When Collateral Isn’t Enough: Why ECL Could Redefine Indian Banking

The Expected Credit Loss regime may well prove to be one of the most transformative changes in Indian banking since liberalisation. But its true disruption will not lie in impairment numbers, which is the most visible outcome, but in a deeper, more consequential shift: a cultural overhaul of how banks perceive and manage credit risk.

For years, Indian banking has cared more about what a borrower owns than what they earn and can earn. Collateral has often stood in for credibility, as though property could replace judgement. The shift to ECL—India’s version of the IFRS-9 accounting standard—forces a change in mindset. It asks banks to look ahead, not behind, recognising possible losses from the day a loan is granted, and to provision for risk before it turns into regret. 

In essence, ECL demands that Indian banks learn to anticipate rather than apologise.

Risk Culture
Globally, internal credit rating systems have long formed the backbone of modern banking. They are central to capital adequacy calculations and risk assessments. In India, however, internal ratings have largely existed as a compliance formality--a box to be ticked under Basel-II guidance rather than a living, data-driven discipline.

ECL changes that. Under the Basel framework, the same risk parameters used for a bank’s internal management, particularly the Probability of Default, must now inform ECL computations as well. In other words, banks can no longer maintain one set of assumptions for internal comfort and another for accounting compliance.

The Reserve Bank of India has done commendable work in creating this convergence, explicitly linking risk management and accounting PDs in its draft ECL framework for banks. In the NBFC version, this was left implicit, creating room for ambiguity. This insistence on alignment should, over time, push banks to build more transparent, data-based, and analytically rigorous credit models.

Lending remains the heartbeat of Indian banking, accounting for roughly four-fifths of its capital allocation. The arrival of ECL will formalise the way credit risk is assessed, disclosed, and modelled. It brings the conversation out of the credit committee’s conference room and into the auditor’s report; and soon, into the investor’s radar.

Steep Climb
The road ahead, though, is not without hurdles. First, the formalisation of credit rating systems will mean that the quality of a bank’s internal models can now directly impact its profitability. Impairments will be tied to these models’ outputs, forcing managements to defend their assumptions before regulators, investors, and analysts alike.

Second, the new framework raises the bar on evidence. Banks must now demonstrate, through data and experience, that their internal credit models and ratings genuinely reflect and track borrower performance. The robustness of link between ratings and real outcomes will decide how credible a bank’s risk management truly is.

That needs to be taken seriously because history offers warnings. For instance, Credit Suisse’s collapse in 2023 exposed deep flaws in its credit risk and model validation processes, showing how even sophisticated systems can fail when complacency seeps in.

Third, the RBI’s draft also signals that supervisory reviews will become more granular. Banks will be judged not only on compliance but also on how their risk models perform under stress and how closely they mirror economic reality.

Reconsidering Relaxations
Yet some proposed relaxations deserve a second look. For instance, the RBI draft allows banks to use overdue payment information to assess credit risk levels when it is deemed “unduly costly or difficult” to make such estimations otherwise. This carve-out mirrors a similar clause in IFRS-9, designed mainly for non-financial entities for whom credit assessment is peripheral.

But for banks, credit risk is the business itself. If assessing borrower risk entails “undue cost or effort”, it raises a fundamental question: how was the loan justified in the first place? Basel’s own guidance recognises this, blocking such exceptions for internationally active banks, while not permitting definitively the exception for others. India, aspiring to global standards, should consider doing the same.

The same prudence applies to post-model adjustments or management overlays—manual interventions that allow management to override model outputs when they believe the models fail to capture certain realities. These are acceptable only as temporary measures. Basel envisages them as stopgap tools until models are refined, not as permanent levers of discretion. If used as a general practice, PMAs could undermine precisely the data-driven discipline ECL seeks to instil.

Financial Implications
The initial financial shock of ECL adoption is inevitable. When NBFCs transitioned to the regime a few years ago, many saw an immediate one-off hit to equity and asset values as provisions rose. The same will likely happen in banking, though the RBI has sensibly allowed a multi-year adjustment window.

The impact won’t be uniform. Much will depend on how clean a bank’s books are and how much capital it already has set aside. Those that have been prudent with provisions will find the shift manageable. Others, weighed down by old lending habits, could feel the hit more sharply.

But over time, this upfront pain could yield longer-term stability. A system that anticipates losses early is less likely to be blindsided by them later.

Across the world, regulation is leaning towards quantification and consistency, and India is no exception. The RBI’s ECL framework reflects that evolution — replacing qualitative comfort with measurable rigour. For an industry long haunted by bad loans and uneven governance, such discipline is overdue. ECL will not just change the way banks provision; it will change the way they think. The shift from collateral to credit culture is more than technical.

India’s banks have long been lenders of security. It is time they became lenders of trust.