DIGITAL FINANCE

UNIT 1: INTRODUCTION TO FIN-TECH

1-THE ROLE OF FINANCE: IMPROVING YOUR LIFE

FINANCE: Area of Economics that analyses investment and financing decisions and the
exchange of funds that these decisions generate.
INVESTMENT: Technology that offers future payments in return of today’s ones. Types
of investments:
- Real Assets: Tangible and intangible goods that, if put to productive use, generate
future wealth.
- Speculative Assets: Unproductive goods whose value changes over time.
- Financial Assets: Contracts that give the right to receive a future payment in
exchange for a payment today.
FINANCING: Obtaining funds to pay for present investments, in exchange
for future payments.
Financial markets: markets where financial assets are issued and exchanged

Consumption allows people to satisfy their wants, but wealth is limited. Each agent
chooses its consumption in each period considering his current wealth, investment
opportunities in real and speculative assets and his financial assets and liabilities.
Decisions in each period affect decisions in subsequent periods.

2-GENERAL PRINCIPLES IN FINANCE: NO FREE MEALS AND OTHER UNCONFORTABLE

TRUTHS

2.1. Law of One Price; Nothing is for free.

Law of one price: If two portfolios have the same payoffs (in every state of nature),
then they must have the same price. It works through arbitrage.
Arbitrage: If payoff A is always at least as good as payoff B, and sometimes A is better,
then the price of A must be greater than the price of B.
Arbitrage opportunity: taking advantage of price differences between two or more
“markets” (places – timing). Opportunity that allows investors to obtain riskless
profits at no initial cost with no initial capital outlay.
- If an asset is quoted in two different markets at a different price, there is an
arbitrage opportunity consisting of ¿?
- If the market works correctly, the arbitrage opportunity gradually disappear.
Why?

2.2. People have to make a rational lifetime resource allocation.

Maximize welfare subject to the constraint that the present value of lifetime
consumption cannot exceed the present value of lifetime labour earnings and earnings
from other investments adjusted for bequests of wealth. Financial markets allow you
to move money around time periods allowing for a better allocation of resources over
your lifetime.
2.3. Use a rational model to maximize expected welfare and minimize risks.
Moving money across time requires borrowing, saving and investing. Borrowing allows
you to spend on consumption more than your current income.
Saving means not spending current income on consumption. Money that is saved has
to be invested (which is the alternative??).
Investing means deciding how to choose investment assets.
- Use NPV criterium.
- Be careful! Without barrier to entry competition drives NPV to zero.
- Use risk-return preferences for assets with NPV=0.
- In well diversified portfolio only relevant risk is market risk (CAPM’s b)
- Careful! No free lunch: high return high risk.

When investment, risk can be reduced through 3 strategies:

Insuring Buying a contract to compensate you for a loss that is much larger than the
insurance premium. Insuring is cost-effective when potential loss is large and
likelihood relatively low.
Diversifying “Not putting all your eggs in one basket.” Some of the gains will cancel
some of the losses, so the riskiness of a portfolio will be lower than the risk of an
investment of equal size in single asset.
Matching To eliminate risk of falling short of a savings goal at a specific future date,
you must match the maturity of your investment to the goal (example: pensions!).

2.4. Imperfect capital markets (because of uncertainty, asymmetric information,

agency) imply taxes and transaction costs. To save-invest, insure, diversify or match
you have to use the capital markets. Capital markets are imperfect and because of this
using them is costly. Take this into account when computing expected returns. Costs of
imperfect capital markets very clear in borrowing and lending (due diligence,
collateral).

3-A SPECIAL ASSET: MONEY (CURRENCY, BANK DEPOSITS AND CENTRAL BANK
RESERVES)

Money is a financial asset, only it pays no interest. To be useful money must be:
1. Durable - it needs to last
2. Portable - easy to carry around, convenient, easy to use
3. Divisible – it can be broken down into smaller denominations
4. Hard to counterfeit – it can’t easily be faked or copied
5. Must be generally accepted by a population
6. It must have a relatively stable and liquid value i.e. can hold value over time and can
be bought and sold fast. People will not accept (offer) in an exchange something that
can lose (win) lots of value fast. When money has a stable value it is also used as a unit
of account: providing a common measure of the value of goods and services being
exchanged. Comparability facilitates transactions, investment and economic growth.
Why then fiat money?
- Paper is worth nothing.
- Value of fiat money is in the promise of the issuer to pay.
 - The value is backed by states ability to tax and promise not to manipulate.
- (Swizerland vs. Venezuela)
- Lots of manipulation is bad: hyperinflation destroys usefulness of money as store-
of-value (e.g. Weimar and Venezuela) or as a medium of exchange. It contracts
economic activity. But, ¡surprise! Fiat money exits because some “manipulation”
(management, not possible with gold) is good.

Growth rate should follow economic growth to avoid both deflation and inflation.
- Cost of deflation: it discourages investment (e.g. silver mines and fall of the
Roman Empire, great financial crisis). Ø Prevented if central bank prints more
money.
- Cost of inflation: menu costs, inefficient adjustments across industries. Ø
Prevented if central bank reduces supply of money.
Central banks “manipulate” very carefully to balance macroeconomic objectives with
credibility and stability. It is important that they are independent from the
government.

4-TECHNOLOGY AND THE FINANCIAL INDUSTRY: A LOVE STORY THAT MAY END UP IN
DIVORCE

Technology: collection of skills, methods and processes used in the production of

goods and services or in the accomplishment of objectives. Can be embedded in
machines.
FinTech refers to the application of technology to finance (at the broad level).
Therefore, FinTech is not new for the financial services industry. Going back in time we
start with the first financial contracts in writing in Mesopotamia. Writing, abacus,
double entry accounting, joint stock companies with limited liability, option pricing
formula where new technologies that had a huge impact on finance.
- 1453 printing machine -> 1661 first bank note -> 1776 Adam Smith
FinTech1. 1866 to 1967. The financial services industry, heavily interlinked with
technology, but remained largely analogue. The introduction of the telegraph (first
commercial use in 1838) and the laying of the first successful transatlantic cable in
1866 (by the Atlantic Telegraph Company) provided the fundamental infrastructure for
the first major period of financial globalization in the late 19th century until the First
World War.

FinTech2. 1967 to 2008. Finance became an industry based upon transmission and
manipulation of digital information.
Introduction of the Automatic Teller Machine (ATM) in 1967 by Barclays Bank marks
the commencement of the transformation from analogue to digital industry (although
calculators, computers, faxes and credit cards are even earlier). Today, the ATM is
often the only point for most consumers at which finance transitions from a purely
digital experience to one that involves a physical commodity (i.e. cash). Financial
services industry single largest purchaser of IT.
Payments are digital: Inter-Computer Bureau established in UK in 1968 (basis of
today’s Bankers’ Automated Clearing Services (BACS)). Interconnected domestic
payments systems across borders through the Society of Worldwide Interbank
Financial Telecommunications (SWIFT) established in 1973.
Trading is digital: Establishment of NASDAQ30 in the US in 1971, transition from
physical trading of securities dating to the late 1600s to today’s fully electronic
securities trading.
Banking services are digital: Online banking development during the 1990’s with direct
banks without physical branches (e.g. ING Direct)

Regulatory issues where brought about by these new technologies:

- Easy virtual bank runs
- Increased competition and deterioration of credit standards
- Large unexpected movements in interlinked global stock exchanges caused by
automated trading (Black Monday 1987).
- Data privacy and security (data centres, back-office security)
1990 Internet -> 2008 the bitcoin -> 2020 digital yuan -> 2025 digital euro -> next?

Quotes: “The most important financial innovation that I have seen the past 20 years is
the automatic teller machine, that really helps people and prevents visits to the bank,
and it is a real convenience.” Paul Volcker (former chairman of the US Federal Reserve
(1979-1987), The only thing useful banks have invented in 20 years in the ATM, NEW
YORK POST (Dec. 13, 2009). “One third of Goldman Sachs’ 33,000 staff are engineers –
more than LinkedIn, Twitter or Facebook.” Jonathan Marino, Goldman Sachs is a tech
company, BUSINESS INSIDER AUSTRALIA (Apr. 13, 2015, 04:20 AM).

