Credit Risk

1
 Basel Norms and its connection with the Banks

Basel Norms are named after a city in Switzerland. This city hosts the Head Quarters
of Bank of International Settlements(BIS).

Established in 1930, the BIS is owned by 60 central banks, representing countries

from around the world that together account for about 95% of world GDP.

BIS’ mission is to serve central banks in their pursuit of monetary and financial
stability, to foster international cooperation in these areas and to act as a bank for
central banks.

https://www.bis.org/about/index.htm 2
Capital Adequacy Ratio (CAR)

CAR = Capital Required

Risk Weighted Assets
Credit Market
Operational

3
Minimum Capital Requirement
Banks need to maintain a minimum capital so that even in stressed times they are
able to function in a business as usual manner.

Insolvency of a bank has far reaching consequences for many counterparties

associated with it including its customers, investors, shareholders and even other
banks.

The questions that arise:

qHow would a bank decide how much capital is required?
qWhy minimum capital, why not more?

4
Basel I
qPrimarily focused on Credit Risk
qIntroduced norms related to Risk Weights based on the Asset Class

Asset Class Risk Weight

(a) Cash and Gold in bank 0%
(b) Claims on OECD* banks and Government bonds 20%
(c) Residential mortgages 50%
(d) Corporate bonds, claims on non-OECD banks, equities 100%

Risk Weighted Asset (RWA) = 0 x (a) + 0.2 x (b) + 0.5 x (c) + 1 x (d)

*Organization for Economic Co-operation and Development

5
Basel I
qIntroduced guidelines related to Capital Adequacy Ratio

Capital Required ≥ 8% of Risk Weighted Assets (RWA)

Example:
For 1 million residential mortgage RWA = 1 x 0.5* = 0.5 million
Minimum capital required = 0.5 x 0.08** = 0.04 million or 40,000

Note:
*Assuming that residential mortgages carry a risk weight of 50% or 0.5
**Minimum capital required is 8% or 0.08 of RWA

6
Basel I: Opportunities
Limited classes, moreover, it focused only on Asset Class but there was no focus on
the counterparty’s credit worthiness e.g. not all corporate bonds are the same, still it
treated them equally.

Easy to misuse
Example: A Corporate with better credit worthiness will always have more banks will
to lend them money, hence they can bargain when it comes to the interest rates they
are willing to pay.
On the other hand, a corporate with poor credit worthiness will have limited options
hence they will be willing to pay a higher interest rate. Now for the same amount of
risk Bank can make more money by making a riskier investment. Is that right?

7
 Basel I vs. Basel II

Supervisory Market
Min. Capital Min. Capital
Review discipline

Operational
Credit Risk Credit Risk Market Risk
Risk

Foundation Advanced
Standard Standard
Internal Internal
Approach only Approach*
Rating Based Rating Based

*Even the standard approach in case of Basel II was further refined e.g. it incorporated
counterparty ratings and didn’t treat them all as same. 9
Basel II: Opportunities

§Internally developed Risk Models often didn’t perform as great as expected and
particularly performed poorly in predicting the probability of high impact but rare
occurrence events.

§Optimism and lack of transparency of Banks often led to low capital to asset ratio

§Market disclosure though incorporated wasn’t really paid much heed

§External Credit Rating Agencies did a good job at an overall level but still didn’t
provide a custom fit specific to a counterparty’s operations

§Lack of validation and stress testing

10
Measurement Examples
While lending money to a business there are a number of factors that are considered
by a bank such as industry, credibility of the client, past performance of the business
through financial statements and many more. All these factors can be used to create
a model that predicts the PD which is always a value between 0 and 1.

If we are a bank, and one of our clients goes bankrupt and defaults. Let’s say the
outstanding debt of our client at the time of default is 50 million. This is EAD.

Let us assume, when the client defaults, we can obtain 20 million, by selling some
collateral. This means, we still lose 50 – 20 = 30 million. This corresponds to 60% of
the EAD (30'000'000/50'000'000 = 0.6), therefore, our LGD is 60%.

11
Approaches to assess Credit Risks

Standardized Approach

In case of standardized approach, the Capital Requirements are computed using

standard calculations i.e. it is 8% of the RWA. Though simple, it lacks sophistication as
the calculation of RWA is less scientific.

It relies on the inputs from External Credit Rating Agencies

Risk Weights for claims on Sovereigns/Central Banks
S&P*/Fitch AAA to AA A BBB BB to B Below B Unrated
Moody’s Aaa to Aa A Baa Ba to B Below B Unrated
Risk Weight (%) 0 20 50 100 150 100

Likewise there are ratings available for each Asset Class. 12

Approaches to assess Credit Risks

Standard Approach Example:

Let’s say, a bank has following Assets:

§80 million of AAA rated Government Bonds
§100 million of AA rated Corporate Bonds
§50 million of residential mortgages
§20 million of other mortgages
Calculate RWA and Capital Requirement, assuming the risk weights for AAA
Government Bonds = 0%, AA Corporate Bonds = 25%, Residential Mortgage = 50%,
Other mortgages = 100%

Solution:
RWA = 0 x 80 + 0.25 x 100 + 0.5 x 50 + 1 x 20 = 70 million
Capital Requirement = 70 x 0.08 = 5.6 million
13
Approaches to assess Credit Risks

Foundation Internal Rating Based Approach

In Foundation approach the RWA is calculated with the help of Probability of Default
or PD. Inputs to PD are Credit Ratings and other supporting predictive techniques.
Once the PD is arrived at, the bank uses the formula given by the regulators to
compute RWA. Again, the capital requirements are 8% of this RWA.

