RBI Draft ECL Directions 2025

Executive Summary

The Reserve Bank of India (RBI) has issued Draft Directions introducing a forward-looking Expected Credit Loss (ECL) provisioning framework, replacing the current incurred-loss-based approach. Effective April 1, 2027, scheduled commercial banks (excluding Regional Rural Banks (RRBs), Small Finance Banks (SFBs), and payments banks) will be required to estimate ECL across all financial assets using robust models for Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD).

This transition aligns Indian banking regulation with globally accepted standards such as IFRS 9 and CECL, promoting greater comparability across institutions, strengthening credit risk management practices, and ensuring prudential floors to safeguard stability. The ECL framework fundamentally changes both the timing and magnitude of credit loss recognition, encouraging earlier identification of credit deterioration and reducing volatility in provisions during stress periods. While the impact on performing assets is expected to be modest, exposures showing early signs of stress will attract higher provisions due to lifetime loss recognition.

The Draft Directions also underscore strong governance requirements, mandating Board and senior management oversight of staging decisions, model assumptions, and judgmental inputs, supported by independent validation and audit trails. Uniqus’ global experience demonstrates that institutions that invest early in data quality, systems, and governance are better positioned to manage this transition smoothly and maintain resilience.

Why the shift matters

Under the existing incurred-loss Income Recognition, Asset Classification and Provisioning (IRACP) regime, banks recognize impairment only after objective evidence of loss (default or restructuring) is established/demonstrated. It could be argued that this delay potentially led to under-provisioning during good times and sharp spikes during stress cycles. ECL replaces these lacunae with probability-weighted estimates of future losses – embedding foresight into capital planning.

Example: If an INR 100 crore MSME portfolio carries a 1% 12-month PD and 40% LGD, the expected loss is INR 0.40 crore.

Under IRACP, the bank would hold a standard general provision of 0.25%, i.e., INR 0.25 crore, primarily prescribed by regulation rather than borrower-level risk.

Under ECL, the provision is risk-based (PD × LGD × EAD), leading to an allowance of INR 0.40 crore at origination itself – not only after default triggers – ensuring earlier and more risk-aligned loss recognition.

Why this shift matters (Reaction to Readiness)

In January 2023, the RBI considered implementing ECL norms through a Discussion Paper titled Reserve Bank of India (RBI) – Discussion Paper on Introduction of Expected Credit Loss Framework for Provisioning by Banks (issued 16 January 2023). In October 2023, a new external working group was formed to establish principles for credit risk models, propose methods for their external validation, and recommend prudential provisioning floors. Through the currently proposed draft guidelines, the RBI has laid a roadmap that will not only facilitate the shift to ECL-based provisioning but also replace the existing IRACP norms. These guidelines apply to Scheduled Commercial Banks (SCBs), excluding RRB, SFB, and Payment Banks. They are scheduled to take effect on 01 April 2027, with a gradual implementation timeline that extends to 31 March 2031.

The shift away from the incurred-loss model is not cosmetic. Under the earlier approach, provisions were recorded after evidence of loss appeared, usually once a borrower had already defaulted. That delay often distorted earnings and weakened capital buffers.

ECL changes the reporting requirements by embedding anticipation into credit management. Losses are recognized based on expected outcomes, not actual defaults. Through probability-weighted modelling and macroeconomic overlays, it embeds foresight into risk management and financial reporting and reduces cyclicality across the system. Over time, this should make Indian Banks balance sheets more transparent, less volatile, and more globally comparable.

Scope, Start Date & Transition

Scope: Includes Scheduled Commercial Banks and All India Financial Institutions (excludes RRBs, SFBs, Payments Banks).

Effective: 1 April 2027

Transition relief (Common Equity Tier (CET) 1 add‑back): The difference between ECL and current provisions can be added back to CET1 on a declining schedule, as shown below, with additional disclosure requirements to be addressed separately.

Core Components of RBI’s ECL Framework (Staging + Parameters)

The RBI’s proposed design rests on three pillars: 

ECL Staging Framework

The ECL model divides all credit exposures into three distinct stages that reflect the progressive deterioration of credit quality. Each stage determines the time horizon for recognizing expected losses and the quantum of provisioning.

To read more, click here

Model Parameters and Conceptual Underpinnings

The Expected Credit Loss (ECL) framework is based on three interlinked model parameters: Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD). Each parameter captures a distinct element of credit risk, and together they determine the expected loss under both baseline and stressed macroeconomic conditions.