FinTech3. 2008 onwards. The current change arises not so much from the new
technologies but from the fact that they come together with a change in who is (or is
not) applying these technologies to finance. From supervised financial institutions to
new start-ups and established technology companies that have begun to deliver
financial products and services directly to businesses and the public. Concerns of
policy-makers and industry is not only with technology but with trust and
intermediaries. Technology is changing in five major areas of finance:
(1) Investment: Crowdfunding and P2P lending as an alternative to venture capital,
private equity, private placements, public offerings, listings etc. Also program trading,
high frequency trading and dark pools. Robo-advisory services.
(2) Internal operations and risk management: Automatic systems based upon VaR and
other quantitative techniques to manage risk. Massive IT spending by financial
institutions.
(3) Payments and infrastructure: Internet and mobile communications payments
(specially in developing countries). § Infrastructure for securities trading and
settlement and for OTC derivatives trading.
(4) Data security and monetization: “Big data” can be applied to enhance the efficiency
and availability of financial services. Monetary value of data can be exploited.
Vulnerability to cybercrime for consumers and for the system (after the GFC, clear that
stability of financial system is a national security issue).
(5) Customer interface: Opportunities for products bundling.
Now days: BBVA USA announce an agreement with Google to offer digital accounts in
the USA. Orange Bank get the operating approval from the Spanish National Bank.
Movistar Money provided more than 16.000 loans. Google get the the first licence in
order to be able to compete with the European banks. Thousands of specialized
boutiques are appearing offering a wide range of services. Traditional money is slow,
big and inefficient. Fiscal and monetary policies converge. Mirar pp

5-FINANCIAL REGULATION: THE “MORAL” IN FINANCES

Is finance good? Is it helpful to people? YES. Why then is public perception not so
good? Many people think of finance as a zero sum game with no growth
and no wealth creation (e.g. laws against usury, religious views against borrowing,
courts judge derivatives from ex-post perspective). This is incorrect. Finance generates
growth and wealth. This is good. Humanity started off all equal and living in poverty,
and we don't particularly like that. Wealth creation can reduce inequality but it can
also increase inequality or its perception. Deng Xiaoping's famous statement in the late
1970's, when China was adopting modern financial methods. And some people were
getting rich, and someone asked Deng, isn't this inconsistent with our ideology? He
said: “we're all going to get rich, but somebody has to get rich first.”

How to combat the inequality generated by financial activities? Communism and

kibbutz have failed (Why?). Taxation and the welfare state work to some extend. But
problems of efficiency (and even corruption) abound because it is difficult to design
good incentives with asymmetric info and agency problems (e.g. unemployment
insurance). Regulation: Regulators design and enforce rules both at the
microprudential and the macroprudential level.
- Microprudential level: Protect non informed consumers of financial services and
products from misbehavior by insiders. Facilitate competition, giving more
consumers access to finance. Requires auditing, regulation of self-dealing
operations, disclosure of inside information, etc. Regulation requires trusted
public and private information intermediaries. This is, institutions that provide
the financial markets with credible information about financial institutions:
Central banks, SEC, Competition authorities, Auditors, analysts, rating agencies,
the press.
- Macroprudential level: Ensuring financial stability (crisis usually hit harder lower
income citizens). Main issue is the ''too big to fail'' problem. Time inconsistency in
enforcement of regulation for big firms and bail outs make the system risky and
can produce systemic crisis. Regulators (like referees) don't attract attention,
except when they make mistakes.
Deciding when and what to regulate: Wait-and-see. But new technologies can imply
fast transition from Too-small-to-care to Too-large-to-ignore and Too-big-to-fail.

Current public perceptions of financial regulation in developed countries:

Financial crisis (CDO, subprime lending, etc.) undermined credibility and trust of
regulated financial institutions. Survey Shows Americans Trust Technology Firms More
Than Banks and Retailers. It has lead to increased regulation (Dodd Frank Act, Basel 3)
raising the costs of banks vis a vis other players (compliance obligations, ring-fencing
obligations, increased regulatory capital, Recovery and Resolution Plans (RRPs) across
jurisdictions, stress tests to evaluate their viability.) Most of this regulations do not
apply to alternative providers of finance (Ana Botín denuncia en Bruselas la
competencia desleal de los gigantes tecnológicos de EEUU). For over 1000 million
unbanked individuals, banking may be a commodity that can be provided by any
institution, regulated or not.

UNIT 2: DISTRIBUTED PLATFORM LEDGERS AND BLOCKCHAINS:

BITCOINS, TOKENS AND BEYOND
1-BASIC INSIGHTS: TRUST AND DISTRUST

Moving wealth around (money, financial assets or other assets as real state) requires a
central trusted intermediary (institutions that intermediate trust) that can prevent
double-spending through their signature.
TRUST/DISTRUST -> CETAINTY/UNCERTAINTY -> LESS BRAKES/MORE BRAKES
Common money is worth what is worth because there ́s an authority we trust
(Central banks) that say so. Therefore, trust is a value that can be monetized.
In the past, pacts were made between individuals, then there were intermediary
institutions. Now people trust in this order more in a person like them, the big
corporations and finally the official government regulators.

DPL and Blockchain provide a technological solution that (supposedly) can do away
with these intermediaries and their signature replacing them with competition among
interested non-trusted parties that produces a consensus based on proof of work.
So far they have been used as money (bitcoin) but its promise to extend into corporate
governance, democratic participation, capital markets... (BLOCKCHAIN : A NEW
MEANING FOR “TRUST”).
TECNOLOGY allow us to trust in people we don’t know (Airbnb, blablacar). From
“INSTITUTIONAL” trust (Opaque, closed, centraliced, formal, top-down) to
“DISTRIBUITED” trust (transparent, inclusive, decentraliced, informal, bottom-up).
Blockchain is a big information chain fed massively with in-puts.
- Inalterable: the transactions record is inalterable
- Verifiable: updates must be approved by the majority of connected computers.
- Distributed: each block is delivered in real time to all connected computers
- Decentralized: records are decentralized (each record has a replica in each comp)
- Secure: blocks are encrypted
- Direct technology: communication among computers is direct
Its benefits: boost transaction growth, in a fast way, and more open way (accessible)
Blockchain: registry of assets and transactions (1991: Stuart Haber y W. Scott
Stornetta). Transaction history is stored in blocks of data, secured by cryptography,
and get locked in blocks of data over the time with cryptography as well. Each block
contains hash from the previous block, data and hash block. Hash:
- It’ unique and unrepeatable
- Identifies the block and all its content
- Blocks are sequentially connected (like plugs)
- If something change inside the block, the hash change. Therefore, the block is not
the same anymore: it’s detected by the following block
- The record is replicated in all computers connected to the network
What happens if someone manipulates/tampers the hash? All the following blocks
became invalid became invalid. However, due to nowadays computer capacity you
could eventually recalculate hashes to amend the situation and make the blockchain
valid again. To mitigate this, Blockchain has the “proof-of-work”: Slow down the
creation of new blocks: new blocks are created every 10 minutes · P2P network
(everybody has the full copy of the blockchain). TO TAMPER WITH BLOCKCHAIN YOU’LL
REQUIRE: to tamper with all blocks in the chain, to redo the proof-of-work for each
block, and to take control of more than 50% of peer-to-peer network (impossible)
Bitcoin genesis block: It took 6 days from creating to launching the genesis block (first
block origin of the chain). Many theories about it’s authenticity, but only Nakamoto
could clarify this. Most accepted theory is “The creation of the Genesis block was
properly checked and verified”. BLOCKCHAIN is not an app or a company, is a data
open infrastructure that store assets (history of custodianship, ownership, location,
assets like bitcoin, certificates, contracts, objects, identifiable info…). How Blockchain
lower the uncertainty?
1. Who are we dealing with? -> Identity ratings, profiles, etc for any individual from any
source
2. How is the transaction going (each actor have a different data base)? -> Providing
the same data base for all of them
3. What happens if things go wrong? -> Settling down a kind of escrow (you don t́ pay
until everything is ok)
General applications overview (outside money and finance): ver videos en pp
1. Digital signage
2.“Smart contracts” Codify legal procedures in order to allow its decentralized,
autonomous, execution (Williams 2016).
3. Traceability
4. Marketing: Unify all records of each client (contact, purchase history, search habits,
complaints, etc.), cross promotions among different brands, full track of consumers
behaviour after a marketing campaign, traceability to create value. Example: Facebook
watch. Watch is personalised automatically according viewer behaviour when
watching. Your Watch history is never visible to others on Facebook. What means for
advertisers? To be able to address specific targets in a very cost effective way.
Facebook also planned to create their own cryptocurrency, Facebook’s libra network,
and VISA, Mastercard and others were considering to participate in this network.
5. Healthcare: Unified full medical history a person (yet being anonymous). Traceability
of medicines to verify its origin. Traceability of food to know its sanitary control
process.
6. Internet of Things: Optimizatión of energy consumption in homes and factories
7. Logistic: products tracking across its distribution chain, traceability of products to
verify its origin, follow up each product component.
8. Retail
9. Automotive industry: Full integration of vehicle sensors data. Warning of anticipated
vehicle failures. Build up of a full track record of individual driving.
10. Real estate
11. Taxes: full taxes management in an automatized way. Ongoing follow up of tax
paid destination. Transparency of donations made to a cause.
12. Banking: Currencies exchanges without intermediaries. Creation and use of
decentralized money. International transfers in real time
13. E-Government: Instantaneous and simultaneous access to a distributed database
that stores public records important for all government agencies (Identity
management, Regulatory & Taxation applications, Foreign Aid, Voting systems).

