Exploring the Future of Auditing & ECL in Modern Practices

Why auditing & ECL matters for financial institutions

IFRS 9 moved the industry from incurred loss to forward‑looking expected credit loss measurement. Auditors now need to validate models and challenge management judgments across PD, LGD and EAD models, macroeconomic scenarios and staging rules such as the Three‑Stage Classification. For institutions, robust auditing improves reliability of provisions, reduces restatement risk, strengthens governance, and supports investor confidence.

Example: a mid‑sized bank with €10bn gross loans that underestimates lifetime PDs by 25% can under‑provision by tens of millions €, materially affecting CET1 ratios and dividend policies. Timely, high‑quality audits reduce those risks and ensure compliance with IFRS 7 Disclosures and related governance expectations.

Core concepts: auditing expectations and ECL building blocks

ECL methodology and model components

ECL calculation is typically composed of probability of default (PD), loss given default (LGD), and exposure at default (EAD), combined with forward‑looking macroeconomic scenarios and weights. The audit must evaluate methodology, data, assumptions and controls across these components. When auditing & ECL, auditors will verify that the ECL Methodology aligns with accounting policy and is consistently applied.

Three‑Stage Classification and staging judgments

Assess whether transfers between Stage 1, 2 and 3 are supported by objective evidence of significant increase in credit risk (SICR). Auditors should test sample cases, review trigger models and back‑test staging decisions against actual defaults to evaluate classification accuracy.

Historical data and calibration

Historical Data and Calibration underpin PD, LGD and EAD models. Auditors must confirm data lineage, completeness, treatment of outliers and the calibration process (e.g., scaling to current portfolio characteristics). Typical checks include comparing estimated vs observed default rates over 1–5 year horizons, and assessing overlays applied for non‑stationary data.

Model validation vs audit scope

Model Validation is an ongoing technical activity often performed by an independent validation team; audit relies on validation evidence and performs substantive testing, review of model governance, and confirmation that outputs feed into the financial statements correctly. For complex models consider commissioning external experts for technical model validation while internal audit focuses on control design and operating effectiveness.

Auditors should also liaise with the finance team to confirm that IFRS 7 Disclosures are complete and consistent with model outputs.

Practical use cases and audit scenarios

Recurring audit: year‑end ECL provision review

Typical steps for year‑end audit over ECL:

1. Obtain ECL methodology documentation and model validation reports.
2. Test key controls: model change approval, data extraction, and parameter sign‑off.
3. Recalculate ECL for selected segments (e.g., mortgages, corporate) using independent tooling or spreadsheets.
4. Reconcile ECL to financial statements and IFRS 7 Disclosures.
5. Report findings and required adjustments.

When you need a deep dive into methodology and model audit steps, consider reading our detailed guide on auditing ECL models for sample test scripts and audit evidence templates.

Ad hoc scenario: major macroeconomic shock

In an economic downturn auditors must assess whether macro scenarios, scenario weights and overlays reflect management’s best estimate. Evaluate how scenario forecasts change PD curves and whether stress scenarios produce plausible LGD/EAD changes. Ensure IFRS 7 Disclosures explain sensitivities and the impact of macro assumptions.

Regulatory and internal audit interactions

Internal and external audit activities should be coordinated to avoid duplication. Internal audit teams will often perform process and control testing; external auditors focus on financial statement assertions. For guidance on structuring that cooperation see our article on internal audit of ECL and how it complements the external audit under IFRS 9.

Model updates and validation lifecycle

When a model is updated (new segmentation, new macro drivers), auditors must assess change management: change request logs, back‑testing of new vs old models, and approval by model risk committees. Also verify that historical re‑calibration was performed correctly (see Historical Data and Calibration checks above).

Impact on decisions, performance and reporting

High‑quality auditing of ECL affects several organizational outcomes:

• Profitability and capital planning — accurate provisions reduce unexpected hits to profitability and capital buffers.
• Regulatory compliance and reputation — robust audits lower the risk of regulatory censure and investor concern.
• Management information and decision making — audited ECL outputs improve confidence in pricing, portfolio management and credit policy.
• Operational efficiency — standardized audit procedures and tooling reduce audit cycle time (example: reduce model sign‑off time from 6 weeks to 3 weeks with automated test suites).

Beyond reporting, auditors play a strategic role by challenging assumptions and recommending improvements in model governance and ECL Methodology—read more on evolving auditor roles in ECL as responsibilities shift from purely historical validation to governance of forward‑looking inputs.

Common mistakes in auditing & ECL and how to avoid them

Common issues we see across institutions, with practical mitigations:

• Weak data lineage: Missing documentation of data sources and transformations. Mitigation: require data dictionaries, sampling of ETL processes, and automated reconciliation scripts.
• Insufficient scenario governance: Scenario selection and weighting lack board oversight. Mitigation: formalize scenario governance with board/minimum committee approvals and store scenario versions.
• Overreliance on a single model: No diversity of evidence (e.g., no back‑testing or benchmarking). Mitigation: mandate independent model validation and perform sensitivity analyses.
• Poor treatment of overlays: Judgemental overlays are undocumented. Mitigation: require supporting analysis, trigger criteria and retrospective reviews.
• Disclosure gaps: IFRS 7 Disclosures miss material sensitivities or qualitative explanations. Mitigation: cross‑check disclosures against model results and board minutes.

Practical, actionable tips and checklists for auditors and ECL teams

Use this step‑by‑step checklist when planning your next audit of ECL:

1. Scoping: identify high‑risk portfolios (top 80% of ECL). Sample by size and risk type.
2. Obtain model documentation: purpose, methodology, limitations, validation reports.
3. Test data: run completeness and accuracy checks on inputs used for PD, LGD and EAD Models.
4. Recalculate outputs: independently compute ECL for a representative sample and reconcile to reported numbers.
5. Review staging decisions: validate SICR triggers and test for bias across segments.
6. Evaluate macro scenarios and weights: compare to external forecasts and check governance approvals.
7. Verify disclosures: ensure IFRS 7 Disclosures are consistent and complete.
8. Document findings and remediation timelines: assign owners and follow up during the next cycle.

Audit teams should leverage modern tooling to automate reconciliation and sampling. For recommendations on technologies and tooling that speed up validation, see our analysis on technology’s impact on ECL and practical guidance on internal audit tools for ECL.

KPIs / Success metrics for auditing & ECL

• Time to close ECL audit cycle (target: reduce by 30% year‑on‑year)
• Number of material adjustments to ECL at year‑end (target: zero or decreasing)
• Coverage of independent model validation reports across significant models (target: 100% of high‑risk models)
• Percentage of controls tested and effective (target: ≥95% for key control set)
• Accuracy of PD/LGD/EAD estimates vs realized outcomes over 1–3 years (back‑testing tolerance thresholds)
• Timeliness and completeness of IFRS 7 Disclosures (no disclosure restatements)
• Audit findings closed on time (target: 90% within agreed remediation periods)

FAQ

Reference pillar article

This article is part of a content cluster expanding on broader trends — see the pillar piece The Ultimate Guide: How IFRS 9 has changed the accounting and finance profession for historical context and the overall evolution of ECL and model governance. For perspectives on auditor strategy looking forward, read about where ECL is headed.