Credit risk losses are bifurcated into two types: expected loss and unexpected loss. Expected loss refers to the amount of loss anticipated on a loan or portfolio of loans and is mitigated using policies, risk pricing and provisions. Unexpected loss is addressed through regulatory capital.

Currently, banks in India follow the incurred loss approach – IRAC norms as prescribed by the Reserve Bank of India (RBI) – in accounting for losses on loans and other financial assets. This approach, however, is not aligned with IFRS 9 and Ind AS 109.

Incurred loss approach is based on the principle that losses are uncertain and difficult to predict, and that they should be recognised when they are certain to have occurred. Accordingly, even if a loan has a high risk of default and expected to result in a loss, the loss will not be recognised until the borrower actually defaults and the loss is incurred.

The expected credit loss (ECL) approach, on the other hand, recognises losses on loans as soon as they are expected to occur, regardless of whether the borrower has actually defaulted. The other key differentiator in the computation of provision for credit loss is that ECL factors in the historical credit quality of the lender, while incurred loss approach does not consider the same. It is used to estimate the expected loss for a portfolio of loans and to make decisions about lending, pricing and risk management.

What is ECL?
 

ECL is a method of accounting for credit risk that is based on the loss that is likely to occur on a loan or portfolio of loans. It is used to estimate potential future losses on financial assets and to recognise those losses in the financial statements. It represents the probability weighted estimate of the present value of all cash shortfalls from an instrument.

In simple terms, ECL is calculated by estimating the forward-looking probability of default for each loan, and then multiplying that probability by the likely loss given default, which is the percentage loss that is expected to occur if the borrower defaults. The resulting value multiplied by the likely exposure at default is the expected loss for each loan, and the sum of these values is the expected loss for the entire portfolio.

What does the regulator propose?
 

Through its discussion paper published on January 15, 2023, the Reserve Bank of India (RBI) proposes to adopt the ECL approach used in IFRS 9 for prescribing guidelines for loss provisioning by banks. The central bank has indicated that the proposed approach would be to formulate principle-based guidelines supplemented by regulatory backstops wherever necessary. While the discussion paper is open for comments, the RBI has, in its initial draft, proposed to align with the IFRS 9 approach.

Here is how it will pan out:

• Classify the financial instruments into three stages: Stage 1, Stage 2, and Stage 3
• Modify revenue recognition and risk measurement based on effective interest rate, and ascertain whether significant increase in credit risk has occurred on a reporting day as compared to initial recognition
• Provide for ECL subject to regulatory backstops

This method will apply to financial assets such as loans and advances, lease receivables, irrevocable financial guarantee contracts, and investments classified as held-to-maturity or available-for-sale.
 

What will be the quantum of credit loss at the reporting date?
 

What are the proposals regarding ECL models?
 

ECL is to be measured as a probability weighted estimate of credit loss (the present value of all cash shortfalls) over the expected life of the financial instrument. The bank will be permitted to design and implement its own models for measuring ECL for the purpose of estimating loss provisions.

The ECL models adopted by banks will be subject to rigorous validation as well as process-based checks, and to prudential regulatory floors.

What should banks do now?
 

In short, plan for life with ECL. The journey of adoption may not be straightforward as it involves developing, testing, validating, and implementing ECL models, while upskilling credit risk management teams. Following are some challenges that the banks can plan for and start addressing:

• Quality and availability of data on loans and other financial assets needed to calculate ECL
  
• Challenges in implementing ECL models for complex products and portfolios
  
• Banks will need to validate the ECL model to ensure that it is accurate and reliable
  
• Banks may have difficulties in finding internal expertise with knowledge of ECL, which needs to be supplemented by risk solutions providers with expertise and strong credentials
  
• Banks will need a tool to compare ECL and incurred loss at account and portfolio level enable them to understand additional capital requirements and underlying risks
  

The CRISIL approach to ECL
 

CRISIL offers consulting and automation services for computation of ECL in line with IFRS 9 regulations, addressing the challenges arising from minimal availability of historical data. Furthermore, our proprietary ECL module, which is an add-on to our existing Asset Classification and Provisioning solution, enables banks and NBFCs perform loan-level incurred loss vs expected loss analysis, provide for credit loss, and help in effective credit risk management at account and portfolio level.