Unlocking Insights with Effective Quantitative Disclosures

1. Why this topic matters for IFRS 9 reporters

Quantitative disclosures are the numeric backbone that validates qualitative narratives about credit risk. Regulators, auditors, analysts and internal stakeholders (Risk Committee, Finance, Internal Audit) rely on detailed tables to reconcile model outputs with financial-statement figures and to test governance claims. The importance of ECL disclosure extends beyond compliance: accurate numbers reduce earnings volatility surprises, support capital planning, and build investor confidence.

Well-structured quantitative disclosures directly support IFRS 7 Disclosures and the IFRS 9 disclosure requirements, and they feed monthly Risk Committee Reports where board members expect concise tables showing stage migrations, movements by vintage, and sensitivity testing summaries.

2. Core concept: What to include in quantitative disclosures

Definition and required components

At minimum, quantitative disclosures should allow a third party to understand: portfolio exposures (EAD), probability of default (PD), loss given default (LGD), modelled ECL (12-month and lifetime), reconciliations of opening and closing provisions, and movements between Three‑Stage Classification buckets (Stage 1 / Stage 2 / Stage 3). Include segmentation, measurement bases and key assumptions.

Essential tables and example layout

Suggested core tables (with sample columns):

• Summary table by portfolio: Gross exposure (EAD), Average PD, Average LGD, ECL (12-month), ECL (lifetime), Coverage ratio.
• Stage migration reconciliation: Opening ECL (Stage 1/2/3), Transfers (S1→S2, S2→S3, etc.), New financial assets originated, Net derecognitions, Changes due to model updates, Closing ECL.
• Movement drivers table (numeric drivers): Impact of changes in PD, LGD, EAD, and forward-looking macro adjustments (e.g., -20bps PD = +X ECL).
• Sensitivity matrix: ECL under baseline and alternative macro scenarios (e.g., base, -50bps GDP, +100bps unemployment).
• Backtesting and calibration summary: Historical default rates vs model PDs, calibration factors applied, and a note on Historical Data and Calibration decisions.

Concrete example (simplified)

For a retail mortgage book with total EAD 1,000,000:

• Average PD (12m): 0.5% → expected defaults = 5,000
• Average LGD: 20% → expected loss given default = 1,000
• 12-month ECL provision ≈ 1,000 (or 0.1% of EAD)
• Lifetime ECL (Stage 2 + Stage 3) additional provisioning may add 2,500 → total provisions 3,500

Tables should show both the numeric ECLs and the assumptions used to compute them (PD curves, LGD models, forward macro scenarios).

3. Practical use cases and reporting scenarios

Monthly Risk Committee Reports and board pack inclusion

Risk directors need concise stage-level tables showing net movement drivers. A one-page summary with three visuals (stage split donut, migration waterfall, sensitivity table) plus an appendix of detailed tables works well for boards.

Regulatory submissions and supervisory reviews

Regulators expect reconciliation to general ledger balances, documentation of Historical Data and Calibration choices, and evidence of backtesting. Cross-reference the regulatory disclosure requirements with your quantitative tables to demonstrate comparative completeness.

External financial statements and investor presentations

When presenting ECL in financial statements, combine a headline table in the notes with detailed movement tables in a supplementary disclosure pack. Investors appreciate a short reconciliation: opening provisions → closing provisions with drivers quantified (new originations, credit deterioration, model changes).

For guidance on how to handle the narrative side, ensure your numeric disclosure is complemented by qualitative ECL disclosures that explain significant judgments and sensitivity assumptions.

4. Impact on decisions, performance and governance

Accounting Impact on Profitability

Quantitative disclosures illuminate the link between provisioning and reported profit. Detailed tables showing provisioning by stage and movement drivers allow CFOs to explain quarter-on-quarter swings and to forecast future profitability under scenario analyses — a direct intersection of Accounting Impact on Profitability and capital planning.

