ECL Governance – Board and Audit Committee agenda for Indian Banks

Regulatory Context: From IRACP to ECL

The ECL Directions – what changes

The Draft Directions subsume and modernize the existing Income Recognition, Asset Classification and Provisioning (IRACP) norms that have historically been the cornerstone of prudential regulation for banks in India. They seek to:

• Introduce staging criteria for asset classification under an ECL approach, while retaining the extant norms for Non-Performing Asset (NPA) classification.
• Replace the incurred-loss based provisioning framework with an ECL approach using PD, LGD, EAD building blocks, and a three-stage model (Stage 1, 2, 3) linked to SICR and credit impaired status.
• Update the principles of income recognition, including the use of Effective Interest Rate (EIR) and treatment of credit-impaired assets.

Key features relevant for Board-level attention include:

RBI’s Broader Governance Expectations

The ECL Directions must be read together with RBI’s governance expectations for banks, which emphasize:

• Board independence and composition: independent Chair, strengthened Audit, Risk Management and Nomination and Remuneration committees (NRC) and minimum independent director presence in Board and key committee meetings.
• Clear separation: the Board sets strategy, risk appetite and oversees management; three lines of defence (business, risk/compliance, internal audit) must work effectively and distinctly.
• Strengthened assurance functions: internal audit, risk and compliance to have direct access to Board committees, authority, resources, and independence to challenge management.
• Compensation and conduct: pay structures aligned to prudent risk taking and long term outcomes, and sharper focus on customer conduct, ethics, data privacy and cyber security.

In RBI’s own words, governance is the backbone of an institution, a well qualified, engaged, and vigilant Board is essential to prevent management failures and to sustain depositor and investor trust. The ECL regime raises the stakes further as provisioning becomes more judgement driven and model based, increasing both prudential risk and supervisory scrutiny.

What RBI Expects from Governance Under the ECL Directions

The ECL Directions explicitly state that:

• The credit policy of a bank shall cover all aspects of the ECL lifecycle, and
• The Board of Directors shall be responsible for the implementation and functioning of the ECL Framework on an ongoing basis.

Further, banks must establish a sub committee of the Board or Board approved committee, including the Chief Financial Officer (CFO) and the Chief Risk Officer (CRO), specifically for ECL purposes, with a mandate to:

• Review and challenge management’s ECL implementation strategy;
• Ensure ECL computation methodologies and assumptions are consistent with risk management practices;
• Oversee data integrity, governance, and internal controls across the ECL lifecycle;
• Safeguard independence of internal model validation;
• Define and monitor key performance indicators (KPIs) related to ECL processes and outcomes; and
• Ensure high quality disclosures and regulatory compliance.

This is a clear elevation of expectations: ECL is not simply an accounting change delegated to Finance; it is a risk and governance change directly anchored in Board accountability.

Key ECL Governance themes for Boards and Audit Committees

The following themes, aligned to the ECL lifecycle and RBI’s governance expectations and setting out practical questions for Board and Audit Committee members are as follows:

1. Governance and Oversight:

A robust ECL framework starts with clear accountability. The RBI’s Draft Directions emphasize Board approval for ECL policy, overseeing staging decisions, and reviewing management overlays and model assumptions. The Board should look beyond compliance checklists and test whether governance arrangements can withstand supervisory and investor scrutiny.

Key questions for Board / Audit Committees to evaluate:

