The RBI Master circular dated September 2012 details out the Prudential norms on

Income Recognition & Asset Classification (IRAC) and emphasizes on the need for an
effective mechanism and granular level data monitoring for NPA. The finance
ministry also echoed this sentiment when it directed all Public Sector Unit Banks
(PSU) to automate calculation and monitoring of NPAs by September 2011, and later
extended the deadline to March 2012. The basic objective is to increase
transparency and consistency in financial reporting while creating an effective
mechanism to monitor NPAs.
In early 2013, Moodys had downgraded the rating of the Indian Sovereign and
Banking industry from Stable to Negative. Key reasons for this were the rising
NPA levels and quantum of restructured loans both of which were considered,
symptoms of deteriorating asset quality. Some other studies, also predicted that
NPA levels for Banks were expected to reach 4% by March 2013 and 4.4% by March
2014. Overall, the current NPA status was grim and checks and balances to monitor
asset quality and NPAs in particular were the need of the hour.
The NPA data below shows trends across Indian Banks. While Gross NPA gives an
overview of gross NPAs/Gross advances, Net NPAs are calculated after taking into
account the provisioning already in place. NPAs peaked in 2008, post which there
has been a gradual decline. However, for PSU Banks which have a greater exposure
to risky sectors like power and agriculture, an increasing trend can be observed.
Private sector and foreign Banks appear to have been fairly successful in controlling
NPAs.

Source: RBI report on Trends & Progress of Banking in India 2011-12

Asset Quality:

The gross non-performing advances (GNPAs) of SCBs as a percentage of the total

gross advances increased to 4.5 per cent in September 2014 from 4.1 per cent in
March 2014. The net non-performing advances (NNPAs) as a percentage of total net
advances also increased to 2.5 per cent in September 2014 from 2.2 per cent in March
2014. Stressed advances increased to 10.7 per cent of the total advances from 10.0 per
cent between March and September 2014. PSBs continued to record the highest level
of stressed advances at 12.9 per cent of their total advances in September 2014
followed by private sector banks at 4.4 per cent (Chart 2.6)

The gross NPA ratio at the aggregate level stood at 3.6 per cent at end-March 2013 up
from 3.1 per cent at end-March 2012. The deterioration in asset quality was most
perceptible for the SBI Group with its NPA ratio reaching a high of 5 per cent at. With
the gross NPA ratio reaching about 3.6 per cent by end-March 2013, the nationalized
banks were positioned next to the SBI Group. There were also signs of a deepening
deterioration within NPAs with an increase in the proportion of doubtful loan assets.
The slippage ratio, defined as additions to NPAs during the year as per cent of
standard advances at the beginning of the year, also showed an increase during 201213. At the aggregate level, the ratio of restructured standard advances to gross
advances stood at 5.8 per cent at end-March 2013. It was the highest for nationalized
banks (at 8.3 per cent) followed by the SBI Group (at 4.7 per cent).
Although there was a rise in the gross NPA ratio in 2012-13, the provisioning
coverage ratio (PCR), defined as provisions for credit loss as per cent of gross NPAs,
showed a marginal decline during the year at the aggregate level. The decline was
most perceptible for nationalized banks.

NPA MANAGEMENT:
Asset quality is the key to understanding the financial health and soundness of the
banking system. The literature identifies credit cycles as an important determinant of
banks asset quality. Cyclicality/pro-cyclicality has been defined as dynamic
interactions (positive feedback mechanisms) between the financial and real sectors of
the economy (FSF, 2009). Financial institutions tend to over-stretch their lending
portfolio during economic booms and tend to retrench the same during economic
downturns. It has been argued that an expansion in credit growth is associated with the
deterioration in asset quality because when banks overexpand their lending, they tend
to lower their credit standards. This behaviour translates itself into greater slippages in
asset quality at matured stages of the credit cycle. The literature identifies various
reasons for such pro-cyclical risk-taking behaviour of banks, viz., herd behaviour

(Rajan, 2005), principal-agent problem between shareholders and managers (Borio

et al, 2001), disaster myopia or shortsightedness in underestimating the likelihood
of high-loss low-probability events (Guttentag and Herring, 1986), among others.

Asset quality has surfaced as an important concern for the Indian banking sector in the
recent years. In the period immediately following the global financial crisis, when
asset quality of banks in most advanced and emerging economies took a beating, the
asset quality of Indian banks was largely maintained, partly on account of the policy
of loan restructuring. However, between March 2009 and March 2012, the gross NPAs
ratio has shown an increasing trend albeit a fall in 2010-11.

A cursory look at the growth in bank credit and gross NPAs reveals a cyclical pattern
(Chart 1). An empirical analysis to model asset quality of Indian banks as illustrated in
equation (I), taking quarterly data from June 2000, suggests a lagged statistically
significant positive relation between deviations from trend in credit to GDP (C-GDP)
ratio (worked out using Hodrick-Prescott filter) and growth in gross NPAs for the
second lag.1 The deviations from trend in C-GDP ratio has been recommended as a
principle guide by the Basel Committee on Banking Supervision (BCBS) for
determining economic and financial cycles under its Basel III framework (BIS, 2010).

