PSU Bank Stocks Fall Up to 3% Today: The RBI Just Changed the Rules on Bad Loans

The Nifty PSU Bank index fell as much as 2% intraday, touching 8,684 on the NSE on Tuesday, April 28, 2026. Bank of India, Bank of Baroda, Canara Bank, Bank of Maharashtra, Union Bank of India, and Punjab National Bank were all down between 2% and 3%. State Bank of India, Indian Bank, Central Bank of India, and UCO Bank fell around 1% each. The broader Nifty 50, by contrast, was roughly flat at the same time. This was not a market-wide selloff. It was a targeted reaction to a single regulatory announcement.

The trigger: The Reserve Bank of India, on Monday evening (April 27, 2026), issued its final guidelines for transitioning to an Expected Credit Loss framework for loan provisioning. Effective April 1, 2027. Banks had been hoping for more time, but they did not get it.

What Is the ECL Framework and Why Should You Care

To understand why this spooked the market, you need to understand how Indian banks currently account for bad loans — and how that is about to change fundamentally.

Right now, Indian banks follow what is called the Incurred Credit Loss model. A bank only sets aside money as a provision against a loan when there is clear evidence the borrower has actually defaulted or is very close to doing so. In short, you wait for the problem to show up before you book it.

The Expected Credit Loss model flips this logic entirely. Under ECL, banks have to estimate the probability of a loan going bad in the future and start provisioning for it from day one, even when the borrower is still paying on time. It is forward-looking, not backward-looking. Think of it as buying fire insurance when you move in, not after the house catches fire.

This is not a new idea globally. The International Financial Reporting Standard 9 (IFRS 9), which most of the world has followed since 2018, is built on ECL. Indian banks have simply been operating on an older framework. The RBI has decided that alignment with international norms can no longer wait.

The Three Stages

The RBI's final guidelines use a three-stage provisioning structure.

Stage 1 covers loans performing normally with no significant credit deterioration. These attract a minimum 0.40% provision for secured retail and corporate loans.

Stage 2 covers loans where credit risk has increased significantly. The key automatic trigger is any loan overdue by more than 30 days, which the RBI terms a Significant Increase in Credit Risk (SICR). Stage 2 requires lifetime expected loss provisioning, with a minimum floor of 5%.

Stage 3 covers credit-impaired loans already classified as Non-Performing Assets. These attract progressively higher provisioning, up to 100% for unsecured exposures.

The number that caught everyone's attention is the 5% floor for Stage 2. Under the current framework, standard assets attract just 0.4%. Stage 2 under ECL requires at minimum 5% — more than twelve times the current requirement for loans that are not yet technically bad but are showing early warning signs. For a banking system carrying a significant pool of Stage 2 assets, this is not a cosmetic change. It is a structural increase in credit costs.

ICICI Securities, in a note on the subject, put it plainly: provisioning requirements will structurally rise under ECL, with Stage 2 assets attracting a minimum 5% provision versus 0.4% currently, implying a meaningful increase in credit costs over time.

Why Banks Were Hoping for More Time

The ECL framework was first proposed in draft form on October 7, 2025. Banks submitted feedback asking for two things: more time to build the data infrastructure needed to model forward-looking credit risk and a lower provision rate for Stage 2 assets than the proposed 5%.

They got neither.

The RBI's final norms are almost identical to the draft. The 5% Stage 2 provision stands. The April 1, 2027 implementation date stands. The message from the regulator is clear: the banking system has had time to prepare, and alignment with international standards is non-negotiable.

Building an ECL model is genuinely complex. Banks need historical loan-level data going back several years, economic scenario models, and the ability to estimate probability of default and loss given default for every borrower category. Smaller public sector banks, which have historically had less sophisticated risk management infrastructure than their private sector counterparts, face a steeper climb here. This is one reason PSU bank stocks reacted more sharply than private bank stocks today.

The Transition Relief

The final framework does contain meaningful transition relief. The incremental provisioning impact at the point of adoption does not have to be absorbed as a one-time P&L hit. Instead, it can be adjusted against opening retained earnings and spread over four years, until March 31, 2031. CET1 capital addbacks are also permitted during the transition period to soften the impact on capital ratios.

ICICI Securities described the transition framework as largely supportive, noting that it mitigates near-term earnings and capital volatility through phased provisioning and reserve-based adjustments.

That is the good news. The less good news is that the relief is temporary. Once the transition period ends, the structurally higher provisioning requirement is permanent. Banks will carry higher credit costs as a baseline going forward, compressing return on assets and return on equity over time, which is what the market is trying to price in today.

What This Means for PSU Banks Specifically

Public sector banks have had a remarkable few years. After nearly a decade fighting the ghost of bad loans from the infrastructure lending cycle of the 2010s, they cleaned up their balance sheets dramatically. Gross NPAs across scheduled commercial banks fell to approximately 2.3% by March 2025, the lowest in over two decades. Profits surged. Dividends recovered. The market rewarded them.

The ECL framework does not undo any of that. The improvement in asset quality is real. But ECL raises the cost of maintaining that improvement. Banks can no longer simply wait for a loan to turn bad before provisioning; they have to anticipate it. That requires capital, data, and a more conservative approach to loan book management, all of which have cost implications.

There is also a competitive dimension. Private sector banks — particularly HDFC Bank, ICICI Bank, and Axis Bank — have generally maintained more granular data systems and more sophisticated credit risk models. The transition to ECL plays to their existing strengths. For PSU banks, the gap in risk management infrastructure is real, even if it is narrowing.

The Bigger Picture

Here is the part that often gets lost in the noise of a down day: the ECL framework is fundamentally the right move.

The incurred loss model has a well-documented flaw. Because it requires visible evidence of default before provisioning kicks in, it is backward-looking. During good times, provisions stay low because loans look fine on the surface. When the cycle turns, banks scramble to provision for a wave of deteriorating assets all at once, exactly when they can least afford to. This is called provisioning procyclicality, and it amplifies financial system stress rather than cushioning it.

ECL forces banks to provision continuously and forward-lookingly. It smooths the cycle. When the next credit downturn arrives, Indian banks will be in a structurally better position to absorb losses without needing emergency capital or government bailouts. The 2010s bad loan crisis cost the Indian government hundreds of thousands of crores in bank recapitalisation. ECL is, in part, a regulatory effort to make sure that does not happen again.

The Bottom Line

Today's fall in PSU bank stocks is a rational market reaction to a concrete increase in future provisioning costs. It is not panic. It is pricing. The RBI has made a policy choice that will modestly compress bank profitability over the medium term in exchange for a structurally more resilient banking system. That is a reasonable trade for the financial system as a whole, even if it is not comfortable for bank shareholders in the short run.

The transition relief, the four-year phase-in, and CET1 addbacks mean the near-term damage is manageable. But the structural reality of higher credit costs is here to stay.

For long-term investors in PSU bank stocks, the question is not whether to sell in a panic today. It is whether the earnings trajectory from here, adjusted for higher provisioning, still justifies current valuations. That is a more nuanced question and one that deserves more than a 2% single-day decline as an answer.