The Reserve Bank of India (RBI) has come out with a discussion paper on implementation of expected credit loss (ECL)-based provisioning for the banking system.

The move, first proposed seven years ago, was put on the backburner before the Covid-19 pandemic. Now, with asset quality and capital available at banks looking better, RBI seems intent on implementing it.

The regulator has invited the views of stakeholders on the new framework even though various non-banking finance companies (NBFCs) already use the existing IndAS 109 framework to compute ECL and report provisions.

To be sure, the proposed ECL-based provisioning norms diverge a little from the IndAS 109 norms for NBFCs. The discussion paper also highlights some important aspects of ECL-based provisioning for banks and poses a few important questions on classification of assets. Key among these:

Provisioning norms may continue to diverge for banks and NBFCs

RBI has highlighted that since ECL-based approaches allow for significant variability of methodology as well as internal data being consumed, a regulatory backstop will serve as a floor for provisioning. Regulatory provision backstops established for large NBFCs (UL) are similar to the ones prescribed under the incurred loss-based approach of Income Recognition and Asset Classification (IRAC) norms.

The discussion paper, however, highlights that regulatory provisioning backstops for banks will be higher as they will consider an ECL approach rather than an incurred loss approach assumed under IRAC. 

Provisioning requirements will go up for banks

Banks will have to increase provisioning as regulatory backstops are expected to be higher under new norms. Even if the asset quality of a specific bank is on the better side, provisioning will go up as the base provisioning itself is likely to increase.

The regulator does plan to allow a five-year window to normalise the impact of the transition on financials by adding incremental provisions (net of tax) back to common equity tier 1 (CET1) capital, which will provide some relief to financial institutions (FIs).

That said, whether the regulatory backstops will be applicable at the loan level, the portfolio level, or the organisation level needs to be considered. Regulatory backstops at the organisation level will be the most aggressive approach and lead to the lowest overall provision numbers, while those at the loan level will be the most conservative approach and lead to higher provisioning.

Other than these, key questions raised in the paper that need deeper stakeholder interactions include:

• Should continuous low (<90) days past due (DPD) for over 90 days be considered default owing to losses on account of time value of money?
• Should there be a 6-month cooling period for transitioning assets from Stage 3 to Stage 1?
• Should restructured assets be considered as Stage 3 until they are under monitoring?
• Would a 60 DPD backstop for Stage 2 be considered instead of a 30 DPD backstop under IFRS9?

“The views of the author are solely based on Reserve Bank of India’s discussion paper on Expected Loss (EL)-based Approach for loan loss provisioning by banks dated January 16, 2023 and does not take into account any modifications made to the discussion paper thereafter. This whitepaper is being published for academic/ research purpose only and is not intended to provide any legal/ regulatory/ statutory advice. The opinions expressed should strictly be read in conjunction with the aforementioned discussion paper and should not be inferred in any other context. CRISIL shall not be responsible for the results obtained from any use of or reliance on the contents herein and especially states that it has no financial liability whatsoever to the subscribers/ users/ readers/ transmitters/ distributors of this white paper.”