IFRS 9

IFRS 9 Staging: Significant Increase in Credit Risk

Practical approaches to identifying SICR events, including quantitative thresholds and qualitative backstops

Paula Kruly Feb 03, 2026 12 min read

1. The Core SICR Principle

IFRS 9 requires migration from 12-month ECL (Stage 1) to lifetime ECL (Stage 2) when credit risk has increased significantly since initial recognition. Unlike binary default classification, SICR captures deterioration before credit-impairment, enabling earlier loss recognition and reducing cliff effects in provisioning.

The standard intentionally avoids prescriptive thresholds, empowering institutions to design frameworks reflecting their risk appetite, portfolio characteristics, and data availability—but this flexibility creates implementation challenges and audit scrutiny.

2. Quantitative Indicators: PD-Based Approaches

Most institutions anchor SICR assessment on changes in Probability of Default (PD) between origination and reporting date:

  • Relative PD change: Stage 2 trigger when current lifetime PD exceeds origination PD by threshold (common: 100-200% increase, adjustable by segment).
  • Absolute PD threshold: Backstop rule assigning Stage 2 if lifetime PD crosses absolute level (e.g., 5% for retail, 3% for corporate).
  • Credit rating migration: Downgrade of X notches (typically 2-3) on internal rating scale.
  • Risk band movement: Transfer from "low risk" to "moderate risk" bucket per institution risk appetite statement.

Key consideration: Use point-in-time PD incorporating current macro conditions, not through-the-cycle estimates that dampen sensitivity.

3. Qualitative Backstops and Overlays

Quantitative metrics alone miss behavioral signals and portfolio-specific risks. Essential qualitative triggers include:

  • Forbearance or restructuring granted due to financial difficulty (automatic Stage 2 under most policies).
  • Watchlist or special mention classification by credit officers.
  • Breach of financial covenants or failure to provide required documentation.
  • Adverse material events: litigation, regulatory sanctions, key management departure.
  • Significant decline in collateral value exceeding haircut buffers.
  • Industry or geographic concentration affected by systemic stress (sectoral overlays).

4. The 30-Day Past Due Presumption

IFRS 9 includes a rebuttable presumption that >30 days past due (DPD) indicates SICR. Institutions may overcome this with evidence that delinquency does not correlate with credit deterioration for specific portfolios (e.g., technical delays in payment processing).

In practice, most banks adopt the 30 DPD rule as a minimum safeguard while layering additional PD-based or qualitative criteria. Rebuttal arguments require rigorous statistical support and documented governance approval.

5. Segmentation and Threshold Calibration

One-size-fits-all SICR thresholds fail to reflect heterogeneous portfolio risk profiles. Best practice segmentation considers:

  • Product type: Mortgages may tolerate higher PD variance before SICR versus unsecured personal loans.
  • Origination quality: Near-prime borrowers show different deterioration patterns than prime or subprime.
  • Collateralization: Secured exposures with strong LTV ratios justify higher PD tolerance.
  • Maturity profile: Long-tenor loans accumulate more lifetime risk, requiring adjusted thresholds.
  • Economic cycle: Stress periods may trigger temporary threshold tightening via management overlay.

6. Low Credit Risk Exemption (Practical Expedient)

IFRS 9 permits entities to assume no SICR for instruments with "low credit risk" at reporting date, typically interpreted as investment-grade rating (BBB- or higher). This expedient applies primarily to:

  • Interbank exposures and reverse repos with highly rated counterparties.
  • Government bonds and central bank placements.
  • Trade finance backed by strong corporates with external ratings.

Retail and SME portfolios rarely qualify; institutions must demonstrate ongoing monitoring and rating refreshment even when using this exemption.

7. Cure and Return to Stage 1

Assets migrate back to Stage 1 when credit risk is no longer significantly higher than at origination. Common policies require:

  • Minimum probation period (3-6 months) of sustained performance without delinquency.
  • PD improvement returning below origination baseline or crossing de-escalation threshold.
  • Removal from watchlists and completion of forbearance observation period.
  • Resolution of covenant breaches and normalization of financial metrics.

Cure provisions prevent artificial provisioning volatility but require discipline to avoid premature optimism—auditors scrutinize cure logic intensely.

8. Practical Implementation Checklist

  1. Define SICR framework in formal credit risk policy approved by board or credit committee.
  2. Calibrate PD thresholds using historical migration analysis and portfolio performance data.
  3. Establish governance for qualitative overrides with documented rationale and approval authority.
  4. Build staging engine calculating all indicators daily with automated flags and exception reports.
  5. Integrate staging outputs with ECL calculation platform and Cosif/IFRS reporting modules.
  6. Perform sensitivity analysis showing provision impact under alternative threshold scenarios.
  7. Document model validation covering data quality, threshold appropriateness, and system controls.
  8. Monitor staging migration rates, compare against peers, and recalibrate when deviations emerge.

References and Further Reading

  • IFRS 9 Financial Instruments paragraphs 5.5.3-5.5.11 (SICR assessment)
  • IASB Educational Material on Expected Credit Losses
  • EBA Report on IFRS 9 implementation across EU banks
  • Big 4 technical guides on SICR threshold calibration