• How is responsibility for ECL defined across the Board, Audit Committee, and Risk Committee charters, and how often will each committee receive focused ECL reporting (beyond standard financial reporting packs)?
• Has the Bank constituted a dedicated ECL Steering or Oversight Committee, and what decisions are reserved for that committee vs the Board?
• How are responsibilities for ECL mapped across the three lines of defence (Business/Risk owners, independent control functions, Internal Audit), and are there any overlaps or gaps?
• Does there exist a Board-approved ECL Governance Policy that covers methodology, model lifecycle, approach to overlays, and documentation standards?  How often will the policy be reviewed?
• How is ECL embedded in the Bank’s risk appetite statement (e.g., limits on Stage 2 exposures, tolerance for provisioning volatility, boundaries around model risk, and overlays)?
• What is the minimum content the Board or Audit Committee would expect in each ECL dashboard? E.g.
  • Stage distribution and migration – exposures by stage and how exposures move between Stage 1, Stage 2, and Stage 3;
  • SICR trends and credit quality signals – how many accounts are showing a significant increase in credit risk over time, breakdown of SICR triggers (rating downgrades, sector stress, EWS triggers, etc); 
  • ECL movement analysis – how and why ECL moved from the previous period to the current period, i.e., due to model parameter changes, macroeconomic scenarios, and weight changes, management overlays;
  • Model performance – how actual defaults and recoveries compare with model estimates;
  • Scenario analysis and sensitivities – how ECL would change if key assumptions or economic scenarios move;
  • Management overlays – list of all overlays with quantifications and how that sits within the Board-approved risk appetite and model risk limits;
  • Data quality, systems, and control indicators – any material data or system issues, workarounds, and remediation timelines, confirmation of period end reconciliations between ECL engine, sub ledgers, and GL;
  • Status of model validation and any open model risk or audit findings related to ECL (internal audit, external audit, regulator); and
  • Explicit statement of any limitations or significant judgments management wishes to draw to the Board / AC’s attention.

2. Model development:

Under ECL, models for PD, LGD, and EAD are the core parameters driving provisioning outcomes. Choices made today about how portfolios are segmented, what data is used, which variables are selected, and what assumptions are set will affect credit costs and capital for the current and future reporting periods.

Key questions for Board / Audit Committees to evaluate:

• How has the Bank segmented its portfolios for PD, LGD, and EAD modelling (e.g., by product, collateral, customer type, geography, industry), and what business logic supports these choices?
• For each segment, does the Bank have sufficient internal history of defaults, recoveries, and utilizations to justify the modelling approach, or are they relying on proxies and judgements?
• What PD methodologies are used (e.g., rating based, scorecard), and how are Point in Time (PIT) PDs derived and adjusted using macroeconomic variables?
• How do LGD models incorporate collateral type, coverage, recovery costs, and time to recovery, and how do LGD assumptions differ between secured and unsecured portfolios?
• How are EAD and Credit Conversion Factors (CCF) modelled for term loans, revolving credit, and off-balance sheet exposures such as guarantees and letters of credit, and what behavioural analysis underpins those models?
• In which portfolios does the Bank depend on regulatory PD/LGD floors, pooled data, or external benchmarks, and what is the roadmap to move towards internally calibrated models?

3. Model implementation:

Even well-designed models can lead to poor outcomes if not implemented properly. The rules for staging, SICR triggers, the rebuttal mechanism, and the Effective Interest Rate (EIR), therefore, require careful implementation and effective controls.

Key questions for Board / Audit Committees to evaluate:

• How has the Bank operationalized the three stage framework (Stage 1, Stage 2 – SICR, Stage 3 – credit impaired) for different portfolios?
• What quantitative and qualitative SICR triggers (e.g., PD multiples, rating downgrades, early warning signs (EWS) indicators, sectoral stress) are built into the framework in addition to the >30 days past due rebuttable presumption?
• What is the Bank’s policy for SICR rebuttals – who can propose and approve them, what objective evidence is required, and over what observation window are accounts monitored before confirming the rebuttal?
• How has the Bank ensured that the models and staging rules documented in methodology papers are faithfully implemented in production systems (including independent code review, parallel run testing, reconciliation to general ledger (GL))?
• How is EIR implemented, especially the shift to interest recognition on amortized cost for Stage 3 assets, and what controls ensure that reversals of uncollectable interest are complete and timely?

4. Model review and performance monitoring:

ECL is not a “one-time build and forget” regime. The framework requires regular management review, back testing, and recalibration. The Board should focus on whether management sees the right diagnostics and acts on them in a timely manner.

Key questions for Board / Audit Committees to evaluate:

• What is the management review process for ECL each reporting period – who participates, what analytics are reviewed, and how are challenges and decisions documented?
• How is the Bank monitoring stage migration trends (Stage 1 → Stage 2, cures,
  Stage 2 → Stage 3) by product and segment, and how quickly are unusual movements escalated to senior management and the Board?
• How are realized default rates compared against PD estimates by grade or score band, and what criteria trigger PD recalibration or segmentation changes?
• How are realized recoveries, time to recovery, and utilizations compared to LGD and EAD/CCF assumptions, and how are persistent biases (over or under estimation) identified and addressed?
• How often are management overlays reassessed, what evidence is required to retain or release them, and how is the impact of overlays communicated to the Audit Committee and investors?

