviser
would have to demonstrate that he/she conducted a fair and independent analysis
of the relevant market for investment products. The relevant market would
include all retail investment products that are capable of meeting investors needs.
Since the range of products is potentially very wide, it would be preferable to
elaborate this definition in order to fully implement it. Investment advice that does
197 For instance, the periodic review of the performance (FSA, 2010a, p. 27).
not meet these requirements would be named restricted, or non-independent.

This service will anyway need to meet the suitability test under MiFID.
This new compensation scheme will impact consumer access to advice, possibly by
both reducing the number of advisers and increasing its cost. The proposal,
however, aims at enhancing the markets reputation and build consumer
confidence, improving sustainability of the sector in the longer term (FSA, 2010a,
p. 18). Anyway, the proposal would not prohibit the provision of restricted
investment advice but only force its disclosure. The authority also commissioned
an impact study, which concluded that the introduction of this charging system
could result in an 11% reduction in the number of advisers; a 9% reduction in
total revenues across all advisers, and an 11% reduction in the number of clients
advised, assuming that other firms did not expand and there were no new
entrants (Oxera, 2010). New entrants would require high initial investments to
build sufficient reputational capital. This situation may create high barriers to
entry the market of advice services. However, it carries high potential benefits in
terms of more efficient allocation of resources and reduced opportunity costs by
avoiding investors being redirected towards more expensive investment products
against their interest. However, both costs and benefits are difficult to quantify, so
the overall effect is unclear.
New proposals. The Commission (2010b, p. 56) has also proposed a

number of measures to improve investment advice:
1) Enhancing disclosure to clients by stating whether the service is
based on an independent and fair analysis;
2) Defining independent and fair analysis as one that considers a wide
range of investment products and is carried by a firm that does not
accept any payments or benefits from product providers;
3) Increasing reporting to clients about performance and market value
of the financial instruments recommended, and to keep updated
information about the client; and
4) Clarifying, when applicable, if the investment advice is also provided
online.
This proposal does not clarify whether investment advice that does
not comply with the definition of independent and fair advice would
remain available to investors. The UK proposal from the FSA does allow
so-called restricted advice, that is, advice under conflicts of interest, to be
offered as an alternative to independent advice. The proposal from the
Commission does not clarify whether the fee can be embedded in the costs
of the investment product and if both remuneration systems can coexist. In
addition, no proposals have been formulated where the adviser is a
company owned by a group that also offers the sale of other investment
products. In this case, the conflict of interest may also be high, since even
though the investment firm does not receive any monetary or non-
monetary benefit for the advice the whole group benefits from the activity
of the adviser.
Conclusion # 17
Some market participants claim that besides the importance of improving disclosure and
the role of the suitability test there should be a thorough review of the mechanisms of
incentives, which prevent advice from being independent. In particular, they argue that to
be considered independent, investment advice should be based on an independent system
of remuneration, by receiving only fees from the clients. No commission should be set by
product providers. Other market participants argue that obliging investors to pay for the
advice would increase the access costs to these services and dramatically reduce the use of
advisory services, with potential long-term costs for end investors. They suggest keeping
the possibility to remunerate distribution through commissions, but with the requirement
to improve disclosure on the nature of the services, in particular disclosing if the advisor is
solely remunerated by the client or whether he is remunerated by a product provider. Full
disclosure of all costs items and remuneration arrangements should be made before signing
the contract. This would improve the ability of investors to choose the service that best suits
their own interest. In the area of pre-contractual information, changes in MiFID will need
to be reconciled with other regulatory initiatives at European level (such as PRIPs, IMD
review, and Prospectus Directive).
6.4 Conduct of business rules

The final piece of the investor protection puzzle under MiFID is a set of
rules to regulate the conduct-of-business (CoB). CoB rules provide some
contractual definitions for the set of fiduciary duties that apply in a
transaction in investment services.
Accordingly, the Directive recognises a fiduciary obligation between
clients and service providers through its general clauses and specific CoB
rules, such as:
i) Best execution (Art. 19.1 and 21);198
ii) Conflicts of interest rules (Art. 13.3 and 18);199
198 Art. 44-46, Impl. Dir.

199 Art. 21-25, Impl. Dir.
iii) Suitability and appropriateness tests (so-called know-your-customer

rules; Art. 19.4 and 19.5)200 and
iv) Other relevant requirements (in particular, information to clients,201
client agreements,202 orders handling,203 and marketing rules).204
In addition, MiFID rules specify exemptions for eligible
counterparties (Art. 24) and define client classification (see Section 6.4.2).205
The constant growth in the last decades of asset under management and
the widespread provision of investment services and products have
brought regulators to set the rules of the game on how investment services
are actually provided to end users. In effect, as portfolios become more
tailored to the needs of investors and to more volatile market trends, the
exposure to complex financial products may increase. Against this
background, CoB rules are becoming more and more important. In
addition, by regulating CoB, MiFID has managed to harmonise an
important part of investor protection in Europe. The Commission is also
looking at the extension of CoB rules to firms exempted by MiFID ex Art. 3
(EU COM, 2010b, p. 52).
6.4.1 Conflicts of interests under MiFID

Definition. Conflicts of interest (CoIs) arise where the interests of a market
intermediary are inconsistent with, or diverge from, those of its clients,
investors, or others (IOSCO, 2007, p. 6). In this respect, it is in the nature of
the business of an intermediary to face CoIs since it needs to exercise
discretion in the provision of investment services and has to deal with
numerous kinds of clients. These aspects make CoIs ubiquitous in financial
services.
In effect, CoIs need to be seen as part of the deal, inherent to the
provision of investment services, since they cannot be suppressed
completely, but regulators can work to minimise both the conflicts
themselves and their negative effects. This regulatory task is performed by
analysing whether CoIs are exploited and are imposing further agency
200 Art. 35-37, Impl. Dir.

201 Art. 19.8, MiFID, and Art. 40-43 Impl. Dir.
202 Art. 19.7, MiFID, art 14.3 and 39 Impl. Dir.
203 Art. 22, MiFID and Art. 47-49 Impl. Dir.
204 Art. 19.2, MiFID, Art. 24 and 27, Impl. Dir.
205 See also Art. 28 and 50 Impl. Dir.
costs (Walter, 2006). Divergence of interests in an agency relationship may

be driven by two factors: i) information asymmetry and ii) transaction
costs. In addition, these two elements depend heavily on market frictions,
and in particular on the level of competition between market players and
their reputational capital. Notably, reputational risk in financial services
plays an important role, since its uncertainty and unpredictability may
drastically reduce expected revenues and put pressure on the management
to commit to the service provided. However, reputational capital is difficult
to measure and may not represent an asset for those managers who do not
commit to a long-term relationship with the investment firm.
There are three types of conflict (IOSCO, 2010d):
1) Firm/client
2) Client/client and
3) Infra-group conflicts.
The first concerns a direct conflict between the provider and the
client. For instance, the broker may favour its own proprietary desk at the
clients expense. The client/client conflict refers, for example, to an
intermediary that favours a group of institutional investors over its retail
clients. The third is the activity of an international dealer group that
favours one of the groups investment firms in one member state to the
detriment of a firm of the group in another member state. These conflicts
are more difficult to detect since there are exogenous variables, such as
legal and fiscal cross-border implications, which can severely raise the costs
of monitoring and detecting the potential wrongdoing. At retail level, these
conflicts are even stronger since information asymmetry and transaction
costs are much higher. Controls should therefore be devised according to
the nature of the counterparty.
MiFID has designed a system of internal controls to prevent, identify,
manage and eventually disclose CoIs (Art. 18), in line with the general
clause to act in clients best interests (Art. 19.1). With this sound statement,
the Directive attempts to secure two important objectives: consumer
protection and market integrity. In particular, the Directive designed a
system of organisational (disclosure) and CoB requirements, which aim to
prevent, manage, identify and disclose CoIs in sensitive areas, such as
brokerage and/or proprietary trading, securities offerings, advice, and
asset management services (Recital 26, Impl. Dir.).
Controls over CoIs can be essentially either internal or external.
Internal controls may consist of (IOSCO, 2007; 2010d):
An internal committee to check conflicts and the functioning of

internal controls,
CoIs policies (to design control tools and compliance mechanisms,
even infra-group, Art. 22, Impl. Dir.),
Clear guidance to refrain from acting in case of conflict,
Information barriers and restrictions (e.g. Chinese walls; Art. 22.3(a),
Impl. Dir.) and
Disclosure to clients (caveat emptor; Arts. 18-19).
Besides disclosure requirements, internal measures are designed to
work primarily ex ante as active and passive organisational arrangements.
Moreover, once the conflict has been identified but cannot be eliminated, it
should be disclosed in line with the principle of caveat emptor, which
foresees that the client must be informed about the CoI but still bears the
ultimate responsibility. In this regard, disclosure is effective if it represents
a sufficient warning, which prevents the client from pursuing the
transaction in some cases. Ineffective disclosure can be counterproductive
and can induce self-interested agents to behave more aggressively
(protected by the disclosure requirements). In addition, if disclosure is not
clear enough (e.g. examples), investors may experience a cognitive bias and
ultimately increase their trust in the intermediary (Enriques, 2006).
MiFID leaves some flexibility to member states to design their own
rules to control CoIs at organisational level (Art. 22.3, Impl. Dir.). This
provision aims to ensure that rules and procedures are designed in line
with the specificities of the country where CoIs emerge. However, this
flexibility has boundaries set by MiFID Level 1 (Art. 31), which do not
allow member states to draw up requirements additional to those defined
by the Directive. On the one hand, too much flexibility could unleash a
race-to-the-bottom among member states to attract more business and
capital. On the other hand, a rigid set of rules would not allow member
states to design the factual rules and procedures to fight conflicts of interest
in line with the social, cultural, legal and economic context of the country.
It is unquestionable that there are countries in which CoIs are met with a
higher level of tolerance, which calls for stronger controls and sanctions.
Under sound and effective organisational arrangements, supervisors would
be able to bring legal suits on behalf of clients, especially where the clients
are retail investors who are deterred by the costs of litigation. Currently,
however, the system of sanctions across member states is highly
fragmented. This situation may create space for supervisory arbitrages in
favour of countries applying low administrative fines.
Finally, external controls may also help to limit the side effects of
CoIs. In particular, regulation in related areas, civil litigations, supervisory
powers and a highly competitive environment can increase the level of
prevention, identification, and management of CoIs. In effect, the way CoIs
manifest themselves may differ depending on the selling practices
involved. For instance, steering practices are frequently used in services
that offer advice on investment products, whenever the portfolio of options
presented to the client includes the advisers own products. Steering
consists of stressing to the investor the advantage of subscribing to a more
costly product because of the re-edit this generates for the advisor, which is
not clearly disclosed to the client. Sometimes this amounts to an intentional
misjudgement of the individuals risk by the financial intermediary to
extract a rent in a specific period, by imposing specific requirements or
additional fees. Other examples of CoIs arise with the provision of
corporate finance services, such as pricing the value of a company that is
also a borrower from the same intermediary. This situation may induce the
intermediary to overprice the value of the company to be able to raise
enough money in a future security offering (e.g. IPO).
New proposals. As mentioned above, the Commission (2010b) is
looking at the possibility of harmonising and strengthening the current
regime of sanction, if there is a violation of MiFID rules, as well as
introducing the principle of civil liability for investment services providers
in order to level investor protection across the EU. Managing conflicts and
structuring incentives for the distribution of financial products are major
concerns for the definition of new implementing measures for CoIs.
Conclusion # 18
MiFID rules on conflicts of interest represent a first step in the introduction of a common
approach across Europe for the prevention, identification, management, and disclosure of
such conflicts. Further initiatives to strengthen the current regime and align supervisory
practices would enhance the treatment of these conflicts and benefit financial markets. A
harmonised set of sanctions, however, should be combined with flexibility for member states
to adapt rules and procedures in line with their national contexts, in order to ultimately
guarantee the goal of a sound and safe environment of protection for investors.
6.4.2 Client categorisation

To complement the recognition of fiduciary duties, MiFID introduced a
regime of client and product classification through which it assigns
different levels of protection to clients. According to MiFID, the level of
protection offered by CoB requirements should be related to the clients

classification. The Directive, therefore, defines two main categories of
clients:
i) Retail (Art. 4.1 (12)) and
ii) Professional (annex II, Sec. 1, MiFID).
Retail. All investors that are not classified as professional fall under
the retail category (definition by exclusion) and enjoy the highest
level of protection afforded by MiFID, typically through the
application of all ORs and CoB rules. However, they may ask to be
treated as professional, if they pass two tests: a basic test (which
assesses the expertise, experience and knowledge of the client); and a
specific test (which requires the investor to meet two out of three
wealth-related thresholds, as defined by Annex II, Section 1, MiFID).
Professional investors are those who have the expertise and are able to
make their own investment decisions. To them, a softer regime applies
that takes into account their more sophisticated nature, e.g.: they only
receive client information on request. MiFID rules are the same overall for
both retail and professional investors, but are tailored to the nature and
needs of each group of investors. Professional investors are, amongst those
listed in Annex II, Section 1.1 MiFID, investment firms, credit institutions
or insurance companies. They can ask to be treated as retail clients but the
other party has to agree. They may also request to become an eligible
counterparty.
Eligible counterparties. The Directive, in effect, introduced a third
special category, called eligible counterparties (ECP; Art. 24).206 This
group of entities includes the most sophisticated clients, such as investment
funds either alternative investment funds (AIFs) or UCITS. Only few CoB
rules and ORs apply to them, such as client agreement rules (Art. 18) and
order handling rules (if acting as broker; Art. 19.8). Professional investors
may request to be treated as ECPs if their member state has decided to
implement this possibility. These investors need to meet a different test
206ECPs are investment firms, credit institutions, insurance companies, UCITS and their
management companies, pension funds and their management companies, other financial
institutions authorised or regulated under Community legislation or the national law of a
member state, undertakings exempted from the application of this Directive under Article
2.1(k) and (l), national governments and their corresponding offices including public bodies
that deal with public debt, central banks and supranational organisations., Art. 24, MiFID.
from the one adopted for retail investors to be considered professional.

Professional investors must meet two out of these three criteria (Annex II,
Section 1.2):
1) Balance sheet of more than 20 million,
2) Net turnover of more than 40 million and
3) Own funds of more than 2 million.
The light touch regime applies to ECPs when they offer services
such as execution of orders; dealing on own account; reception and
transmission of orders; or any related ancillary service.
Client classification should be defined before providing the service
and should be included in the contract or anyway stated in writing. Retail
investors cannot be classified as ECPs, while financial institutions
(including ECPs) can be classified on request as non-professional (retail;
Art. 24.2, MiFID). All other combined classifications are possible. For
instance, the management company of an investment fund may ask a
dealer, executing an order on its behalf, to provide best execution following
retail instead of the professional standards, since the fund mainly collects
investments from retail investors. The firm/dealer also needs to agree to
this request and if it so does, it would have to apply the relevant retail or
professional tests when offering services to such eligible counterparties.
Some segments of the market suggest that portfolio managers should be
entitled to require reclassification as retail clients, since there is a
mismatch in their rights and obligations. Portfolio managers are eligible
counterparties under MiFID but they manage money on behalf of
professional and retail clients, and they have the obligation to act in the
best interest of their clients when placing orders for execution. Others,
however, consider this request to be inappropriate because it may also be
used for activities unrelated to retail clients, and most notably because this
would no longer justify that portfolio managers gain directly (through
commission on profits) from rules that should go directly to serving retail
clients. In effect, their role goes beyond the execution on behalf of investors.
New proposals. It is a commonly held view that the client
categorisation regime should not be radically changed (CESR, 2010c; EU
COM, 2010b). Current proposals, however, concern (EU COM, 2010b): the
possibility to extend the general clause of Art. 19.1 to ECPs; the limitation
of the ECP regime only to non-complex products and the exclusion of non-
financial institutions from the ECP regime; the abolishment of the
presumption that professional investors have the necessary expertise and
knowledge; and the exclusion of municipalities207 from being classified as

