Abstract

This article examines the proposed transformation of India's credit risk framework following the 'Reserve Bank of India (All India Financial Institutions-Asset Classification, Provisioning and Income Recognition) Directions, 2025 – Draft for Comments' issued by RBI in October 2025. The objective of the article is to analyse the need for a forward-looking provisioning system, evaluate its theoretical foundations, and interpret regulatory expectations outlined in recent policy directions.

The study for this article was conducted by reviewing regulatory drafts, accounting standards, global practices, and risk-measurement methodologies. The analysis also assessed the structural components of the Expected Credit Loss model, including probability of default, loss given default, and exposure at default, along with scenario-based modelling frameworks.

The research evaluated classification stages, governance mechanisms, prudential floors, transitional capital adjustments, disclosure requirements, and the broader alignment with international standards. The findings indicated that the Expected Credit Loss system represented a significant shift toward early identification of credit deterioration, improved transparency, and enhanced resilience.

The analysis concluded that, despite challenges related to data quality, modelling complexity, and operational readiness, the Expected Credit Loss framework offered long-term benefits that strengthened financial stability and brought the credit system closer to global regulatory norms.

Introduction

India's financial system is entering a pivotal phase as the Reserve Bank of India (RBI) works toward adopting the Expected Credit Loss (ECL) approach for credit impairment. For decades, Indian banks and financial institutions have grappled with the limitations of delayed recognition of stress, clustering of provisions during downturns, and inconsistent representation of underlying credit risks.

ECL promises to change this landscape by incorporating future economic expectations, borrower-specific indicators, and probability-based models. As India prepares for this transition, the financial sector stands on the threshold of a more robust, transparent, and globally aligned risk ecosystem.

Recently, the Reserve Bank of India also came up with Reserve Bank of India (All India Financial Institutions-Asset Classification, Provisioning and Income Recognition) Directions, 2025-Draft for Comments. As per the same, the ECL provisioning is tentatively scheduled to be implemented from 01.04.2027 with a glide path up to 31.03.2031.

Why India needs a new provisioning framework

The incurred loss model depends heavily on observable evidence of impairment, such as overdue payments, adverse financial results, or a clear decline in financial performance. As a result, provisions often begin only when credit quality has already deteriorated significantly. This reactive approach has several drawbacks, like stress not getting captured early, a spike in provisions, real weakness not getting detected early, issues in credit differentiation, etc.

The proposed ECL norms, being forward-looking in nature, will be able to capture the potential losses in advance. Following the 2008 global financial crisis, several jurisdictions recognised the inadequacy of incurred-loss models. The introduction of IFRS 9 marked a worldwide shift toward forward-looking risk assessment. Many countries have already adopted ECL, and India's move is both timely and necessary to remain consistent with global best practices.

What exactly is the loss from the ECL perspective

While lending, banks have to take care of both the cost and rewards in a loan, the rewards being the interest income, processing fee, etc., and the cost being the cost of funds, operational expenses, which are visible in nature and potential losses, which are invisible in nature, etc. These losses can be expected losses, or they can be unexpected losses. Expected loss is the average rate of loss expected from a loan portfolio. Losses above the expected levels are usually referred to as unexpected losses (UL).

To read more, please subscribe.