Locating Risks Before It Becomes Reality: How IFRS 9 ECL Supports the Resolution of Bad Loans

IFRS 9’s forward‑looking expected credit loss (ECL) regime has strengthened the link between provisioning and non‑performing loan (NPL) strategies by pulling loss recognition forward and reducing the accounting disincentives to write down, restructure and influencing remediation of bad assets. Global evidence suggests that where IFRS 9 is combined with strong supervisory pressure and functioning NPL markets, banks clean up legacy NPL stocks faster and limit the build‑up of new bad loans, turning oft-cosmetic accounting change into real balance‑sheet repair.

From Incurred Loss to Expected Loss

IAS 39’s incurred‑loss model allowed banks to delay provisioning until after a loss “trigger” event, which contributed to slow and incomplete recognition of problem loans before and during the global financial crisis. IFRS 9 reversed this logic by requiring 12‑month ECL from day one (Stage 1) and lifetime ECL once credit risk has increased significantly or the exposure is non‑performing (Stages 2 and 3), based on past events, current conditions and forward‑looking macro information.

This shift raises provisions earlier in the life of a deteriorating loan, making it less attractive to evergreen troubled exposures merely to avoid a P&L hit and more attractive to restructure, sell or write off when the economics justify it. It also forces much closer integration between risk modelling, data, and finance, aligning NPL management with the way risk appetites and stress‑testing frameworks are set.

How ECL Changes Bad‑Loan Incentive Structures

Under IFRS 9, loans that exhibit a “significant increase in credit risk” (SICR) migrate from Stage 1 to Stage 2 and attract lifetime ECL even if they are not yet formally in default, while Stage 3 loans are both credit‑impaired and provisioned on a lifetime basis with interest income recognised on a net basis. Forbearance and cosmetic restructuring typically trigger at least a Stage 2 classification, which blunts the old incentive to repeatedly roll over or lightly restructure loans simply to avoid NPL status.

Higher and more risk‑sensitive coverage ratios on under‑performing and non‑performing assets, in turn, reduce the incremental P&L pain of executing a genuine resolution strategy—whether that is a deep restructuring with haircuts, a portfolio sale, a securitisation or a write‑off—because much of the loss has already been reflected through provisions. Supervisors in Europe and elsewhere have reinforced this effect with “calendar provisioning” for NPEs, add‑on prudential filters and targeted NPL guidance, ensuring that banks cannot indefinitely offset the accounting discipline of IFRS 9 through optimistic modelling or aggressive assumptions.

Empirical Evidence: What the Data Show

A growing body of empirical work confirms that IFRS 9 has both raised provisions and altered banks’ behaviour in ways that support earlier recognition and resolution of bad loans. Key findings from major studies can be summarised as follows.

Table 1 – Selected Empirical Findings on IFRS 9, Provisions and NPLs

Study / Source (Year)	Scope and Sample	Key Empirical Finding (stylised for blog use)	Implication for Bad‑Loan Resolution
ECB Working Paper “Same same but different: credit risk provisioning under IFRS 9” (2023)	Euro‑area banks	IFRS 9 increased the risk‑sensitivity of provisions; banks with weaker capital spread provisioning over several quarters but still recognised expected losses earlier than under IAS 39.	More forward‑looking provisions and tighter capital–provisioning links supported more orderly NPL resolution, avoiding abrupt cliff‑effects when loans turned non‑performing.
Salazar et al., “IFRS 9, banking risk and COVID‑19: Evidence from Europe” (2023)	European listed banks during COVID‑19	ECL‑based provisions better anticipated future credit losses than the old incurred‑loss model, especially in sectors hit hardest by the pandemic.	Anticipatory provisioning allowed banks to absorb pandemic‑related losses while retaining room to restructure or dispose of emerging NPLs, instead of simply delaying recognition.
Bojar, “The Impact of IFRS 9 on the Link Between Lending and the Capital Ratio” – Poland (2022)	Publicly traded Polish banks, 2013–2018	IFRS 9 strengthened the sensitivity of loan growth to capital ratios; only well‑capitalised banks could significantly expand lending, with others becoming more conservative.	Tighter capital‑risk linkage encouraged earlier recognition of problem exposures and constrained riskier loan origination, limiting future NPL build‑up and supporting clean‑up.
Arfé, “Real effects of the IFRS 9 adoption in banking” – PhD thesis (Europe)	Loan‑level data from major European banks	After IFRS 9, banks increased pricing for riskier corporates and raised collateral standards for retail portfolios, reflecting the lifetime cost of ECL on under‑performing borrowers.	Tougher origination standards reduce inflows into NPL status over time, complementing resolution measures on the existing stock of bad assets.
EBA Report on the results of the EBA impact assessment of IFRS 9 (2017–2018)	EU banks’ quantitative impact study	The assessment found significant increases in credit‑loss provisions and some negative impact on CET1 at transition, but no broad‑based capital shortfall and clear improvements in risk‑sensitivity.	Stronger starting coverage ratios and supervisory overlays enabled authorities to push ambitious NPL reduction plans without generating widespread capital breaches.
GCC panel study “A Comparison Study between IFRS 9 and IAS 39 in GCC Countries” (2022)	53 GCC banks, 2012–2020	Implementation of IFRS 9 is associated with a statistically significant increase in provisions for credit losses, write‑offs and changes in non‑performing credit ratios relative to IAS 39.	Higher provisions and write‑offs under IFRS 9 indicate more active recognition and clean‑up of bad loans, reducing evergreening in GCC banking systems.