2-BITCOIN HISTORY

2008: Nakamoto described how a network of users could engage in secure peer-to-
peer financial transactions, eliminating the need for financial intermediaries and
reducing the cost of payments. This requires a structure (blockchain) + a
communication protocol. It solves the Byzantine Generals’ Problem (a group of
Byzantine Generals are camped around an enemy city in different locations. If they all
attack simultaneously, then they have superior firepower to their enemy. The problem
is that they need to agree to attack at the same time. The additional complication is
that they can only communicate by sending messages and there may be a traitor
amongst their ranks and may send false messages, Lamport 1982), enabling the
network to achieve consensus without requiring knowledge of users’ identities, or
trust relationships.
2009: creation of bitcoin (many other cryptocurrencies followed).

3-THE MECHANICS OF BITCOIN AND DPL

3.1. THE KEYS

If you own bitcoins, what you own is the private cryptographic key to unlock a specific
address. Addresses are anonymous. Each bitcoin address has a public + private key Ø
Anyone can send to a public key (e.g. account number shared with many people)
Ø However, you need a private key to extract a bitcoin from any particular address
(e.g. only you can order a transfer from your account, through your signature and ID
card at the bank branch) Ø Payments are irreversible. The private key is a long string of
numbers and letters generated by a wallet software the first time you buy a bitcoin.
(Example: E9873D79C6D87DC0FB6A5778633389_F4453213303DA61F20BD6
7FC233AA33262). Anonymity of private keys is the libertarian but also the weak part of
the system

3.2. THE BLOCKCHAIN (OR LEDGER)

The history of every anonymous bitcoin (all the anonymous addresses it had visited) is
publicly known and can be consulted by anyone in a giant ledger existing in
cyberspace: the blockchain. If one address wants to transfer a bitcoin (or some
satoshis, which are the smallest unit of a bitcoin (Bitcoin = 100 MM satoshis)) everyone
can see whether the address that wants to spend it has it. Anyone can verify a
transaction. Blocks of new transactions are added to the chain as new transactions
happen. Currently, new block added every 10 minutes (transactions are in memory
pool waiting to be added). The blockchain is secured. To alter the blockchain you
would need an unrealistic amount of computing power.

3.3. THE MINERS

They create new blocks. To add a new block (transaction) to the chain, it’s necessary to
solve a cryptographic problem very complex and unique. To do so it’s necessary a big
computational power. The problem is considered solved when the community say so.
The winner miner gets a reward in bitcoins (+-12). The miners compete to add the new
transactions to the ledger. As a result, bitcoin mining has been called “competitive
book-keeping”. For mining there are programs as Nicehash (Services provided: wallet,
mining, exchange). Issues to. Consider: APP cost and commision on transactions cost.
The miner reward is in bitcoins (+- 12). But every four years approx. the HALVING take
place. HALVING is an automatized process that cut by half the reward for block
creation. The Bitcoins number is not infinite; the maximum number fixed by its
software is 21 MM. This tool aim to reduce the issuance level which implies a
deflationary model that makes Bitcoin value rises steadily. In absence of this tool,
miners would have miner by now all Bitcoins and its reward would have be left to just
a commission on transactions. For mining it’s necessary:
- A software compatible with our hardware (Ej.: CGMiner, Claymore, etc.)
- A program to track performance of hardware (temperature, stability, etc.) ej.:
ASIC, Antminer, etc.
- A wallet: hardware type (ej. Trezor, Ledger, Opendime, etc.) or software APP type
(ej. Coinomi, Wasabi, Exodus, etc.), or in line (ej. Mycrypto, Blockchain, Bitcoin,
etc.). Watch out, the later can be hacked more easily.
Most of the mining occurs in pools (people join forces to reduce uncertainty and
obtain a kind of insurance). You can join a mining pool if you have some computing
power in the form of graphics processors (GPUs) (rather than standard computer
processors (CPUs)) and “application-specific integrated circuits” or ASICs. (GPU:
processor made of several specialized cores that work jointly SPEED)
Miners’ computing power: 9,800,000 petaFLOPs. Currently, the world’s fastest
supercomputer is China’s NUDT Tianhe-2 which runs at ‘only’ 33.9 petaFLOPS. Hence,
to match the network, you would need 300,000 of the world’s fastest supercomputers.
But the key issue that needs to be solved is not related to floating point operations
(speed) but to hashing power. There is a purpose for the bitcoin computing power: in
order to hack the chain, you need match the hashing capacity of the network
(estimated at a cost of $2000 million).

3.4 THE HASHES ISSUE

To understand the security of the block chain, it is important to understand a
cryptographic hash function (commonly used to provide communication security on
the Internet). There are many hash functions. SHA-256 (“secure hash algorithm”) is
open source code originally developed by the US National Security. You can feed a
string of text or numbers to this program and it will spit out a 64 hexadecimal code
(this means it will use numbers 0 to 9 and letters a to f, 16 characters).
Properties of hashes:
- Changing input slightly produces completely different output;
- The output is always 256 bits but the input could be any text (maximum length
for input is 264-1).
- Hashes only go one way (like broken eggs). Most important property.
- Only security risk with hashes is a “collision”: two different inputs deliver same
output. Ø Practically impossible (the hash produces 2256 -1 different outputs). Ø
 Even if it were to happen, problem is not collision but the possibility to Identify it
(that someone can compute two inputs that give the same output). Not possible
in terms of current computational power.
How does the blockchain use hashes?
- Bitcoin transactions are validated by network checking and changing the ledger.
- If the sender has the bitcoins, transaction and movement of the bitcoins to the
receptor is added to the ledger.
- To secure the ledger, the miners seal it behind computational code: The hash is
used to require proof-of-work to do the sealing, it has nothing to do with the
transaction or with verifying it. It is an extra activity that allows competition. The
miner rewarded is the first one to do the sealing. Competition ensures lots of
people are watching the transaction and this visibility is what avoids
manipulation.

3.5. PROOF-OF-WORK: NONCES

To seal the block you have to solve a cryptographic problem. The solution to the
problem (the “proof of work”) is added at the end of the block by the fastest miner to
seal the block and get rewarded. Problem to solve: Ø You have to discover an input
that will produce a particular Hash output (i.e. to repair a broken egg). Ø Miners take a
list of data from the contents of the block they are working on (transactions, time
stamps, reference to previous block) and then add a random number called a “nonce”.
Ø The list of data is the input of the hash function, trying to find the number at the
end that converts that input into an output with a certain number of leading zeroes
(the more zeroes the greater the difficulty). Ø Success can be easily checked by the
rest of miners. Difficulty is reset taking into account number of miners and
transactions adding more zeros (in year 2016 the number of leading zeroes was 17), so
that it takes a maximum of 10 minutes for someone to find the solution.
When the miner finds the nonce that works, he provides the nonce with the block and
everyone (not just miners) verifies simply plugging it into the hash creator. After a
number of correct verifications (not many are necessary) the miner that found the
nonce first “wins” the block and can add and seal it and get remunerated for doing it.

3.6. PUTTING EVERYTHING BACK TOGETHER AGAIN

1. Alice wants to use Bitcoin. She opens account on Coinbase.com (for example). This is
the link to where she still has to discover her identity and to link to her bank or to real
world currencies and it is not anonymous. But this step would disappear if everyone
used bitcoin for everything. Receives a private key.
2. Alice wants to buy from Bob using bitcoins. Bob also sets up a digital address and
sends his public key to Alice. Alice adds Bob's address and the amount of bitcoins to a
'transaction' message. Alice signs the transaction with her private key. Alice broadcasts
the transaction (not the key!) to every Bitcoin node on the Internet. If the transaction
is validated, it is added to the ledger.
3. Transaction between Alice and Bob is added to the chain. The ledger is broken up
into 10 minute "blocks“ with several transactions in it. The first miner to solve the
problem that allows the block to be sealed gets 12 bitcoins for doing it. Every block
contains a reference to the block before it so you can trace every transaction all the
way back to 2009. (If now Bob wants to use his bitcoin you can see he has one because
you can go back to his transaction with Alice). Blockchain is updated and stored on the
internet and in every computer of the network (running bitcoin).