To read more, click here

Governance and Validation

Robust governance and independent validation are what separate technical compliance from credible implementation. The ECL framework introduces multiple judgmental elements, including SICR triggers, scenario weights, overlays, and LGD assumptions, that directly affect financial statements and capital. To ensure integrity, institutions must establish a formal governance ecosystem that encompasses ownership, oversight, validation, and audit.

Governance Framework

1. Board and Senior Management Oversight: The Board of Directors holds ultimate accountability for approving the ECL policy, model architecture, key assumptions, and scenario design. It must periodically review the end-to-end provisioning outcomes, i.e., stage movement trends, macroeconomic overlays, and deviations between model and actual loss. The Audit Committee should be briefed each quarter on provisioning drivers and management overlays.
2. ECL Oversight Committee: Banks should constitute a cross-functional ECL Oversight Committee (typically comprising Risk, Finance, Internal Audit, and Information Technology) to supervise day-to-day governance. The committee should review model inputs, staging rationale, and exception cases, and ensure consistency between regulatory and financial reporting.
   • Risk owns methodology and parameter calibration.
   • Finance ensures accounting alignment, disclosures, and GL integration.
   • Internal Audit independently tests design effectiveness and control adherence.
   • IT to support the data governance and system-level support.
3. Policy and Documentation: All ECL methodologies, model hierarchies, and overlay policies must be captured in an ECL Governance Policy, approved by the Board. The policy should detail the process for:
   • Model change management and re-approval.
   • Handling of overrides or expert judgment.
   • Escalation and documentation of material deviations.
   • Frequency of recalibration and validation.

Independent Validation and Model Risk Management

1. Independent Validation Units: A dedicated Independent Validation Unit -functionally independent of model developers-should assess both conceptual soundness and empirical accuracy. Validation must cover:
   • Theoretical soundness: underlying statistical logic, model design, parameter assumptions, and treatment of outliers.
   • Data integrity: completeness, consistency, and representativeness of the datasets used.
   • Performance metrics: discriminatory power (e.g., Gini, KS), calibration accuracy, and stability over time.
   • Back testing: comparing predicted versus realized default and recovery outcomes over multiple periods.
2. Validation Reporting and Frequency: Results of validation should be reported to the Board Risk Committee or a designated Model Risk Committee, with all findings tracked through remediation plans and closure timelines. Models should undergo full validation at least annually, or earlier if there is a material change in data, methodology, or portfolio composition.

Audit and Data Lineage Controls

1. End-to-End Traceability: Audit reviews must ensure that every ECL computation is traceable from source data to general ledger entries. This includes validation of:
   • Data extraction logic from core systems.
   • Data transformations, adjustments, and reconciliation steps.
   • Final journal postings and disclosure reconciliations.
2. Internal Audit Role: Internal Audit should perform periodic thematic reviews focusing on:
   • Compliance with governance policy and approval protocols.
   • Adequacy of controls over data lineage, parameter updates, and version management.
   • Verification of management overlays and scenario selections.
3. Governance Forums and Continuous Assurance: Leading institutions are instituting ECL Review Forums comprising Risk, Finance, and Internal Audit to sign off on each quarterly computation. These forums serve as a “four-eyes” control mechanism before results are finalized, strengthening accountability and preventing bias in judgmental assumptions.

Supervisory Expectations and Good Practices

• Establish an ECL model inventory with ownership, version history, and validation status.
• Maintain comprehensive model documentation covering conceptual design, calibration data, validation outcomes, and governance sign-offs.
• Embed challenge culture – periodic independent reviews by Model Risk or Internal Audit teams to test assumptions.
• Disclose governance arrangements transparently in Pillar 3 and financial statements, including details of oversight bodies and frequency of validations.

In essence, strong governance transforms ECL from a compliance requirement into a credibility signal. Banks that operationalize clear accountability, independent challenge, and full audit traceability will not only meet regulatory expectations but also inspire investor and supervisory confidence.

Prudential floors & Stage‑3 ladder

To avoid under‑provisioning and improve consistency, the draft guidelines prescribe minimum floors by product and stage as follows:

To read more, click here

Income Recognition (EIR) Alignment

The Effective Interest Rate (EIR) method ensures that interest income reflects the true yield on a financial asset, consistent with its credit risk stage. Under the ECL framework, income recognition is dynamically aligned with the collectability of the underlying assets.

Operational note: core banking and financial reporting systems should automatically shift interest recognition from gross basis to amortized-cost basis upon Stage 3 classification and reverse any unrealized accruals. This ensures reported earnings remain aligned with economic substance and regulatory expectations.

Alignment with Model Risk Management (MRM) Standards

Confidence, not just models

The credibility of a bank’s Expected Credit Loss (ECL) outcomes depends less on statistical complexity and more on the strength of its Model Risk Management (MRM) framework.

