Businorem ipsum dolor sit amet cons interdum quam duis variuy time honored tradition etting .

Sydney Harbour Circular City of Sydney,Australia.

img

Banks Move to ECL in April 2027: What NBFCs’ Experience Can Teach Them

For decades, Indian banks have provisioned for loan losses only after the damage became visible. Under the incurred-loss approach, a loan attracts meaningful provisions once it turns non-performing at 90 days past due, by which point the borrower’s stress is often months old. The Reserve Bank of India’s Expected Credit Loss (ECL) framework, which takes effect for scheduled commercial banks on 1 April 2027, reverses this logic. Banks will now have to estimate losses before they occur, based on how each borrower’s risk is changing and where the economy is heading. 

This is a significant change in mindset, but it is not untested in India. NBFCs with net worth above ₹500 crore have provisioned under Ind AS 109, which follows the same expected-loss principles, since FY2018-19. Their experience over the last eight years, which includes a pandemic, offers banks a practical guide to what the framework demands beyond building models. 

The most important lesson is that ECL shifts attention from default to deterioration. The framework places performing loans in Stage 1, loans whose credit risk has risen significantly in Stage 2, and impaired loans in Stage 3. The move from Stage 1 to Stage 2 is where the cost lies, because provisions then cover expected losses over the loan’s entire remaining life rather than only the next twelve months. The RBI treats a loan that is more than 30 days past due as presumptively in Stage 2, and sets a minimum provision of 5% for most such loans, many times what banks hold on standard assets today. A loan that slips into early delinquency therefore carries a real cost even if it later recovers, which makes collections and early-warning systems part of provisioning rather than just recovery. 

The second lesson is that ECL depends heavily on judgement, and judgement needs governance. Models learn from historical data, so they struggle when conditions change suddenly. NBFCs saw this during COVID-19, when most relied on management overlays, adjustments made on top of model outputs, to capture risks their models could not see. Overlays are legitimate, but without discipline they can also become a way to smooth reported profits. This is why the RBI requires every NBFC overlay to be approved by the Audit Committee, with its rationale documented. Banks would benefit from going a step further and deciding in advance what would justify creating an overlay and what evidence would justify releasing it. 

Third, ECL rarely produces a single, simple answer. NBFCs calculate provisions under Ind AS 109 alongside the RBI’s traditional norms, and set aside the difference in a separate reserve whenever the regulatory requirement is higher. Their Stage 3 assets can also differ from reported NPAs, because accounting rules allow impairment to be recognised before the 90-day mark. Banks will face similar gaps between model estimates, regulatory minimums and NPA classification. Boards, investors and rating agencies will need these differences explained clearly, which makes disclosure as important as the calculation itself. 

Finally, ECL changes the economics of lending, not just its accounting. Because provisions are recognised from the day a loan is made, and minimum levels differ by product, faster growth in riskier segments will show up as credit cost much earlier than before. Where the regulatory floor exceeds what a bank’s own model estimates, the floor effectively sets the cost of that loan. Banks should therefore build ECL into pricing, product limits and capital planning, rather than leaving it to the year-end close. 

The shift matters for NBFCs as well. Banks are significant lenders to NBFCs and their partners in co-lending, and once these exposures are provisioned on an expected-loss basis, banks are likely to scrutinise the quality and timeliness of partner data far more closely. NBFCs that can share reliable loan-level information and early-warning signals will be better placed in funding and co-lending discussions. 

Ultimately, ECL rewards lenders that treat it as a way of managing credit risk rather than a reporting requirement. NBFCs have spent eight years learning this lesson; banks now have less than five months to put it into practice.

Banks Move to ECL in April 2027: What NBFCs’ Experience Can Teach Them