4-BLOCKCHAIN/DLPS GOVERNANCE

Mechanics seem automatic and independent of any intermediary, central authority or

human decision, but there are three decisions that should require a central authority:
1. Allowing people to set up accounts and access keys. Verification of identity may
be imposed under KYC or anti-money laundering legal requirements.
2. Implementing changes to the blockchain protocol (i.e. how much is paid by
transaction or how many transaction are in a block)
3. Permissions for mining
Who decides which changes are made? Different DLP offer different approaches to this
political problem (Hsieh, Vergne and Wang, 2017):
- Power in the hands of core developers.
- Mixed model: Non-controversial issues decided by core developers and
controversial issuers decided by miners (bitcoin model).
- Miners democracy through voting.
Issues related are: 1. Core developers have more information. 2. Incentives of core
developers and miners may be different. Miners may distrust developers who may
manipulate the currency. 3. Changes implemented through changes in code. Open-
source code means if group disagrees with leadership they can “fork the software” by
copying the current version and then evolving the copy as a separated project.
Developers improve the software, but they can not orchestrate a change in the bitcoin
protocol, as users are free to choose the software and version they want. In order to
keep being compatibles among them, they all need to have a common and agreed
software. Bitcoin can only work as long there’s consensus among all users. Multiple
currencies, such as Litecoin (Rizzo 2017) or Ethereum (Yermack 2017), have been
affected by disagreements between groups of miners and the core developers. There
are two governance models regarding (1) users access (2) changes execution (3) miners
permits.
- Free access model (Bitcoin model): 1 Public access: anyone can submit and read
transactions. 2 Mix governance arrangements important to give correct
incentives to miners. 3 Anyone with computational power can participate in the
verification process.
- Permissioned model (conventional model): 1 Restricted access to read and submit
transaction data. Access depends on a central authority (KYC requirements). 2
Standard governance arrangements and standard incentive mechanisms: central
authority. 3 Verification nodes preselected by central authority or consortium.
Permissioned blockchains eliminate all exciting and interesting novelty form
blockchain, but make it useful for most (legal) financial purposes. Increasing number of
crypto-currencies use the Permissioned Model (conventional governance frameworks
e.g. Ripple) (Evans 2014): They have “owners”, they are usually developed or have
close links to the financial industry, they centralize verification (see discussion on
proof-of-stake versus proof- of-work below), use stocks shares to provide incentives.
Hsieh, Vergne and Wang (2017) analyse the correlation between the return of
investment in various cryptocurrencies and their governance structure and find that
centralized governance seems to offer higher returns, contradicting innitial libertarian
ideals.

5-VULNERABILITY AND RISKS

5.1. VULNERABILITY OF THE USERS

The vulnerable part of the blockchain are the keys and their storage, which happens
outside the blockchain. The problem is made worst because of anonymity inside the
blockchain. If users cannot be traced (no KYC policies) and they use stolen coins only
within the blockchain (do not convert to other currencies) they can get away with
theft. How to steal a bitcoin:
1. COPYING THE KEYS: If you keep your key in a paper printout, it can be easily
copied. Large scale robberies are conducting attacking online services that
store the private keys for a large number of users (e.g. Sheep Marketplace,
already closed).
2. KEEPING ANONIMITY: The fact that the blockchain is public means that anyone
can see to which address the coins were transferred next. You should be able
to track the thief as he moves the stolen coins from address to address. But,
the thief’s identity is still unknown. In any case, to be safer, the thief may
launder his stolen bitcoins using “mixers” also called “tumblers,” which
randomly crisscross your bitcoins with other users’ bitcoins so that you get a
clean address. The tumbler makes over time small transfers from other dirty
addresses to a new “clean” address owned by thief. The tumbler is only
accessible through the anonymizing Tor network.
3. EXCHANGING BITCOINS FOR EUROS: Once the thief has clean bitcoins he may
want to exchange them for standard currency. Most exchanges require
identification.

5.2. ENVIRONMENTAL VULNERABILITY

Mining takes a lot of energy: in 2018 the electricity consumption of mining network is
comparable to that of Denmark. Some cryptocurrencies now use more
environmentally friendly proof-of-stake system rather than proof-of-work for miners.
Proof-of-work: algorithm rewards miners that solve mathematical problems aiming to
validate transactions and to create new blocks. Environmental vulnerability:
Verification power proportional to computational power of each user. High fix costs.
Secure rewards for miners but possibility to be low.
Proof-of-stake: a new block creator is selected in a deterministic way, considering his
wealth (as participation). Without block reward. Environmental vulnerability:
Verification power depends on some posted collateral that each user has, which
usually is the number of the coins he/she owns (Mazieres 2015). Since computational
power is not required, low operating costs and low transaction fees (Davarpanah et al.
2015). Reward not fixed, miners earn a rate of return on the number of coins being
staked and the amount of time they have been idle (i.e. an interest rate).

5.3. VULNERABILITY OF THE BLOCKCHAIN DATA.

51% attack: If a mining pool gains a large amount of computing capacity, they can
attack the network, rewrite all the blocks and create a new block chain. This is not
impossible but it is not clear what is the incentive: if it ever happened, the value of the
Bitcoin would disappear.
Not enough miners: Competition among many miners is what keeps ledger secure. In
2018 there were only 5,000 to 10,000 validating nodes for 5.8 to 11.5
million wallets in active use. Most attempts at creating alternative coins (“altcoins”)
fail because of this problem. You need to have enough miners with the right long-term
incentives.

5.4. LEGAL VULNERABILITY

Threat 1: State monopoly on money issuance: Laws protect state monopoly on money
issuance and could shutdown cryptocurrencies.
Threat 2: Tax laws: Bitcoins could be treated like Euros or Yens to speculate on
appreciation. If you bought a bitcoin for $100 but it appreciated to $500, you are taxed
on the capital gain when converted back into $. Capital earnings.
Threat 3: Capital markets laws across jurisdictions: Tokens could be regulated as
trading of any other security. All regulatory compliance costs and disclosure rules
should apply.
Threat 4: Money laundering laws: Financial institutions required to operate under KYC
laws. This is biggest legal threat.
Is anonymity socially desirable? Fundamental social debate. The Silk Road case:
Website Silk Road, was the Amazon of the black market. Accessed via a Tor browser (a
distributed anonymous network). All prices for illegal drugs and other illicit materials
were quoted in bitcoins. Silk Road has since been shut down. Bitcoin is used for illegal
activities, but this is also true for cash for any other currency.

5.5. VULNERABILITY AS MONEY

INFLATION: Not a problem for bitcoin because supply is determined from start. Mining
rewards stop after ceiling of 21 million bitcoins is reached. From this point forward,
Nakamoto (2008) argues that transaction fees can become a sufficient incentive for
miners. You pay a transaction fee to get your transaction first on the queue for the
miners to solve. Price paid to the miners for a new block: 2009: 50; 2012: 25; 2016:
12.5; 2022: 6.25; 2028: 3.125.
STABILIZATION ISSUE: Lack of central authority avoids inflationary pressures but it
makes it impossible to manage the currency. What happens if there is a strong
deflation? Remember why gold standard was abolished and countries moved towards
fiat money (Gold standard was abolishes after 1929 crack to avoid prices and salaries
will drop as a result of a global demand reduction).
THE VOLATILE ISSUE: Money is only accepted among population as long as it can
storage value and it is useful as an account unit. Bitcoin is far too volatile to satisfy
these requirements. Bitcoin es approx. 8 time more volatile that the average assets
market and close to 20 times more against US dollar; it is also more volatile that most
of developed countries currencies. What does it mean? We can expect that bitcoin
fluctuates between -12% and +12% any time. Instead, the asset market fluctuates
between -2% a +2% on an average day. Worst ever return for S&P 500 since 1957 was -
20,47% on October 19th 1987, “Black Monday”. Second worst was -9,03% on October
15th 2008. In last four years Bitcoin dropped more than 20.47% in 13 days, and 9,03%
in 78 days.
Why Bitcoin is so volatile? Information (news, rumours, etc.), Its value related to the
evolution of fiat money as an alternative, “Fork” incident, Movements from big bitcoin
players(knock on effect), Liquidity in a certain moment, Problems about it’s valuation.
Volatility because bitcoin more than money is a speculative asset (like tulips in the
Netherlands during XVII c.)
About Bitcoin value: Currencies are difficult to value in general terms: For shares we
can use their value discounting future estimate cash flows. For a currency, (i.e.
yen/dollar Exchange) we can use NIP forecasts and inflation. Bitcoin case is harder as it
is not related to those issues. When looking thoroughly to Bitcoin, we can see that its
value is related to its capacity for transactions. This is affected by: Competition among
platforms, competition with standard currencies, the economic prospective and the
expected transaction volume. All very difficult to assess. People have suggested to use
assessments based on the computer power required to get a certain number of
Bitcoins, or the savings in transaction costs when using Bitcoin along time.

