Discover IFRS 9 impact on the profession of accounting today

Why this matters for financial institutions and preparers

IFRS 9 altered how credit risk is measured and reported — shifting from incurred-loss to expected-credit-loss accounting, which has far-reaching consequences in provisioning, capital planning, product pricing and model governance. Understanding the IFRS 9 impact lets CFOs, CROs, model risk managers and finance teams align reserve policies, satisfy auditors and supervisors, and avoid earnings volatility surprises.

Practical pain points

• Estimating lifetime losses for stage 2 / 3 exposures and documenting forward‑looking assumptions.
• Integrating credit models (PD, LGD, EAD) across risk and finance systems to produce consistent ECL outputs.
• Meeting IFRS 7 Disclosures expectations while maintaining commercial confidentiality and operational efficiency.

The shift also created demand for specialized skills. For a deeper historical perspective see the Evolution of IFRS 9, which explains how the standard emerged from lessons learned after the global financial crisis.

Core concepts: PD, LGD, EAD and the ECL methodology

Definition and components

Start with a precise Definition of IFRS 9 for your documentation, then build your ECL model around three driver models and forward-looking adjustments:

• PD (Probability of Default) — the likelihood a borrower defaults within a defined horizon (12-month or lifetime). Example: A corporate borrower with a 2% one-year PD translates to a 0.02 probability of default in the next 12 months.
• LGD (Loss Given Default) — expected percentage loss if default occurs, after recoveries and collateral. Example: An LGD of 40% on a USD 1,000,000 loan implies an expected loss of USD 400,000 conditional on default.
• EAD (Exposure at Default) — expected outstanding balance at the time of default. For a revolver this requires drawdown modeling (e.g., 60% drawdown factor).

ECL calculation formula and examples

ECL is the probability-weighted present value of expected credit losses over the relevant time horizon. A simplified one-period formula:

Expected Loss = PD × LGD × EAD

Illustrative example: A retail mortgage portfolio with average PD = 1.5% (0.015), LGD = 20% (0.20), EAD per account = 100,000 → ECL per account = 0.015 × 0.20 × 100,000 = 300. For 10,000 similar accounts ECL = 3,000,000.

Forward‑looking adjustments and macro scenarios

IFRS 9 requires scenario-weighted PD, LGD and EAD inputs when forward-looking information materially affects credit risk. That means constructing at least three scenarios (base, upside, downside), assigning probabilities (e.g., 50/30/20) and applying macro-sensitivity factors to driver models.

Principles and objectives

Design models and governance around the IFRS 9 principles and the IFRS 9 objectives: unbiased, probability-weighted, and supportable forward‑looking estimates that align with business practices and risk management information.

Practical use cases and scenarios

Monthly provisioning cycle for a mid-size bank

Scenario: A regional bank calculates ECL monthly for financial reporting. Workflow example:

1. Extract aging, collateral and exposure data from the loan servicing system.
2. Run PD model (12-month PD for performing exposures; lifetime PD if significant increase in credit risk).
3. Estimate EAD for drawn and undrawn facilities using contractual terms and behavioural curves.
4. Apply LGD based on collateral type and cure assumptions; run scenario adjustments using macro variables (GDP, unemployment).
5. Aggregate by portfolio and produce IFRS 7 Disclosures for management and external reporting.

Stress testing and capital planning

Use ECL outputs directly in stress tests. A representative exercise: impose a 3% GDP contraction and reweight PDs and LGDs — observe credit reserve increases (e.g., +30%) and quantify capital impacts for strategic decisions.

Loan pricing and product design

Link PD, LGD and EAD models to pricing engines so new products incorporate expected credit cost. Example: If lifetime ECL for a credit card product rises from 0.8% to 1.2% of exposure under revised PD calibrations, pricing adjustments or underwriting tightening may be required.

