Financial Services
Insights
Local roots. European reach. Tailor-made solutions.
September 2025
EBA's "new definition of default" (DoD): a deep dive for banks – With NPV test front and center
An overview of the endorsed changes, how it will impact banks and the underlying financial service sector.
Europe's supervisors are keeping the 1% NPV loss cap for distressed restructurings under Article 178(3)(d) ("diminished financial obligation"). That sounds small, but it's a big deal: even a pure timing deferral, with no haircut and no interest-rate cut, can blow through 1% fast. At a 5% EIR, pushing the remaining cash flows out by two years produces a ~9-10% NPV hit.
Share:
Author
01
Introduction
The European Banking Authority revisited the Definition of Default (DoD) Guidelines under CRR3 Article 178(7). The latest consultation (published in early July) centers on distressed restructurings/ forbearance and the NPV test, and addresses moratoria, factoring, and several technical alignments with CRR3. The aim is greater consistency across banks and closer alignment with accounting treatment, while preserving a tight materiality lens.
This article converts that rulemaking, and the surrounding industry debate, into practical consequences for institutions. It examines impacts on capital and Risk-weighted Assets (RWA), IFRS 9 provisioning and staging, operating model and process design, reputational considerations, and business strategy.
The article also includes a simple NPV comparison that discounts both “before” and “after” cash flows at the original Effective Interest Rate (EIR) by using a plain 5-year bullet loan to showcase how a timing-only concession can drive a ~9–10% NPV loss even when nominal cash flows are unchanged. It also illustrates compensation levers (rate step-ups, fees, maturity tweaks) needed to keep the ΔNPV ≤ 1% threshold and provide templates and governance checklists, so teams reach consistent, defensible decisions.
02
The Regulatory core: what’s changing and what’s not
When a borrower receives forbearance due to financial difficulty, banks must compare the NPV of expected cashflows before and after the change, discounting both at the original Effective Interest Rate (EIR). If the NPV loss exceeds the bank’s internal threshold (capped at 1%), the exposure is classified in default under Article 178(3)(d) of the CRR, i.e. the unlikely-to-pay due to Diminished Financial Obligation (DFO) trigger. Despite industry participants’ push back, the EBA intends to retain the 1% cap, with the rationale to ensure alignment with accounting logic, harmonization with material-past-due thresholds (also 1%), and avoiding arbitrage and RWA variability.
Other proposals and non-changes
Although a shorter probation/cure period for low loss restructurings was considered initially, it was not adopted in the recently published consultation to avoid divergence from NPE cure rules (≈12 months). With respect to Moratoria, EBA proposes no blanket carve out as a moratorium does not automatically mean default and institutions still assess whether it is true forbearance and, if so, perform NPV test. Furthermore, for non-recourse Factoring, EBA proposed ease of “technical past due” noise by moving the invoice level tolerance from 30 to 90 days and clarifying scope and edge cases. Finally, several CRR3 clean-ups are included, such as removing references to the former 180 Days Past Due (DPD) discretion and related technical updates.
The framework effectively pushes banks to structure and price concessions to be NPV-neutral (or close), for example via step-ups, fees, or maturity tweaks, rather than simply pushing payments out and triggering default through time-value effects.
03
The NPV Tet made simple, and why 1% is easy to breach
In order to apply the test, institutions need to first establish and document that the modification is forbearance, i.e. a concession granted because the borrower is (or is likely to be) in financial difficulty (as prescribed in CRR Art. 47b). Only then should the NPV test be performed, and it must be run at the modification date (the “test date”). From that date forward, institutions should construct two cashflow paths:
the original-remaining schedule, which captures every amount still due under the pre-existing contract through legal maturity, and
the new schedule, which reflects the concession exactly (including any holiday, capitalization of interest, step-ups, fees or maturity changes) at the contract’s actual frequency.
Once done, Institution should discount both streams at the loan’s original EIR (the effective interest rate set at origination, consistent with IFRS 9 modification logic) and compute the loss percentage:
The result needs to be compared with internal policy threshold, which must be no higher than 1%. If the loss exceeds the threshold, the exposure needs to be classified as defaulted under Art. 178(3)(d); if it does not, the outcome needs to be documented, and any other Unlikely-To-Pay (UTP) indicators need to be assessed before concluding.
Even when nominal amounts are unchanged, pushing payments later reduces their present value at a positive discount rate. If all remaining cashflows are deferred by Y periods, with annual EIR r and f payments per year, the timing-only loss is approximately:
As an example, with r = 5%, monthly f = 12, and a 24-month deferral is calculated according to the below:
At annual granularity, a two-year push-out gives:
Both figures far exceed the 1% cap, even though the borrower ultimately repays in full.