6-APPLICATIONS IN FINANCE

6.1. PAYMENT SYSTEMS

Blockchain with bitcoin is being used as a payment service. Would blockchain be
efficient as a payment system using standard currencies system? Standard payment
systems: (e.g. Western Union) perform two different functions:
1. Computerized network for transferring money from the receiving agent to the
sending agent.
2. Perform KYC checks and handing cash.
How does blockchain perform these two functions?
1. Computerized transfer of money: Requires incentives for the miners. Traditional
platforms such as Visa, PayPal, etc. are extremely efficient at transferring funds.
Pure transmission cost are near zero. Therefore, there ́s not a clear advantage by
using Bitcoins.
2. Perform KYC checks and handing cash: Blockchain does not do this. It relies on
exchange services for buying and selling bitcoins for cash (e.g. Bit2me) . Those
exchange services are paid beside. Prices paid to traditional payment services
(commissions) result from services related to regulation and fraud protection.
These should apply also to Blockchain.
6.2. CLEARING AND SETTLEMENT OF FINANCIAL ASSETS
In a trade a buyer and a seller agree to trade a particular asset (security) at a particular
price. Clearing implies the two counterparties updating their accounts and arrange for
the transfer of the asset and the associated moneys. Usually takes 2-3 days. Several
participants are involved in the clearing process, including: Clearing members, who
have access to the clearing facility in order to settle trades (the banks
of trader and seller). CCP (central counterparty in central clearing), an intermediary
that enters into bilateral contracts with the two counterparties. This allows for higher
contract standardization, lower settlement risk and reduction in the capital required to
be maintained with the clearing facility due to multilateral netting of cash and fungible
securities.
Can blockchain be applied to clearing and settling trades? Could be. For example, a
consortium of clearing members could set up a distributed clearing house, eliminating
CCP (central counterparty). Bilateral clearing, but contract stipulations can be
administered through a smart contract, thus reducing settlement risk. Funds can be
placed in escrow, not allowing release until each party is satisfied with the
performance of the other as reflected in a digital signature. Reporting, compliance and
collateral management handled through the Blockchain, thus reducing back-office
costs.
But, payment systems face credit risk, liquidity risk, operational risk, and legal risk.
-Assuming blockchain would solve credit and liquidity risk it would still pose new
operational risks.
Operational risk is defined by the Federal Reserve as “‘the risk that deficiencies in
information systems or internal processes, human errors, management failures, or
disruptions from external events will result in the reduction, deterioration, or
breakdown of services provided by the [financial market infrastructure] ... include[ing]
physical threats, such as natural disasters and terrorist attacks, and information
security threats, such as cyberattacks. Further, deficiencies in information systems or
internal processes include errors or delays in processing, system outages, insufficient
capacity, fraud, data loss, and leakage.’”Because payment systems pose huge systemic
risks, decentralized decision making does not seem appropriated. Resolution of a crisis
requires a rapid response by an informed party. A permissioned block-chain used by
the clearing members and central clearing facility is likely to work well.

6.3. SMART FINANCIAL CONTRACTS

Smart contracts: computer protocols that facilitate, verify, or enforce the negotiation
or performance of a contract. No dependence on intermediaries to execute or courts
to enforce could increase security and reduce transaction costs. VER VIDEO PP.
Functioning: Contractual clauses are made self-executing and/or self-enforcing: clauses
automatically executed when pre-programed conditions are satisfied. If contract’s
conditions depend upon real world data (e.g., a rental payment having been met) then,
agreed-upon outside systems called “oracles” monitor and verify prices, performance,
or other real world events. Multi-signature (or “multi-sig”) functionality requiring
approval of two or more parties before some aspect of the contract can be executed.
Example of smart contracts in finance:A smart derivatives contract could be
preprogramed with all contractual terms (i.e., quality, quantity, delivery) except for the
price, which could be fed from market data. Margin could be automatically transferred
upon margin calls and the contract could terminate itself in the event of a
counterparty default. Blockchain would perform the recordkeeping, auditing and
custodial functions traditionally performed by intermediaries, resulting in transactional
cost savings for the contracting parties (no exchange, clearing house, back office).
VER VIDEO PP

Cryptocurrencies taxation in Spain

- IRPF: Income taxation take place when the transmission occur (patrimonial
earning). % Taxation varies according earning ranges. If losses, they get
compensate by other patrimonial / capital earnings. The duty is extended to
cryptocurrencies exchange to conventional money, and also to the Exchange
between other cryptocurrencies (in this case as a barter).
- Tax on Patrimony: same treatment as a currency capital.
- VAT: transmission is exempt from VAT. Same applies to mining (no deduction
allowed for VAT supported).

UNIT 3. ONLINE ALTERNATIVE FINANCE: P2P LENDING, REWARD

CROWDFUNDING AND ICOS.
1-INTRODUCTION

FinTech: new technologies (usually Internet-enabled or large scale computing) applied

to financial services such as online payment systems, online banking, online brokers,
online financial advice, digital wallets....
Online alternative finance: Technology enabled online channels or platforms that act
as intermediaries in the demand and supply of funding to individuals and business
outside the traditional financial system.
• Balance sheet lending: A platform entity provides a loan directly to a consumer
or business borrower.
• Market place/P2P/Crowdfunding: Collective grouping of individual investors
(the “crowd” or “peers”) provide funds directly.
Alternative finance: understood as investment in alternative asset classes, private debt
placements funded by institutional investors, shadow banking activities or micro-credit
(not the same)

Regulation In the US: The JOBS Act substantially changed a number of laws and
regulations making it easier for companies to both go public and to raise capital
privately and stay private longer. It includes an increase in the number of shareholders
a company must have before becoming a publicly reporting company to 2000
shareholders, and an exception to register publicly with the SEC (Security and
Exchange Commission) if the annual aggregate limit on the amount each person may
invest in offerings of this type is below $2,000 for non-accredited investors. The SEC
also has minimum standards for P2P leading and Crowdfunding, including an annual
income and net worth of at least $70000 or $250000 of equity with no minimum
income for accredited investors.
In Spain: Organic Law 5/2015, of 27 April to promote business financing establishes the
requirements: to be a platform of indebtedness or financing by loan, on the activities
they can do (not be intermediaries, give loans or recommendations), limits of funds
per projects of 200000 fully funded, not worth loans with mortgage guarantee, and the
annual limit is 10000 are for accredited investors (3000 per project).
Trad loan ->P2P Trad investment -> crowdfunding IPOS->ICOS

2-TRADITIONAL LENDING

A debt implies that the beneficiary must make fixed payments including interest and a
claim in case of default. Gale and Hellwig (1985) and Hart and Moore (1998) prove that
the debt contract is the optimal contractual arrangement when it is costly for outsiders
to verify cash-flows. The best would be internal financing, but if external funds are
required, with debt, if it is repaid, there is no need to prove compliance with the FCs,
and if it is not repaid, but it is costly for the lenders. The types of debt are:
Bonds: for companies, fixed income negotiable by investors. Only large public
companies. They can be held by small investors.
Bank loans: companies and individuals. Financial intermediaries: intermediary helps to
borrow money or to invest. Commercial Banks: offers deposits and credits from their
own funds.
Commercial credit line: debts with suppliers
Bank loans are important because intermediaries keep track of before, to avoid
adverse selection, and after to avoid moral hazard. As both are costly, a service
premium is charged. Commercial banks also transform illiquid long-term loans
demanded by borrowers into short-term deposits preferred by lenders (Gurley and
Shaw 1960). To control ex ante, intermediaries reduce the duplication of monitoring
costs caused by asymmetric information, which are worse for small firms more prone
to adverse selection and moral hazard (Holmstrom and Tirole 1997). To control expost,
they reduce duplication of verification costs when verification of cash flows is costly
(Diamond 1984).

3-PEER-TO-PEER LENDING AS AN ALTERNATIVE

3.1. BUSINESS MODEL IN P2P LENDING AND KEY QUESTIONS

These are platforms that link individuals willing to lend with individuals in need of
loans, with a minimal interest rate (0-1%, compared to 8% for banks), directly
connecting investors with borrowers without intermediaries. Currently, P2P accounts
for only 1-2% of total consumer leading. Examples: prosper, lendingclub.

3.2 HOW DOES IT WORK?

The platforms earn money from the fees they charge individuals and allow them to
post debt requests with an amount and a cause. The platform sets the interest rate
and provides a credit rating. Money is only released if the full performance is met,
which almost never happens. Borrowers may never get the money. They pay interest
between 6-36% depending on their history, as well as default, origination or closing
fees (depending on risk). They can borrow up to 35,000-40000 with terms of 30-60
months. Lenders can invest a minimum of $25 and obtain risk information from the
platform and other anonymous personal data. They decide whether to lend money to
that particular borrower, or semi-automatic provision by the platform. If the borrower
does not pay they receive nothing, if he pays, the principal and interest. They pay 1%
annual commission to the platform.

3.3. DETERMINANTS OF BORROWING SUCCESS ON P2P LENDING

The biggest problem is asymmetric information leading to adverse selection and moral
hazard. For lenders it is important:
- Financial determinants: (Klafft 2008) the rules that apply in these loans are very
similar to those in the traditional banking system. The credit rating of the
borrower is what influences the interest rate of the loan, as well as the debt-to-
income ratio of the borrower. With a weak credit rating you don't get loans, and
58% of propers, of which 5.5% are successfully financed, compared to 54% in the
case of a good rating.
- Demographic characteristics: Discrimination can be of two types (Pope & Sydnor
2008): Taste-based discrimination (according to the similarity between borrower
and lender, Ravina 2007) and statistical discrimination (when it is by default rate).
Also by age (young people are more likely to get a loan and older people less
likely), gender and race (blacks are less likely than whites to get a loan, pay more
interest, and have lower profitability).
- Social characteristics (small but significant effects): military-only platforms that
have better performance and behavior. Chinese who return books in libraries and
graduate from good universities get more loans and are less likely to default.
Photo publications are more successful. Loans between friends/acquaintances
work better.