Impact on decisions, performance and outcomes

Understanding the Impact of IFRS 9 on financial statements changes incentives across the firm:

• Profitability: Early recognition of expected losses increases provisioning volatility but reduces surprise credit hits later; impacts reported ROE.
• Capital planning: Higher forward-looking reserves can affect regulatory capital ratios and dividend policies.
• Risk appetite: Transparent, scenario-based ECL encourages tighter underwriting or portfolio rebalancing where expected credit cost is material.
• Investor communication: Consistent IFRS 7 Disclosures and scenario narratives improve market confidence and reduce information asymmetry.

Strategically, robust ECL models create better alignment between risk and finance and enable data-driven decisions about new products, customer segments, and geographic exposure.

Common mistakes and how to avoid them

1. Weak governance and unclear ownership

Problem: Model changes without finance sign-off lead to inconsistent ECL. Fix: Establish a RACI for model development, calibration, validation, and reporting; require senior finance and risk approvals.

2. Poor historical data and improper calibration

Problem: Using short or unrepresentative history biases PD and LGD. Fix: Use at least one full credit cycle where possible, supplement with proxy data, and document adjustments clearly in model change reports.

3. Ignoring forward‑looking components

Problem: Treating ECL as backward-looking leads to under-provisioning. Fix: Build macro linkages and scenario frameworks; perform sensitivity testing to quantify key drivers.

4. Over-reliance on single-model outputs

Problem: Blind reliance on a single PD model obscures model risk. Fix: Enforce independent IFRS 9 tools for benchmarking, back-testing and validator reviews; maintain alternative models or overlays for material portfolios.

5. Failing to reconcile risk and finance systems

Problem: Differences between risk and finance exposures create reporting gaps. Fix: Standardize definitions (e.g., exposure definitions, default definitions), reconcile monthly, and log reconciliation items.

Also consider trade-offs described in IFRS 9 balancing when making pragmatic modeling decisions (e.g., simplicity vs. accuracy, transparency vs. complexity).

Actionable tips and an implementation checklist

Follow this step-by-step plan to operationalize IFRS 9 ECL models effectively.

Quick implementation checklist (first 90 days)

1. Inventory all credit portfolios and map current PD, LGD, EAD models.
2. Identify data gaps and agree remedial data collection (e.g., collateral valuations, behavioural drawdowns).
3. Define scenario set (at least base, adverse, optimistic) and assign probabilities.
4. Set governance: model owners, validators, and reporting cadence.
5. Run parallel ECL calculations and reconcile with current accounting reserves.
6. Prepare documentation for auditors and regulators, including sensitivity tests and rationale for overlays.

Model validation and sensitivity testing

Perform independent model validation covering conceptual soundness, data quality, calibration, back-testing and stability. For sensitivity testing, quantify the effect of ±1% GDP, ±50bps unemployment, or a 10% collateral valuation shock on portfolio ECL. Document all outcomes and incorporate them into capital planning.

Disclosure and stakeholder communication

Draft your IFRS 7 Disclosures early: include methodology, key inputs, scenario probabilities and reconciliations between opening and closing ECL balances. This reduces audit queries and supports investor transparency.

KPIs / Success metrics

• Model accuracy: PD back‑testing coverage ratio > 90% over a rolling 24‑36 month window.
• Calibration drift: annual change in average PD within expected tolerance (e.g., ±20% of benchmark).
• Data completeness: percentage of exposures with complete collateral and behavioural fields ≥ 98%.
• Validation backlog: < 2 outstanding high-severity validation items at any time.
• Reporting timeliness: ECL reports produced within X+5 business days of period close (define X as your close cycle).
• Disclosure quality: No material audit adjustments to ECL or disclosures for two consecutive years.

Frequently asked questions

Reference pillar article

This article is part of a content cluster that expands on the broader effects of IFRS 9. Read the pillar piece for a comprehensive view: The Ultimate Guide: How IFRS 9 has changed the accounting and finance profession – from historical models to forward‑looking models and higher specialization in financial accounting.