5. Model validation and model risk management:

The RBI’s Draft Directions and global practice converge on one message: independent validation and robust model risk management (MRM) are central to ECL credibility. For a listed Bank, weaknesses here can quickly lead to regulatory findings and market concerns.

Key questions for Board / Audit Committees to evaluate:

• Does the Bank have a dedicated Independent Validation Unit (IVU) with sufficient skills and independence from model developers?
• What quantitative and qualitative criteria define a satisfactory model?
• How often will full validation be performed for each model (initial, annual, ad hoc) and what events trigger out of cycle validation (material portfolio shifts, methodology changes, performance deterioration)?
• How are validation findings classified (e.g., high/medium/low) and what governance exists to track remediation plans, deadlines, and overdue actions, including escalation to the Board where needed?
• Is there a central ECL model repository capturing ownership, purpose, validation status, and risk rating, and how is this repository used to prioritize model review and investment?
• How has the Bank articulated its model risk appetite for ECL (e.g., limits on use of unvalidated models, constraints on overlay size where models are weak, thresholds for provisioning volatility)?
• What is the model change policy for ECL, how are material vs non material changes defined, tested, approved, and disclosed and how is version history maintained for supervisory and audit review?

6. Data, Systems, and control environment:

Strong ECL results depend on clean data and robust systems. If information is spread across systems, relies on manual fixes, or cannot be traced back to its source, the outputs become hard to rely on. 

Key questions for Board / Audit Committees to evaluate:

• Does the Bank have a centralized ECL data mart with clearly documented data lineage from origination systems, collateral, and collections modules to the ECL engine and GL?
• What are the key data quality KPIs (completeness, accuracy, timeliness, reconciliations) tracked for ECL, who owns them, and how often are breaches reported to senior management and the Audit Committee?
• Where are the most material historical gaps (defaults, recoveries, collateral valuations, EWS indicators), what workarounds are in place, and what is the timeline to close those gaps?
• Which parts of the ECL process are fully automated (data extraction, staging, parameter application, calculation, posting), and where do manual spreadsheets or uploads still exist? What controls mitigate operational risk in those areas?
• What access controls, segregation of duties, logging, and audit trails apply around ECL models, code, parameters, and data, and what have Internal audit and external auditors observed in their latest Information Technology General Controls (ITGC) reviews?
• Has the ECL engine been stress tested for performance and resilience (multiple scenarios, full portfolio runs at quarter ends), and what are the recovery time objectives if the system fails close to reporting dates?

7. Capital, business strategy, and disclosures:

ECL outcomes will directly influence Common Equity Tier 1 (CET 1), pricing, portfolio choices, and market communication. The Board should ensure there is a coherent story linking ECL numbers to the Bank’s business strategy and disclosures.

Key questions for Board / Audit Committees to evaluate:

• What does the latest bridge from IRACP provisions to ECL look like by portfolio and stage, and what are the main drivers (staging, PD/LGD assumptions, prudential floors, overlays)?
• How will the Bank use RBI’s phased CET1 add back for the difference between ECL and current provisions, and how is the glide path reflected in capital plans, dividend policy, and growth aspirations?
• How are ECL parameters and stage migration expectations being embedded into pricing frameworks and RAROC, particularly for high-risk or unsecured segments (e.g., MSME, credit cards, personal loans )?
• Which portfolios are expected to see the largest uplift in Stage 2 and Stage 3 provisions, and how is this influencing origination standards, sectoral limits, collateral requirements, and collection strategies?
• What additional disclosures (in financial statements, Pillar 3, investor presentations) will the Bank provide on staging, ECL movements, overlays and sensitivities and how will we explain one-off transition effects versus ongoing volatility?
• How will the Bank ensure consistency and reconciliation of ECL-related figures across RBI filings, stock exchange disclosures, annual reports, earnings calls, and rating agency interactions?

Time Sequenced Agenda for Boards and Audit Committees

To make the ECL transition executable, Boards will benefit from a time sequenced agenda integrated with RBI’s effective date of 1 April 2027 and the transitional capital regime through FY 2030–31. 

