Banks Credit Risk

Risk management
and Indian Banking:

Opportunities and Challenges
Susan Thomas
http://www.igidr.ac.in/susant
susant@mayin.org
IGIDR
Bombay
Risk managementand Indian Banking:Opportunities and Challenges p. 1
Background
The problem
Banking is risky business!
1. Very high leverage:
India: Food 1.1, Machinery 0.6, Automobiles 1.1, Auto
Ancillaries 1.19, Banking 17.6.
US: Manufacturing 0.25-0.35, Utilities 1.4-1.5, Trade 0.3-0.4,
Banking 15.
2. Opaque assets: Loans, OTC derivatives. Non-transparent OTC
trading.
3. Moral hazard: Deposit insurance.
Nobel Laureate Merton Miller: Banking is a disaster

prone 19th century industry.
The goals of Basel norms

In the late eighties, there was a lot of cross-border
lending particularly by the Japanese banks.
Japanese banks grew enormously and gathered
market share; Western banks complained about
Japanese banks being regulated badly.
Basel I was an attempt to standardise the regulation
governing the global banking industry.
The heart of the Basel I norms defined minimum
required equity capital, i.e. an attempt to contain
leverage.
Basel I, contd.
Required equity capital was a single number
calculated as a fraction of the risk weighted assets
(RWA).
RWA = w1 x1 + w2 x2 + . . ., where x1 was corporate
exposure, and w1 = 1.
The weights for all the other classes of assets was set
at less than 1.
The main focus appeared to be on addressing credit
risk.
The minimum equity requirement was set through a
minimum Capital Adequacy Ratio (CAR), at
typically 8% of RWA.
What was right about the Basel Capital Accord

The CAR requirement did reduce the extremely high
levels of leverage in the banking industry.
What was wrong about the Basel

Capital Accord
The calculation of RWA is incorrect.
Risks in the banking portfolio are not linear.
Assets were classified on very broad lines.

(Eg. OECD government bonds.)
The focus on credit risk gave banks incentives to

find new ways of bearing risk.
(Eg. higher exposure in interest rate risk, OTC derivatives.)
Ignored the problem of opacity - loans, OTC

derivatives, OTC trading.
Ignored differences between countries.
(If 8% works for the OECD, what is correct for India?)
Negative consequences
Even though these were broad recommendations,
they became rigid in the hands of weak banking
regulators.
This became especially problematic countries where
the regulatory framework was not strong enough to
develop their own risk management rules.
The focus shifted from taking risks with a clear
understanding of the returns, to blindly using BIS
rules.
Basel II
An attempt to move away from linear rules of thumb.
Some of the implementation involves
Trying to improve upon the linear formula.
Reliance on credit ratings.
Exploit internal models of risk measurement in banks.
Taking more interest in incentives - of banks, of securities
markets.
This is still playing the game of Basel I - but trying to

find a better formula for equity capital.
What India should do with Basel-II

We should not repeat the mistakes about Basel-I
that of blindly adopting some externally supplied set
of rules.
We should treat Basel-II as a set of interesting ideas,
and craft a new framework of banking regulation
based on genuine understanding of risk.
Improving Indian banking regulation

Need to develop models for interest rate VaR.
Need to develop models of credit risk.
Banks must be given incentives to create such models
internally. One proposal:
The bank must present their internal risk models to the
regulator, and if they are good enough, be used for the
calculation of CAR.
Need to move towards more transparent assets bonds, not
loans; exchange traded, not OTC derivatives.
These require sound market design.
Importance of market discipline; models of bank failure
probability based on the stock price.Risk managementand Indian Banking:Opportunities and Challenges p. 11
Progress in India on
interest rate risk modelling
Importance of fixed income risk

There is some evidence that banks in India
substituted credit risk by interest rate risk when RBI
lay down a common risk management framework
for the Indian banking sector based on Basel-I.
Patnaik & Shah (2002): Roughly two-thirds of
banks in India would lose more than 25% of equity
capital when faced with a 99% shock.
Existing rules about interest rate risk regulation are
wrong: i.e. 2.5% risk weightage, and IFR.
Fixed income VaR