ECP or professional per se. In particular, abolishing the presumption of
expertise and knowledge for professional investors would require
investment firms to treat them as retail clients. This situation, on the one
hand, increases the general level of protection in the market, but it boosts
overall costs of dealing with clients, on the other. To reduce the imbalance
of this trade-off towards costs, an opportunity to opt out of the protection
for retail clients should be left available to be unilaterally exercised at any
time for those who consider themselves to be professional investors.
However, there is also a bias that comes into play. A homogeneous
application of retail investors protection rules may need the
implementation of sophisticated and burdensome protection tests, which
may be perceived as high sunk costs for investors to comply with standard
procedures (such as complex and long questionnaires). They may
ultimately decide to opt out to avoid those time-consuming obligations.
Abolishing such distinctions may therefore generate unclear effects on the
provision of investment services.
Conclusion # 19
Although the crisis showed that some of the eligible counterparties (ECPs) were not able to
understand risk properly, the client categorisation regime should not be subject to a major
overhaul. Since portfolio managers are eligible counterparties under MiFID but manage
money on behalf of professional and retail clients and have the obligation to act in the best
interest, some market participants suggest that portfolio managers should be entitled to
unilaterally require the reclassification as retail clients. Others, however, advocate the
inappropriateness of this request since their role is more than execution on behalf of
investors, but they would directly gain (commissions on profits) from rules that should
theoretically only benefit retail investors.
6.4.3 Suitability and appropriateness tests

The duty of suitability rejects the prevailing paradigm of caveat emptor and
forces providers to internalise the harm that they cause when they exploit
information asymmetries to the detriment of customers. (Engel & McCoy,
2002, p. 1334).
The fiduciary relationship between the provider of investment
services and the client implies an obligation to act in the best interests of the
latter (Art. 19.1). In accordance with this general clause, MiFID CoB rules
207 CESR (2010c) referred instead to local authorities.

include a duty to make an adequate investigation of investors suitability208

through a specific suitability test (know your customer obligation)209 and
an appropriateness test (know the security),210 to ensure that the investor
has a professional knowledge of the risk involved in the service provided
(Hazen, 2006).
The suitability and appropriateness tests are respectively articulated
in Articles 19.4 and 19.5. Investment firms must acquire three types of
information about their clients (or potential clients), to be able to decide if
the service is suitable for them:
1) Knowledge and expertise
2) Financial situation and
3) Investment objectives.
The suitability test applies to discretionary portfolio management
and advisory services. In other words, the suitability test applies when the
service offered includes an element of recommendation or discretion. Since
the concept of suitability is fairly abstract, in practice there is a need to
translate it into a meaningful service for clients. For professionals, these
requirements apply less stringently since the provider can presume that the
investor has enough knowledge and expertise and, particularly for advice,
that the investor is able to sustain any financial risk (Art. 35.2, 36, Impl.
Dir.). Implementing legislation may need to further specify several aspects
of the suitability test, which is already described more extensively than the
appropriateness test in the Implementing Directive. Simple risk profiles
may not be enough, so firms may need to look at the cultural and
behavioural attitudes of their clients, as well as at other risk diversification
measures (e.g. geographical aspects). There is a lot of
subjectivity/hindsight in every transactions involving investment services
208The suitability rule was originally defined in USA as an antifraud device (violation of the
Rule 10b-5 under section 10(b) SEC Act 1934, 17 C.F.R. 240.10b-5, 2001). See Clark v. John
Lamula Investors, Inc., 583 F.2d 594 (2d Cir. 1978). In the EU, instead, the suitability rule was
created as a rule to protect investors confidence and to foster market integrity; see Art. 19.4,
MiFID and Art. 35-37, Impl. Dir.; Moloney (2008).
209US Courts consider unsuitable an investment that is incompatible with the investors
objectives and if the broker recommended it, even though she knew or reasonably believed
that the investment was inappropriate. See Kreenan v. D.H. Blair & Co., 838 F. Supp. 82, 87
(S.D.N.Y. 1993).
210 See Alton Box Bd. Co. v. Goldman, Sachs & Co., 560 F.2d 916, 922 (8th Cir. 1977).
that may push people to give less consideration to aspects related to the
performance (through risk profile assessment) than it was originally
thought. Hitherto, assessing other aspects than risk profiles would improve
the reliability of the suitability test. It is also worth noting that the
implementation of suitability requirements among member states may
have raised barriers to market products in some countries. It would be
worthwhile to explore the consequences of a fragmented implementation.
Benefits and costs. The introduction into the regulation of very specific
requirements to assess risk profile and suitability may become a mythical
search for a risk-free solution for consumers. The last pioneers who left for
this search never returned. Rather than focusing only on the appropriate
level of risk for investors, trying to mimic investors choice, a more
workable alternative would be to make sure investors are fairly informed
about the risks they decide to incur, which is one of the original objectives
of MiFID. Responsibility should ultimately fall on investors. However, a
loose definition of suitability that shifts the whole burden onto investors is
not desirable either. Above all, service providers need to assist investors in
their decisional process. Notably, an array of aspects (behavioural and
financial) should be considered in a suitability test. Most of all, firms
should look at the risk attitude and the decision style of consumers when
they create a personalised investment portfolio. Only the combination of
these aspects will allow a meaningful rating process of products and
clients, thereby leading firms to provide a more suitable service. Finally,
supervisors should make sure that the implementation of suitability
requirements is uniform since differences in national implementation can
create costly barriers to the cross-border provision of services, as well as
important fractures in the level of protection for end investors. The
suitability and appropriateness tests are among the main pillars of the CoB
rules introduced by MiFID, and are meant to strengthen the protection of
investors and the quality of the services provided to them. This protection
is particularly relevant where firms are providing cross-border services
through an EU passport. However, the roll-out of the infrastructure to
perform these two tests has required uniform implementation and has
come at a significant cost. One-off costs to upgrade systems and update
clients information are among the most important, next to the costs of
assessing the compliance of services and products with the MiFID
framework (FSA, 2006b, p. 18).
Conclusion # 20
The suitability test is a crucial aspect of the provision of investment advice and portfolio
management services. Views diverge slightly on how to assess the knowledge, financial
situation and objectives of investors, and in particular, on how deeply the suitability test
should look into investors habits and their willingness to undertake risk. Some members
regret the lack of harmonisation in the implementation of suitability requirements for
discretionary portfolio management. These members would welcome action by ESMA to
improve legal certainty and reduce barriers to market investment products across the EU.
However, other members do not think any intervention is needed in this regard and are
satisfied with the current level of harmonisation.
The appropriateness test applies for non-advised services, such as

reception and transmission of orders.211 Whenever the firms realise that
according to the information provided the product or service offered is not
appropriate for the client, the firm shall warn the client. Such warnings
could be provided in a standard format. The appropriateness test applies
when the client does not rely on the firms recommendation or advice.212
The amount of information the investment firm needs from the client
concerns the clients knowledge and experience in the investment field
relevant to the service offered (Art. 19.5) and this information depends on
the complexity and risk of the product. As mentioned above, the
investment firm can assume that professional clients have enough
knowledge and experience to understand risks. Besides, the information
about clients that needs to be collected refers to (Art. 37, Impl. Dir.):
1) The types of service, transaction and financial instrument;
2) The nature, volume, and frequency of the clients transactions and
3) The profession and level of education.
For non-complex financial instruments, clients can opt out of these
safeguards (appropriateness test), asking for an execution-only service.
Non-complex instruments (Art. 38, Impl. Dir.), currently under MiFID,
are:
1) Shares admitted to trading on RMs,
2) Money market instruments,
3) Bonds and other securitised debt (with no embedded derivative),
4) UCITS and
211 See Casey & Lannoo (2009, pp. 46-53).

212 Giraud & DHondt (2006, p. 52).
5) Other non-complex financial instruments (Art. 38, Impl. Dir.).

Remaining instruments are classified as complex, so investors using
those them cannot opt out of the appropriateness test. In case of non-
complex instruments, for execution-only services provided without an
appropriateness test, retail clients should be adequately warned, the service
should be provided at the initiative of the client, and the firm has to comply
with CoIs rules (Casey & Lannoo, 2009).
New proposals. The definition of complex vs non-complex financial
instruments is under discussion (CESR, 2010c and EU COM, 2010b, p. 55).
Two options have been proposed (EU COM 2010b, p. 55):
1) Excluding all financial instruments that embed a derivative and all
UCITS that adopt complex strategies (to be defined) from the non-
complex category;
2) Removing the distinction between complex and non-complex
financial instruments, and applying the appropriateness test to all
financial instruments.
The first proposal requires a revision of the UCITS Directive in order
to define which UCITS follow complex strategies. In this regard, it is hard
to split the complexity factor from the risk factor and, for UCITS in
particular, units are typically sold to investors under the latters
presumption that the product is non-complex and moderately risky given
the security image of the brand. If more complex and perhaps riskier
strategies are pursued through UCITS, it should be left to the UCITS
Directive (2006/65/EC) to differentiate complex and non-complex UCITS.
The second proposal may create higher costs for those market
participants whose offer concentrates on non-complex products. Moreover,
the impact of competition among investment firms on investor protection
rules is uncertain, since typically those protection rules prescribed by
MiFID are not actually priced by the investors, who do not tend to require
these services if not unilaterally offered. With strong competitive pressures
and insufficient supervision, incentives to apply these requirements might
consistently be insufficient.
Conclusion # 21
MiFID foresees a different regime for execution-only services based on the product
classification between complex and non-complex financial instruments. Any change should
take into account that complexity does not necessarily mean more risk. The objective should
be to verify if the product is in line with investors understanding of the ultimate risk that
they are going to bear. Regulation should not decide the level of risk investors want to take.
Some market participants believe that certain UCITS might have become too complex to be
easily understood by investors and so to skip the appropriateness test, at least for retail
clients. Others, instead, argue that classification, in particular for UCITS, should remain
as such, since a change in classification could damage the UCITS brand outside the EU.
Box 16. MiFID and asset management at a crossroads
MiFID has had widespread implications for several areas of financial markets, in
particular for the asset management business. Besides the application of MiFID to
portfolio management services, it is the interaction between MiFID and UCITS that
raises concerns. In particular, MiFID sets a formal exemption (Art. 2.1 (h)) that has
not been de facto applied. In effect, UCITS are currently exempted from the
appropriateness test as non-complex financial instruments, which means that they
are part of the Directive. UCITS management companies are then classified as
ECPs under MiFID, therefore best execution or suitability tests are not applied,
unless agreed otherwise. Besides these formal exemptions, there are several aspects
that put UCITS under the MiFID spotlight. UCITS typically apply and get the
status of professional investor, but they also ask to be treated as retail investors,
since the investment product is mainly designed for retail investors. The UCITS
Directive also provides for similar organisational requirements as those in MiFID
(Recital 1, UCITS Implementing Directive 2010/43/EU) to create a level playing
field between UCITS managers and MiFID investment firms undertaking portfolio
management services.
Overall, there are other aspects that emerge with the revision of MiFID as
points of discussion that may affect the debate on UCITS. Among others, the
following aspects would have an impact on the asset management industry:
Unilaterally granting retail best execution to managers of retail funds
(without the need for the agreement of dealers);
Modifying current definitions of complex versus non-complex financial
instruments;
Investigating the status of the implementation of suitability requirements for
investment advice and discretionary portfolio services; and
Changing the charging structure for investment advice and its basic
requirements, and promoting greater inducement disclosure.
The interaction between these areas may have important implications for the
future of the industry, in particular for the convergence between alternative and
traditional investments (if no differential status would be granted to traditional
investments) and on the traditional distribution channels (by changing the role and
nature of investment advice).
6.4.4 Grasping the definition and economics of best execution obligations

Setting the scene. MiFID (Art. 19.1) requires that an investment firm act in
accordance with the best interests of its clients to ensure a high level of
investor protection. Besides the general clause, which de facto sets an
important fiduciary duty, the article sets a dynamic obligation of results
(Giraud & DHondt, 2006), which calls for investment firms to ensure best
execution when they offer:
Execution of orders on behalf of the clients (Art. 21, MiFID),
Reception and transmission of orders in relation to one or more
financial instruments (Art. 45.4, Impl. Dir.) and
Portfolio management (Art. 45.4, Impl. Dir.).
These duties apply to all financial instruments, except spot foreign
exchange instruments. Best execution obligations, should therefore be
achieved by:
1) Taking all reasonable steps to obtain the best net result for the
client, taking into account price, costs, speed, likelihood of execution
and settlement, size, nature or any other consideration relevant to the
execution of the order. (Art. 21.1, MiFID);
2) Drafting an effective and detailed order execution policy that
includes the execution venues and the parameters of choice, and then
getting the clients agreement (Art. 21.2, 21.3, MiFID); and
3) Demonstrating, at the request of the client, that execution has been
carried out in accordance with the existing execution policy, which
allows the achievement the best possible result (Art. 21.3, 21.4,
MiFID).
From these detailed requirements, some argue that the Directive aims
at imposing an obligation of means. So far, however, regulators have not
clarified yet whether MiFIDs original spirit was towards results (dynamic
approach) or means (static approach).
Definition. Firstly, MiFID draws up a comprehensive definition of best
execution, which takes into account all potential factors that can affect a
financial transaction (as there is not a universal definition of best
execution). Such factors can have different meanings that should be then
considered.213 For instance, costs can be seen as implicit or explicit and
213 Please, see in general, FSA (2006), p. 15-18.

they can assume different forms according to multiple factors, such as the
nature of the client.
In this regard, the criteria to be considered by investment firms when
performing best execution of an order are the following (Art. 44.1, Impl.
Dir.):
i) Specific instructions given by the client,
ii) Nature of the client (retail or professional),
iii) Nature of the financial instruments and
iv) Nature of the execution venues.
What should be considered as best execution for retail clients
(typically price and costs) may not be the same for professionals (Art. 44.3,
Impl. Dir.). The difficulty of drawing up a precise definition of what is best
execution is obvious in todays financial markets. Accordingly, the
Directive does not give an easy-to-enforce definition of best execution
either. The legal text tries to grasp all factors influencing a financial
transaction, but does so in a general manner. However, this situation is not
necessarily a detriment to final investors for two reasons, which should
always be verified. On the one hand, general clauses are typically set to
define a fiduciary duty, which should by definition fill gaps in incomplete
contracts, i.e. when it is too costly and difficult for the contracting parties to
specify all relevant terms. This situation does not call for detailed
provisions in the law, but rather requires stronger enforcement tools and
judicial review and the flexible application of best execution requirements
(Macey & OHara, 1997 & 2005). On the other hand, when transaction costs
are sufficiently low, a broad definition allows counterparties to bargain and
contract around the duty of best execution or customise it (even opt out)
. In addition, technological developments (Smart Order Routers, SORs)
support the customisation of best execution policies, even in a context of
great market fragmentation where best execution cannot be easily verified.
Nevertheless, a strict regulatory definition can increase verifiability
but the quality of the execution (investment) service would be potentially
lower, with ultimately potential negative effects on demand. A too precise
definition of best execution may exacerbate ex ante commitment and so
inefficient investments. Current legal text sets the boundaries in which
competing investment firms tailor execution policies in line with clients
interests. More should be done to increase verifiability by improving the
content of execution policies and the quality and depth of market data.
Execution policies. Secondly, as said, a loose contract may provide

enough flexibility to contract around (or opt out) at the risk of distorting ex
ante commitments. Pre-contractual investments, however, are primarily
based on execution policies. By improving the quality of execution policies,
the contract therefore becomes less vague. In particular, by promoting:
Quality of execution policies and
Observability and verifiability of best execution (through better and
more data).
On the quality of execution policies, MiFID obliges investment firms
to make a list of execution venues where the intermediary is going to
execute the orders, and a list of the factors affecting the choice of a specific
trading venue (19.3 and 21.3, MiFID).214 Best execution policies should
receive the clients consent, and firms must review their policies once a year
or whenever a material change occurs, for instance when a significant
new execution venue emerges (FSA, 2006a, p. 34), and assess them on a
regular basis (Art. 45.6 and 46, Implementing Directive). In addition,
member states should require investment firms to monitor the effectiveness
of execution arrangements (Art. 21.4, MiFID) and national authorities may
expressly ask investment firms to be able to demonstrate to clients that the
transaction is compliant with their execution policy and so with best
execution requirements (Art. 21.5, MiFID). To facilitate disclosure to clients
about the execution policy, the Commission (2010b) has therefore proposed
to introduce a template for policies.
MiFID also allows for different types of best execution policies,
especially in relation to the selection of the venue for diverse asset classes
and clients. Execution policies do not need to be defined on an order-by-
order basis (dynamic approach) but in general (static approach).215
214 This list must be reviewed periodically, taking into account execution venue fees,
clearing and settlement fees and any other fees paid to third parties involved in the
execution of the order (Art. 44.3, Impl. Dir.). Effective barriers to add new venues can be
(among others, CESR, 2010h, p.5): (1) The lack of interoperability between clearing houses,
so the inability to choose a clearer of choice; (2) Local rules which prevent remote
membership of exchanges or complex re-registration processes; (3) Different regulatory
regimes and clearing and settlement requirements in several countries, creating obstacles to
competition between venues; (4) Absence of a central counterparty continues to be a
clearing problem for some venues; (5) Distance speed (or latency), mitigated by new
facilities, e.g. hosting facilities, IT developments etc.; and (6) System costs, integration to
back office and settlement systems.
215 See, in particular, FSA (2006), p.12.
However, the regular review of best execution policies currently provides

