Credit Risk is defined as a potential risk that a bank borrower or counterparty will fail to meet its obligations in accordance with agreed terms. It can also be due to the deterioration of creditworthiness of a debt instrument issuer or even in down gradation of credit rating of a pool of credit portfolio. The goal of credit risk management is to maximize a bank’s risk-adjusted rate of return by maintaining credit risk exposure within the acceptable levels. If we closely watch any Bank’s composition of risk-weighted assets we can find that major contributor is credit risk.
The Basel II Framework presents two approaches for calculating credit risk capital charge in a continuum of increasing sophistication and risk sensitivity:
- Standardised Approach and
- Internal Rating Based (IRB) Approach:
- Foundation Internal Rating Based (FIRB) Approach,
- Advanced Internal Rating Based (AIRB) Approach.
WHY INTERNAL RATING BASED APPROACH?
In the standardized approach, the risk weights for different exposures are specified by the regulator. To determine the risk weights for the standardized approach, the bank can take the help of external credit rating agencies that are recognised as eligible by Reserve Bank of India. In this approach Banks were relying mechanically on external rating and the granularity, as well as risk sensitivity with respect to the expertise of individual bank, was absent. On the contrary in IRB each bank can have its model with approval from the regulator and a bank with robust credit monitoring mechanism can save precious capital with lower risk-weighted assets.
Internal rating Based (IRB) Approach
The IRB Approach allows banks, subject to the approval of RBI, to use their own internal estimates for some or all of the credit risk components [Probability of Default (PD), Loss Given Default (LGD), Exposure at Default (EAD) and Effective Maturity (M)] in determining the capital requirement for given credit exposure.
The IRB approach allows it to use internal models to calculate credit capital, enabling more sensitivity to the credit risk in the bank’s portfolio. Furthermore, incorporating better risk management techniques on its portfolio will show its effect on minimizing the regulatory capital required. Another incentive to move to the IRB approach is that the IRB-based regulatory capital is “lower than” the standardized approach for higher credit-rated banks and “higher than” for lower credit-rated banks, thus providing a better alternative for investment-grade banks.
The IRB approach is again classified into:
- a) Foundation IRB (FIRB) approach: Banks estimate Probability of default (PD) using internal models, while the other parameters take supervisory estimates.
- b) Advanced IRB (AIRB) approach: Banks provide their own estimate of PD, Loss Given Default LGD, and Exposure at default (EAD) and their calculation for Maturity (M) is subject to the supervisory requirements.
Under the IRB approach, banks are required to categorize their banking book exposures into the following asset classes:
KEY CONCEPTS IN IRB APPROACH
- PROBABILY OF DEFAULT (PD)
It is the probability that a borrower with certain credit rating will fail to pay the interest or repayment obligation on the due date. It aims to measure the probability of borrower assigned with a rating other than default rate defaulting over a specific time horizon. Banks which have an internal rating methodology use the same to differentiate the corporate borrowers into different rating grades corresponding to varying credit risk profile. The borrowers in different rating grades will have different likelihood of default.
Bank would need to compute its internal rating wise PD for the corporate portfolio. PD may be estimated based on the historical default data.
EXAMPLE: From the above-mentioned table if we try to see the no of accounts in the rating grade UBC 4 was 200 at the beginning of the year and at the end, 5 accounts migrated to default so the PD works out to be 2.5%.
The PD estimate arrived using the above method is used for 12 months ECL computation. For Lifetime ECL, LIFE Time PD is required to be estimated from 12 months PD using matrix multiplication method.
|RATING GRADES||TOTAL NO AT THE BEGNING OF YEAR||MIGRATION DURING THE YEAR||PD IN %|
- EXPOSURE AT DEFAULT (EAD)
It’s the gross exposure under a facility at the time of default. Normally it is the total outstanding in case of fixed exposures like term loans. For running accounts or revolving facility, we can divide the facility into drawn and undrawn exposure. The undrawn commitment is arrived at by multiplying it with a conversion factor.
Example: If there is an exposure of Rs 1 crore financial guarantee which is a non-fund based facility then a CCF of 100% will be applied to it and is converted to funded exposures 1 crore.
Under internal rating-based approach the database should be for minimum 7 years.
- LOSS GIVEN DEFAULT (LGD)
It is the proportion of exposure that will be lost if a default occurs in an exposure. It normally indicates the magnitude of loss and expressed in percentage norms. It depends primarily on the type of collateral, value of collateral and security coverage ratio. LGD is facility-specific and different facilities to the same borrower may have different LGDs. The LGD may be floored to zero (RBI floored at min 20%) and capped at 100%. Under (Indian Accounting Standards) Ind AS 109 Expected Credit Loss (ECL) is computed to arrive at the provisioning requirement for loans and advances. LGD is one of the key inputs for ECL computation.
LGD = 1- RECOVERY RATE
Recovery rate is the amount that can be recovered through foreclosure or bankruptcy procedures in the event of default. It is generally expressed in percentage norms.
- EXPECTED LOSS (EL)
Expected loss of an asset is average loss that the bank can expect to loss over the period up to a specific horizon.
Where PD and LGD is expressed in terms of percentage and EAD in amount.
EXAMPLE: XYZ PRIVATE LIMITED has a term loan of Rs 100 crore with the bank. The PD for 1 Year is estimated at 2.5% and LGD at 65%
EL= 2.5%*65%*100CRORE = 1.625 CRORE
- UNEXPECTED LOSS (UL)
It is defined as a risk on a specific time horizon around the expected loss. This is measured by standard deviation of the asset value or loss incurred in the case of default. It’s the volatility of potential loss around expected loss. The Standard Deviation of PD about the expected loss shall generate Unexpected Loss. Normally Banks depend on their comfort requirement or as prescribed by regulatory authorities, multiply the unexpected losses arrived as above by the sigma number for the desired confidence level as mentioned below, to arrive at the economic capital requirement for credit risk of the Bank.
1.0 sigma – 68% confidence
1.65 sigma – 95% confidence
2.33 sigma – 99% confidence
3.00 sigma – 99.87% confidence
The baseline for all these concepts is to effectively predict how much capital is required to effectively manage a healthy asset portfolio keeping a view on regulatory as well as economic capital. The main focus of Indian banks is on the regulatory capital which alone is enough to take care of the expected losses. It may or may not be the actual risk and should not be a base for pricing a loan asset. On the contrary economic capital is based on statistical model and aims to absorb the unexpected losses to a certain confidence level. The economic capital should be the yardstick to price a certain asset portfolio as it factors into the unexpected losses and provides a clear picture. Another model which is now gaining importance and Banks are taking into account beyond a threshold level is Risk Adjusted Return On Capital (RAROC). It is calculated by adjusting net return from an asset with the expected amount of unexpected losses arising from it and discounting it by economic capital.
RAROC = (Net income-operating expense-Expected losses)/Economic capital