The yield curve fluctuates - this generates price risk
for every fixed income portfolio.
We seek statements like VaR for the portfolio at a
99% level on a one-day horizon.
This is the rupee loss which will be exceeded
tomorrow with a 1% probability.
Fixed income VaR

Make a model about fluctuations of the yield curve.
Simulate 10,000 draws from it.
Fixed income VaR

Reprice the full portfolio at each of these draws.
Fixed income VaR

So we get 10,000 outcomes for the profit/loss on the

portfolio on a one-day horizon.
Read off the 100th worst loss after sorting these
10,000 numbers.
Fixed income VaR

faces serious hurdles

10,000 numbers.
Fixed income VaR

This requires a good model telling us how the entire
yield curve fluctuates.

10,000 numbers.
Fixed income VaR

This requires a sound pricing technology whereby
the impact of alternative yield curves is clearly
known.
10,000 numbers.
Fixed income VaR

This requires a sound pricing technology whereby
the impact of alternative yield curves is clearly
known.
10,000 numbers.
This is easy.
Difficulties in testing
VaR methodologies must be backed by testing.
Banking applications require VaR over long
horizons.
Here the tests of VaR are particularly weak.
Data in India is weak.
We should be careful in knowing what we do not
know.
Fixed income derivatives

So far we have only talked about bonds.
What about interest rate futures, interest rate
options?
Jayanth Varmas Risk Management Committee has
worked on models for computing VaR on a one-day
horizon.
(The committee report is on the SEBI website.)
This is fundamentally easier, since we seek only a
one-day horizon, not a one-year horizon.
Committee observes lack of scientific knowledge,
but is confident about conservative approximations.
This work will drive collateral requirements for
interest rate futures and options.
Progress in India on
credit risk modelling
Credit risk requires three steps

What is the failure probability of a bond?
This is about predicting default.

What is the loss given default?
This is about creditors rights.

The above two give how much risk premium to
charge for a loan.
This has logic like the CAPM - betas, systematic
risk.

The above two give how much risk premium to
charge for a loan.
This has logic like the CAPM - betas, systematic
risk.
How do we put the pieces together to think about
portfolio VaR?
What you cant diversify has to be priced over and
above simple risk-neutral reasoning.
Situation in India
Problem
Failure of a bond
Loss given default
Portfolio credit risk
What we know
Quite a bit
Little.
Very little.
Progress on modeling failure of one

firm
Requirement
1.
Underlying firm-level accounting database
2.
Defaults database
3.
Model predicting default
4.
Working through stock

market data (Merton
model, KMV model)

firm
Requirement
1.
Underlying firm-level accounting database
2.
Defaults database
3.
4.

market data (Merton
model, KMV model)
How we get there

firm
Requirement
How we get there
1.
Underlying firm-level acCMIE - available in 1989.

counting database
2.
Defaults database
3.
4.

market data (Merton
model, KMV model)

firm
Requirement
How we get there
1.

counting database
2.
Defaults database
3.
4.

market data (Merton
model, KMV model)
CMIE - available in 2002.

firm
Requirement
How we get there
1.

counting database
2.
Defaults database
3.
4.

market data (Merton
model, KMV model)

CMIE Credit Model.

firm
Requirement
How we get there
1.

counting database
2.
Defaults database
3.
4.

market data (Merton
model, KMV model)

CMIE Credit Model.
Shah & Thomas, 2000; Thomas, Shah
& Karandikar (2002).
From the CMIE Credit Model:

1.0
0.8
0.6
0.4
CMIE Credit Model

Moodys
Altman Z-Score 1
Altman Z-Score 2
Model with no prediction capability
Model with perfect prediction capability
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
Risk managementand
Indian Banking:Opportunities
and Challenges p. 22
From Thomas, Shah & Karandikar,

2002:
Downgrades
Upgrades
Reaffirmations
Avg. DfD
2.0
1.5
1.0
-200
200
Number of days
Further reading
VaR
S USAN T HOMAS and A JAY S HAH. Risk and the
Indian economy. In: India Development Report
1999-2000, (editor) K IRIT S. PARIKH, chapter 16,
pages 231242. Oxford University Press, 1999
A JAY S HAH and S USAN T HOMAS. Rethinking
prudential regulation. In: India Development
Report 1999-2000, (editor) K IRIT S. PARIKH,
chapter 17, pages 243255. Oxford University Press,
1999
M ANDIRA S ARMA, S USAN T HOMAS, and A JAY
S HAH. Selection of Value at Risk models. Journal
of Forecasting, 22(4):pages 337358 (2003)
Recent work on fixed income VaR