Approach PD LGD EAD

FIRB Internal Regulator Regulator

14
Approaches to assess Credit Risks

Advanced Internal Rating Based Approach

In Advanced approach, in order to compute RWA the banks are given freedom to
develop their own probabilistic models through thorough research without any
compulsion to stick to the formulae given by the regulators. In addition to Probability
of Default (PD), here banks also calculate Exposure at Default (EAD) and Loss Given
Default (LGD). However, the regulator approves this model and once approved the
Capital Requirement is calculated as 8% of the RWA.

Approach PD LGD EAD

FIRB Internal Regulator Regulator
AIRB Internal Internal Internal

15
Credit Scoring

16
Credit Scoring Approaches

Credit Exposures

Non-retail
Retail Exposures Exposures

§Application Scorecard §Prediction based

§Behavioral Scorecard §Expert rating based
§Rating agency based

17
Credit Scoring: Retail

18
Application Scorecard: Application variables
Sr# Application Variable
1 Age
2 Gender
3 Income
4 Existing Customer
5 Nature of employment
6 Years at current residence
7 Years with current employer What all does your bank
know about you?
8 Relationship tenure with Bank
9 Marital status
10 Spouse’s income
11 Number of dependents and more… 19
Credit Bureau insights
Credit Bureaus provide credit score as they have visibility to the overall financial well-
being of an individual based on his/her credit history not limited to just one bank.
Therefore, it represents the creditworthiness of a person.

Sr# Application Variable

1 Outstanding debt – from all sources
2 Number of Credit Cards Credit Bureaus
Equifax | TransUnion | Experian
3 Credit limit utilization
4 Past delinquencies
5 Number of credit inquiries
and more…
20
Credit Score

Fair, Issac & Company (FICO)

Sr# Application Variable Weight

1 Payment history 35%
2 Current debt burden 30%
3 Credit tenure 15%
4 Credit diversity 10%
5 Recent credit applications 10%

Image source: https://www.experian.com/blogs/ask-experian/infographic-what-are-the-different-scoring-ranges/

21
Behavioral Scorecard
q Generally performed over a longer time horizon
of around 24 months.

q Involves many more variables and derived ratios.

q Generated for existing credit customers.

q Involves reassessment of the Credit Risk based

on behavior e.g. change in job status, payment
delays, balance maintained in the account,
holding period.

q Observes trends with higher weights assigned to

the recent behavior.
23
Credit Scoring: Nonretail

24
 Prediction based approach

Based on the availability of historical data about the corporations which includes:
q Financial statements and various ratios as derived/reported
q Stock prices for listed companies

Altman Z score is a very popular measure for predicting bankruptcy. It is a linear combination of
5 financial ratios listed below:

R1 = Working capital/Total assets (measures liquidity ratio)

R2 = Retained earnings/Total assets (measures the extent to which a company relies on debt)
R3 = Earning before interest and taxes (EBIT)/Total assets (measures general profitability)
R4 = Book value of equity/Total liabilities (measure of market’s reaction to company’s position)
R5 = Net sales/Total assets (measures asset turnover)

25
Altman Z Score
It was originally developed for the manufacturing companies but has now been extended
to non-manufacturing sector as well. The values of coefficients corresponding to each ratio
is derived using a statistical technique known as Linear Discriminant Analysis.

For public industrial companies:

Z = 1.2R1 + 1.4R2 + 3.3R3 + 0.6R4 + R5

Where Z > 2.99 is considered healthy, Z < 1.81 is considered poor

For private industrial companies:

Z = 6.56R1 + 3.26R2 + 6.72R3 + 1.05R4

Where Z > 2.60 is considered healthy, Z < 1.1 is considered poor 26
 Expert Rating
Prediction approach is a good choice when data is available, but in the absence of data
the scorecard is developed based on Expert Rating. It is a qualitative approach which considers
the following factors:

qPosition within the business sector - market share, size, product portfolio, geographic spread

qManagement capability - experience, corporate governance

qExternal liquidity access - debt spread over time, diversified financing options

27
Rating Agency based
Credit rating agencies assign credit ratings based on a debtor's
ability to pay back the debt by making timely principal and
interest payments and the likelihood of default.

These rating agencies use variations of an alphabetical

combination of lowercase and uppercase letters, with
plus/minus signs and even numbers.

Moody's, S&P and Fitch Ratings control approximately 95% of

the credit ratings business.

Reference: https://en.wikipedia.org/wiki/Credit_rating
28