Contact@esgweise.com | +971585075373
IFRS 9 ECL Model & Governance
Meeting supervisory priorities across CBUAE, CBB, CBO and SAMA
Central banks and prudential supervisors across the GCC — including the Central Bank of the UAE (CBUAE), the Central Bank of Bahrain (CBB) and the Saudi Central Bank (SAMA) — place strong emphasis on forward-looking provisioning, model governance and capital resilience. IFRS 9 ECL frameworks sit at the intersection of accounting, credit risk and regulatory capital discipline; supervisors expect institutions to produce transparent, auditable and well-governed ECL outcomes that reflect both historical performance and plausible future scenarios.
IFRS 9 ECL Model
IFRS 9 Expected Credit Loss (ECL) programme combines quantitative modelling, accounting logic, forward-looking adjustments and strong governance. ESG WEISE builds transparent, auditable ECL frameworks that are defensible to auditors and regulators and practical for finance and risk teams to operate. We support the clients in,
-
Policy and Governance Framework
-
ECL Model Development
-
ECL Model Independent Review
Probability of Default (PD)
Estimates the likelihood an exposure will default (12-month and lifetime). Techniques include scorecards, logistic regression and transition matrices, segmented by product and customer cohort.
Loss Given Default
Estimates post-default loss as a percentage, using recovery curves, collateral haircuts, cure behaviour and timing of recoveries.
Exposure at Default
Predicts the exposure amount at default, accounting for utilisation patterns for revolvers, amortising schedules and off-balance items.
Staging & SICR logic
Accounting rules that determine provision scope
-
Define quantitative and qualitative Significant Increase in Credit Risk (SICR) indicators.
-
Set thresholds and triggers to transition exposures across Stage 1, Stage 2 and Stage 3.
-
Establish governance for overrides, migration rules and documentation of judgemental decisions.
Forward-looking overlays & scenario design
Integrating macroeconomics and judgment
-
Build scenario frameworks (baseline, adverse, upside) and assign weights.
-
Translate macro drivers into model adjustments (elasticities linking GDP, unemployment, commodity prices to PD/LGD).
-
Apply overlays or judgmental adjustments where models do not capture emerging risks or structural changes.
Governance, validation & testing
Assurance that models are robust and reproducible
-
Independent model validation and back-testing
-
Code and implementation review, unit tests and reconciliation to source data.
-
Documentation: methodology, assumptions, data lineage, version control and change logs.
-
Management reporting and audit-ready deliverables (validation reports, KPI packs, remediation roadmaps).