The RBI’s draft Directions explicitly expect banks to demonstrate full life-cycle governance, from model development to independent validation, approval, implementation, and ultimately, performance monitoring, ensuring that model outputs are not only accurate but also trusted and repeatable.

Life-Cycle Governance and Accountability

1. Development and Design:
   • Models should be conceptually sound, using statistically defendable methodologies that are explainable to management and auditors.
   • PD, LGD, and EAD models must use representative datasets spanning at least one credit cycle, adjusted for current and forecasted macroeconomic conditions.
   • Model development documentation should clearly capture data sources, segmentation logic, variable selection, treatment of outliers, and justification for model form (logistic regression, survival, decision tree, etc.).
2. Validation and Approval:
   • All models must undergo independent validation before use – testing both conceptual soundness and empirical performance.
   • The Independent Validation Unit (IVU) should review development notes, assess statistical metrics, replicate results, and challenge business assumptions.
   • Validation findings and remediation actions must be approved by a Model Risk Committee (MRC) chaired by the CRO or CFO, ensuring separation between model owners and approvers.
3. Implementation and Monitoring:
   • Production models should be implemented in controlled environments with version management, parameter logs, and automated reconciliation to the general ledger.
   • Ongoing performance monitoring should evaluate model drift, override patterns, and back-test results at least quarterly, with defined thresholds for recalibration triggers.
   • Model governance dashboards should highlight: (a) predictive accuracy trends, (b) override rates, (c) floor bindings, and (d) parameter stability indices.

Core Validation Metrics

RBI expects banks to employ quantitative validation standards consistent with international MRM practices. Each ECL parameter is evaluated through both discriminatory power and calibration accuracy:

Validation tests above are quantifiable statistical metrics, not subjective assessments, and must be documented and reproducible under audit conditions.

Model Documentation and Change Control

• Comprehensive Documentation: Every model must maintain a Model Dossier containing model inventory, use of model, development methodology, data lineage, validation reports, management sign-offs, and performance logs.
• Version Control: Each update (parameter change, dataset refresh, or recalibration) should carry a unique version ID with date, reason, and approval record.
• Change Governance: Material changes trigger re-validation; minor adjustments (e.g., input refreshes) require validation sign-off but may follow fast-track review.
• Retention: Historical versions and validation evidence must be archived for at least seven years for supervisory inspection.

Oversight and Organizational Integration

• CRO / CFO Accountability: The CRO owns model risk governance; the CFO ensures accounting alignment and disclosure integrity. Joint stewardship reinforces independence and accountability.
• Model Risk Committee (MRC): Should meet quarterly to review performance dashboards, override logs, floor utilization, and validation status of all models.
• Integration with ICAAP: Model risk findings should feed into other risk management activities such as ICAAP and stress-testing processes, ensuring capital buffers account for potential model uncertainty.

Building Confidence Beyond Compliance

Strong Model Risk Management turns ECL from a regulatory requirement into a signal of institutional maturity. Banks that embed rigorous MRM practices – combining technical validation, transparent governance, and continuous challenge – build confidence among regulators, investors, and rating agencies.

In an environment where model outputs drive both provisioning and capital, confidence is the real currency.

Impact at a glance

Impact Bridge – From IRACP to ECL

Under the current IRACP framework, provisioning is largely rule-based and backward-looking, linked to overdue status or restructuring.  ECL introduces a forward-looking, probability-weighted approach, applying lifetime expected-loss recognition even to performing assets once a significant increase in credit risk (SICR) is detected. This transition materially alters both the timing and magnitude of provisioning.

To read more, click here

Pricing Under Floors

As more exposures migrate to Stage 2, RBI’s 5% provisioning floor effectively raises the cost of risk, requiring higher net interest margins (NIMs) to preserve returns.

A simple relationship illustrates the linkage:

Using realistic assumptions (16% target ROE, 12% CET1, 2% Opex, 70% Stage 1, and 30% Stage 2), the required NIM rises from ~5.0% to ~6.1%, implying the need for a 90–150 bps upward pricing adjustment or efficiency gains.

Illustrative Example: A bank with INR 100 crore in retail loans sees 30 % migrate to Stage 2. Provisioning increases from INR 0.4 crore (IRACP) to INR 2.2 crore (ECL). Unless pricing or collections improve, ROE declines by ~180 bps. Rebalancing the loan mix or enhancing recoveries can restore profitability.

Broader Impact Dimensions

ECL adoption affects not only capital and profit but also governance, systems, and risk culture.

Transition and Capital Relief

To mitigate the day-one impact, RBI proposes transitional CET1 add-backs over a five-year period.