3.4. BACK TO BUSINESS MODEL. CAN IT WORK?

The P2P model will work as long as the platforms make a better (using machine
learning algorithms, or alternative data that gives more accuracy) or cheaper (taking
advantage of borrowers that the bank does not want because of fixed costs)
assessment than the banks.

4-TRADITIONAL FINANCING FOR ENTREPRENEURS

In the case of new companies, the lack of information makes financing difficult, which
means resorting to: internal capital, FFF (family, friends, fools), venture capital,
wealthy individuals or venture capital funds. The latter are specialized by sector, are
expert advisors and intervene in the management of the company, provide partial and
staged financing. It is contrary to crowdunding.
5-CROWDFUNDING AS AN ALTERNATIVE

Some say it has the potential to revolutionize small business financing and generate
capital for small, job-creating businesses. However, it is also said that it can become an
efficient online weapon of investor fraud, that the good part of crowdfunding is
coming to an end, and that investors are ignoring those who really know about it.

5.1. INTRODUCTION TO CROWDFUNDING

Charity crowdfunding is a way of raising money for non-for-profit projects, to donate
Equity and debt crowdfunding are used to finance for-profit ventures using standard
equity and debt contracts that commit the creator to pay the small and disperse
financial investors: Debt (capital + interest), equity (profit on shares).
Reward crowdfunding raises money in exchange for some project outcome, which may
range from a mere acknowledgement of the funds received to the promise of delivery
of one unit of the final product or service, with the creator committing to making
"best-efforts" to deliver.
Examples of CF: fundingcircle, circleup, kickstarter, artistshare, verkami, indiegogo,
kiva

5.2. HOW DOES IT WORK?

1) Creator ask platform to accept a project
2) Platform screen for fraudulent projects
3) Creator determines funding goal and campaign time spam
4) Platform gets the funds and releases it to creator: All or nothing scheme
5) Platform gets 4 to 5% of the total funding raised

5.3. HOW DO LENDERS CHOSE PROJECTS AND DEAL WITH ASYMMETRIC

INFORMATION?
Funders are not experts and oversight costs are multiplied, but they rely on: active
feedback through the platform (requesting information from the creator), all-or-
nothing reduces coordination problems, "Wisdom of the crowd" effect, sequential and
observable decisions of backers and on estimating the viability and capacity of the
creator to deliver the product. Success of a CF campaign depends on signals of project
quality (including quality of videos, frequent information updates during the campaign,
absence of spelling mistakes…). Projects that succeed in their CF campaigns are more
likely to gain afterwards outside funding.

5.4. SPECIFIC PROBLEMS WITH EQUITY AND DEBT CROWDFUNDING

Uninformed small investors are at a disadvantage, since in the event of having to sell
their shares they would have a high information search cost. The IPO process
protected them: through minimum size, age and track record requirements; full
disclosure of the company and its status; an intermediary bank; marketing the issue to
both institutional and small investors; appropriate pricing; cancellation of the offering
if institutional investors tranch not fully subscribed; and subsequently the shares can
be traded on a liquid and regulated stock exchange.

5.5. SPECIFIC PROBLEMS WITH REWARD CROWDFUNDING.

The question arises as to whether funders should be protected as consumers, since
they are paying for the outcome of a project and are early adopters of the product or
service.
A. IS REWARD CROWDFUNDING SOMETHING NEW? OR NOT?
Example projects: Pebble, by Migicovsky, the case of an entrepreneur who designed an
interactive watch. Apart from the funding he got from investors, he used the RCF to
ask for an additional 100000$ promising a watch for every 120$ contributed. In the
end he raised $10 million. The watches took longer than expected to come out. In a
second round he raised $15 million.
Extremoduro, the Spanish rock band, managed to finance the recording of their first
studio album by selling ballots at 1000 pesetas (6€) that could be exchanged later for
their album if it was recorded. They got 250,000 pesetas.
B. TRADITIONAL CONSUMER FINANCE: PAYMENT OF THE GOOD OR SERVICE IN
ADVANCE
Advance sales occur when the consumer is paying for something that the seller will
give him for his enjoyment in the future, such as a concert. Why raise money from
consumers? When cash flows are difficult to exceed and/or to appropriate, or when
projects have high uncertainty about future cash flows, since it is difficult how much
debt to borrow otherwise.
C. NOVELTY IS IN THE “BEST-EFFORTS CONTRACT” USED IN REWARD CROWDFUNDING.
Many times RCF projects, for example on Kickstarter, do not make it. If this happens,
the money is refundable to the investors. In the event that delays occur or the product
does not come out, investors can contact the project owners. 37% of the projects go
over budget, 5-14% do not complete the project and 50% are late, 90% later get access
to funding through other means, and less than 25% do not get the funds.

6-THE IPOS (INITIAL PUBLIC OFFERING) MARKET

Asymmetric information in the IPOs market leads to underpricing (on the day of issue
of the shares, the market price rises, this is to attract small investors) and
underperformance (IPO market returns are below the average market returns for a
period of 3 to 5 years after the issue (which usually starts after a honeymoon of 6
months). Insiders take advantage of this to sell shares that will be overvalued. These
asymmetric info problems may be much larger in the case of ICOs....

7-ICOS

7.1. INTRODUCTION
An initial coin offering (ICO) is the cryptocurrency industry's equivalent to an initial
public offering (IPO). In CF and normal financing, a financing contract is made in
exchange for shares, debt or products and services. The difference with ICOs is the
nature of the tokens or crypto-assets that are obtained in exchange for the financing,
and the use of code to make the contract between the buyer and the issuer. ICOs are
at the intersection of crowdfunding and DPL and are a new asset class used to finance
technology start-ups. What are tokens? You pay cryptocurrencies and get these
crypto-assets, with value related with the final success of the user. It is a medium of
exchange that allows future use of services of the proyect and participation in
governing of the future ecosystem. Canusuallybetradedlaterinon-lineplatforms.

Omnicoins/Mastercoins: el primer intent de utilizar blockchain para financiar la

creación de una nueva empresa en 2013. La idea de crear esta nueva capa de
protocolo la sugirió J.R.Willet en una conferencia de Bitcoin. A cambio de Bitcoins, los
inversores reciben nuevas monedas, que son una capa de protocolo ejecutada sobre la
red de blockchain de BTC y permite los contratos inteligentes en esta red. Este
consiguió con un solo anuncio en bitcointalk casi 5000 BTC de 550 personas a las que
después proporciono estos mastercoins, llamados después Omni. Los fondos se usaron
para desarrollar el software para esta idea y los tokens se pudieron negociar
automáticamente en la bolsa de mastercoins y después en muchas otras.
Etherum (2014): first formal and most famous ICO. Ethereum needed money to
develop and implement "smart contracts" on the Ethernel platform. They ran a 40-day
campaign and raised $18 million in exchange for Etherum tokens and a detailed
description of the project and future plans. The funds were raised by a Swiss
foundation. Investors were registered compulsorily but rather informally, getting more
secure contribution addresses. There was an escrow account under third party control
for transparency and reduction of moral hazard. A legal terms of sale document was
made to receive new Etherum coins. This software is available on Etherum and new
ICOs use this technology to create tokens in a short time.

7.2. THE ICO PROCCESS

1. Information: the fundraising company publishes a white paper with information
about the project. Its legal status is not clear and less than half of these papers give
clear information about the promoters of the idea and contact details. More
information is available on Slack, Telegram and Blockchain forums.
2. Token issuance number: there is usually a hard (upper) cap (limit) and a soft (lower)
cap (limit) on the number of tokens that can be issued. If the soft cap is not reached,
the campaign can keep the reduced funding or return the money.
3. Incentives for developers: developers keep approximately 39% of the tokens,
without the possibility of selling, as an incentive, with an average maturity of 1.1 years.
If the VC provided the initial funding, they also receive tokens.
4. Segmented sale by type of investor: first stage where a 10% of supply is offered to
relevant investors, signal for small investors. Second stage is for main public, including
discount for early participants and price list for the rest.
5.Legal issuance: Sold tokens are distributed automatically by smart contract, during
the crowd sale or immediately after. Unsold tokens are burned or destroyed.
6. Secondary market: The issuers arrange for listing of the tokens on various online
crypto-exchanges.
7. Legal compliance: Many ICOs choose Switzerland or Singapore as governing
jurisdictions for token sales. Some prohibit participation by US residents following the
SEC's decision to treat ADO tokens as securities (federal securities laws apply to those
offering and selling securities in the US)(with all legal tracking requirements) and
Chinese residents as China banned all domestic ICO activity. Most ICOs perform KYC
checks on all investors, allowing only accredited investors from selected jurisdictions to
participate and warning investors from other jurisdictions.

