Will Indian banks’ credit loss, NPAs surge as new accounting standard sneaks around the corner?

Jaya Vaidhyanathan, CEO, BCT Digital

COVID-19 has sent the Indian economy spiralling. With close to 74% of small industries and start-ups scaling or shutting down, and more than 2 crore jobs cut back since April 2020, the economy is stopping just short of freefall. With even their most reliable borrowers succumbing to the economic devastation of COVID-19, financial institutions need to heighten diligence to protect themselves against the risks of bad loans and NPAs. IND AS 109 and its highlight – the expected credit loss approach to provisioning – have been heralded as the silver bullet to counter the collapse of India’s banking sector. Will they live up to expectations, or will the pandemic complicate things?

The global financial crisis (GFC) of 2008 changed our perception of risk management in a profound manner. The excessive pro-risk mentality of financial institutions (FIs) was replaced with concerns founded on increasing market volatility and rising risk exposure levels. One of the key outcomes of the crisis was how it highlighted systemic flaws that resulted in the late recognition of credit losses. This was later attributed to inadequate provisioning for unexpected losses caused by unpleasant economic events, credit downturns, change in the market value of the asset, and so on.

The wake-up call: Lessons from the Lehman fall and GFC

In order to enhance predictability and credit health, reduce risks and ensure better preparedness for contingencies, global accounting standard boards, such as IASB (International Accounting Standards Board) and FASB (The Financial Accounting Standards Board – US) published new forward-looking standards for loan loss provisioning, which would come into effect between 2018 and 2021. The new models rely on ECL (Expected Credit Losses) and CECL (Current Expected Credit Losses) standards that, rather than look for red flags or distress signals as traditionally done, accommodate probable credit losses in the future, irrespective of the current state of the asset or the lending market.

The main similarity is that both of the standards recognize and promote a forward-looking stance in calculating lifetime losses. The differences are many, but the primary variation is in terms of the degree to which the credit losses are recognized over an asset life. CECL calculates lifetime ECL for all financial assets, going back to the beginning. Meanwhile, ECL measures loss allowance at either a) a 12-month ECL for stage 1 financial assets, or b) the lifetime ECL for stage 2 and 3 assets (those that are classified under high or impaired risk).

IND AS 109 vs global accounting standards

When compared with its global counterparts like the IFRS 9, IND AS 109 is superficially similar. The underlying principle and objective are the same; they are forward-looking and highly effective. Many countries in the European Union and across South-East Asia have adopted IFRS 9 and are in the transition period. India too will join this league with IND AS 109 coming into effect in future – but the wait is far from over.

ECL vs Incurred Loss

Nearly an immediate offshoot of the financial crisis, regulators began to rethink the way banks measure and account for credit risk. The erstwhile Incurred Loss Model warranted a close re-examination. The model works on the assumption that all loans will be repaid, unless evidence to the contrary – a trigger event. It records credit losses as of the balance sheet date when loss is triggered by an observable event, like past-due, decrease in collateral value, and so on. The application of diverse and inconsistent provisions in the case of loss incurred but not reported is a particular challenge to overcome.

One of the biggest arguments against this approach is that – as illustrated during the Lehman Brothers crisis – its backward-looking nature can result in underestimation of losses and higher provisions during a recession. On the other hand, ECL calculates the expected present value (PV) of credit losses that are expected to arise if the borrower defaults during the life of the financial asset. Key components of ECL are Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD). This methodology requires measuring provision based on the deterioration of credit risk and expected events in a more systematic manner to mitigate risk impact, making its adoption more justifiable in the current scheme of things.

The COVID-19 impact on ECL

Things have been further complicated by the recent changes in global economics as well as the COVID pandemic. Early reports show that the fallouts of COVID-19 have had a challenging effect on ECL computation, at least in most instances. ECL is calculated taking into effect historical losses as well as current and future economic scenarios and predictions. With COVID coming into the picture, future predictions are complex, to say the least. Any default assumptions regarding the current and future scenarios, too, are no longer valid.

The Basel Committee on Banking Supervision agrees that banks should take extraordinary support measures into account for their ECLs and advises amendments to the transitional arrangements for the regulatory capital treatment. In India, the RBI is clear about the treatment of credit loss during the moratorium period while allowing the deferment on instalments, including loan repayments, bullet repayments and equated monthly instalments. The banks were permitted to defer the ageing of asset classification to default/NPA, until 30 September 2020.

The moratorium further disrupts risk modelling and accuracy of predictions. As the moratorium gave the borrower the liberty not to pay until 30 September 2020, any default in payment during the period cannot be considered non-payment of dues. Any payment overdue during this period and the outstanding overdue before March 2020 will come under reclassification of risk grade (or marked as an NPA) only from 30 September 2020. The implication here is that during the moratorium period, the asset cannot be classified “at risk” or flagged for SICR, solely based on non-repayment. This has led to NBFCs requesting the RBI permission to draw-down from their reserves to make additional provision for expected losses due to the pandemic.

The effective interest or EIR remaining the same during this period, there is no significant loss or profit on record for the purpose of short-term calculation. However, we are focusing on the post-moratorium scenario when the banks start to recognise the overdues and account slippages to NPA for all the accounts, including the ones overdue before the moratorium period. Hence the focus is on long-term ECL due to insolvency, which would have larger impact on the provisioning. Banks need to analyse the gap of the current solvency scenario to potential insolvency at the aggregate and large exposure levels.

The road ahead

The current global scenario will prove to be an acid test for IND AS 109, and ECL in particular. The whole concept of ECL is based on the principle that it promotes prudency in lending during the short-term, which fortifies FIs for better performance in the long-term, irrespective of the market scenario. However, the pandemic and the looming uncertainty around it will call this claim to task. One may very well have to wait and watch if ECL is able to deliver on its high promises to the global economic landscape.

Jaya Vaidhyanathan is CEO at BCT Digital. Views expressed are the author’s personal.