The introduction of Expected Credit Loss (ECL) under IFRS 9 marked a fundamental shift in how credit risk is assessed. What was once a backward-looking, event-driven process has become a forward-looking exercise—one that requires institutions to anticipate risk rather than react to it.

In an environment of economic uncertainty and rapid portfolio change, ECL is no longer just an accounting requirement. It has become a key metric into how credit portfolios are priced, monitored, and managed.

Understanding the Shift to Expected Credit Loss

Under IFRS 9, credit losses must be recognized based on expected outcomes, even when no default has occurred. This approach relies on three core components:

• Probability of Default (PD)

• Loss Given Default (LGD)

• Exposure at Default (EAD)

These estimates must also reflect forward-looking macroeconomic information, making ECL sensitive to changes in the broader economic environment.

Rather than relying solely on historical performance, institutions are now required to continuously reassess risk as conditions evolve.

How the Three-Stage Model Works in Practice

IFRS 9 classifies financial assets into three stages based on changes in credit risk:

• Stage 1: Performing assets with no significant increase in credit risk, requiring a 12-month ECL
• Stage 2: Assets that have experienced a significant increase in credit risk, requiring lifetime ECL
• Stage 3: Credit-impaired assets, requiring lifetime ECL with interest recognized on a net basis

While defaults often draw the most attention, movement from Stage 1 to Stage 2 frequently has a larger impact on provisions, well before actual credit deterioration becomes visible.

Why ECL Feels More Complex Today

Several factors have increased the complexity of ECL implementation:

• Faster-changing portfolios and product innovation
• Limited historical data for newer asset classes
• Greater reliance on assumptions and scenario design
• Increased scrutiny from auditors and regulators

As a result, ECL outcomes can become volatile and difficult to interpret without a well-structured framework.

Modelling and Validation

• PD and LGD drive provision sensitivity

Small shifts in PD term structures or LGD recovery assumptions can materially impact Stage 1 and Stage 2 provisions. Robust modelling ensures that changes in ECL reflect genuine movements in underlying credit risk rather than artificial volatility arising from weak assumptions or model instability.

• Forward-looking calibration is essential

Both PD and LGD must incorporate macroeconomic expectations, portfolio mix changes, and evolving borrower behaviour. Models that fail to adapt to current conditions risk understating or overstating emerging credit risk.

• Validation strengthens credibility

Independent validation, back-testing against realized defaults and recoveries, sensitivity analysis, and regular recalibration are critical. Strong governance over overlays and expert judgment enhances transparency with auditors, regulators, and senior management.

Common Implementation Challenges

Despite widespread adoption of IFRS 9, many institutions continue to face challenges such as:

• Treating ECL as a compliance-only exercise
• Heavy dependence on manual processes
• Inconsistent criteria for identifying significant increases in credit risk
• Limited alignment between risk, finance, and business teams

These issues reduce transparency and limit the usefulness of ECL for decision-making.

From Regulatory Requirement to Management Insight

More mature ECL frameworks go beyond minimum compliance. They help institutions:

• Detect emerging risk segments earlier
• Support pricing and portfolio strategy decisions
• Improve capital planning and stress testing
• Provide clearer explanations of provision movements

When embedded into regular risk discussions, ECL becomes a forward-looking management tool rather than a reporting obligation.

Conclusion

IFRS 9 has fundamentally changed expectations around credit risk management. Expected Credit Loss is no longer just about recognizing losses earlier—it is about understanding risk better.

Institutions that invest in strong data, transparent assumptions, and cross-functional alignment will be better positioned to navigate uncertainty and make informed credit decisions.