To read more, click here

Global Experience and Transition Lessons

IFRS 9 and CECL Experience – The Global Transition Story

According to the Care Edge rating report on ECL Implementation: Limited Impact on Banks’ Capital, published on October 29, 2025, the ECL framework was first operationalized under IFRS 9 for European banks in 2018. To ensure a smooth transition of capital, regulators allowed a four-year phase-in, which ended in 2021.

Similarly, the U.S. introduced its Current Expected Credit Loss (CECL) framework in 2020 with a five-year glide path, offering temporary CET1 relief for the first three years, followed by a phased phase-out till 2024.

During this period:

• European banks experienced an average CET1 impact of 10–50 bps, manageable due to phased provisioning and strong capital buffers. 
• U.S. banks faced a higher 30–70 bps impact, largely due to lifetime-loss recognition across all asset classes.

Both jurisdictions deployed transitional add-backs and enhanced disclosure requirements to safeguard confidence.

GCC Case Study – The Early-Adopter Experience

• As per the Transition impact on banks in the Gulf Cooperation Council report published in 2018, Saudi Arabia and the UAE were at the forefront in implementing IFRS-9, driven by SAMA and CBUAE supervision.
• The impact of ECL on CET1 capital led to a reduction ranging from 0.5 per cent to 3.6 per cent in these regions.
• Early adoption fostered closer alignment between Finance, Risk, and Credit functions, with regulators emphasizing IFRS 9 governance policies, model validation, and data integration

According to the “Transition Impact on Banks in the Gulf Cooperation Council” report published in 2018, Regulators have issued prescriptive guidance for local consistency and alignment of risk definitions; however, this has also led to different applications of some aspects of ECL among the GCC countries. This experience demonstrated that regulatory engagement and data discipline are as critical as model sophistication

When implementation is backed by strong supervision, ECL acts as a stabilizing mechanism, not a disruptor.

Global Benchmark Comparison

A structured table summarizes how major jurisdictions managed their transition, illustrating both the capital impact and supervisory relief mechanisms as per the Care Edge rating report on ECL Implementation (October 2025) and the Transition Impact on Banks in the Gulf Cooperation Council report (published in 2018).

Global Implementation Lessons

Across these diverse experiences, three consistent success factors emerged:

These lessons collectively show that the credibility of ECL adoption lies not in the model, but in the governance that surrounds it.

Implications for India

India’s proposed transition aligns strongly with these global precedents.

The RBI’s transition period (2027–2031) is consistent with both IFRS 9 and CECL frameworks, providing banks with adequate time to build capacity and absorb the impact gradually.

Key implications include:

• Minimal Stage 1 impact: Floors for performing assets (0.25–1 %) are close to current IRACP norms.
• Contained Stage 3 effect: While Stage-3 uplift will indeed be modest due to already high PCR levels, the overall impact remains comparable to global experience because the bulk of the increase comes from Stage-2, driven by lifetime ECL recognition and the binding 5% prudential floor.

This mirrors the global narrative: ECL is evolutionary, not disruptive, provided governance and data readiness mature in parallel.

Global experience validates that while ECL adoption front-loads provisions, it ultimately stabilizes earnings and strengthens capital resilience. India’s framework, which combines forward-looking provisioning, a structured glide path, and regulatory transparency, embodies global best practices.

Key Implementation Challenges and Recommended Mitigations for ECL Adoption 

To read more, click here

ECL adoption is a multi-dimensional transformation. Governance and data form the foundation; modelling and technology provide the operational core; capital alignment, human capability, and supervisory transparency ensure long-term sustainability.

A bank that builds maturity across all the above dimensions will not only achieve compliance by the 2027 timeline but also strengthen its overall risk and capital management framework.

Closing thought

ECL replaces reaction with readiness. The question every bank must answer now is simple: “Can we evidence the risk before it materializes – consistently, transparently, and at scale?” If the answer is yes, ECL won’t just protect capital, but it will unlock it.

Banks have been preparing for the transition to ECL norms for several years, and the recent draft guidelines, which require banks to adopt a new framework for asset classification and income recognition, will be implemented from 01 April 2027, with a transition period of up to March 2031.

RBI’s ECL framework reimagines credit risk provisioning as foresight, not hindsight.

It strengthens India’s financial stability agenda and aligns with G20 and IMF commitments on transparency.

Early adopters will gain first-mover advantage through more stable earnings, predictable capital ratios, and enhanced investor trust. Those who delay may face steeper transition shocks and heightened supervisory scrutiny.

Ultimately, ECL is a discipline that connects capital, risk, and strategy, helping Indian banks’ balance growth ambition with prudential stability for the decade ahead.