IFRS 9 ECL
Calculator

De Nederlandsche Bank (DNB) has issued specific guidance on ECL implementation for banks and other financial institutions supervised under Dutch prudential regulation. DNB expects banks to use granular, portfolio-specific historical data rather than top-level industry benchmarks, and requires documented governance over the ECL model including model validation, back-testing, and board-level sign-off on key assumptions such as forward-looking macroeconomic scenarios.

The AFM (Autoriteit Financiële Markten) has made IFRS 9 impairment a recurring inspection focus for PIE audit firms. Common AFM findings include insufficient challenge of management’s SICR (significant increase in credit risk) criteria, over-reliance on days-past-due as the sole SICR indicator without considering qualitative factors, and inadequate documentation of forward-looking information used to adjust historical loss rates. The AFM expects auditors to document their independent assessment of whether management’s forward-looking overlay is reasonable and supportable.

For non-financial entities applying the simplified approach, the NBA (Koninklijke Nederlandse Beroepsorganisatie van Accountants) applies NV COS 540, requiring auditors to evaluate management’s provision matrix methodology including the appropriateness of aging buckets, the sufficiency of historical data periods, and the basis for forward-looking adjustments. Dutch statutory auditors should pay particular attention to the disclosure requirements under IFRS 7.35F–35N, which the AFM reviews as part of its financial reporting enforcement programme.

The FRC (Financial Reporting Council) has conducted multiple thematic reviews on IFRS 9 implementation since the standard became effective. Key FRC findings include: banks and insurers using insufficiently severe downside scenarios in their probability-weighted ECL calculations, entities failing to update forward-looking information between the interim and year-end reporting dates, and inadequate sensitivity disclosures that do not allow users to understand the range of reasonably possible ECL outcomes. The FRC expects preparers to clearly document the linkage between macroeconomic scenarios and the resulting ECL impact.

The PRA (Prudential Regulation Authority) sets expectations for banks and building societies on ECL modelling under its supervisory framework. PRA expectations include robust model governance with independent validation, use of at least three probability-weighted macroeconomic scenarios (base, upside, and downside), and explicit consideration of post-model adjustments (management overlays) with documented rationale. The PRA has highlighted that management overlays became a significant component of total ECL during and after the COVID-19 pandemic, and expects firms to demonstrate a clear path to unwinding overlays as conditions normalise.

For auditors, ISA (UK) 540 (Revised) applies to the audit of ECL estimates. The FRC’s Annual Quality Inspection reports have identified ECL as a recurring area of audit quality concern, with findings including: auditors not adequately challenging management’s SICR criteria and staging thresholds, insufficient testing of the completeness and accuracy of data inputs to ECL models, and over-reliance on management’s specialists without applying sufficient professional scepticism to model outputs.

APRA (Australian Prudential Regulation Authority) has issued APS 220 (Credit Risk Management) which sets expectations for authorised deposit-taking institutions (ADIs) on ECL modelling under AASB 9 (the Australian equivalent of IFRS 9). APRA expects ADIs to maintain ECL models that are independently validated, use a minimum of three probability-weighted economic scenarios, and incorporate forward-looking information that reflects current economic conditions rather than relying on through-the-cycle averages. APRA also requires ADIs to maintain adequate data infrastructure to support granular ECL calculations at the individual exposure level.

ASIC (Australian Securities and Investments Commission) has made IFRS 9 ECL a focus area in its financial reporting surveillance programme. ASIC has identified common issues including: entities not adequately considering forward-looking information when determining ECL under the simplified approach, use of aging-based provision matrices without adjustment for macroeconomic conditions, and insufficient disclosure of the methods, assumptions, and sensitivity of ECL estimates as required by AASB 7 (IFRS 7). ASIC expects entities to disclose how they have incorporated forward-looking adjustments, including the specific macroeconomic variables used and the sensitivity of ECL to changes in those variables.

For auditors, the AUASB applies ASA 540 (Revised) to the audit of ECL estimates. ASIC inspection findings have noted instances where auditors did not sufficiently challenge management’s forward-looking assumptions, particularly when the entity used a single-scenario approach rather than probability-weighted scenarios. Auditors should ensure they develop an independent expectation of ECL and evaluate whether management’s estimate falls within a reasonable range, considering the full spectrum of reasonably possible outcomes.

The Central Bank of the UAE (CBUAE) has issued specific IFRS 9 implementation guidance for banks and finance companies operating in the UAE. CBUAE circulars require banks to apply IFRS 9’s general model with full staging assessment, use a minimum of three probability-weighted macroeconomic scenarios, and maintain documented SICR criteria that go beyond simple days-past-due triggers. The CBUAE has set minimum provisioning requirements that may exceed IFRS 9 ECL in certain circumstances, requiring banks to hold the higher of the IFRS 9 ECL and the CBUAE minimum provision.

CBUAE circulars on provisioning require banks to report ECL by stage (Stage 1, 2, and 3) and by portfolio segment, with quarterly reporting to the regulator. The CBUAE also expects banks to disclose the macroeconomic variables used in their ECL models, the probability weights assigned to each scenario, and the sensitivity of ECL to changes in key assumptions. For forward-looking information, the CBUAE expects banks to consider UAE-specific indicators including oil price forecasts, real estate market indices, trade activity volumes, and government spending plans.

For auditors, the UAE adopts ISA as issued by the IAASB. Given the CBUAE’s additional provisioning requirements, auditors must assess compliance with both IFRS 9 and the CBUAE circular requirements. This dual compliance obligation means the auditor needs to evaluate whether management has correctly identified situations where the CBUAE minimum provision exceeds the IFRS 9 ECL and has appropriately accounted for the difference. Financial free zones (DIFC and ADGM) follow IFRS 9 as endorsed by their respective regulators (DFSA and FSRA).

The United States does not apply IFRS 9 for domestic reporting. Instead, US GAAP applies ASC 326, Financial Instruments — Credit Losses, which introduced the Current Expected Credit Loss (CECL) model. The fundamental difference between CECL and IFRS 9 ECL is that CECL requires recognition of lifetime expected credit losses from Day 1 for all financial assets measured at amortised cost, with no staging assessment. Under IFRS 9, Stage 1 assets only recognise 12-month ECL; lifetime ECL is triggered only when credit risk has increased significantly (Stage 2) or the asset is credit-impaired (Stage 3). This means CECL generally produces higher Day 1 provisions than IFRS 9 for performing loan portfolios.

The PCAOB has identified credit loss estimation as a critical audit matter (CAM) in a significant majority of bank audit reports since CECL implementation. Common PCAOB inspection findings include: insufficient testing of the qualitative adjustment factors applied to historical loss rates, failure to evaluate the reasonableness of management’s economic forecasts used in CECL models, and inadequate assessment of model risk including the sensitivity of CECL estimates to changes in key assumptions. The SEC has also issued comment letters challenging CECL disclosures, particularly the adequacy of sensitivity analysis and the transparency of qualitative adjustment methodologies.

For multinational groups that report under IFRS but have US subsidiaries (or vice versa), the CECL vs IFRS 9 difference creates consolidation complexity. IFRS reporters with US subsidiaries must convert CECL-based provisions to IFRS 9 staging for group reporting, which requires mapping the subsidiary’s loan portfolio into IFRS 9 stages and recalculating ECL under the general or simplified model. Auditors of such groups need to understand both frameworks and evaluate whether the conversion adjustments are complete and accurate.