Risk-based decision making

Credit origination policies, pricing and portfolio remediation strategies are influenced by ECL outputs. A bank that reports a concentration of Stage 2 exposures in a particular region should adjust underwriting or tighten covenants. Presenting drillable tables enables business units to identify hotspots quickly.

Auditability and model governance

Well-constructed quantitative disclosures support model validation and audit processes by providing clear inputs and outputs for testing. Include the assumptions and calibration approaches used so validators can reproduce results or run alternative scenarios. For material model changes, quantify the ECL impact in a dedicated table; this aligns with expectations in the ECL impact on disclosures guidance.

5. Common mistakes and how to avoid them

• Unclear segmentation: publishing aggregate numbers without portfolio breakdowns. Remedy: adopt a consistent segmentation that mirrors risk models and internal management reporting.
• Non-reconciled totals: closing provisions in disclosures don’t match the GL. Remedy: include a reconciliation table and the methodology for cut-offs and timing.
• Insufficient historical data disclosure: not documenting sample periods or adjustments. Remedy: add a Historical Data and Calibration appendix showing sample sizes, vintage analyses and smoothing techniques.
• Poor sensitivity presentation: only qualitative statements about sensitivity without numeric matrices. Remedy: include a sensitivity table (e.g., ±50/100/200 bps PD) and show absolute ECL changes.
• Ignoring IFRS 7 Disclosures linkages: failing to connect ECL tables to broader risk disclosure narratives. Remedy: present cross-references to IFRS 7 tables and ensure consistency across notes (IFRS 9 disclosure requirements and IFRS 7).

6. Practical, actionable tips and checklists

Step-by-step for preparing a quantitative disclosure pack

1. Define the perimeter and segmentation used by models (retail, corporate, C&I, sovereign).
2. Extract GL balances and model outputs for the reporting date and ensure timing alignment.
3. Populate the core tables (summary, stage reconciliation, sensitivity, calibration notes).
4. Reconcile closing provisions to the GL and provide a variance analysis for material differences.
5. Review with Model Risk, Finance and Compliance; obtain risk committee sign-off on key assumptions.
6. File the external disclosure and retain a reproducible workbook for auditors and supervisors.

Short checklist to include in your working papers

• Segmentation map and mapping table from ledger to models.
• Vintage and historical sample size table.
• Model change impact table (quantified).
• Macro scenarios and probability weights used in Sensitivity Testing.
• Board-approved minutes or Risk Committee Reports that reference the main assumptions.
• Cross-reference table that links quantitative tables to qualitative narratives and to external regulatory requirements such as regulatory disclosure requirements.

For implementation, many teams follow a documented checklist that includes production steps, validation sign-offs and archive procedures — see our practical ECL implementation checklists.

Best practices for formatting and readability

• Start each note with a short summary of the key message (one or two lines).
• Use consistent units (thousands, millions) and disclose them prominently.
• Provide a machine-readable annex (CSV/XLSX) for analysts and supervisors to re-run simple reconciliations.
• Keep a version history of disclosure tables and annotate retrospective changes.
• Follow best practices for ECL disclosures such as table numbering, consistent labelling, and a single source of truth for input data.

KPIs / Success metrics

• Provision coverage ratio: Total ECL / Gross carrying amount, by portfolio and stage.
• Stage distribution: Percentage of exposures in Stage 1 / Stage 2 / Stage 3.
• PD calibration error: Average difference between realized default rate and model PD over 12–36 months.
• Sensitivity delta: Absolute ECL change for a +100 bps PD shock and for the base macro downside.
• Reconciliation variance: Difference between disclosed closing ECL and GL provisions (target: 0).
• Time-to-produce: Hours required to prepare the full quantitative pack (target: reduced over time with automation).
• Audit findings: Number of audit or supervisory findings related to disclosure accuracy (target: zero).

FAQ

Reference pillar article

This article is part of a content cluster that expands on the transparency principles covered in our pillar piece: The Ultimate Guide: The importance of disclosure about expected credit losses – why IFRS 9 places great emphasis on transparency and how disclosure enhances investor confidence.