Below, we outline Uniqus’ perspective on Immediate (0–12 months), Medium term (12–24 months), and Long term (24+ months) priorities.

Immediate priorities (now to 12 months from finalization of Directions)

Establish programme governance and Board engagement

• Constitute or refine the Board level ECL Committee mandated by RBI (including CFO and CRO), with a clear charter, reporting lines to the Board, and linkages to the Audit Committee and Risk Management Committee of the Board (RMCB).
• Clarify committee boundaries and handshakes:
  • RMCB to oversee ECL from a risk appetite, model, data, and macro scenario perspective.
  • Audit Committee to focus on financial reporting, internal controls, assurance, and external audit coordination.
  • NRC to review skills, experience, and time commitment required at the Board and senior management level for ECL governance.
• Schedule regular Board and Committee deep dives on ECL, including external education sessions where necessary to raise the collective understanding of modelling concepts, prudential floors, staging, and disclosure expectations.

Approve an institution wide ECL transition roadmap 

Define foundational policy choices

Boards should prioritize a first wave of policy decisions, including:

Strengthen data, systems, and control foundations

• Commission a data quality and availability review for ECL critical fields (e.g., days past due, behavioural data, collateral attributes, recovery cash flows, restructuring flags), including completeness, accuracy, and lineage from source systems.
• Approve investments to enhance data infrastructure, including data warehouses, data governance tooling, and interfaces between credit, risk, finance, and collections systems.
• Direct internal audit and risk functions to design an ECL specific internal control framework, spanning data capture, model implementation, manual adjustments, management overlays and disclosure processes.

Align assurance and external audit expectations

• Ask the Audit Committee to ensure early coordination with statutory auditors on ECL methodologies, staging, data, controls, and disclosures, recognizing that auditors will be required to opine on ECL driven financial statements.
• Mandate internal audit to develop a phased ECL audit plan, starting with design stage reviews of governance, policies, and data, and moving to operational testing during parallel runs.

Medium term priorities (12–24 months)

Oversee model development, validation, and post model adjustments 

During this period, Banks will typically design, develop, test, and refine their ECL models.

Boards and Audit Committees should:

• Review and approve the model risk management framework for ECL, including model inventory, tiering, validation standards, performance metrics, and escalation thresholds.
• Ensure that independent model validation is properly resourced, has the requisite skills, and reports functionally to a Board level committee rather than to model owners.
• Scrutinize proposals for Post Model Adjustments (PMAs) and management overlays, insisting on documented rationale, quantitative support and clear de scoping / roll off criteria.
• Challenge management on sensitivity analysis and back testing plans—how will the bank test model outputs against realized losses, and how will it interpret and act on deviations?

Run robust parallel runs and assess capital impact

RBI’s transitional arrangement allows banks to add back a fraction of the difference between ECL and existing IRACP provisions to CET1 capital from FY 2027–28 through FY 2030–31, with a reducing factor each year.

Boards should therefore:

• Oversee extended parallel runs, where ECL provisions are computed alongside IRACP provisions, using real portfolios and data.
• Require clear analysis of P&L and capital impacts under different scenarios, including:
  • Base case macro economic environment
  • Stress scenarios consistent with ICAAP and supervisory stress tests
  • Management overlays for emerging risks
• Use these insights to revisit the bank’s risk appetite, growth plans, product mix, and pricing, ensuring that business strategies are viable under likely ECL-driven provisioning patterns.
• Consider the optimal use (or non use) of transitional relief, balancing short term capital flexibility with market expectations of “fully loaded” capital strength.

Embed ECL into business, risk, and finance processes

To realize the full benefits of ECL and meet RBI’s expectation that ECL inputs are aligned with risk management practices, Boards should push management to go beyond a “compliance build” and integrate ECL into key processes:

• Credit underwriting and pricing: using ECL outputs (e.g., lifetime PD, LGD) in risk-adjusted return metrics, product pricing, limit setting, and credit approval authority structures.
• Early warning and collections: leveraging SICR indicators and staging movements to inform proactive remediation and restructuring strategies.
• Budgeting and planning: embedding ECL in financial planning, including forecasts of credit costs, capital consumption, and returns.
• Risk appetite and limits: refreshing risk appetite statements and limits to be consistent with ECL based loss and capital expectations.