*Significant at 1 per cent level.

This exercise brings out the cyclicality in the behaviour of asset quality of Indian
banks. Further, it justifies the countercyclical prudential regulatory policy, as pursued
by the Reserve Bank, and corroborates the need to further strengthen such a policy by
basing it on a more systematic and rule-based footing to effectively address the
concern of asset quality.
NPA MANAGEMENT:
Predictive Analytics has been effectively used in some cases to score customers, based
on behavioral and demographic /psychographic parameters. These scores are
considered to be a fairly accurate indicator of expected repayment behavior and
determine credit eligibility. This can be used effectively to decide whether or not an
asset product is to be cross-sold to a customer or to determine the terms. Credit
Information Bureau India Limiteds (CIBIL) scores, attempt to provide this kind of
data even at an inter-bank level.
Once assets are sold, preemptive analytics come into play. Early warning signals are
created that flag off assets that are delinquent or tending towards delinquency. This
provides a head start to Banks that then work with customers, to ensure they dont
become full blown NPAs. In this manner, preemptive and predictive analytics play a
key role in the NPA management strategy of a bank.
Analysts and experts today consider NPAs to be one of the key indicators determining
the health of the balance sheet along with other measures like CASA ratio and NIM.

The regulatory guidelines, on the topic, classify assets into 4 broad categories
standard, substandard, doubtful and loss assets with progressively increasing
provisioning levels. There are also specific norms governing collateral apportionment,
treatment of restructured assets and complex credit facilities like syndicated loans.
Banks face many challenges on the path to achieve NPA automation and monitoring
goals set by RBI. First and foremost, is the presence of multiple source systems,
housing data pertaining to different asset products/facilities availed by the same
customer. While, some of these systems may be equipped to automatically, perform
NPA calculations, others are not. RBI guidelines advice Banks to take a customer
view rather than a facility view, which implies that data from multiple systems needs
to be integrated to provide a unified NPA view at customer level. This is in tandem,
with the cross default clause, that features in many loan agreements and monitors
repayment behavior. Default on one facility will have consequences on all the other
facilities granted to the customer by the same Bank.
Secondly, the process of NPA computation needs to be automated to be able to handle
different product types and the huge data volumes. Additional layers of complexity are
also introduced by the presence of an underlying asset (collateral) whose value needs
to be appropriately apportioned before calculating provisioning levels. Automation
promotes consistency and a relatively error-free calculation while potentially reducing
resource requirements for the activity.
Finally, once the key metrics pertaining to NPAs are calculated, they also need to be
incorporated, in the relevant financial statements. Additionally, a lot of analyses can
also be performed at this stage to understand process issues, discrepancies, which led
to NPAs in the first place. The insights obtained at this stage, would provide critical
inputs, to the preemptive asset quality management strategies followed by Banks and
go a long way in managing the challenge of balancing growth with quality. Overall, a
key feature, any Bank will look for while automating NPA management is the
flexibility to handle new product launches, new source systems and external changes
like modifications in regulatory guidelines. An asset management cycle is depicted in
the following diagram.
Preparation: This stage encompasses obtaining data from multiple source systems to
ensure a single view of each customer. The challenge is also to handle and process a
variety of special cases like securitized assets and syndicated loans. Early warning
signals can also be configured at this stage on the basis of which reports will be

generated which list out cases that fall into the potential NPA category. This is detailed
in an earlier section of this paper as a typical preemptive step to ensure assets do not
become NPAs.
Processing: NPAs are first classified into the appropriate buckets after considering a
variety of factors like vintage, product type, availability of collateral. Calculating
provisioning levels requires collateral data also to be integrated. Also, restructured
loans have more stringent provisioning criteria which need to be taken into
consideration. There could also be some facilities which are exempt from NPA
computation. Though NPA calculation and provisioning is automated, Banks will still
probably need to cross check the calculations. In some cases, discretionary calls may
also be taken to reclassify. At this stage, it is crucial ensure changes are auditable and
this requires having in place a review mechanism. An accounting interface would also
be needed, to seamlessly integrate the figures thus calculated, into the balance sheet,
income statement and other relevant financial statements.
Presentation: NPA and provisioning levels once calculated need to be reported. After
meeting the obligatory regulatory requirements, Banks will also typically like to study
the NPA data along multiple dimensions like product type, branch, geography and
industry/sector and do a root cause analyses, to identify weak links in the asset
lifecycle.
Smart Management of the asset lifecycle can enable Banks to not just be compliant,
but over the long term, also help adjust their credit policy, product portfolio and
lending processes in a bid to reduce bad loans. From a regulatory perspective, NPA
data helps in building an accurate picture of asset quality which in turn becomes a
useful input in macroeconomic policy making.