some space for introducing dynamicity in the way investment firms
provide best execution. The interpretation of what dynamic means is
perceived differently by the market. Some argue that dynamism refers to
the obligation to modify execution policies on an order-by-order basis (led
by a general obligation of results). In this sense, dynamic order execution
may be more efficient and convenient for clients than static approaches,
which only meet the minimum legal requirement (Ende et al., 2008). Some
others believe that dynamic generally means reviewing the execution
policy if major events occur and assessing the effectiveness of execution
arrangements periodically as required by MiFID. In this respect, in their
view, there are no pending issues with regard to execution policies, besides
a need for a proper implementation of current MiFID requirements and the
lack of high quality data to verify best execution.
Observability and verifiability. Thirdly, broad legal definitions call for
greater enforcement of execution policies by national financial authorities.
However, current legislation leaves gaps that make enforcement difficult,
such as the difficulty to assess factors other than price and speed, and the
current design of best execution duties, which allows exemptions and do
not make directly accountable specific trading practices.216 Moreover, the
lack of quality data on execution for financial instruments makes
comparability of post-trade information complicated.217 MiFID requires
firms to act in their clients best interest (Art. 19.1), which constitutes a
good-faith clause that should allow judges the necessary discretion to
decide whether best execution has effectively taken place. However, since
MiFID also defines what constitutes the clients best interests, it impairs
judiciary discretion and rather favours the principle of caveat emptor. This
principle is not satisfactory either, due to the lack of verifiability, which
means investors have to choose among execution policies that may not
216For instance, the US case law recognises as industry practice - therefore not considered
violation of the best execution duty - accepting a rebate for directing order flow to a
particular broker or trading venue; see Macey & OHara (1997), pp.195-196.
217CESR is also looking to define an appropriate execution metric in order to measure
execution (on a quarterly basis) in shares (for other financial instruments it has been
postponed to a future discussion) and it may ask execution venues to produce reports on
execution quality using metrics set by CESR. CESR (2010), Investor Protection and
Intermediaries, op. cit., pp. 18-26.
always be implemented properly. Observability and verifiability are

considered minimal informational requirements for an event to define a
contractual contingency (Hermalin et al., 2007). Best execution is
characterised as a contractual term but it is not verifiable, even though
observable (monitoring costs that are too high). It is therefore fundamental
to improve the available quantity and quality of execution data (CESR,
2010b), which can also reduce the transaction costs of contracting around
the best execution requirements, as set by the Directive. Bargaining
something that is not verifiable does not make any sense contractually.
The Commission (2010b) also believes that better and more data should be
combined with an effective civil liability for the investment firm.
Market structure. Best execution duties influence two different areas:
market structure and investor protection (Macey & OHara, 1997). Best
execution de facto is only the one that the market is willing to provide, thus
order flow markets must be competitive. Best execution is synonymous
with low transaction costs. It is achieved when the least number of
resources are lost in intermediation, for the execution services industry
(Harris, 1996). Best execution policies must ensure that, when the
abovementioned conditions are met, the investment firm screens the whole
market, in particular those execution venues that can offer best execution to
clients. On the one hand, the duty of best execution may sustain a
competitive environment between trading venues, as incumbents are
indirectly forced to deliver better quality of execution (in terms of lower
implicit and explicit trading costs), due to the investment firms
commitment to deliver best execution. This pressure reduces the incentives
to define an oligopolistic setting with newcomers to the market, and among
investment firms (as they ought to compete around quality and dynamicity
of execution policies). On the other hand, this duty comes from a more
important recognition of a fiduciary relationship between client and
investment firm and aims at protecting end investors. Strong information
asymmetries between client and investment firm may impede the
definition of what precise factors should be taken into account to achieve
best execution for a specific client (McCleskey, 2004).
The lack of an immediately applicable definition of best execution
and the fear of undermining the structure of the US national market system
led the SEC to impose a strict definition of best execution (around the best
available price, i.e. National Best Bid and Offer, NBBO) and require brokers
to report execution performance statistics (SEC, 1962). In this regard, the US
regulators decided to fix an enforceable requirement (price) under
securities regulation and to leave to markets, under competition, the
provision of related services that help to achieve best execution (such as

speed, multi-market trading, etc).
Non-equity markets. For non-equity financial instruments, traded OTC
or on inter-dealer platforms, best execution may rely on other relevant
aspects such as national legal rules, counterparty risk, products and market
feasibility. In particular, it may be relevant to consider the volumes traded
on a specific platform, as well as the type of organised venue or the details
of the dealer that represents the counterparty in an OTC transaction. In line
with the type of venue where the transaction is executed, execution policies
should consider the most appropriate approach.
Nevertheless, it may be difficult to find comparable data for some
complex derivative or credit product, or to see pre-trade information on
prices. The kind of transparency obligations adopted for equities may not
be as beneficial as for other types of markets and products (see Section 1).
For instance, buying a bond that is only traded OTC for a retail client may
not be compliant with the price factor of best execution since the
counterparty would only agree to sell at a specific price set unilaterally
(due to the low size). There is no space for negotiation and it is not
verifiable if the trade could have occurred with a different dealer (the
information is simply not available). In a RFQ trading model there is more
chance that the investment firm can choose between few competing
bid/offer prices and therefore to apply best execution obligations.
Conclusion # 22
Best execution duties lie at the foundation of the fiduciary relationship between service
providers and clients. Execution policies and data are essential aspects of the effectiveness of
these legal requirements. MiFID tries to grasp all factors influencing the best execution of a
financial transaction, and does so in a very general manner. A strict legal definition (price-
only) would assume that other services needed to achieve best execution (such as speed,
transaction cost analysis, etc) would be efficiently provided by the market itself under
competition, with no need of a formal legal protection. A broad definition, however, is not
necessarily detrimental for final investors as long as execution policies are properly
implemented and data allows sufficient verifiability of execution. Conflicting views emerge
between those who argue that execution policies should be designed around a dynamic
obligation of result, i.e. complying with the minimum legal requirements is not enough in
their view to achieve best execution. Other participants however challenge the view that
issues with best execution come from execution policies, which are MiFID-compliant.
Investors themselves receive full information about their execution policies. Both recognise
that those issues emerge from a consistent lack of data on execution quality from trading
platforms, which are under discussion within the debate on the new transparency regime.
Overall, MiFID requires firms to act in their clients best interest (Art. 19.1), which
constitutes a good-faith clause that should allow judges the necessary discretion to decide
whether best execution has effectively taken place. However, since MiFID also defines what
constitutes the clients best interests, it impairs judiciary discretion and rather favours the
principle of caveat emptor. This principle is not satisfactory either, due to the lack of
verifiability, which means investors have to choose among execution policies that may not
always be implemented properly. Observability and verifiability are considered minimal
informational requirements for an event to define a contractual contingency.
Best execution also impacts on market structure. Best execution de facto is
synonymous of low transaction costs and is achieved when the fewest resources are lost in
intermediation. Best execution policies must ensure that, when the abovementioned
conditions are met, the investment firm screens the whole market, in particular those
execution venues that can offer best execution to clients. Hitherto, the duty of best
execution sustains a competitive environment between trading venues, as incumbents are
indirectly forced to deliver better quality of execution, due to the investment firms
commitment to deliver it.
7. A NEW REGIME FOR COMMODITY
DERIVATIVES
Commodities markets play a fundamental role for our economies. They

ensure that demand and offer of commodities find their best market
clearing price. MiFID acknowledged the different context into which firms
may need to enter in a financial transaction in commodity derivatives by
giving full exemption to those firms providing investment services in those
instruments, if their main business or the core activity of the group is not
the provision of investment services or the transaction is done for hedging
purposes (when dealing on own account; Art. 2.1 (d) (i) (k) (l), MiFID). The
Directive de facto grants a broad exemption to groups such as energy
companies, which frequently use derivatives to cover exposures (see figure
below). For instance, energy companies providing electricity need to cover
their exposures since electricity cannot be stored and its availability is
subject to many endogenous (e.g. infrastructure, alternative renewable
energies) and exogenous variables (e.g. weather). The need to hedge those
risks requires energy companies to be heavily involved in trading activities
onexchange and overthecounter.
In order to minimise regulatory burdens on such hedges, the text of
the Directive currently grants full exemption for the provision of
investment services related to commodities, except if this is part of a group
whose main business is the provision of investment services (e.g. dealers).
The exemption is not by instrument (commodity derivatives are included
in the definitions of Annex I, Section C, MiFID) but by use (hedging) or
main business/activity of the firm. No sound definition of how hedging
activities can actually be classified is currently available.
| 205
206 | A NEW REGIME FOR COMMODITY DERIVATIVES
Figure 33. Exemptions for commodity instruments and firms
Source: RWE Supply & Trading.
The financial crisis. The recent increase in market prices and volatility
of such goods has diverted attention away from a potential link between
prices on spot markets and volumes and prices of transactions done on
underlying future markets. The alleged financialisation of commodities
and their prices formation process has put many derivatives transactions
under the regulatory spotlight, in an attempt to address the historical high
price volatility in commodity markets. It needs to be said that derivative
markets (in particular, future markets) are just one of the variables that
may affect the price formation processes of commodities. Amongst other
variables, there are fundamental aspects such as: demand (e.g. the GDP is a
valid measure); supply constraints; transportation costs; storage costs; and
other exogenous factors (weather, political stability, etc). Uncertainty about
the global economic outlook, testified by the rise in the prices of

commodities such as gold, does not help to control volatility either. The
growing uncertainty has increased producers' needs to find protection by
increasingly accessing sophisticated financial transactions, particularly in
derivatives markets. These markets provide a formidable mean of hedging
against price changes in the physical product. Since markets, however,
would not be liquid with only investors hedging in the same direction
(there would be no counterparty to their trades), speculation, intended as a
way of trading by investing in gathering private information, becomes
indispensable to create a buffer of 'noise' trading that can give markets
enough liquidity.
New proposals. Besides fragmented and conflicting views on the role
of financial markets on commodity prices, regulators are asking for more
powers and transparency over positions in commodity derivatives in order
to avoid risks of manipulation and better control market participants
positions (EU COM, 2010b, pp. 37-39). One option to reach these objectives
implies a partial or total removal of exemptions under MiFID. This
situation would extend the scope of MiFID and bring the energy and
energy trading markets into the scope of financial market regulation for the
first time. Additional regulatory and capital requirements will apply to
business firms that will need to get authorisation as MiFID investment
firms. For instance, energy companies running a trading desk internally
will have to set up a licensed investment firm. Such firms would be then
subject to the full regime of the Capital Requirements Directive (CRD)
and need to hold regulatory capital if the parallel exemption for
commodity firms under the CRD is not extended beyond 2014 (capital
requirements applying to commodities firms will be most likely subject to a
separate review under CRD).
Instead of regulating market participants, another viable option
would be to strengthen regulation on trading activities through new
market integrity and transparency rules (e.g. transaction and trade
reporting).
The Commission (2010b) has suggested introducing new reporting
obligations for positions in commodity derivatives, covering all official
venues (RMs, MTFs, OTFs). The proposal in line with the review
requested by MiFID in Art. 65.3 requires investment firms and banks to
disclose the counterparty, which can be an entity classified under EU
legislation (such as UCITS, credit institution, etc) or a commercial trader
(not classified under any EU legislation), and then decide if more

information is needed. An alternative option would require reporting only
for transactions not qualifying as hedges, using current international
accounting standards such as IFRS IAS 39. A combination of the previous
two options is also on the table.
In line with the approach to the review, the Commission will most
probably narrow the breadth of the exemptions in order to cover
commercial firms that are not sophisticated clients under MiFID ORs and
CoB rules. To achieve this, Art. 2.1(i) and 2.1(d) would clarify that the
exemption should be granted only to the business of hedging physical and
price risks (EU COM, 2010b, p. 41). To verify if activities are ancillary to
the main business, quantitative (e.g. a certain percentage of the main
business) and qualitative measures should be introduced (e.g. dedicated
personnel). Finally, the exemption of Art. 2.1(k) seems to overlap with the
others, so it could be deleted. However, the original intention of this
exemption was to enable a review of the capital requirements for
investment firms dealing with commodities before applying any MiFID
rules and CRD regime, and giving the time to define an appropriate
prudential regime for this business. Finally, the Commission decided that
capital requirements for these firms will be consistently defined in a
specific review of the Directive 2006/49 (EU COM, 2010b, p. 41).
As a result of these multiple changes, investment firms dealing with
commodities not for hedging purposes will need to apply licensing,
transparency, conduct of business rules and organisational rules as well as
potential requirements under the CRD regime.218 They will therefore need
to set up MiFID licensed firms and apply organisational requirements and
CoB rules.
Market views. On transparency and legal exemptions, market views
are split between those who believe that exemptions should be kept as they
now stand or allow for a less onerous regime for these companies, and
those who favour a comprehensive harmonisation of rules in this area. On
the one hand, some market participants (in particular, energy companies)
believe that the removal of the exemption will harm their business
218This approach does not seem to be in line with the legal text proposed for central
counterparty (CCP) clearing of over-the-counter derivatives in the recent proposal of
European Market Infrastructure Regulation (EMIR; COM (2010), 484/5). In that proposal,
non-financial counterparties are not in principle covered by the scope of the regulation if
they do not breach a certain clearing threshold.
(especially for small companies) and indirectly affect underlying prices

(e.g. energy prices). The inclusion of these companies in the MiFID original
regime will mean:
1) Treating specialised commodity firms as financial institutions, even
though they do not access central bank money or other credit
facilities at favourable terms, do not offer products to retail investors
(or are not retail investors themselves), and do not add up to the
systemic risk of the financial system U;219
2) Imposing commodity firms to transfer their business to a
MiFIDlicensed firm (high compliance costs); and
3) Complying with all MiFID requirements (CoB and ORs).
In their view, the potentially positive effects of being classified as a
MiFID-licensed firm may not offset the actual costs of this proposal. In
particular, benefits would include:
i) The availability of an EU passport for MiFID investment firms, which
may not be useful for some participants, such as energy trading
companies, as they do not offer cross-border investment services;
ii) The assurance of continuity of the trading business under the MiFID
framework (ORs); and
iii) A harmonised set of rules for investor protection (CoB).
As a result, the imbalance between benefits and costs may boost
hedging costs, either causing energy price for end-consumers to go up or
reducing the overall use of hedging tools. The proposal may also reduce
liquidity and competition by raising barriers to entry, in contrast with
widespread processes of market liberalisation in many sectors (e.g. energy).
On the other hand, some market participants and regulators stress the
importance of levelling the playing field in terms of market transparency
among different classes of traders (also for financial stability purposes).
219The joint advice of the Committee of European Banking Supervisors (CEBS) to the EU
Commission (CEBS-Advice) in 2008 clearly states that systemic risk concerns [of
commodities business firms] appear significantly smaller relative to the systemic risks
posed by banks and ISD financial investment firms. In the commodities case studies
examined in this report, systemic concerns were limited and contained. See, joint CESR-
CEBS Advice (ref. CESR/08-752), 15 October 2008, see ref. 12, ref. 38 et seq., 213 et seq., 282
et seq., (http://www.c-ebs.org/getdoc/ee9b85fa-4d64-48dc-9f45-a7350881ddac/2008-15-10-
CESR-CEBS-advice-on-Commodities.aspx).
Greater transparency may also improve price formation processes in

commodity related markets (such as future markets), ultimately ensuring
better price formation in the underlying commodity market. Finally, small
firms (e.g. farmers) whose main business is not the provision of investment
services, will in this way be granted a greater set of investor protection
rules, which should reduce the possibilities of fraud and market
manipulation.
Definition under MiFID. Furthermore, the definition of a commodity
derivative as a financial instrument under MiFID is also under discussion
in order to have a larger coverage of over-the-counter derivatives (EU
COM, 2010b, pp. 41-42). In addition to the requirement for the physical
delivery of the commodity to take place on at least two trading days, an
OTC derivative contract would be considered MiFID financial instrument
if:
(1) Traded or equivalent to a contract traded on RMs, MTFs, or similar
facility and
(2) Standardised in terms of price, lot, date of delivery, etc.
A third requirement concerning the need to be cleared on a CCP or
being subject to the payment of margins is currently under discussion.
However, the requirement to be centrally cleared would restrict the
number of OTC derivatives that would fall under MiFID rules. In addition,
by doing so, the eligibility requirements for CCP-clearing would be
indirectly applied to trading, where different conditions and a regulatory
approach under MiFID applies. Moreover, current legal text may diverge
from the US Dodd-Frank Act, which exempts physically settled OTC
forward transactions from the definition of a swap. The level playing field
between the EU and US regulation should be preserved.
Position limits and management. With the review of MiFID, regulators
and competent authorities advocate strengthening and harmonising
supervisory powers to control the build-up of systemic risk and disorderly
markets arising from market manipulation. These powers should allow
setting position limits to minimise the risks of price and market
manipulation (Langen, 2010; EU COM, 2010b, pp. 82-83).220
Curbing speculation is a vague objective, since how someone can
actually distinguish between hedging and speculative trading remains
220In line also with the US Dodd-Frank Wall Street Reform and Consumer Protection Act,
P.L. 111-203, H.R. 4173.
highly controversial. Speculation involves the use of private information