I LA PATNAIK and A JAY S HAH. Interest-rate risk in
the Indian banking system. Technical report,
ICRIER, New Delhi (December 2002)
Jayanth Varmas committee report on interest rate
derivatives - came up on SEBI website on 19/3/2003.
Gangadhar Darbha has done work on extreme value
theory for interest rate VaR.
Credit risk
CMIE Prowess manuals.
S UBRATA S ARKAR and S USAN T HOMAS.
Assessing default probabilities using accounting
data: a case of firms in india. Technical report,
IGIDR (2003)
A JAY S HAH and S USAN T HOMAS. Systemic
fragility in Indian banking: Harnessing information
from the equity market. Technical report, IGIDR,
Bombay, India (December 2000)
S USAN T HOMAS, A JAY S HAH, and R AJEEVA L.
K ARANDIKAR. Does the stock market get it before
the rating agencies? Some evidence on the Merton
model. Technical report, IGIDR and ISI, Delhi
(June 2002)

Banks Credit Risk

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Banks Credit Risk

Uploaded by

Copyright:

Available Formats

Risk management

and Indian Banking:

Risk managementand Indian Banking:Opportunities and Challenges p. 1

Risk managementand Indian Banking:Opportunities and Challenges p. 2

Nobel Laureate Merton Miller: Banking is a disaster

Risk managementand Indian Banking:Opportunities and Challenges p. 3

The goals of Basel norms

Risk managementand Indian Banking:Opportunities and Challenges p. 4

What was right about the Basel Capital Accord

Risk managementand Indian Banking:Opportunities and Challenges p. 6

What was wrong about the Basel

Assets were classified on very broad lines.

The focus on credit risk gave banks incentives to

Ignored the problem of opacity - loans, OTC

Risk managementand Indian Banking:Opportunities and Challenges p. 8

This is still playing the game of Basel I - but trying to

Risk managementand Indian Banking:Opportunities and Challenges p. 9

What India should do with Basel-II

Risk managementand Indian Banking:Opportunities and Challenges p. 10

Improving Indian banking regulation

Risk managementand Indian Banking:Opportunities and Challenges p. 12

Importance of fixed income risk

Risk managementand Indian Banking:Opportunities and Challenges p. 13

Fixed income VaR

Risk managementand Indian Banking:Opportunities and Challenges p. 14

Fixed income VaR

Risk managementand Indian Banking:Opportunities and Challenges p. 15

Fixed income VaR

Reprice the full portfolio at each of these draws.

Risk managementand Indian Banking:Opportunities and Challenges p. 15

Fixed income VaR

Reprice the full portfolio at each of these draws.

So we get 10,000 outcomes for the profit/loss on the

Fixed income VaR

Reprice the full portfolio at each of these draws.

So we get 10,000 outcomes for the profit/loss on the

Fixed income VaR

So we get 10,000 outcomes for the profit/loss on the

Fixed income VaR

Fixed income VaR

Risk managementand Indian Banking:Opportunities and Challenges p. 16

Fixed income derivatives

Risk managementand Indian Banking:Opportunities and Challenges p. 17

Risk managementand Indian Banking:Opportunities and Challenges p. 18

Credit risk requires three steps

Risk managementand Indian Banking:Opportunities and Challenges p. 19

Credit risk requires three steps

Risk managementand Indian Banking:Opportunities and Challenges p. 19

Credit risk requires three steps

Risk managementand Indian Banking:Opportunities and Challenges p. 19

Credit risk requires three steps

Risk managementand Indian Banking:Opportunities and Challenges p. 20

Progress on modeling failure of one

Underlying firm-level accounting database

Model predicting default

Working through stock

Progress on modeling failure of one

Underlying firm-level accounting database

Model predicting default

Working through stock

How we get there

Risk managementand Indian Banking:Opportunities and Challenges p. 21

Progress on modeling failure of one

How we get there