7.3. QUALITY OF ENFORCEMENT BY CODE

ICO proponents claim that in ICOs investor protection is delivered through code.
Cohney, Hoffman, Sklaroff and Wishnick (2018) compare the actual code to 3 types of
promises made to protect investors in whitepapers for the top 50 ICOs of 2017.
Promises vs facts: 1. Restriction of the supply of the crypto asset: 41 promised and 31
coded. 2. Restriction of the crypto assets assigned to developers: 37 promised and only
8 coded. 3. Developers retain the power to modify the smart-contract code governing
the tokens: only 10 permit modification in their code.

7.4. CHARACTERISTICS PREDICTING ICO SUCCESS

Boreiko & Sahdev (2018) and Fahlenbrach & Frattaroli (2019) find that ICO success and
number of investors is positively correlated to:
1. Better information (Massive marketing campaigns in all media; listing in the ICO
aggregation websites for greater publicity and quality certification, ICOs size).
2. Less asymmetric information (quality with prior VC-participation or with a pre-sale
for large investors).
3. Better incentives (sale limitation for developers for a longer period of time).
4. Better legal compliance (KYC policies)
5. Fair treatment of investors (offering less discounts).
6. Lower expropriation threats (smaller % of tokens allocated to founders).

7.5. CHARACTERISTICS OF ICO INVESTORS

The mean of the median contribution per investor in the crowdsale is $1,203. Majority
of investors only contribute to one ICO. Investors seem more interested in the
speculative returns than in LT project: On average 49.3% of all investors sell some or all
of their tokens within 90 days of the ICO. Within those 90 days, the original ICO
investors sell 42% of the issued tokens in the secondary market.
Why are ICOs attractive to retail investors despite the lack of transparency and
investor protection? They take risks on volatile stocks with high correlation, on
companies in developing markets with high potential and the promise of companies
using blockchain.
It is not yet known whether ICO tokens are in a speculative bubble.
 UNIT 4. BIG DATA AND MACHINE LEARNING IN FINANCE:
APLICATIONS, PROBLEMS AND LIMITATIONS.
1-WHAT IS BIG DATA?

Big data consists of the collection, management and analysis at high speed of large
volumes of heterogeneous data generated by users and machines, which we transform
into useful information that helps us make business decisions to solve problems. Large,
complex and growing data sets or combinations of data sets that cannot be handled
with conventional technologies, such as relational databases or conventional statistics
or visualization packages, in the time necessary to be useful. Characteristics (8Vs):
volume, value, veracity, visualization, variety, velocity, viscosity and virality. The
complexity comes from unstructured data, which Big Data combines with structured
data (usually from a relational database) usually taken from an ERP or CRM. The data
comes from millions of transactions that are made every day. Most of it has been
recorded in the last few years and almost all of it is unstructured data. We individuals
contribute to generating that data, as do enterprise customers and their internal
processes. Big data process: 1.- Collection 2.- Storage 3.- Search 4.- Sharing 5.- Analysis
6.- Visualization. Big data source: (Telefonica, VISA, BBVA, insta, Google, FB, Carrefou..)
- Web and social networks: information available on the Internet (web content,
generated by users…)
- Machine-to-Machine (M2M): data generated from communication between
intelligent sensors integrated into everyday objects.
- Transactions: Includes billing records, calls, or transactions between accounts.
- Biometrics: data generated by technology to identify people by means of
recognition Facial, fingerprint or genetic information.
- Generated by people: emails, messaging services or call recording.
- Generated by both public and private organizations: data related to the
environment, government statistics on population and economy, electronic
medical records, etc.
Data tipology: 1st level data (data provided consciously by the user), 2nd level data
(provided unconsciously by the user through web browsing) and 3rd (collected from
user for marketing with consent). Also structured (organizable in rows and colums,
easily ordered and processed, labelled and easily accessible) or unstructured (no
identifiable structure, binary data, massive and disorganized with no value until they
are organized, not in a database)
Big data applications: tourism (using customer data), health care (diagnoses or
treatments almost immediately),administration (maintaining quality and productivity
With tight budgets), world of sports, retail (understand customers, predict trends,
recommend new products and increase profitability), manufacturing companies
(deploy sensors in their products to receive telemetry data) and advertising (target
consumers near stores).
Challenges of a Big Data: we have information from the ETL process, ERPs, CRMs and
additional information from outside the company:
1. Difficulty of integration: because of so many sources (internet mobile data, IoT
data, sectoral data and experimental data), data types and complex structures.
 2. Quantity and fast data: the huge volume of data makes running it a time
consuming process. It is difficult to collect, clean, integrate and obtain high-
quality data quickly. It takes a lot of time to transform unstructured types into
structured types and process that data.
3. Huge volatility: data changes quickly and losses value. High processing power is
needed. A bad procedure leads to erroneous conclusion for decision-making
4. No unified data quality standards: in 2011 the ISO published data quality
standards, that need to be mature and refined.
Big Data in the financial services industry: used to define profitable trading strategies,
set exchange rates, creditworthiness (based on rating), selection of investments,
portfolio construction and idiosyncratic and systemic risk and predict trends.

2-WHAT IS MACHINE LEARNING?

Machine Learning (or ML) refers to the set of algorithms that learn complex patterns
(structure identified may not be representable as a finite set of equations, they figure
out variable dependencies from data) in a high-dimensional space (solutions often
involve a large number of variables and interactions btw them) without being
specifically directed (the algorithm derives structure from the data, searching for
patterns). ML is inside AI. Learn from experience beginning from data.
How is machine learning different from regression analysis? Researchers use
traditional regressions to fit a predefined functional form to a set of variables.
Regressions are extremely useful when we have a high degree of conviction regarding
that functional form and all the interaction effects that bind the variables together.
ML and Big Data: ML techniques are especially adept at analyzing big data. They infer
non-linear relationships and causal effects from data which has a lot of noise
and little information. The biggest misconception is the black box. ML can be
extremely insightful for development of economic and financial theories.

3-APPLICATIONS IN FINANCE

Big Data, AI and ML are useful for tasks people can’t do because of absence of
resources (time, interest, memory…), for games of complete information (problems or
tasks with all information needed). Are financial problems like this? ML is used to
convert data into other data (financial items into ratios), for complex and non-linear
relationships. The three problems of its application in finance are: low signal-to-noise
ratio of financial data (bc of arbitrage forces and non-stationary systems, ML might
confound noise and signal and will always find a pattern even if it is fluke and not
persistent), humans are not inherently good at solving financial problems and complex
(if human use algorithms with black box, they don’t understand and validate the
solution) changing and incomplete information environments. Applications in finance:
1. Quantitative modelling and stock market prediction: Examine the potential of
social media signals for predicting S&P 500 index returns and volatility, identifying
directional sentiment and non-directional disagreement signals. Before market
crashes, there is a period when investors are looking for market information to
buy or sell, which can be used to make profitable trading strategies. It can also be
used to rank photos according to the investor sentiment index, as photo
 pessimism predicts market return reversal and increase in trading volume. Finally,
predicting market reversal using online news is not as valid and gives less income.
2. Bankruptcy warning modelling: use data mining methods to design a financial
distress early warning system for small- to medium-sized enterprises. There is no
significant difference between models based on BD analytics and prediction
models. Individual financial outcomes, which seem to be associated with
diversity, loyalty and regularity, are associated with consumption habits
3. Financial fraud modelling: language-based tool that relies on data to identify
important indicators of fraud. It has an initial training period, using a decision-
three, with inputs of fraud firms to rank order list of words that distinguish fraud.
Then, it uses vector order machines to predict the status of financial reports and
assign a truth probability.
4. Auditing: Financial information could be made available in real-time (no quarterly
or annual reports). The role of the auditor changes from providing assurances
about numbers to assurances about real-time systems (choices for data
collection, requirements for display and visualizations, validation of dashboards,
etc)

4-BENEFIS AND DANGERS FOR CUSTOMERS 5-BIG DATA FOR FINANCIAL REGULATION

Benefits:
1. ML can produce cost reductions and efficiencies, consumer get this efficiency in
financial products and welfare increases.
2. Low cost personalized treatment, limiting options for investors, facilitating their
choice and avoiding inaction, offering advice.
Dangers (information asymmetry):
1. Firms can identify our willingness to pay a product or service and charge us close
to our limit price, increasing their profit. Consumer can try to use programs,
software or hardware options to achieve anonimaty (not useful if we want
financial advice or KYC checks) or firms should be forced to disclose the
personalization and offer a “opt out buttom”.
2. Discrimination because of age, gender, religion, etc, which is forbidden by law.
3. Exploitation of human biases in decision making: regulations include use of
“visceral notices” and withdrawal rights with a cooling-off period (rights for one
party to withdraw from certain types of contract within a specified period
(cooling-off period) without having to give reasons for the withdrawal).