based on greater market knowledge, which is different from inside
information (manipulation).221 Informed trading activities based on more
or less sophisticated expectations that market prices may go in a specific
direction may be also done for hedging purposes. For instance, an
electricity company willing to cover its future exposure by purchasing a
future price contract based on the expectation that prices may go up/down,
even though the company does not currently own the commodity (energy
is not storable), which would be hardly defined as speculation, even
though it looks that way. In addition, with no investors investing in
information there would be no space for hedgers to find a counterparty
that actually prices and is willing to trade the probability that the event to
be hedged in the end will not occur. Regulators, therefore, should shed
light on the risks of price and market manipulation in commodity markets,
potentially arising from the accumulation of dominant net positions in
derivatives markets (in particular, futures markets). The attempt to squeeze
the market by collecting a dominant share of a commodity future for a
specific settlement date may have disruptive effects on spot prices, since
the dominant owner of a long position for that date may exercise
unreasonable upward pressure on futures prices and thereby indirectly on
spot market prices as well.
To reduce the harmful effects of settlement squeezes, regulators and
supervisors typically adopt two tools: position limits and position
management. Position limits may be imposed not only for exchange-traded
products but also for OTC contracts. In addition, counterparties entering
the derivative contract could be asked to provide a full explanation of the
position and other relevant documentation, besides the regulatory
requirements currently in place. Position limits may need to be further
investigated to determine the size of a position that can be deemed to be
manipulative. It is essential to ensure that limits are neither set at
unsustainable (too low) nor at irrelevant levels (too high), which would
mean either an excessive shift in market structure or no impact at all. In
addition, the extension to OTC contracts, on the one side, may reduce the
incentives to move some of the trading activities from open markets to
221 The border between private legal information and inside information is often so
unclear that regulators may be willing to ban trading a product rather than monitoring
activities on an ongoing basis. See Art. 1, Market Abuse Directive 2003/124/EC.
OTC. On the other side, position limits for OTC contracts may increase the
number of trades and eventually the costs that the requesting
counterparty should pay, since it could not benefit from the freedom to
privately negotiate an important risk exposure. However, it greatly
depends on how the position limit would be formulated, whether or not it
would be an obligation to reduce size of the position below the limit, or
whether it would only require disclosure to regulators. Position limits
generally impose on commodity traders a cap on the size of transactions
based on broader indicators. Some argue that these measures may be easily
circumvented by trading more frequently with smaller sizes, which makes
supervision a more complex and costly activity. By contrast, the use of
position limits may be indispensable in markets where the single
transaction can directly manipulate prices, such as physical markets or
markets for non-storable commodities (e.g. electricity exchanges).
Position management allows the detection of dominant net positions
at the end of the trading day and a more accurate monitoring of systemic
risk. If positions create unreasonable upward or downward pressures on
prices, market operators can require traders, at the beginning of the next
trading day, to reduce their positions. Position management may thus be a
more effective tool for tackling this issue, since manipulation in commodity
markets does not usually come from the impact on prices of the availability
of investors to transact a security at a specific price, but from the
availability of counterparties to bargain a future position. Since the
availability of settlement dates is actually quite limited, it would be more
meaningful to collect all trading reports and calculate the total net position
of an investor in that specific market. In securities markets, however, the
availability of securities is typically wider and manipulation comes
frequently from the misuse of a single or multiple transactions by
exploiting inside information or the accumulated size (trading with
knowledge). Some have also cast doubt on the improper use of leverage.
These worries also concern securities markets, but the issue itself should be
confined to the reforms on capital requirements for financial institutions
(Basel III), as it may be more appropriate to tackle them in that context.
Conclusion # 23
Narrowing exemptions for commodity derivatives under MiFID may have a substantial
impact on the business of non-financial companies. Some market participants advocate the
need for a level playing field between financial and non-financial firms when they come to
trade financial instruments. In their view, it would represent an important step towards
greater transparency and investor protection, while others ask for further investigation of
the unintended consequences in terms of higher costs of hedging relevant exposures in the
market, as well as lower competitiveness, liquidity, and competition. The need for
consistency across several regulations in the commodity business may need a more fully
articulated answer and coordination with initiatives to define capital requirements for
investment firms.
Curbing speculation is a vague objective, since how someone can actually
distinguish between hedging and speculative trading remains highly controversial.
Regulators instead should shed light on the risks of price manipulation potentially arising
from the accumulation of dominant net positions in derivatives markets (futures).
Strengthening supervisory powers may in principle be effective to control price
manipulation, but its role in controlling systemic risk may be doubtful, in particular
through position limits. The use of position management tools to monitor the size of net
positions would be more effective.
8. CONCLUSIONS
The MiFID review is an opportunity to boost investor confidence and

strengthen the resilience, efficiency and transparency of financial markets
and instruments. Investor protection and market efficiency should remain
the guiding principles of the Directive but be reconciled with broader goals
brought about by the financial crisis, such as market safety and financial
stability. Regulation, as a consequence, will become more prescriptive so
the market will be left with less freedom to self-regulate. Supervisors will
carry on the difficult task to keep rules up-to-date with fast-changing
market developments. Regulators should also make sure that the revised
text will be coherent with other forthcoming legislations, which will also
touch upon areas primarily falling under MiFID.
MiFID has changed the landscape of European capital markets for the
better in many ways. Most notably, the Directive has led to a more
competitive environment, huge investments in technologies, and greater
investor protection. The revision of the Directive, however, should clarify
intended scopes and fill regulatory gaps in the legal text, as well as create a
more harmonised framework of supervisory practices among member
states. It should make sure that the benefits of a new competitive
environment are spread along the value chain and passed on to final users,
retail and wholesale investors, as appropriate.
Transparency plays a crucial role in the smooth functioning of
financial markets and the monitoring of systemic risk. It also ensures that
the process of price formation works well, through efficient price discovery
mechanisms. However, transparency is not a panacea for market failures.
Ill-defined transparency requirements would harm market efficiency in
less-liquid markets with no increase in investor protection or reduction of
systemic risk. Hence, regulatory intervention should be proportional to the
214 |
nature of each market, whether auction or bilateral, and take into account
the dynamics through which orders find their market clearing price.
Pre-trade transparency supports the functioning of venues trading
mechanisms, as well as efficient price discovery and the implementation of
best execution policies. Under certain conditions, however, pre-trade
transparency may impair market liquidity.
For equity financial instruments, waivers of pre-trade transparency
should be retained. A move towards a sounder rule-based approach,
however, should be balanced with flexible application and ongoing
supervision in order to meet market needs. Conflicting views in the market
emerge when discussing the breadth of these exemptions. Regulators need
to devise a new set of rules that promotes the efficient and stability of
Europes capital markets and meets investors needs with no adverse
impact on market structure, market liquidity, efficiency, or investor
confidence. In addition, the consistent and uniform application across
Europe should be ensured.
For non-equity financial instruments, a strong push towards more
pre-trade public disclosure would require, in some cases, a rethink of the
current market structure for less-liquid asset classes, and a shift from its
mainly institutional demand to a more retail and smaller professional one.
Clashing positions in this area emerge as a result of different views around
the most efficient market structure for these products.
Liquidity in non-equity markets, such as markets for bonds,
derivatives and structured products, is mostly handled through quote-
driven auction markets, inter-dealer platforms or purely bilateral
negotiations through the direct commitment of dealers capital. For auction
markets, whether led by dealers/market-makers (quote-driven) or directly
by demand (order-driven), pre-trade transparency is strictly needed. For
inter-dealer platforms (request-for-quotes model) or bilateral negotiations,
where dealers commit capital by being non-neutral counterparties, less pre-
trade transparency than in order-driven ones (e.g. equity) could enable
them to function properly. Executable prices might thus not always be
consistently available. Current market structure, however, does not impede
future market developments in the years to arrive at a different structure of
intermediation and nature of the demand.
The alternative to a shift in market structure and demand, which may
not necessarily occur, is to design a different transparency regime from the
one applied to equities. However, an appropriate level of pre-trade
216 | CONCLUSIONS
transparency may also be beneficial for non-auction markets, as it reduces

investors search costs and promotes greater competition among dealers.
Turning to post-trade disclosure, the financial crisis called for a
further layer of transparency requirements. A new regime should include
the disclosure of aggregate data on capital markets to monitor systemic risk
and increase market integrity and efficiency. The extension of trade
reporting to both shares admitted to trading only on MTFs, or to organised
trading facilities and to equity-like instruments would be helpful, since all
these instruments serve similar purposes. However, for financial
instruments other than shares and the like, the mere extension of the rules
for equities would most probably generate inconsistencies, given their
diverse nature.
For equity markets, transparency issues remain with the quality of
OTC market data and the costs of consolidated solutions. In the post-MiFID
environment, several aspects have contributed to reduce the quality of data
and hindered its consolidation. The MiFID review should promote a
greater standardisation of both data formats (code identifiers, etc) and
flags. Market initiatives should consistently reduce the number of trade
flags, currently around 50, to fewer than 10 across Europe. In this regard,
ESMA should rather support current industry-led initiatives to improve
standardisation and reduce inconsistencies. However, either the
Commission or ESMA should be able to impose consistency if commercial
initiatives do not lead to a satisfactory solution in a reasonable timeframe.
Finally, on time limits for trade reporting, reducing the maximum
allowed delay for equity transactions from three minutes to one may prove
immaterial since this delay cannot be exploited by trading platforms in
favour of their members. In any case, the legal obligation is to report as
close to real time as possible and this duty should be duly enforced. All
market data that is not subject to delays should be freely available after 15
minutes, in line with ESMAs recommendations. Delays should be
permitted in specific circumstances, with appropriate calibration for trades
done at the end of the day.
For non-equity markets, post-trade transparency should be
consistently applied both to auction markets and purely over-the-counter
bilateral transactions. For this purpose, the post-trade transparency regime
for equities should be extended, with appropriate changes, to all financial
instruments admitted to trading on regulated markets, multilateral trading
facilities or organised trading facilities. A transparency regime disclosing
meaningful information would stimulate price discovery.
Exemptions and due calibrations should be based on dynamic

liquidity measures, to be defined at Level 2, in order to preserve efficient
price formation and to guarantee an effective monitoring of systemic risk.
Calibrations should take into account the nature of these markets and of
each financial instrument, rather than a division into broader categories
(e.g. by asset classes). Finally, dynamic measures of liquidity can be
designed around aspects such as frequency of trades, overall turnover or
prospective liquidity, product standardisation, or transaction size.
Confidential disclosure to regulators (transaction reporting) should
be extended to all financial instruments admitted to trading on RMs, MTFs
or OTFs. In addition, to monitor the build-up of systemic risk, aggregate
data on net exposures, in particular for financial instruments whose value
is linked to an underlying asset (e.g. structured financial products or OTC
derivatives) should be disclosed, leveraging current infrastructures such as
trade repositories.
Priority should be given to removing obstacles to the use of
consolidated post-trade data solutions in terms of costs and lack of data
quality. Unbundling of data services and fees would reduce costs and
increase the accessibility to consolidated data solutions. Broader actions by
competition authorities to identify potentially unfair and anticompetitive
practices in the market for data should also be undertaken. Further positive
contribution may come from the standardisation of data formats and the
use of harmonised flags. Consolidated tapes, therefore, can be designed
and offered by competing data operators once the rules of the game have
been clearly defined and duly enforced. These tapes could potentially cover
not only shares but also other financial instruments admitted to trading on
RMs, MTFs or organised trading facilities, once an appropriate post-trade
transparency regime is in place. Only if the industry fails to deliver, should
ESMA adopt the necessary arrangements to set a single consolidated tape
in the EU.
Narrowing exemptions for commodity derivatives under MiFID is an
action coherent with the post-crisis approach to regulation. However, this
action may have a substantial impact on the business of non-financial
companies by requiring them to become MiFID-licensed investment firms
and eventually to be subject to capital requirements. The consequences in
terms of the cost of hedging relevant exposures under narrower
exemptions should be further investigated.
218 | CONCLUSIONS
Curbing speculation is a vague objective, since how to distinguish

between hedging and speculative trading remains highly controversial.
Regulators should instead shed light on the risks of price manipulation that
arise from the accumulation of dominant net positions in future and
derivatives markets. Supervisory powers should be strengthened through
position limits, for physical markets or markets for non-storable
commodities (e.g. electricity). There should also be position management
mechanisms to monitor whether operators reach a net dominant position in
the market, which may lead to manipulative actions.
MiFID should ensure a harmonised approach in the application of
regulatory requirements for official trading venues. Regulated markets and
multilateral trading facilities should be subject to convergent legal
obligations and supervisory oversight across member states.
On the classification of broker-dealer crossing networks (BCNs), there
is no agreement between market participants on prospected proposals.
However, the review should clarify what kind of trades are subject to OTC
requirements under MiFID (e.g. child or parent orders). Current
proposals would bring further confusion to current definitions. For
instance, multilateral does not mean that third parties can enter an order; t
it rather means that the platform does not act on its own account where
matching transactions, and therefore does not exercise any discretion on
how trading interests should interact.
Financial market infrastructures, e.g. trading venues, are competing
networks. Stronger action by competent authorities is needed to keep
barriers to entry and exit low, giving due attention to economies of scale
and scope. Legislation and supervisory practice should work together to
realise a more open market architecture. The industrys Code of Conduct
was a positive first step in this direction. However, more needs to be done
to solve existing commercial and technical challenges. Increased
accessibility to the post-trading infrastructure could be achieved through
unbundling and better interoperability of services, together with clearer
legal definitions (e.g. legitimate commercial ground).
The systemic importance of modern capital markets highlights the
inner tensions among financial stability, market efficiency and
technological innovation. A well-functioning market must balance
efficiency and safety. A coherent set of emergency procedures in case of
market disruptions should be designed in consultation with market
participants (e.g. circuit breakers). There are a several efficient monitoring
systems already in place, which could serve as model systems.
Best execution duties lie at the foundation of the fiduciary