Boards and Audit Committees should request concrete examples of how business units are using ECL insights, not just how Finance is producing ECL numbers.

Strengthen people, skills, and culture

The shift to ECL demands specialized skills across risk modelling, data science, credit analytics, accounting, and regulatory reporting. Boards should:

• Ask NRC to review whether the Board and senior management composition adequately cover these skill sets, especially in the Audit Committee, RMCB, and ECL Committee.
• Encourage targeted training and certification for directors and key executives on ECL concepts, data governance, model risk, and advanced credit risk management.
• Promote a culture of challenge and transparency around models and assumptions—where raising concerns about data or model robustness is encouraged, not penalized.

Long term priorities (24+ months)

Institutionalize ECL in governance and decision making

As ECL becomes business as usual, Boards should ensure that:

• ECL metrics and staging distributions, PD/LGD trends, and back testing results are embedded in regular Board and committee dashboards.
• The Board evaluation process explicitly assesses ECL related oversight effectiveness—e.g., sufficiency of challenge on models, scenarios, PMAs and disclosures.
• Compensation frameworks for senior management include metrics that reflect risk-adjusted performance, quality of underwriting, and effectiveness of credit risk management, not just short-term profitability.
• The bank’s governance policies (including Board charters, committee terms of reference, and internal control frameworks) are updated to permanently embed ECL responsibilities.

Evolve model and data capabilities

Over time, regulatory expectations and business needs will likely push banks toward more sophisticated modelling and data analytics, including:

• More granular segmentation and model variants by portfolio, geography, and customer profile.
• Enhanced use of behavioural and alternative data (within regulatory guardrails) to refine PD and SICR assessments.
• Advanced macro scenario generation and stress testing capabilities that align ECL, ICAAP, and business strategy.

Boards should treat ECL as a stepping stone toward a broader data and analytics driven risk management framework, rather than a static end state.

Enhance disclosures and stakeholder communication

The ECL Directions prescribe detailed disclosures on credit risk management practices, inputs and assumptions, reconciliations of loss allowances, and credit risk exposures by stage and product.

Boards and Audit Committees should oversee:

• The development of clear, investor friendly narratives explaining ECL methodology, governance, sensitivities, and trends, in addition to the required tabular disclosures.
• Consistency and connectivity between ECL disclosures and other public communications, including annual reports, ratings presentations, and investor calls.
• Engagement strategies with RBI, rating agencies, and investors on how the bank’s ECL framework reflects its risk profile, underwriting standards, and governance strength.

Well crafted disclosures will not only meet regulatory requirements but can also help banks differentiate themselves on risk transparency and governance quality.

Closing Thought

Based on the ECL Directions and RBI’s governance pronouncements, we see the following actions for Boards and Audit Committees of Banks:

• Treat ECL as a strategic transformation, not a compliance project: Anchor ECL at the Board level, with a dedicated committee, integrated with the RMCB and Audit Committee. Insist on a holistic roadmap that covers governance, models, data, systems, people, and culture—not just accounting outcomes.
• Invest early in data, controls, and independent assurance: RBI’s expectations on data quality, forward looking information, and model risk management leave little room for weak foundations. Early investment here will pay dividends in smoother supervisory reviews and lower remediation costs.
• Own the hard policy choices—SICR, segmentation, macro scenarios, and overlays: These are not purely technical decisions; they fundamentally shape the bank’s risk posture, earnings volatility, and capital trajectory. Boards should actively debate and approve these choices, informed by scenario analysis and external perspectives.
• Align ECL with business strategy, risk appetite, and incentives: ECL should influence which segments the bank grows in, how it prices risk, and how it measures performance. Over time, Boards should see ECL as a decision support tool for strategic choices, not just a provisioning mechanism.
• Leverage governance as a competitive differentiator: RBI has been explicit that banks with robust governance frameworks are better positioned to access financial resources and support India’s growth ambitions. Boards that demonstrate credible oversight of ECL implementation backed by strong assurance functions and transparent disclosures will likely enjoy greater regulatory confidence and market trust.

As the ECL Directions move from draft to final form, Boards and Audit Committees have a narrow but important window to shape their institution’s response. Those that act early, think strategically, and embed ECL into the fabric of governance and decision making will be better placed to navigate both supervisory expectations and market cycles in the years ahead.