Taken together, these studies show that IFRS 9 raises coverage on risky and problematic assets, tightens the link between capital, provisioning and lending, and pushes banks to recognise and resolve bad loans earlier, especially where supervisors lean against optimistic modelling. They also highlight that ECL is not just an accounting change but a behavioural shock to loan pricing, underwriting standards and NPL strategies.

Country and Regional Experience

Cross‑country experience confirms that IFRS 9 works best when embedded in a broader NPL resolution framework that includes legal reforms, NPL markets, asset management companies and transparent supervisory expectations. The country evidence below illustrates how ECL has interacted with these tools to support the clean‑up of legacy NPLs and prevent future build‑ups.

Table 2 – Country and Regional Experience with IFRS 9 and NPL Resolution

Jurisdiction / Region	Stylised NPL Trend and IFRS 9 Timing	IFRS 9 / ECL Features and Supervisory Overlay	Observable Effects on Bad‑Loan Resolution (stylised)
Euro Area	System‑wide NPL ratios fell from high post‑GFC peaks to low single digits by the early‑mid 2020s; IFRS 9 introduced in 2018 alongside strong ECB/SSM NPL guidance and secondary‑market initiatives.	Forward‑looking ECL, ECB calendar provisioning for NPEs, harmonised NPL definitions, and policy efforts to develop NPL secondary markets and data hubs.	Higher coverage and earlier recognition facilitated portfolio sales, securitisations and write‑offs of legacy corporate NPLs, significantly accelerating balance‑sheet repair.
Greece	Double‑digit NPL ratios after the euro‑area crisis declined to mid‑single‑digit levels by the early 2020s, though still above the EU average.	IFRS 9 combined with the “Hercules” state‑guaranteed securitisation scheme, legal/judicial reforms and intensive supervisory pressure on NPL reduction.	ECL‑driven high coverage and state guarantees enabled large‑scale NPL securitisations and sales, transforming an inherited stock problem into a more manageable flow.
Italy	NPL ratios peaked in the mid‑teens and fell to low single digits by the early 2020s	IFRS 9 adoption coincided with the GACS guarantee scheme for NPL securitisations, judicial reforms and more conservative provisioning guidance by national and ECB supervisors.	Enhanced provisions under IFRS 9, coupled with guarantees, supported aggressive NPL portfolio disposals and rapid de‑risking of bank balance sheet
Spain and Portugal	Elevated NPL ratios after the property and sovereign crises converged toward EU averages, reaching low single digits by the early 2020s.	IFRS 9 plus national coordination platforms, asset management vehicles and clear supervisory targets for NPL reduction	Forward‑looking coverage improved pricing and investor appetite for NPL transactions, allowing banks to compress real‑estate and SME NPL stocks.
Poland (bank‑level evidence)	NPL ratios moderate by EU standards, with some cyclical variation; IFRS 9 implemented in 2018 under conservative local supervisory expectations.	IFRS 9 heightened the sensitivity of lending decisions to capital ratios and raised the cost of riskier borrowers through lifetime ECL.	Evidence shows tighter lending and stronger reliance on capital buffers, encouraging earlier problem recognition and limiting future NPL accumulation.
GCC (Bahrain, Saudi Arabia, UAE, Qatar, Kuwait, Oman)	NPL ratios generally low‑to‑moderate, with pockets of legacy corporate stress; IFRS 9 phased in from 2018, often with transitional capital relief.	IFRS 9 ECL models required higher provisions and refined staging criteria, with regional regulators issuing guidance and FAQs to avoid excessive procyclicality.	Panel evidence shows higher provisions, write‑offs and changes in NPL ratios post‑IFRS 9, indicating more active recognition and clean‑up of problem loans.
UK and non‑euro EU (e.g., Nordics, Netherlands)	Already low NPL ratios remained low; pandemic‑related upticks were contained quickly and coverage ratios stayed robust.	Early adoption of robust ECL models, strong stress‑testing, and conservative supervisory expectations on NPE provisioning.	High starting coverage and robust risk management meant that when stress hit, banks could restructure viable borrowers and dispose of non‑viable NPLs without destabilising capital.

These country experiences underline that IFRS 9 is a necessary but not sufficient condition for effective NPL resolution: its benefits emerge when accounting discipline is paired with credible resolution tools, legal frameworks and supervisory resolve. Where those complements are missing, higher provisions may still show up, but the system can remain clogged with hard‑to‑resolve bad loans and weak enforcement.

Moving Forward: Using IFRS 9 as a Resolution Enabler

For jurisdictions still grappling with high NPL stocks or embarking on IFRS 9 implementation, the lesson from global experience is that ECL should be treated as a resolution enabler, not merely a compliance exercise. This means aligning transitional capital relief with explicit NPL reduction plans, ensuring tax and legal regimes do not penalise write‑offs, and building secondary markets so that higher coverage can translate into actual disposals rather than just accounting entries.

Equally important is the micro‑level integration of ECL with early warning systems, restructuring protocols and collateral management, so that loans are moved to Stage 2 and 3 in time for meaningful action rather than as a belated acknowledgement of losses already incurred. When those pieces are in place, IFRS 9’s core promise—seeing credit risk before it becomes reality—translates into faster, cleaner and more durable resolution of bad loans.

References

SUNANDO ROY – On Banking, Finance and Society

Modernizing Bank Regulation and Supervision: Messages from the Federal Reserve

FSAP Management in Central Banks: Governance, Operational Arrangements and Coordination

Your cart (items: 0)