relationship between service providers and clients. MiFID should clarify if
execution policies should follow an obligation of means or a more general
obligation of results. More accurate execution policies and better quality
for execution data should allow sufficient verifiability of execution.
Investment advice must always be suitable under MiFID. However,
conflicts of interest may affect the quality of investment advice. These can
arise from the remuneration structure and/or from disclosure levels.
Mandating a purely fee-based remuneration structure, however, may
increase service costs and reduce access to advice by investors, with unclear
long-term effects. It might be preferable to make both remuneration
solutions available and, in addition, to mandate disclosure of all adviser
fees. In this way, investors themselves would be able to choose one that
best suits their needs.
The distinction between complex and non-complex financial
instruments should be reviewed under the caveat that complexity does not
necessarily mean more risk. The objective should be to distinguish
products according to the investors ability to understand the ultimate risk
they carry. Regulation should not decide the level of risk investors
ultimately want to take.
MiFID rules on conflicts of interest are not only disclosure
requirements but provide rules for the identification and management of
these conflicts. However, this set of rules should be coupled with
harmonised supervisory practices and a strong common set of sanctions
that would allow some flexibility to member states to adapt rules and
procedures in line with their national contexts.
REFERENCES
AFM (Authority for the Financial Markets) (2010), High frequency trading.
The application of advanced trading technology in the European
marketplace, Report, December (http://www.afm.nl/en/
professionals/afm-actueel/rapporten/2010/hft-rapport.aspx).
Akerlof, G.A. (1970), The market for lemons. Qualitative uncertainty and
the market mechanism, Quarterly Journal of Economics, Vol. 84, No. 3,
August, pp. 488-500.
Amihud, Yakov and Haim Mendelson (1996), A New Approach to the
Regulation of Trading across Securities Markets, New York University
Law Review, Vol. 71, No. 6, December.
Amihud, Yakov, Haim Mendelson and Beni Lauterbach (1997), Market
microstructure and securities values: Evidence from the Tel Aviv
Stock Exchange, Journal of Financial Economics, Vol. 45, No. 3,
September, pp. 365-390.
Angel, James J., Lawrence E. Harris and Chester Spatt (2010), Equity Trading
in the 21st Century, Working Paper (www.ssrn.com).
Areeda, Phillip (1990), Essential Facilities: An Epithet in Need of Limiting
Principles, 58 Antitrust L.J.
Armstrong, Mark (2006), Competition in two-sided markets, RAND
Journal of Economics, Vol. 37, No. 3, Autumn.
Atradia (2010), The cost of access to real time pre- and post-trade order book data
in Europe. A review of content fees within the wider context of data product,
delivery and infrastructure costs, Research Study, August.
Baruch, Shmuel (2005), Who Benefits from an Open Limit-Order Book?,
Journal of Business, Vol. 78, No. 4, July, pp. 1267-1306.
Battalio, Robert H. (1997), Third Market Broker-Dealers: Cost Competitors
or Cream Skimmers?, The Journal of Finance, Vol. 52, No. 1, March.
Battalio, Robert H., Jason Greene and Robert Jennings (1997), Do
Competing Specialists and Preferencing Dealers affect Market
Quality?, The Review of Financial Studies, Vol. 10, No. 4, October, pp.
969-993.
220 |
Baumol, William J., John C. Panzar, and Robert D. Willig (1982), Contestable
Markets and the Theory of Industry Structure, New York, NY: Harcourt
Brace Jovanovich, June.
Bessembinder, Hendrik, William Maxwell and Kumar Venkataraman
(2006), Market transparency, liquidity externalities, and institutional
trading costs in corporate bonds, Journal of Financial Economics, Vol.
82, May, pp. 251-288.
Biais, Bruno (2010), Algorithmic Trading. Equilibrium, efficiency and
stability, presentation at Market Microstructure Conference, Institut
Louis Bachelier (ILB), Paris, 6 December.
Biais, Bruno and Fany Declerk (2007), Liquidity, Competition and Price
Discovery in the European Corporate Bond Market, IDEI Working Papers,
Institut dEconomie Industrielle, University of Toulouse
(http://idei.fr/doc/wp/2007/liquidity_competition.pdf).
Biais, Bruno, Fany Declerck, James Dow, Richard Portes and Ernst-Ludwig
von Thadden (2006), European Corporate Bond Markets: Transparency,
liquidity, efficiency, CEPR Working Paper for the City of London,
Centre for Economic Policy Research, London, May
(http://www.cepr.org/PRESS/TT_CorporateFULL.pdf).
Biais, Bruno and Richard C. Green (2007), The Microstructure of the Bond
Market in the 20th Century, IDEI Working Paper, Institut dEconomie
Industrielle, University of Toulouse, 29 August (http://neeo.univ-
tlse1.fr/502/1/bondmarket.pdf).
Bloomfield, Robert and Maureen O'Hara (1999), Market Transparency:
Who Wins and Who Loses?, The Review of Financial Studies, Vol. 12,
No. 1, Spring, pp. 5-35.
Boehmer, Ekkehart, Gideon Saar and Lei Yu (2005), Lifting the Veil: An
Analysis of Pre-Trade Transparency at the NYSE, The Journal of
Finance, Vol. 60, No. 2, April, pp. 783-815.
Brogaard, Jonathan A. (2010), High-frequency Trading and Its Impact on
Market Quality, Working Paper, Northwestern University, July
(www.ssrn.com).
Buccirossi, Paolo (ed.) (2008), Handbook of Antitrust Economics, Cambridge,
MA: The MIT Press.
Burton-Taylor LLC (2010), Financial Market Data/Analysis Global Share
and Segment Sizing 2010 (http://www.burton-
taylor.com/samples/B-T_Global_Market_Data-
222 | REFERENCES
Analysis_3_Year_Segment_and_5_Year_Competitor_Analysis_2010-
Information_Kit.pdf).
Buti, Sabrina, Barbara Rindi and Ingrid M. Werner (2010a), Dynamic Dark
Pool Trading Strategies in Limit Order Markets, Dice Centre Working
Paper, 2010-6, Charles A. Dice Center for Research in Financial
Economics, Ohio State University, March.
(2010b), Diving into Dark Pools, Dice Centre Working Paper,
2010-10, Charles A. Dice Center for Research in Financial Economics,
Ohio State University, June.
Caglio, Cecilia and Stewart Mayhew (2009), Equity Trading and the
Allocation of Market Data Revenue, draft Working Paper
(http://business.rutgers.edu/files/caglio_mayhew_may272009.pdf).
Casey, Jean-Pierre (2006), Bond Market Transparency: To Regulate or not to
Regulate, ECMI Policy Brief, No. 4, December.
Casey, Jean-Pierre and Karel Lannoo (2009), The MiFID Revolution,
Cambridge: Cambridge University Press.
Celent (2008), The Evolution of Direct Market Access (DMA) Trading
Services in the US and Europe Report, March
(http://reports.celent.com/PressReleases/20080313/DMA.htm).
(2009), Demystifying and Evaluating High Frequency
Equities Trading, Report, December.
CESAME Group (Clearing and Settlement Advisory and Monitoring
Experts Group) (2008), Solving the industry Giovannini Barriers to Post-
Trading within the EU, Group Report, 28 November
(http://ec.europa.eu/internal_market/financial-
markets/clearing/index_en.htm).
CESR (Committee of European Securities Regulators) (2007), Best
Execution under MiFID. Questions and Answers, Ref CEST/07-320.
(2008), Transparency of corporate bond, structured finance product
and credit derivatives markets, Consultation Paper CESR/08-1014, 19
December.
(2009a), Impact of MiFID on equity secondary markets functioning,
Report 09/355, 10 June.
(2009b), Transparency of corporate bond, structured finance product
and credit derivatives markets, Report CESR/09-348, 10 July.
(2010a), CESR Technical Advice to the European Commission

in the context of MiFID Review Transaction Reporting, Technical
Advice CESR/10-808, 29 July.
(2010b), CESR Technical Advice to the European
Commission in the context of the MiFID Review Equity Markets,
Technical Advice CESR/10-802, 29 July.
(2010c), CESR Technical Advice to the European Commission
in the context of the MiFID Review Investor Protection and
Intermediaries, Technical Advice CESR/10-859, 29 July.
(2010d), Micro-structural issues of the European equity
markets, Call for Evidence, CESR/10-142, 1 April.
(2010e), Waivers from Pre-trade Transparency Obligations
under the Markets in Financial Instruments Directive (MiFID).
Update., CESR/09-324, 6 January.
(2010f), CESR Technical Advice to the European Commission
in the Context of the MiFID Review: Non-equity markets
transparency, Technical Advice CESR/10-799, 29 July.
(2010g), CESR Technical Advice to the European
Commission in the Context of the MiFID Review Equity Markets.
Post-trade Transparency Standards, Technical Advice CESR/10-882,
October.
(2010h), Summary of responses from investment firms and
execution venues to CESRs 2009 Best Execution Questionnaire
(Sections 1-4), Report, 19 November.
(2010i), Inducements: good and poor practices, Feedback
Statement, CESR/10-296, 19 April.
CFA Institute (Chartered Financial Analysts) (2009), Market Microstructure:
The Impact of Fragmentation under the Markets in Financial Instruments
Directive (http://www.cfapubs.org/toc/ccb/2009/2009/13).
Chaboud, Alain, Erik Hjalmarsson, Clara Vega and Ben Chiquoine (2009),
Rise of the Machines: Algorithmic Trading in the Foreign Exchange Market,
Federal Reserve Board International Finance Discussion Paper, No.
980, October (http://ssrn.com/abstract=1501135).
Cheuvreux (2010), Micro-structural issues of the European equity
markets, Call for evidence CESR/10-142, 30 April
(http://www.cesr-eu.org/index.php?page=responses&id=158).
224 | REFERENCES
Chi-X (2010), Response to CESR call for evidence on market

microstructure issues of the European equity markets
(http://www.cesr-eu.org/index.php?page=responses&id=158).
Chowdhry, B. and V. Nanda (1991), Multi-market trading and market
liquidity, Review of Financial Studies 4, 483511.
Cici, Gjergji, Scott Gibson and John J. Merrick (2008), Missing the Marks:
Dispersion in Corporate Bond Valuations across Mutual Funds, Working
Paper, July (www.ssrn.com/abstract=1104508).
Coase, Ronald (1960), The Problem of Social Cost, Journal of Law and
Economics, Vol. 3, October.
Comerton-Forde Carole and James Rydge (2004), A Review of Stock
Market Microstructure, Equity Market Microstructure Review, CMCRC
Limited and SIRCA, April.
Cooter, Robert D. and Bradley J. Freedman (1991), The Fiduciary
Relationship: Its Economic Character and Legal Consequences,
NYU. Law Review, Vol. 66, pp. 1045-1075.
CRA (Charles River Associates) International (2009), Evaluation of the
Economic Impacts of the FSAP, Study for the European Commission, 7
July.
Credit Suisse (2011), Measuring Dark Pools Impact, Market
Commentary, Portfolio Strategy, 31 January.
David S. Evans and Michael Salinger (2002), Competition Thinking at the
European Commission: Lessons from the aborted GE-Honeywell
Merger, George Mason Law Review, Vol. 10.
De Winne, Rudy and Catherine DHondt (2007), Hide-and-Seek in the
Market: Placing and Detecting Hidden Orders, Review of Finance,
Vol. 11, No. 4, pp. 663-692.
Degryse, Hans, Frank de Jong and Vincent van Kervel (forthcoming 2010),
Impact of OTC trading on invisible market depth of large Dutch
firms, preliminary results.
Degryse, Hans, Mark Van Achter and Gunther Wuyts (2008), Shedding Light
on Dark Liquidity Pools, TILEC Discussion Paper, DP 2008-039,
November.
(2009), Dynamic order submission with competition between
a dealer market and a crossing network, Journal of Financial
Economics, Vol. 91, pp. 319-338.
Demarigny, Fabrice (2010), An EU-Listing Small Business Act, Report for the
French Ministry, March
(http://www.eurocapitalmarkets.org/node/455).
Demsetz, Harold (1968), The Cost of Transacting, Quarterly Journal of
Economics, Vol. 82, No. 1, February, pp. 33-53.
Di Noia, Carmine (2001), Competition and Integration among Stock
Exchanges in Europe: Network Effects, Implicit Mergers and Remote
Access, European Financial Management, Vol. 7, No. 1, March, pp. 39-
72.
Domowitz, Ian (1990), The Mechanisms of Automated Trade Execution
Systems, Journal of Financial Intermediation, Vol. 1, pp. 167-194.
(1996), The Classification and Regulation of Automated
Trading Systems, in Andrew W. Lo (ed.), The Industrial Organisation
and Regulation of the Securities Industry, National Bureau of Economic
Research Conference Report, Chicago, IL: University of Chicago
Press.
Domowitz, Ian and Benn Steil (1999), Automation, Trading Costs, and the
Structure of the Securities Trading Industry, Brookings-Wharton Papers
on Financial Services (http://www.cfr.org/content/publications/
attachments/Brookings-Wharton.pdf).
Dow, James and Gary Gorton (2006), Noise Traders, NBER Working Paper,
No. W12256, National Bureau of Economic Research, Cambridge,
MA, May.
DTCC Deriv/Serv, CDS Data warehouse (www.dtcc.com/products/
derivserv/data/index.php).
Easley, David, Nicholas M. Kiefer, Maureen OHara (1996), Cream-
Skimming or Profit-Sharing? The Curious Rule of Purchased Order
Flow, The Journal of Finance, Vol. 51, No. 3, July, pp. 811-833.
Easterbrook, Frank H. and Daniel R. Fischel (1993), Contract and
Fiduciary Duty, Journal of Law and Economics, Vol. XXXVI, April.
Economides, Nicholas (1993), Network Economics with Application to
Finance, Financial Markets, Institutions and Instruments, Vol. 2, No. 5,
December, pp. 89-97.
(1996), The Economics of Networks, International Journal of
Industrial Organization, Vol. 14, pp. 673-699.
226 | REFERENCES
(2008), Public Policy in Network Industries, in Paolo

Buccirossi (ed.), Handbook of Antitrust Economics, Cambridge, MA:
MIT Press.
Economides, Nicholas and Robert A. Schwartz (1995), Electronic Call
Market Trading, Journal of Portfolio Management, Vol. 21, No. 3,
Spring.
Edwards, Amy K., Lawrence Harris and Michael S. Piwowar (2007),
Corporate Bond Market Transaction Costs and Transparency,
Journal of Finance, Vol. 62, No. 3, June.
EFAMA (European Fund and Asset Management Association) (2010),
Revisiting the landscape of European long-term savings A call for action
from the asset management industry, Report (sponsored by EFAMA),
March.
Ende, B., P. Gomber and M. Lutat (2008), Smart order routing technology in
the new European equity trading landscape, Working Paper E-Finance
Lab, Goethe University, Frankfurt.
Engel, Kathleen C. and Patricia A. McCoy (2002), A Tale of Three Markets:
The Law & Economics of Predatory Lending, Texas Law Review, Vol.
80, No. 6, May.
Engelen Peter and Karel Lannoo (eds) (2009), Facing New Regulatory
Frameworks in Securities Trading in Europe, Mortsel, Belgium:
Intersentia.
Enriques, Luca (2006), Conflicts of Interests in Investment Services: The
Price and Uncertainty Impact of MiFIDs Regulatory Framework, in
Ferrarini Guido and Eddy Wymeersch (eds), Investor Protection in
Europe, Oxford: Oxford University Press.
ESME (European Securities Markets Expert Group) (2009a), Fact-finding
regarding the availability of post-trade data in equities in the EU, 19
March (http://ec.europa.eu/internal_market/securities/docs/
esme/report-data-availability_en.pdf).
(2009b), Fact-finding regarding the developments of certain
aspects of pre-trade transparency in equities under MiFID, 27 July
(http://ec.europa.eu/internal_market/securities/docs/esme/report
_pretrade_11_full_esme_20090727_en.pdf).
EuroCCP (2010), Reducing Risk Among Interoperating CCPs,
Interoperability Update Briefing, 14 January
(http://www.euroccp.co.uk/leadership/index.php).
ECMI (European Capital Markets Institute) (2010), 2010 Statistical Package,

October, (www.eurocapitalmarkets.org).
European Commission (2008), Report on non-equities markets transparency
pursuant to Article 65(1) of Directive 2004/39/EC on markets in financial
instruments (MiFID), DG Internal Market and Services Working
Document, 3 April, (http://ec.europa.eu/internal_market/
securities/docs/isd/nemt_report_en.pdf).
(2009a), Commission Communication on ensuring efficient,
safe and sound derivatives markets: future policy actions, COM
(2009) 563 final, October.
(2009b), The Code of Conduct on Clearing and Settlement: Three
Years of Experience, Commission Staff Working Document to the
ECOFIN, 6 November.
(2010a), Europe 2020. A European strategy for smart,
sustainable and inclusive growth, COM (2010) 2020, 3 March.
(2010b), Public Consultation. Review of the Markets in
Financial Instruments Directive (MiFID), Consultation Report, 8
December, (http://ec.europa.eu/internal_market/securities/isd/
mifid_en.htm).
Fabozzi, Frank J. (2007), Fixed Income Analysis, CFA Institute, New Jersey:
John Wiley & Sons.
Ferrarini, Guido (2009), Market transparency and best execution: Bond
trading under MiFID, in M. Tison et al. (eds), Perspectives in company
law and financial regulation, Cambridge: Cambridge University Press.
Ferrarini, Guido and Eddy Wymeersch (eds) (2006), Investor Protection in
Europe. Corporate Law Making, the MiFID and Beyond, Oxford: Oxford
University Press.
FESE, EACH, ECSDA (2006), European Code of Conduct for Clearing and
Settlement, 7 November (http://ec.europa.eu/internal_market/
financialmarkets/docs/code/code_en.pdf).
(2007), Access and Interoperability Guideline, 28 June
(http://ec.europa.eu/internal_market/financialmarkets/docs/code/
guideline_en.pdf).
Financial Stability Board (2010), Implementing OTC Derivatives Market
Reforms, 25 October (http://www.financialstabilityboard.org).
228 | REFERENCES
Fleuriot, Pierre (2010), The Review of the Markets in Financial Instruments