5-BIG DATA FOR FINANCIAL REGULATION

A well-functioning financial system is essential for economic growth. It is complex

and one of the most heavily regulated parts of the economy. It is necessary to protect
unsophisticated agents and stability of the system. These regulations are expensive
and charged to consumer. However, BD, ML and AI can lower regulatory costs.
Artificial Intelligence is for tasks with requirements stablished, so it is good for
microprudential regulation but worse for macro prudential regulation. Still, there are
problems:
 - Data availability: finance generates almost infinite amounts of data to train
machines but the system continually undergoes structural changes (new types of
participants or new moves from new financial innovations). This reduces the
value of historical data, especially for macroprudential. There are 3 events:
known-knowns (certainties), known-unknown (events we think might happen, do
not usually cause crises, largest stock market crash 1987) and unknown-
unknowns (complete surprise, not predictable, they are never repeated in their
initial form and past data is not useful, financial crisis of 2008).
- Risk evaluation: two types: exogenous (entities are making independent
decisions, AI can be really good at these risk) or endogenous (arises from
interaction of economic agents, requires AI being able to reason and act
strategically, taking into account how market participants will react to hitherto
unseen event).
- Issues of trust: trust in AI and micro prudential regulation: AI will rapidly take over
micro regulation management functions. The trust issue will be how to deal with
“malicious actors”. Systems have to be transparent and AI highly rational, but his
makes an AI regulator more predictable and this could help aggressors more than
the current human-centered setup. AI an macroprudential regulation: crisis
resolution will remain a human task, because a system output developing this
task would be an unexpected behavior. Human regulators cannot anticipate
unknown-unknowns, but neither can AI, and they are reasonably well equipped
to respond (historical, contextual and institutional knowledge, and shared
understanding of values and environment, and AI doesn’t) . EURISKO and Captain
Sully successful landind

UNIT 5. PROBLEMS IN THE REGULATION OF DIGITAL FINANCE.

1-INTRODUCTION

FinTech is growing very fast due to the application of new technologies (AI, BD, ML,
DLT, Smart contracts, etc). Current uses are crowdfunding, digital currencies, ICOs,
touchless e-payments solutions and robo-advisors. These innovations increase
efficiency, reducing costs and intermediaries, and save costs allowing regulatory
arbitrage or regulatory avoidance. Financial regulators should be carefull when
introducing regulation (too much vs too little, too soon vs too late). There are other
risks in Fintech such as data protection, control of self-learning algorithms and
protection from cyberattacks or bugs, that have the potential to be systemic risks. The
need of regulation is driving the traditional financial system to become more
technological because of RegTech (use of technology, Its, for regulation), which forces
financial institutions to spend large amounts in technology to be able to comply the
new information requirements. Examples of RegTech include, stress testing based on
machine learning, electronic Know-Your-Customer (KYC) systems which help keep safe
client data and also enhance market integrity, automated compliance monitoring and
reporting with regard to trading limits, and algorithm-based reviews of trading
patterns in listed stocks to ensure compliance with insider dealing laws.
2-WHAT ACTIVITIES TO REGULATE?

The regulating financial innovations have to (regulatory trilemma):

1. Maintain market integrity: at the micro (combat fraud against consumers and
financial services) and the macro level (for economies to remain healthy,
prevent weaknesses of financial institutions from affecting others, support
financial firms). To achieve this, financial regulation needs: comprehensive
rules, disclosure obligations, antifraud protections, safety measures and sharp
punishments.
2. Encouraging market innovation: to make finance more efficient and accessible,
promoting technological innovations that transmit information more quickly
and fully btw investors and the market , tools designed to increase the
precision to data collection and analysis for a better understanding of risks, and
innovations giving consumers easier and cheaper access to financial products.
3. Providing clear rules: clear and simple rules provide certainty, predictability and
stability. It lowers the barriers to entry in financial marketplaces, promoting
competition, thus is easier for smaller firms to compete and more competition
makes the system safer and more inclusive.
The trilemma arises because only two of these objectives can be achieved at the same
time. If market safety and clear rules is prioritized, then do it through broad
prohibitions, like inhibiting financial innovation. If regulators wish to encourage
innovation and clear rules, they will risk market integrity. Finally, if regulators look to
promote innovation and market integrity, they will assume rule exemptions. This
trilemma is more important than in the past because of the speed of financial
innovations, so when regulating, many important facts are not known yet: lack of an
adequate sample or reliable data, lobbying (distorted identification of relevant and
irrelevant facts because of entrenched interests about new technologies) and some
facts may be “unknown-unknowns”. This brings the risk of doing too little too late
(endangering market integrity) or doing too much too early (killing innovation).
Two tools can help be helpful:
- Data driven regulation: Use of data to determine what and when needs to be
regulated. Data measuring investment in new technology and innovation can be
used as an index or proxy of the necessity of regulation.
- Regulatory sandboxes: they create a safe space for testing innovative products
and services without being forced to comply with the applicable set of rules and
regulations. It helps maintaining market integrity (discovering relevant and
unknown-unknowns facts without affecting consumers negatively) and fosters
innovation (lowering regulatory barriers and costs). If a company applies for a
sandbox, the regulator will verify that it’s a genuine innovation and it carries
consumer benefits. If both conditions are met, the regulator establishes some
parameters at the beginning regarding duration, customers, disclosure and
commitment not to enforcement actions.

3-WHAT PARTICIPANTS TO REGULATE?

Now innovations are being introduced from outside the financial industry, so there are
three actors that regulators should pay attention:
 - Traditional financial institutions (central banks, commercial banks, investment
banks…)
- FinTech’s: small companies that identify a new or already existing financial service
that they can do better than incumbents or that they are not doing.
- Tech fins / Big Techs: non-financial firms whose relationship with customers is in a
non-financial services setting. They collect massive amount of data from those
relationships, and then seek to make use of that data in finance by selling it to
financial services providers, by improving their customer and suppliers’
relationship, and by providing financial services directly itself.
The entrance of a Tech Fin into financial services typically comes in three stages: first,
licensing out aggregate data to financial institutions or FinTechs. Second, using the
datasets to guide its own business decisions improving risk management. And third,
moving into offering financial services and start competing with incumbents.

The conversion of these Big Tech into Tech Fins poses important risks: cybersecurity
risks, they can become new SIFIs (Systematically important financial institutions, a
oligopolistic market with too-big-to-fall risk and too-connected-to-fail risk).This risks
financial integrity if Tech Fins remain outside financial regulation and reduce
innovation. Regulators should impose information requirements on technological
firms in all aspects related to data gathering and analytics, but only in these aspects of
their activities (reporting on data gathering, location of its clients and data delivery to
intermediaries). To exempt small businesses, foster innovation and increase
competition, there should be a minimum threshold exemption (too small to care
threshold). In order to prevent systemic risks there should be a maximum threshold
allowance before financial regulation kicks in (Too big to ignore threshold). Once this
threshold is crossed, macroprudential regulation could imply things such as:
diversifying Tech Fin to diversify data sources, structural requirements for Tech Fin,
and empowering regulators to shut down the activity while preserving the data.

4-APPENDIX: EU REGULATION RELATED TO FINTECH

This regulation was not implemented specifically for dealing with Fintech but its result
has been used for RegTech in the EU.
- Digital regulatory reporting requirements particularly of AIFMD(Alternative
Investment Fund Managers Directive) and MiFID: Impose significant electronic
reporting requirements, which imply continuous investments in IT to ensure the
sufficient data collection. They have also forced regulators and supervisors to
develop data management systems which can receive and process the volume of
data being generated and delivered by the financial services industry. This has led
to a Reg Tech "revolution" in the European financial services industry.
- General Data Protection Regulation (GDPR): GDPR imposes rules that seek to
protect natural persons in relation to the processing of their personal data. GDPR
is restricted to data processing of personal data in connection with a professional
or commercial activity Natural persons should have control of their own personal
data. Consent requirement must be passed to collect data. This data collected
cannot be stored forever and the person has the right to request the forgotten up
of that data and to data portability. GDPR also imposes the use of a pseudo-
anonymization of personal data. Natural persons have the right to be subject to a
decision by humans where the decision produces legal effects. Controllers of data
are under the obligation to undertake data protection impact assessments. Any
person damaged because infringement of GDPR has a right to be compensated
and EU can impose a fine up to 4% of the total worldwide annual turnover to the
culpable company.
- Open banking regime introduced by the Second Payments Services Directive (PSD
2), particularly combined with the data portability requirements in GDPR:
Mandates that banks have to transfer customer data to third parties when
directed to do so by their customers. PSD 2 thereby allows for “open banking”:
broad competition among incumbent and new participants by forcing incumbent
financial intermediaries to share client data with third parties, including
potentially innovative new competitors. This fosters innovation.
- Pan-European digital identity framework built pursuant to eIDAS (Electronic
IDentification, Authentication and trust Services regulation): Adopted in 2014 to
provide mutually recognized digital identity for cross-border electronic
interactions between European citizens, companies and government institutions.
When an eID is ultimately recognized throughout the EU, an individual will be
able to use it in any member state and enables any EU citizen or entity so
identified to enter into transactions digitally and have the same legal status as
traditional paper-based processes. Use cases include submitting tax declarations,
enrolling in a foreign university, remotely opening a bank account, setting up a
business in another member state, and bidding for tenders