Enhancing Coordination between Audit & Regulation Efforts

Why coordination between audit & regulation matters for ECL

Auditors and supervisors both seek confidence that ECL estimates under IFRS 9 are unbiased, supportable and consistently applied. Poor coordination creates repeated requests for documents, inconsistent findings, conflicting remediation priorities and higher operational cost. For firms that run complex PD, LGD and EAD models, this wastes scarce validation resources and delays provisioning decisions that affect capital planning and investor disclosures.

Regulatory and audit overlap

Regulators focus on prudential safety and the integrity of loss estimates; auditors focus on accounting recognition and disclosure. Aligning their activities around a shared evidence base — documented governance around Risk Model Governance, Historical Data and Calibration, and Model Validation — reduces friction. For example, creating a single validated dataset and a joint evidence pack addressing both prudential questions and IFRS 7 Disclosures avoids multiple ad-hoc data pulls.

For practical reading on the oversight context, see a focused review of IFRS 9 regulatory challenges faced by institutions and supervisors.

Core concepts: definitions, components and clear examples

What “coordination” means in practice

Coordination between auditors and regulators means: shared timelines where possible; joint or sequenced requests that reuse work; agreed definitions (e.g., default, forborne, cure period); and common evidence packages covering model assumptions, sensitivity testing and governance. It does not mean merging roles — independence remains crucial — but it does mean reducing duplication.

Key components to coordinate

• Risk Model Governance: documented policies, model inventory, approval and change controls.
• PD, LGD and EAD Models: design assumptions, segmentation, use of macroeconomic scenarios and forward-looking adjustments.
• Historical Data and Calibration: data lineage, vintage analyses, and back-testing results.
• Model Validation & Sensitivity Testing: independent validation reports, stress tests, and parameter sensitivity analysis.
• IFRS 7 Disclosures: methods, key inputs, significant judgments and reconciliations to regulatory measures.

Illustrative example

Consider a mid-sized bank with retail credit portfolios. Auditors ask for calibration evidence on PD curves; the supervisor requests stress transition matrices for capital planning. If the bank prepares a single dossier with: data lineage, vintage PD vs observed default rates, model code snippets, sensitivity analysis of PD under three macro scenarios, and a reconciliation between accounting ECL and regulatory expected loss, both parties can close their queries faster and with fewer follow-ups.

Practical use cases and common scenarios

Scenario 1 — Year-end provisioning

At year-end institutions face parallel requests: auditors require detailed documentation to support impairment amounts and disclosures, while supervisors review model performance and stress outcomes. A shared timeline and an agreed “evidence pack” containing model runs, reconciliations, sensitivity testing and an executive summary reduce last-minute rework.

Scenario 2 — Model change or recalibration

When recalibrating LGD models, notify both the external auditor and the regulator early. Provide the model change request, validation plan and an estimated timeline. This prevents conflicting remediation requests and ensures both parties evaluate the same change under their respective mandates.

Scenario 3 — Data quality event

If a data gap is discovered (e.g., missing historical recoveries for a segment), coordinate a remediation approach that addresses both accounting (impact on ECL) and prudential (impact on capital and risk) perspectives. Joint acceptance of interim adjustments avoids inconsistent reporting between financial statements and supervisory reports.

For more on how supervisors approach these challenges, review discussions about the regulatory challenges for ECL and common supervisory expectations.

Impact on decisions, performance and governance

Good coordination reduces time-to-close findings, lowers external fees, improves the quality of IFRS 7 Disclosures and strengthens confidence in capital planning. Measurable impacts include:

• Faster remediation cycles: target 30–50% reduction in time to close external findings.
• Lower audit/regulatory effort: fewer duplicate data requests and fewer onsite inspections.
• More stable provisioning: fewer restatements and clearer management commentary.

Role-level impacts

CROs and model owners gain better predictability in model approval timelines. Chief Accountants gain clearer evidence for judgmental disclosures in IFRS 7. Regulators and auditors gain more robust, reproducible evidence that supports their supervisory or attestation conclusions.

Where integration of accounting and prudential views matters, institutions often benefit from frameworks aimed at integrating ECL with regulation, which reduces gaps between internal risk and reporting teams.

Common mistakes and how to avoid them

1. Reactive engagement: Waiting until the audit fieldwork or a regulatory visit to compile deliverables. Avoid by maintaining an ongoing “audit/regulatory-ready” evidence pack.
2. Inconsistent definitions: Different definitions of default or cure across teams. Prevent by publishing a single authoritative glossary in Risk Model Governance and enforcing it.
3. Separate data pulls: Preparing slightly different datasets for accounting and supervision. Use a single reconciled dataset and document any jurisdictional adjustments.
4. Insufficient sensitivity testing: Providing point estimates without scenario analyses. Remedy by including sensitivity testing for key PD/LGD/EAD parameters and documenting assumptions.
5. Poor change control: Failing to log model changes or approvals. Maintain a model inventory with version history and validation sign-offs to satisfy both auditors and supervisors.

Auditors and regulators both have distinct duties; to understand auditor-specific expectations around ECL evidence, consider reading material on auditor roles and challenges.

Actionable tips and a coordination checklist

Use this checklist to operationalize coordination:

• Establish a cross-functional coordination forum (monthly) including model owners, validation, accounting, internal audit and the supervisory contact point.
• Create a reusable evidence pack template covering: model specification, validation reports, calibration outputs, sensitivity testing, IFRS 7 Disclosures and reconciliation to regulatory templates.
• Agree on timelines: publish an annual audit & supervisory calendar with expected deliverable dates.
• Standardize definitions and document them in policy; reference the glossary in reports and disclosure notes.
• Perform pre-submission dry-runs: share the evidence pack internally before external submission to identify gaps.
• Automate routine reconciliations between accounting ECL and regulatory ECL inputs where possible, to reduce manual error.
• Maintain an auditable change log for models and data sources linked to approvals and validation outcomes.

Practical example: sensitivity testing outputs

When presenting sensitivity testing include: base ECL (e.g., 1.2% of portfolio), ECL under mild stress (+25% PD) and severe stress (+50% PD), plus the resulting impact on profit before tax. Including simple numeric tables helps both auditors and regulators see materiality quickly.

Where regulatory alignment is required for regional compliance, be aware of differing expectations captured under supervisory requirements in Europe and coordinate local supervisory contacts early.

KPIs / Success metrics for coordination between audit & regulation

• Time to close external findings (days) — target < 60 days for most items.
• Number of duplicated document requests per engagement — target reduction by 50% year-over-year.
• Percentage of models with joint evidence packs — target 100% for material models.
• Frequency of reconciliations completed automatically vs manually — target 80% automated.
• Rate of disclosure restatements related to ECL — target 0% (or reduction to near zero).
• Number of model changes without prior validation sign-off — target 0.

Establish dashboards to track these KPIs and escalate issues in the coordination forum.

FAQ

Reference pillar article

This article is part of a content cluster exploring oversight and implementation of IFRS 9; for the supervisory perspective read the pillar piece: The Ultimate Guide: The supervisory role in applying IFRS 9 – why regulators must monitor ECL implementation and the link between accounting and banking supervision.

For background on the supervisory role itself, see a practical discussion of the regulatory disclosure requirements and coordination implications in that guide, and how supervisors and auditors can complement each other for robust ECL outcomes.