Directive (MiFID), Working Group Report to the French Minister for
the Economy, Industry and Employment, February.
Flood, Mark D., Ronald Huisman, Kees G. Koedijk and Ronald J. Mahieu
(1999), Quote Disclosure and Price Discovery in Multiple-Dealer
Financial Markets, The Review of Financial Studies, Vol. 12, No. 1,
Spring , pp. 37-59.
Foucault, Thierry (1999), Order flow competition and trading costs in a
dynamic limit order market, Journal of Financial Markets, Vol. 2, pp.
99-134.
Foucault, Thierry and Albert Menkveld (2008), Competition for Order
Flow and Smart Order Routing Systems, The Journal of Finance, Vol.
LXIII, No. 1, February.
G20, (2009a), Declaration on strengthening the financial system, London
Summit, 2 April (http://www.londonsummit.gov.uk/resources/
en/PDF/annex-strengthening-fin-sysm).
(2009b), Leaders Statement: The Pittsburgh Summit,
Pittsburgh Summit, 24-25 September
(http://www.pittsburghsummit.gov/mediacenter/129639.htm).
Francioni, Reto and Robert A. Schwartz (2004), Equity Markets in Action. The
Fundamentals of Liquidity, Market Structure and Trading, John Wiley &
Sons, Inc.
Francioni, Reto, Sonali Hazarika, Maritn Reck and Robert A. Schwartz
(2008), Equity Market Microstructure: Taking Stock of What We
Know, The Journal of Portfolio Mangagement, Vol. 35, No. 1, pp. 57-71.
Frankel, Tamar (1983), Fiduciary Law, 71 CAL. L. REV., p. 795.
FSA (Financial Services Authority) (2006a), Implementing MiFIDs best
execution requirements, Discussion paper DP06/3, May.
(2006b), The Overall Impact of MiFID, Report, November.
(2008), Sponsored Access (SA), Market Watch, No. 30,
November.
(2010a), Distribution of retail investments: delivering the
RDR, Feedback to CP09/18 and final rules, March.
(2010b), Trading venue outages, Market Watch, No. 36,
August.
Gajewsky, Jean-Franois and Carole Gresse (2007), Centralised order
books versus hybrid order books: A paired comparison of trading
costs on NSC (Euronext Paris) and SETS (London Stock Exchange),

Journal of Banking and Finance, Vol. 31, No. 9, September, pp. 2906-
2924 .
Giovannini Group (2001), Cross-Border Clearing and Settlement
Arrangements in the European Union, Brussels, November
(http://ec.europa.eu/internal_market/financial-markets/clearing/
index_en.htm).
(2003), Second report on EU Clearing and Settlement
Arrangements, April (http://ec.europa.eu/internal_market/
financial-markets/clearing/index_en.htm).
Giraud, Jean-Ren and Catherine DHondt (2006), MiFID. Convergence
towards a unified European capital markets industry, London: Risk Books.
Glosten, Lawrence R. (1989), Insider Trading, Liquidity, and the Role of
the Monopolist Specialist, Journal of Business, No. 62, pp. 211-236.
(1994), Is the Electronic Open Limit Order Book Inevitable?,
The Journal of Finance, Vol. XLIX, No. 4, September.
Goldstein, Michael A., Edith S. Hotchkiss and Erik R. Sirri (2006),
Transparency and Liquidity: A Controlled Experiment on Corporate Bonds,
Working Paper, 2 March (www.ssrn.com).
Gomber, Peter and Axel Pierron (2010), MiFID. Spirit and Reality of a
European Financial Markets Directive, Report, Goethe University
Frankfurt and Celent, September.
Gomber, Peter, Gregor Pujol and Adrian Wranik (2009), The
Implementation of European Best Execution Obligations. An Analysis
for the German Market, Proceedings of the 4th International
Workshop on Enterprise Applications and Services in the Finance
Industry, Springer LNBIP, pp. 126-144.
Graf, Raoul, Fabien Durif and Mario Belzile (2008), Echo Generation:
Switching costs and the relational approach in the banking industry,
Innovative Marketing, Vol. 4, Issue 1, pp. 77-86.
Gresse, Carole (2010), Multi-Market Trading and Market Quality,
Working Paper, University Paris Dauphine
(http://basepub.dauphine.fr/handle/123456789/3148).
Group of Thirty (2010), Enhancing Financial Stability and Resilience.
Macroprudential Policy, Tools, and Systems for the Future, Special
Report, August.
230 | REFERENCES
Hagerty, Kathleen and Robert L. McDonald (1996), Brokerage, Market

Fragmentation, and Securities Market Regulation, in Andrew W. Lo
(ed.), The Industrial Organisation and Regulation of the Securities
Industry, National Bureau of Economic Research, The University of
Chicago Press.
Harris, Lawrence E. (1996), The Economics of Best Execution, USC Working
Paper, University of Southern California, Los Angeles, CA, 9 March.
(1997), Order Exposure and Parasitic Traders, Report prepared
for Deutsche Boerse AG Symposium Equity Market Structure for
Large and Mid-Cap Stocks, Frankfurt, 12 December.
(2002), Trading and Exchanges: Market Microstructure for
Practitioners, Oxford: Oxford University Press.
(ed.), (2010), Regulated Exchanges: Dynamic Agents of Economic
Growth, 50th Yearbook, World Federation of Exchanges, in
cooperation with the Centre for European Policy Studies, Oxford:
Oxford University Press.
Hazen, Thomas Lee (2006), Principles of Securities Regulation, Second
Edition, Thomas/West.
Hendershott, Terrence and Charles M. Jones (2005), Island Goes Dark:
Transparency, Fragmentation, and Regulation, The Review of
Financial Studies, Vol. 18, No. 3, Autumn, pp. 743-793.
Hendershott, Terrence, Pamela C. Moulton and Mark S. Seasholes (2007),
Market Maker Inventories and Liquidity, AFA 2008 New Orleans
Meetings Paper, 9 May (www.ssrn.com).
Hendershott, Terrence, Charles M. Jones, and Albert J. Menkveld (2008),
Does Algorithmic Trading Improve Liquidity?, WFA 2008 Paper,
March.
Hendershott, Terrence and Ryan Riordan (2009), Algorithmic Trading and
Information, NET Institute Working Paper No. 09-08, September.
Hermalin, Benjamin E., Avery W. Katz, and Richard Craswell, Contract
Law, in Mitchell Polinsky and Steven Shavell (eds.), Handbook of Law
and Economics. Volume 1, Elsevier.
Holmstrom, Bengt (1979), Moral Hazard and Observability, The Bell
Journal of Economics, Vol. 10, No. 1, Spring, pp. 74-91.
Hynes Richard and Eric A. Posner (2001), The Law and Economics of
Consumer Finance, John Olin Law and Economics WP No. 117, 20
February (www.ssrn.com).
IOSCO (International Organization of Securities Commissions) (2004),

Transparency of Corporate Bond Markets, Report of the Technical
Committee of the International Organisation of Securities
Commission, May.
(2007), Market Intermediary Management of Conflicts that
Arise in Securities Offerings, Consultation Report, Technical
Committee, February.
(2009a), Policies on Direct Electronic Access, Consultation
Report, Technical Committee of the IOSCO, February.
(2010a), Objectives and Principles of Securities Regulation,
Report, June.
(2009b), Transparency of Structured Finance Products,
Consultation Report, Technical Committee, September.
(2010b), Principles for Direct Electronic Access to Markets,
Final Report FR08/10, August.
(2010c), Issues Raised by Dark Liquidity, Consultation
Report CR05/10, October.
(2010d), Guidance for Efficient Regulation of Conflicts of
Interest Facing Market Intermediaries, Final Report, October.
International Swaps and Derivatives Association (2003): ISDA Credit
Derivatives Definitions, Supplements and Commentaries.
Jolls, Christine (2007), Behavioral Law and Economics, WP No. 12879,
National Bureau of Economic Research, Cambridge, MA, January.
Jolls, Christine, Cass R. Sunstein and Richard Thaler (2000), A Behavioral
Approach to Law and Economics, in Cass R. Sunstein (ed.),
Behavioral Law and Economics, Cambridge: Cambridge University
Press.
Klemperer, Paul (1995), Competition When Consumers Have Switching
Costs: An Overview with Applications to Industrial Organisation,
Macroeconomics and International Trade, Review of Economic Studies,
62, pp. 515-539.
Korobkin, Russel B. and Thomas S. Ulen (2000), Law and Behavioral
Science: Removing the Rationality Assumption from Law and
Economics, California Law Review, Vol. 88.
Kovtunenko, Boris (2002), Post-Trade Transparency: The More the
Better?, May (www.ssrn.com).
232 | REFERENCES
Kraakman, Reinier H. (1986), Gatekeepers: The Anatomy of a Third-Party

Enforcement Strategy, Journal of Law, Economics and Organization,
Vol. 2, No. 1, Spring.
Kumar, Kiran K., Ramabhadran S. Thirumalai and Pradeep K. Yadav
(2009), Hiding behind the Veil: Pre-trade Transparency, Informed
Traders and Market Quality, Working Paper (www.ssrn.com).
Kyle, Albert S. (1985), Continuous Auctions and Insider Trading,
Econometrica, Vol. 53, No. 6, November, pp. 1315-1335.
Lagan, Marco, Martin Peina, Isabelle von Kppen and Avinash Persaud
(2006), Implications for liquidity from innovation and transparency in the
European corporate bond market, ECB Occasional Paper No. 50,
European Central Bank, Frankfurt, August.
Langen, Werner (2010), Report on derivatives markets: future policy options,
Committee on Economic and Monetary Affairs 2010/2008 (INI),
European Parliament, 7 June.
Lannoo, Karel and Diego Valiante (2009), Integrating Europes Back
Office: 10 Years of Turning in Circles, in Peter Engelen and Karel
Lannoo (eds) (2009), Facing New Regulatory Frameworks in Securities
Trading in Europe, Intersentia, pp. 31-53.
(2011), The MiFID Metamorphosis: A new paradigm for market
structure, Journal of Securities and Custody, Vol. 3, No. 4, January.
Lee, Ruben (1998), What is an exchange? The Automation, Management, and
Regulation of Financial Markets, Oxford: Oxford University Press, New
York.
(2002), Capital Markets that Benefit Investors. A Survey of the
Evidence on Fragmentation, Internalisation and Market Transparency,
Oxford Finance Group, 30 September.
Levin, Mattias (2003), Competition, Fragmentation and Transparency. Providing
the Regulatory Framework for Fair, Efficient and Dynamic European
Securities Markets. Assessing the ISD, CEPS Task Force Report No. 46,
CEPS, Brussels, April.
Llewellyn, David T. (1995), Regulation of Retail Investment Services,
Economic Affairs, Vol. 15, No. 2, pp. 12-17.
Lo, Andrew W. (ed.) (1996), The Industrial Organisation and Regulation of the
Securities Industry, National Bureau of Economic Research, The
University of Chicago Press.
London Economics (2010), Understanding the impact of MiFID in the context of

global and national regulatory innovations, Report prepared for the City
of London, October.
London Stock Exchange (2010), Response to CESR MiFID Consultation
Paper 10-394 equity markets (www.cesr.org).
Macey, Jonathan R. and Maureen OHara (1997), The Law and Economics
of Best Execution, Journal of Financial Intermediation, No. 6, pp. 188-
223.
(2005), From Orders to Markets, Regulation, Vol. 28, No. 2,
pp. 62-70, Summer.
Madhavan, Ananth (1992), Trading Mechanisms in Securities Markets,
The Journal of Finance, Vol. 47, No. 2, June, pp. 607-641.
(1995), Consolidation, Fragmentation, and the Disclosure of
Trading Information", Review of Financial Studies,No. 8, pp. 579-603.
(1996), Security Prices and Market Transparency, Journal of
Financial Intermediation, No. 5, pp. 255-283.
(2000), Market Microstructure: A Survey, Journal of Financial
Markets, Vol. 3, pp. 205-258.
Madhavan, Ananth, David Porter and Daniel Weaver (2005), Should
securities markets be transparent?, Journal of Financial Markets, No. 8,
July, pp. 266-288.
Malinova, Katya and Andreas Park (2008), Liquidity, Volume, and Price
Behaviour: The Impact of Order vs. Quote Based Trading, 11th
Symposium on Finance, Banking, and Insurance, University of
Karlsruhe, December.
Matthews, Barbara (2010), MiFID Review and Best Execution Cross-
Border Macro-Trend Analysis, Presentation to CEPS/ECMI MiFID
Review Task Force, BCM International Regulatory Analytics LLC,
October.
McCleskey, Scott (2004), Achieving Market Integration: Best Execution,
Fragmentation and the Free Flow of Capital, Butterworth-Heinemann.
McInish, Thomas H. and Robert A. Wood (1996), Competition,
Fragmentation, and Market Quality, in Andrew W. Lo (ed.), The
Industrial Organisation and Regulation of the Securities Industry,
National Bureau of Economic Research, The University of Chicago
Press.
234 | REFERENCES
Mendelson, Haim (1987), Consolidation, Fragmentation and Market

Performance, Journal of Financial and Quantitative Market Analysis,
Vol. 22, No. 2, pp. 189-207.
Milgrom, Paul and John Roberts (1992), Economics, Organization and
Management, New Jersey: Prentice Hall Inc.
Moinas, Sophie (2010), Hidden limit orders and liquidity in order driven
markets, Working Paper, Toulouse University and IDEI
(http://idei.fr/doc/wp/2010/hidden_sm_03_2010.pdf).
Moloney, Niamh (2008), EC Securities Regulation (2nd ed), Oxford: Oxford
University Press
Monti, Mario (2010), A New Strategy for the Single Market: At the Service of
Europes Economy and Society, Report to the Present of the European
Commission, Jos Manuel Barroso, 9 May.
Naik, Narayan Y., Anthony Neuberger and S. Viswanathan (1999), Trade
Disclosure Regulation in Markets with Negotiated Trades, The
Review of Financial Studies, Vol. 12, No. 4, pp. 873-900.
Nasdaq OMX (2010), Automated Gains: How to adapt to a High-Speed
Environment, presentation by Hans-Ole Jochumsen, President,
October.
Nelson, Phillip (1970), Information and Consumer Behavior, The Journal
of Political Economy, Vol. 78, No. 2, March - April, pp. 311-329.
Nomura (2010), Market Structure. European Equity Market Structure
Update, Nomura Market Structure Newsletter, October.
Office of Fair Trading (OFT) (2008), Assessing the effectiveness of
potential remedies in consumer markets, OFT 994, April.
OHara, Maureen (1995), Market Microstructure Theory, Cambridge:
Blackwell Publishing.
OHara, Maureen and Mao Ye (2009), Is Market Fragmentation Harming
Market Quality?, Working Paper, Cornell University
(www.ssrn.com).
Oxera (2009), Monitoring prices, costs and volumes of trading and post-trading
services, Report prepared for the European Commission,
MARKT/2007/02/G, July (http://www.oxera.com/cmsDocuments/
Trading%20and%20post%20trading%20report.pdf).
(2010), Retail Distribution Review Proposals: Impact on
market structure and competition, March
(http://www.fsa.gov.uk/pubs/policy/oxera_rdr10.pdf).
Pagano, Marco (1989), Trading Volume and Asset Liquidity, The

Quarterly Journal of Economics, Vol. 104, No. 2, May, pp. 255-274.
Pagano, Marco and Ailsa Roll (1993), Auction markets, dealership
markets and execution risk, in Vittorio Conti and Rony Hamaui
(eds), Financial Markets Liberalisation and the role of Banks, Cambridge:
Cambridge University Press.
(1996), Transparency and liquidity: a comparison of auction
and dealer markets with informed trading, Journal of Finance, No. 51,
pp. 579-612.
Parlour, Christine A. (2010), Limit Order Markets: An Economic
Perspective, presentation at Market Microstructure Conference,
Institut Louis Bachelier (ILB), Paris, 6 December.
Pistor, Katharina and Chenggang Xu (2002), Fiduciary Duty in Transitional
Civil Law Jurisdictions. Lessons from the Incomplete Law Theory, Law
Working Paper, No. 01/2002, European Corporate Governance
Institute.
Posner, Richard A. (1998), Economic Analysis of Law, Aspen Law and
Business, 5th ed., January.
Ready, Mark J. (2009), Determinants of Volume in Dark Pools, AFA 2010
Atlanta Meeting Papers, 12 December (www.ssrn.com).
Renda, Andrea and Diego Valiante (2010), Legal and Economic Approach to
Tying and Other Potentially Unfair and Anticompetitive Commercial
Practices: Focus on Financial Services, Working Paper Series, February
(www.ssrn.com).
Riordan, Ryan, Andreas Storkenmaier, Martin Wagener (2009),
Fragmentation, Competition and Market Quality: A Post-MiFID Analysis,
Working Paper, Karlsruhe Institute of Technology (www.ssrn.com).
Rochet, Jean-Charles and Jean Tirole (2003), Platform Competition in Two-
Sided Markets, Journal of the European Economic Association, Vol. 1,
No. 4, pp. 990-1029.
(2004), Defining Two-Sided Markets, mimeo, University of
Toulouse.
(2006), Two-Sided Markets: A Progress Report, The RAND
Journal of Economics, Vol. 37, No. 3, Autumn, pp. 645-667.
236 | REFERENCES
Sabatini, Giovanni and Isadora Tarola (2002), Transparency on Secondary

Markets. A Survey of Economic Literature and Current Regulation in
Italy, Quaderni di Finanza CONSOB, No. 50, May.
Schmiedel, Heiko, Markku Malkamki and Juha Tarkka (2006), Economies
of scale and technological development in securities depository and
settlement systems, Journal of Banking and Finance, No. 30, pp. 1783-
1806.
Schmiedel, Heiko (2002), Total factor productivity growth in European stock
exchanges: A non-parametric frontier approach, Bank of Finland
Discussion Paper, November.
Schmiedel, Heiko and Andreas Schnenberger (2005), Integration of
securities market infrastructures in the Euro area, Occasional Paper Series
No. 33, July.
Segr, Claudio (1966), The development of a European capital market, Report of
a Group of Experts appointed by the European Commission.
(2009), Regulation of Non-Public Trading Interest, Release
No. 34-60997, File No. S7-27-09, 13 November.
(2010a), Concept Release on Equity Market Structure,
Release No. 34-61358, File No. S7-02-10, 14 January.
(2010b), Findings Regarding The Market Events of May 6, 2010,
Report of the Staffs of the CFTC and SEC to the Joint Advisory
Committee on Emerging Regulatory Issues, September 30.
Shim, Ilhyock and Haibin Zhu (2010), The Impact of CDS trading on the bond
market: Evidence from Asia, BIS Working Papers No. 332, Bank for
International Settlements, November.
Simon, Herbert (1957), A Behavioral Model of Rational Choice, in Models
of Man, Social and Rational: Mathematical Essays on Rational Human
Behavior in a Social Setting, New York, NY: John Wiley.
Smith, Reginald (2010), Is High-frequency Trading Inducing Changes in Market
Microstructure and Dynamics?, Working Paper Series, MIT Sloan
School of Management, MIT, Cambridge, MA, June (www.ssrn.com).
Stoll, Hans R. (1978), The Supply of Dealer Services in Securities Markets,
The Journal of Finance, Vol. 33, No. 4, September, pp. 1133-1151.
Swinburne, Kay (2010), Report on Regulation of trading in financial instruments
dark pools etc., Committee on Economic and Monetary Affairs,
2010/2075(INI), European Parliament, November.
Valiante, Diego (2010), Shaping Reforms and Business Models for the OTC
Derivatives Market: Quo vadis?, ECMI Research Report No. 5, ECMI,
Brussels, April (www.eurocapitalmarkets.org).
Van Cayseele, Patrick and Jo Reynaerts (2010), Complementary
Platforms, under revision, Journal of Network Economies.
Van Cayseele, Patrick and Christophe Wuyts (2006), Measuring Scale
Economies in a Heterogeneous Industry: The case of European Settlement
Institutions, SUERF The European Money and Finance Forum,
Vienna.
(2007), Cost Efficiency in the European Securities Settlement
and Depository Industry, Journal of Banking and Finance, Vol. 31.
Varian, Hal R. (2006), Intermediate Microeconomics. A Modern Approach, New
York, NY: W.W. Norton and Company Inc.
Walter, Ingo (2004), Conflicts of Interest and Market Discipline Among
Financial Services Firms, European Management Journal, Vol. 22, No.
4, pp. 361-376.
(2006), Reputational Risk and Conflicts of Interest in Banking and
Finance: The Evidence So Far, Working Paper, Stern School of Business,
NYU, New York, NY, 9 December.
Williamson, Oliver E. (1975), Markets and Hierarchies: Analysis and Antitrust
Implications A Study in the Economics of Internal Organization, New
York, NY: Free Press.
ANNEX I. LIST OF ACRONYMS
ABCP Asset-Backed Commercial Paper

ABS Asset-Based Security
ADT Average Daily (order book) Turnover
ASO Average Size of Orders
ATS Alternative Trading System
BBO Best Bid Offer
BCS Broker-dealer Crossing System
BCN Broker-dealer Crossing Network
CATS Computerised Automated Trading System
CCP Central Counterparty Clearing
CDO Collateralised Debt Obligation
CDS Credit Default Swap
CEBS Committee of European Banking Supervisors
CESR Committee of European Securities Regulators
CFTC Commodity Futures Trading Commission
CoB Conduct of Business
CoI Conflict of Interest
CRD Capital Requirements Directive
CSD Central Securities Depositories
CSDR Central Securities Depositories Regulation
CUA Customised Unilateral Access
DA Direct Access
DEA Direct Electronic Access
DMA Direct Market Access
DTCC Depository Trust and Clearing Corporation
EACH European Association of CCP Clearing Houses
EBBO European Best Bid and Offer
ECSDA European Central Securities Depositories Association
EEA European Economic Area
EFAMA European Fund and Asset Management Association
EMIR European Market Infrastructure Regulation
ECP Eligible Counterparty
ETD Exchange-Traded Derivative
ESMA European Securities and Markets Authority
ESME European Securities Markets Expert Group
238 |
FESE Federation of European Securities Exchanges

FINRA Financial Industry Regulatory Authority
HFT High-Frequency Trading
IMD Insurance Mediation Directive
IOI Indication of Interest
IOSCO International Organization of Securities Commissions
IPO Initial Public Offer
IRS Interest Rate Swap
ISD Investment Services Directive
LISO Large In-Scale Order
MAD Market Abuse Directive
MiFID Markets in Financial Instruments Directive
MTF Multilateral Trading Facility
NASDAQ National Association of Securities Dealers Automated
Quotations Systems
NBBO National Best Bid and Offer
NYSE New York Stock Exchange
OLOB Open Limit-Order Book
OTF Organised Trading Facility
PRIP Packaged Retail Investment Product
RFQ Request For Quote
RM Regulated Market
SA Sponsored Access
SEC Securities and Exchange Commission
SEF Swap Execution Facility
SFP Structured Financial product
SI Systematic Internaliser
SLD Securities Law Directive
SMS Standard Market Size
SPV Special Purpose Vehicle
STP Straight-Through Processing
TFA Transaction Feed Access
UCITS Undertakings for Collective Investments in Transferable
Securities
VWAP Volume Weighted Average Price
ANNEX II.
THE OBJECTIVES OF THE TASK FORCE
Mandate
The Task Force was composed of 31 members, who represented most
segments of European capital markets.222 It aims at increasing knowledge
around new economic and regulatory issues brought about by MiFID and
to include where agreement was found recommendations that may
support the regulatory and policy-making process. This report, in effect,
offers its contribution based on a coherent and wherever possible
agreed approach to relevant regulatory and economic issues, aimed at
avoiding haphazard or ill-informed decisions.
The Task Force is an independent and non-political group
representing major segments of the industry (see list of participants). The
report is built on the basic tenets of financial markets and it has been
drafted to be understandable not just by technical experts and to address
readers on issues and technicalities of European capital markets. Official
papers of the CESR and the European Commission provided background
documents for each of the five meetings held from June to November 2010.
Priorities
The members acknowledges that the scope of the Task Force is fairly broad,
and hence priority has been given to three main areas:
1. Pre- and post-trade transparency (including extension of post-trade
requirements to non-equity markets, and improved over-the-counter
trade transparency and reporting);
2. Trading venues classification (including regulated markets,
multilateral trading facilities and other trading venues223) and market
microstructure; and
3. Conduct of business rules (selling practices and best execution).
No representatives of issuers associations were formally involved in the work of the Task
222
Force.
223 In particular, organised trading facilities, so-called OTFs (EU COM, 2010b).
240 |
The approach taken aimed to combine theoretical aspects from

diverse disciplines (e.g. competition policy, economics, financial regulation,
etc.) with more empirical aspects (such as market data) and members
direct contributions. One of the objectives of the Task Force report therefore
is to clarify where differences of views are due to the use of different
definitions and also, most notably, to provide data in order to give issues
their real value.
The report will present as many recommendations as possible but it is
intended to include recommendations only when there is general
consensus between all the segments represented in the Task Force,
including both members and observers. Recommendations will represent
the views of the Task Force participants only, and not those of the
rapporteurs, while the content of the Report is mainly their work. The role
of the rapporteurs thus is to set out the views of the Members in a proper
and neutral manner.
The final report takes into account all different views including
minority ones explaining them in a fair and, where possible, neutral
manner. The report aims at having a balanced and well-documented set of
facts. The intention was not to seek to force consensus when views clearly
differed.
Activities and meetings of the Task Force have followed a tight
timeline from June to December 2010. As shown below, the Group held five
meetings, three of which were devoted to a discussion of the core issues of
the MiFID Review.
Figure AII.1 Task Force timeline
Materials and members contributions/presentations have been circulated

in order to enrich the content of each session. Meeting reports have been a
helpful tool to crystallise recommendations in the aftermath of each
meeting.
ANNEX III. OVERVIEW OF REGULATED MARKETS224 AND
PAN-EUROPEAN MTFS225 INFRASTRUCTURE ( SEE NOTES AT END OF THE TABLE)
Settlement
Trading Services Clearing Services Market Model
Services
Platform Type226
SETS (Order-
driven)
SETSqx LCH.Clearnet Ltd
TradElect Euroclear UK
LSE (Hybrid) SIX X-clear Horizontal
(Electronic) & Ireland
SEAQ Eurex Clearing227
(Quote-
driven)
Euroclear
LCH.Clearnet229 SA
NYSE Universal Trading Group231
Order-driven Interbolsa (Lisbon) Semi-Horizontal
Euronext228 Platform (Electronic) Interbolsa
Eurex Clearing230
(Lisbon)
Xetra 11.0
Clearstream
(Electronic)
Banking
Frankfurter
Deutsche Frankfurt233
Wertpapierbrse Order-driven Eurex Clearing232 Vertical
Brse Euroclear Bank234
Frankfurt Stock
SIS
Exchange (Floor
SegaInterSettle235
Trading)
242 |
BME Spanish
SIBE (Electronic) Order-driven Iberclear236 Iberclear237 Vertical
Exchanges
Borsa TradElect Monte Titoli
Order-driven CC&G Vertical
Italiana (Electronic) Euroclear Bank238
Euroclear
EMCF240
Nasdaq OMX INET Nordic Sweden and
Order-driven EuroCCP241 Vertical
Nordic239 (Electronic) Finland, VP
SIX X-clear242
Securities (DK)243
SIX X-clear SIS
SIX Swiss SIX Platform LCH. Clearnet SegaInterSettle246
Order-driven Vertical
Exchange (Electronic) Eurex Clearing244 Clearstream247
EMCF245 Euroclear Bank248
TradElect Oslo Clearing
Oslo Order-driven VPS250
(Electronic) LCH.Clearnet249
OASIS (Integrated
Automatic System Hellenic
Athens Order-driven ATHEXClear S.A. Vertical
for Electronic Exchanges S.A.251
Trading) (Electronic)
WARSET (WARsaw
National
Stock Exchange National Depository
Warsaw Order-driven Depository for Vertical
Trading System) for Securities
Securities
(Electronic)
Oesterreichische
Central
Kontrollbank AG
Vienna Xetra (Electronic) Order-driven Counterparty Vertical
(OeKB)252
Austria (CCP.A)
244 | ANNEXES
MMTS (Multi Central Clearing Central Clearing

Budapest Market Trading Order-driven House & Deposi- House & Deposi- Vertical
System) tory Rt. (KELER) tory Rt. (KELER)
Central Securi-
Central Securities
Prague - Order-driven ties Depository Vertical
Depository Prague
Prague
ISE Xetra 11.0 Euroclear UK &
Irish SE Order-driven Eurex Clearing Horizontal
(electronic) Ireland
BSE Horizon EFA
Bucharest Software Services Order-driven BSE Equator BSE Equator Vertical
(electronic)
OASIS (Integrated
Automatic System Cyprus Stock Cyprus Stock
Cyprus Order-driven Vertical
for Electronic Exchange CSD Exchange CSD253
Trading) (Electronic)
KDD (Central
KDD (Central
Securities
Ljubljana BTS (Electronic) Order-driven Securities Clearing Vertical
Clearing
Corporation)
Corporation)
Nasdaq OMX INET Baltic
Order-driven National CSDs National CSDs Horizontal
Baltic254 (Electronic)
Central
Bulgarian SE Xetra (Electronic) Order-driven Central depository Vertical
depository
Central Securi-
Central Securities
ties Depository
BELEXFIX Depository and
Belgrade Order-driven and Clearing Vertical
(Electronic) Clearing House a.d.
House a.d.
Beograd
Beograd
UTP (Universal
Banks and
Luxembourg Trading Platform Order-driven LCH.Clearnet Horizontal
ICSDs books
SM) (Electronic)
Central
EBOS (Electronic Central Securities
Securities
Stock Exchange Depositary of the
Bratislava Order-driven Depositary of the Vertical
Trading System) Slovak Republic
Slovak Republic
(Electronic) (CSD SR)
(CSD SR)
Malta Stock Malta Stock
Malta SE Horizon (Electronic) Order-driven Vertical
Exchange CSD Exchange CSD
Lit Pan-
Settlement
European Trading Services Clearing Services Market Model
Services
Venues255
EMCF
Chi-X Platform
Chi-X SIX X-clear National CSDs Horizontal
(Electronic)
LCH.Clearnet256
Millennium EuroCCP
National CSDs
Turquoise Exchange platform SIX X-clear257 Horizontal
Euroclear Bank259
(Electronic) LCH.Clearnet258
LCH.Clearnet260
BATS MTF
BATS Europe Order-driven EMCF National CSDs Horizontal
(Electronic)
SIX X-clear261
Universal Trading LCH.Clearnet262
NYSE Arca
Platform - NYSE Order-driven National CSDs Horizontal
Europe EuroCCP
Euronext systems
246 | ANNEXES
(Electronic)
National CSDs
Xetra Xetra 11.0
Order-driven Eurex Clearing Clearstream Vertical
International (Electronic)
Banking Frankfurt
Nomura NX
Nomura NX Order-driven Nomura NX Nomura NX Vertical
(Electronic)
LCH.Clearnet
Equiduct
Equiduct Order-driven Depending on National CSDs Horizontal
(Electronic)
home market263
Euroclear Sweden
BTP (Burgundy
and Finland, VP
Burgundy Trading Platform) Order-driven EMCF Horizontal
Securities
(Electronic)
Denmark
Six X-clear
UBS MTF UBS MTF Order-driven National CSDs Horizontal
EuroCCP264
224 For latest developments in access and interoperability for clearing and settlement of cash equities between trading platform and
CCPs/CSDs and between CCPs and/or CSDs, please see FESE, EACH & ECSDA, Joint Status Update of the Code of Conduct,
(http://ec.europa.eu/internal_market/financial-markets/docs/code/mog/20100315_fese_each_ecsda_en.pdf). There are three types of
access to the infrastructure: standard access (SA); customized unilateral access (CUA or CA); and transaction feed access (TFA).
Interoperability between infrastructures means advanced forms of relationships amongst Organisations where an Organisation is not
generally connecting to existing standard service offerings of the other Organisations but where Organisations agree to establish
customised solutions. Amongst its objectives, Interoperability will aim to provide a service to the customers such that they have choice of
service provider. Such agreement will require Organisations to incur additional technical development. Please, see FESE, EACH and
ECSDA, European Code of Conduct for Clearing and Settlement, 23-24 (http://ec.europa.eu/internal_market/financial-
markets/docs/code/code_en.pdf).
225 Firms that showed interest in entering the market (potential competitors) but have not yet been granted the link with incumbent
infrastructures are represented in Italics.

226 According to different rules, the vast majority of these order-driven markets are supported by liquidity providers, i.e. market-makers
allowed to submit quotes if needed.

227 Requested a TFA to LSE.
228 Includes Paris (CAC 40), Amsterdam (AEX), Brussels (BEL 20), and Lisbon (PSI 20).
229 NYSE Euronext has recently decided to terminate its contract with LCH.Clearnet and it will start soon clearing trades in-house;
(http://www.efinancialnews.com/story/2010-05-24/nyse-euronext-challenges-cme).
230 Requested a TFA to Euronext Lisbon, Paris, Amsterdam and Brussels and a TFA to Euroclear Belgium, France and Netherlands, and
Interbolsa.
231 Euroclear Group comprises International Central Securities Depository (ICSD) Euroclear Bank in Brussels and Central Securities
Depositories (CSDs) Euroclear Belgium, Euroclear Finland, Euroclear France, Euroclear Nederland, Euroclear Sweden and Euroclear UK &
Ireland.
232 TFA request from SIX x-clear not live.
233 Part of Link Up Markets, a joint venture by ten leading Central Securities Depositories (CSDs) Clearstream Banking AG Frankfurt
(Germany), Cyprus Stock Exchange (Cyprus), Hellenic Exchanges S.A. (Greece), IBERCLEAR (Spain), MCDR (Egypt), Oesterreichische
Kontrollbank AG (Austria), SIX SIS AG (Switzerland), STRATE (South Africa), VP SECURITIES (Denmark) and VPS (Norway). Its key
objective is to improve efficiency and reduce costs of post-trade processing of cross-border securities transactions by streamlined
interoperability on the CSD layer.
Launched on 30 March 2009, Link Up Markets has established a common infrastructure allowing for streamlined interoperability between
CSD markets and introducing efficient cross-border processing capabilities. The solution enables CSD customers to significantly reduce
the cost gap between settling and safekeeping domestic and foreign securities.
234 TFA with Eurex Clearing requested.
235 Requested TFA with FWB and Eurex, and SA or CA with Clearstream.
236 Iberclear provides risk management services. A project to set up a CCP is in progress, with a consultation open by the Spanish
Supervisor (CNMV).
237 Part of Link Up Markets. See footnote 233 above.
238 Requested a TFA with CC&G.
239 Includes Copenhagen (OMX Copenhagen 20, or OMXC20), Stockholm (OMX Stockholm 30 Index, or OMXS30), Helsinki (OMX
Helsinki 25, or OMXH25), and Iceland (OMX Iceland 6 PI ISK, or OMXI6ISK).

240 European Multilateral Clearing Facility; EMCF is owned 78% by ABN AMRO Bank N.V. and 22% by NASDAQ OMX AB.
248 | ANNEXES
241 Not live yet. MoU with Nasdaq OMX to clear equities for Copenhagen, Stockholm and Helsinki.
242 Not live yet. Interoperability between CCPs dealing with trades executed on Nasdaq OMX Nordic (EMCF).
244 It requested: TFA and Interoperability with SIX X-clear; interoperability with LCH.CLearnet; and TFA with SIX Platform and SIS
SegaInterSettle.
245 TFA requested.
247 SA or CA with SIS SegaInterSettle and SIX Platform requested.
248 TFA with X-Clear requested
249 Interoperability with Oslo Clearing requested.
254 Includes Tallinn (OMX Tallinn, or OMXT), Riga (OMX Riga, or OMXR), and Vilnius (OMX Vilnius, or OMXV).
255 The biggest 11 venues by turnover (Thomson Reuters data; January - May 2010).
256 Requested a TFA to Chi-X and Interoperability with EMCF.
257 Requested a TFA to Turquoise and Interoperability with EuroCCP.
258 TFA to Turquoise requested.
259 TFA to EuroCCP requested.
260 TFA to BATS Europe requested (for all Irish securities).
261 Not live yet.
262 TFA to NYSE Arca requested.
263 SIX x-clear for German stocks.
264 To be activated when interoperability models have been approved by national authorities.
ANNEX IV. FIGURES AND TABLES
ILLUSTRATING EQUITY MARKETS
Figure AIV.1 Global market domestic capitalisation 1993-2009 ($ million)

25,000,000
20,000,000
15,000,000
10,000,000
5,000,000
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
BRICs US Europe AfricaMiddleEast
Source: WFE.
| 249
250 | ANNEXES
Table AIV.1 Market capitalisation ($ million)

NYSE Euronext (US) 11,837,793.3
LSE Group 3,452,292.6
Tokyo SE 3,306,082.0
NASDAQ OMX 3,239,492.4
NYSE Euronext (Europe) 2,869,393.1
Shanghai SE 2,704,778.5
Hong Kong Exchanges 2,305,142.8
TSX Group 1,676,814.2
BME Spanish Exchanges 1,434,540.5
BM&FBOVESPA 1,337,247.7
Deutsche Brse 1,292,355.3
Source: WFE.
Figure AIV.2 European (10 biggest trading venue) aggregate annual turnover
($ million)
14,000,000
12,000,000
10,000,000
8,000,000
6,000,000
4,000,000
2,000,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
MillionEURAnnualTurnover Average
Sources: BATS Europe, FESE and Thomson Reuters.

Table IV.2 EEA market share by national market
London 22.385%
Frankfurt 16.474%
Paris 15.646%
Milan 9.652%
Zurich 8.443%
Madrid 6.752%
Amsterdam 6.318%
Stockholm 5.409%
Oslo 2.377%
Helsinki 2.065%
Copenhagen 1.128%
Brussels 1.119%
Lisbon 0.467%
Vienna 0.366%
Others 1.399%
Sources: BATS Europe, Thomson Reuters (Jan-Dec 2010, % turnover; lit and auction books).
Figure AIV.3 Trading volumes by trading venue, 2010

0 100,000,000,000 200,000,000,000 300,000,000,000
London
Milan
ChiXEurope
BATSEurope
BolsadeMadrid
NasdaqOMXNordic
Paris
DeutscheBrse
Turquoise
OsloBrs
Amsterdam
Warsaw
SIXSwissExchange
Other
DarkPools
Sources: BATS Europe, Thomson Reuters (Volumes Lit, Auction, Dark Books).
252 | ANNEXES
Figure AIV.4 Turnover by order book, 2010
1.468%
57.22% 34.277%
7.033%
0% 20% 40% 60% 80% 100%
OrderBook Lit OrderBook Auction DarkOB+Hidden MiFIDOTCTradeReported
Source: Thomson Reuters (2010 turnover).

Figure AIV.5 EU27 debt and securitised financial instruments

(outstanding; billion)
16,000,000
14,000,000
12,000,000
10,000,000
8,000,000
6,000,000
4,000,000
2,000,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
GovernmentDebtSecuritiesMarket,Outstanding CorporateDebtSecuritiesMarket,Outstanding
FinancialInstitutionsDebtSecuritiesMaket,Outstanding Securitisedproducts,Outstanding
Source: ECMI (2010), BIS, AFME.

254 | ANNEXES
Figure AIV.6 EU27 issuance of debt and securitised financial instruments

( billion)
3,000,000
2,500,000
2,000,000
1,500,000
1,000,000
500,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
NetIssuanceofGovernmentDebtSecurities NetIssuanceofCorporateDebtSecurities
Securitisedproducts(issuance) NetIssuanceofFinancialInstitutionsDebtSecurities
Source: ECMI (2010); data from BIS, AFME.

Figure AIV.7 EU27 financial highlights (1) ( million)
AggregateSizeofDebt
SecuritiesMarket,
Outstanding
25,000,000
20,000,000
15,000,000
10,000,000 EquityDomestic
GDP
MarketCapitalisation
5,000,000
GlobalOTCDerivatives
EquityTotalValueof
Market,GrossMarket
ShareTrading
Values
2000 2005 2009

Sources: ECMI (2010) data from BIS, Eurostat, WFE.
256 | ANNEXES
Figure AIV.8 EU27 financial highlights (2) ( million)

GDP
25,000,000
20,000,000 AggregateSizeof
Exchangetraded
DebtSecurities
derivatives(th
15,000,000 Market,
contracts)
Outstanding
10,000,000
5,000,000
0
GlobalOTC
AggregateNet
Derivatives
IssuanceofDebt
Market,Gross
Securities
MarketValues
EquityDomestic
EquityTotalValue
Market
ofShareTrading
Capitalisation
2000 2005 2009
Sources: ECMI (2010) data from Eurostat, BIS, WFE,
Figure AIV.9 OTC (notional) vs exchange-traded derivatives (ETDs) ( billion)

500,000
450,000
400,000
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
2002 2003 2004 2005 2006 2007 2008 2009
Exchangetradedderivatives,Outstanding OTCderivatives,Notionalamount
Sources: ECMI (2010), data from BIS, WFE (ETD are estimations by default).
258 | ANNEXES
Figure AIV.10 OTC (gross market value) vs exchange-traded derivatives

( billion)
90,000
80,000
70,000
60,000
50,000
40,000
30,000
20,000
10,000
0
2002 2003 2004 2005 2006 2007 2008 2009
Exchangetradedderivatives,Outstanding OTCderivatives,Grossmarketvalue
Sources: ECMI (2010), data from BIS, WFE (ETD are estimations by default).
Figure AIV.11 Global debt and derivatives markets vs world GDP ( billion)
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Securitisedproducts,Outstanding GlobalOTCDerivativesMarket,GrossMarketValues(outstanding)
GovernmentDebtSecuritiesMarket,Outstanding CorporateDebtSecuritiesMarket,Outstanding
FinancialInstitutionsDebtSecuritiesMarket,Outstanding WorldGDP
Sourcse: ECMI (2010); data from BIS, Eurostat, World Economic Outlook, AFME.
ANNEX V. TASK FORCE PARTICIPANTS
Disclaimer
The views expressed in this report do not necessarily reflect the views and positions of all
members involved in the Task Force. Members do not necessarily agree with all relevant
undertaken positions and do not necessarily endorse any reference to academic and
independent studies. A sound and clear set of principles has guided the drafting process, in
order to preserve a neutral approach to divergent views. All members views have been
heard and if well-grounded incorporated in the final text. When fundamental
disagreement materialised, the Rapporteurs attempted to explain all views in a proper and
fair manner. Finally, members have received enough time to comment and to contribute to
each version of the Final Report, which can be only attributed to the Rapporteurs and no
one else within the Group. Data included in the text have been generally considered as
material and relevant.
Chair
Pierre Francotte
Former CEO, Euroclear
General Manager, PLF International
Rapporteurs
Diego Valiante, Ph.D. Karel Lannoo

Research Fellow CEO and Senior Research Fellow
ECMI-CEPS CEPS-ECMI
Members
Ludovic Aigrot Georg Baur
Head of EU Affairs Chair Securities Working Party
NASDAQ OMX European Banking Federation
George Andreadis Catherine Blake

Head of AES Liquidity Strategy Legal Counsel
Europe CME Group
Credit Suisse Securities
Andrew Bowley
Rosa Armesto Head of Electronic Trading Product
Head of Fixed income and Management
Economics & Statistics Nomura International
FESE - Federation of European
Securities Exchanges
| 260
Bertrand Brhier Huong Hauduc

Deputy Head Banking and Financial Associate Director and Assistant
Regulation General Counsel
Socit Gnrale CME Group
Vanessa Bruynooghe Samuel Hinton-Smith

Compliance Officer Market Director Public Affairs
BNP Paribas Fortis Nomura International Plc
Michele Calderoni Karl-Peter Horstmann

Treasury and Capital Markets Head of Markets Regulation
ENEL RWE Supply and Trading GmbH
Simon Crown Bertrand Huet

Partner Senior Policy Advisor
Clifford Chance Fleishman Hillard
Jrme Desombre Henrik Husman

Director - Strategy Vice President Nordic Cash Equity
Euroclear Products
NASDAQ OMX
Godfried De Vidts
Director of European Affairs Burak Inel
ICAP Deputy Secretary General
FESE -Federation of European
Graham Dick Securities Exchanges
Head of Business Development
Chi-X Europe Denzil Jenkins
Director of Regulation
Andrew W. Douglas Chi-X Europe
Head of Public Affairs
DTCC - Depositary Trust and Sophia Kandylaki
Clearing Corporation Director
Markit Group Limited
Florence Fontan
European Affairs Rene Karsenti
BNP Paribas Securities Services Executive President
ICMA - International Capital Market
Benot Gourisse Association
Deputy Director
French Banking Federation
262 | ANNEXES
Sacha Kumaria Frank Pottie

Deputy Director of European Affairs Director - Public Affairs
ICAP Euroclear
Sonja Lohse Cian Rice

Head of Group Compliance Partner
Nordea Bank Kreab Gavin Anderson
Graziella Marras Donald Ricketts

Senior Policy Advisor Head of Financial Services
EFAMA - European Fund and AM Fleishmann-Hillard
Association
Tillie Rijk
Barbara Matthews Regulatory Affairs Analyst
Founder and CEO Deutsche Bank
BCM International Regulatory
Analytics LLC
Julia Rodkiewicz
Policy Manager
Nick Miller International Swaps and Derivatives
Associate Ass. (ISDA)
Morgan Stanley
Susanne Rompel
Michele Morena EU Liaison Office
Associate Director RWE
Kreab Gavin Anderson
Jos Manuel Santamara
Celia Neves International Affairs
Representative BME - Bolsas y Mercados Espaoles
Nordea Bank
Marcus Schueler
Francesca Passamonti Managing Director
Representative before the European Markit Group Limited
Union
Intesa Sanpaolo
Guy Sears
Director of Wholesale
Jeremie Pellet Investment Management Association
Strategic Analyst (IMA)
BNP Paribas
David Self
Product Manager
Xtrakter
Martin Sjberg Hendrik van Vliet

Manager European Affairs Senior Adviser Markets Industry
CFA Institute ABN AMRO Clearing
Enzo Stingone Juan-Pablo Urrutia

Vice President Connectivity Associate and Counsel
Chi-X Europe Goldman Sachs International
Steven Travers Helena Walsh

Head of Regulatory Law and Manager
Strategy Cicero Consulting
London Stock Exchange Group
Santiago Ximenez
Dermot Turing Director Market Data
Partner BME - Bolsas y Mercados Espaoles
Clifford Chance
Elina Yrgard
Geert Vanderbeke Legal Counsel EU Affairs
Head of Sales Europe NASDAQ OMX
ABN AMRO Clearing
Jack Vensel
Managing Director
Citi
Observers
Hans Degryse Valerie Ledure

Professor of Financial Intermediation Policy Officer, European Commission
and Markets CentER - Tilburg
University Rhiannon Price
Assistant to MEP Kay Swinburne
Salvatore Gnoni European Parliament
Seconded National Expert
European Commission Patrick Van Cayseele
Professor of Economy and Vice-Dean
Stephen Hanks Catholic University of Leuven
Policy Manager, Conduct Policy
Financial Services Authority Hans Wolters
Head of Policy
Hannes Huhtaniemi
Dutch Financial Authority
Policy Officer Securities Markets Unit
European Commission
George Zavvos
Legal Adviser, European Commission

MiFID e Version

Uploaded by

Copyright:

Available Formats

MiFID e Version

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MiFID e Version

Uploaded by

Copyright:

Available Formats

MiFID 2.

REPORT OF THE ECMI-CEPS TASK FORCE

CHAIR PIERRE FRANCOTTE

RAPPORTEURS DIEGO VALIANTE

CENTRE FOR EUROPEAN POLICY STUDIES

The rapporteurs gratefully acknowledge the contributions provided by all

European Capital Markets Institute

The review of the Markets in Financial Instruments Directive (MiFID) is a

trading facilities and to equity-like instruments would be helpful, since all

passed-on to final users. Regulators and competition authorities should

auction markets in order to stimulate investors willingness to bid, since

Derivatives and structured financial products

transparency than order-driven ones, such as equity markets. Executable

Structured financial products (SFPs) and OTC derivatives are good

Financial market infrastructures are networks that can operate in a

The systemic importance of modern capital markets highlights the

detriment of final investors as long as execution policies are properly

All cost items and remuneration arrangements should be fully

Organisational requirements play a crucial role in ensuring business

1.1 The MiFID review

its level of detail, the review of MiFID is a very technical undertaking.

Table 1. The MiFID review

The Committee of European Securities Regulations (CESR) has

Figure 1. Interaction with main EU regulatory actions

2 Replacing CESR as of January 2011.

creation of a new harmonised framework of rules across Europe that had

Figure 2. MiFID timeline

Source: European Commission website and rapporteurs.

2.1 The legislative process

4 See European Council, Presidency Conclusions, 15-16 June 1998, Cardiff

better legislative process for innovative financial regulation. The FSAP

and far-reaching attempt of the Commission to provide the European

2.2 The legal structure

9 See Box 16.

principles set by the Financial Services Action Plan (FSAP): i) a single

Investorprotection Orderlymarketsand Marketefficiency

Organisational Conductofbusiness Marketaccessand

4) Market structure and transparency (abolition of the concentration

extension of the ISD European passport regime (for other investment

12 Art. 19.8, MiFID and Arts 40-43 Implementing Directive.

requirements address typical issues in a fiduciary relationship between

with maarket fragmenttation),

The imp pact on price formation

3.2 A more competitive environment

(MTFs) and systematic internalisers (SIs).

139 Equity DarkMarkets

Box 1. Defining dark in securities trading

On the one side, a more competitive environment produced positive

Figure 6. Bid/ask average spread and implicit volatility on EuroSTOXX, FTSE100

Source: NYSE Euronext (Response to CESR Consultation; CESR, 2010d).

Copenhagen 0.964% Amsterdam 4.292%

25 See, more generally, chapter 1.

These new trading platforms are gradually increasing their market

Figure 8. Most liquid EU shares by execution venue

As shown below, market shares of incumbent-regulated markets

As a result of fierce competition, trading venues are struggling to

3.3 The introduction of new technologies

professional investors in order to improve their trading strategies, often supported by

speed of execution for small, immediately executable orders was 10.1

Table 2. Share of HFT by trading venue (shares of order books)