As seen above, the threshold is very narrow which indicates that merely shift timing will often breach the threshold unless compensation is added (for example through modest rate step-ups, small fees, or calibrated maturity tweaks) to keep ΔNPV ≤ 1%. While the consultation also tidies some CRR3 references, the real operational impact for banks is to design restructurings that are NPV-neutral (or close) rather than simply deferring cash flows.
In Figure 1, it’s clear that the cashflows of the loan does not change in amount but only is shifted into the future. Nevertheless, the PV is impacted and is well above the 1% threshold.
04
Capital, provisioning and RWA – what the 1% cap means in practice
When a default is triggered, the capital impact is felt first at the exposure level. Under the Standardized Approach (SA), defaulted unsecured exposures are shifted into the default exposure class, attracting a 150% risk weight where Specific Credit Risk Adjustments (SCRA) plus value adjustments on the unsecured portion are less than 20%, and otherwise a 100% risk weight. For many retail and SME portfolios, this results in higher risk weights compared to the pre-default position, causing risk-weighted exposure amounts to rise until the exposure cures.
In the Foundation IRB (F-IRB) framework, the mechanics differ as the risk weight on defaulted exposures is set to 0%, probability of default (PD) is fixed at 100% and the Expected Loss (EL) must be fully recognized. Any shortfall between expected loss and provisions or adjustments is deducted directly from own funds in accordance with Article 159, creating a capital drag even though the nominal risk weight is zero.
Advanced IRB (A-IRB) banks face a more nuanced treatment, since risk weights are determined by Expected Loss Best Estimate (ELBE) and LGD-in-default parameters. Here too, PD is immediately set to 100% and provisioning requirements bite upfront, but the direction of the RWA change relative to pre-default depends on the model parameters in place. It should be noted that these effects are temporary as the additional RWA burden unwinds when exposures cure, with the duration of the headwind depending on probationary periods and “material payment” conditions.
At the portfolio level, a stricter definition of default typically increases the number of defaults, which in turn raises observed default rates and drives up PDs across the remaining non-defaulted book. Under F-IRB, this feeds through immediately, whereas under A-IRB the effect filters in via the calibration window. Interestingly, in the A-IRB approach, loss given default (LGD) may actually decline if more low-loss or cure-prone exposures are pulled into the default category, providing a partial offset to the higher PDs. The overall effect on portfolio risk weights is therefore highly bank-specific and data-driven.
Finally, the accounting dimension compounds the picture. A default recognized under Article 178(3)(d) generally triggers a migration to IFRS 9 Stage 3. Many exposures subject to forbearance may already have been classified as Stage 2 due to significant increase in credit risk, but Stage 3 requires recognition of lifetime expected credit losses and restricts interest recognition to the net carrying amount. The result is a front-loaded impact on profit and loss and CET1 capital at the time of default recognition, with potential partial reversals later on cure or recovery.
05
Operational, reputational, business consequences
When a default flag is raised, even if it stems from a purely “technical” breach of the NPV test, it immediately activates the full non-performing exposure (NPE) machinery. Collateral must be re-valued, accounts are shifted into workout workflows, monitoring becomes more intensive, and additional reporting and model flags are triggered. For small retail and SME portfolios, this cascade can be costly and disproportionate due to volumes surge, backlogs build, and teams are diverted from preventative or value-adding tasks into checklist-heavy processes. The resulting hand-offs, from frontline staff to collections, then to work-out and finance, prolong turnaround times and increase operational risk at precisely the moment when speed and stability are most critical.
The consequences extend beyond operations into reputation. Once the default label appears, it often surfaces in credit bureaus or external registers, constraining borrowers’ future funding options. Clients who approached their bank for relief may experience the outcome as punitive, forcing relationship managers into defensive conversations and, in some cases, driving customers to leave. The stigma also radiates outward as suppliers, counterparties, and co-lenders may tighten terms, amplifying stress for firms that remain fundamentally viable.
These operational and reputational effects feed back directly into business incentives. The combination of the default label, higher capital and provisioning costs, and the heavy workflow burden can make institutions reluctant to grant forbearance at all. Faced with disproportionate consequences, some banks may simply leave original terms in place, raising the risk of outright default and higher eventual losses, the very perverse outcome regulators warn against. Put differently, even a narrow breach of the 1% NPV threshold can push both bank and borrower toward a worse end-state if not managed carefully.
The remedy lies in designing support structures that avoid triggering the default label when the borrower remains fundamentally viable. This means engineering NPV-neutral (or near-neutral) solutions delivered through proportionate workflows such as small rate step-ups, modest fees, or micro-extensions that offset timing effects and keep ΔNPV within the 1% limit. Institutions can e.g. formalize a “lite” NPE path for such technical breaches, with automated eligibility checks, streamlined cash-flow recalculations, and pre-packaged, transparent client disclosures.
Clear communication helps also banks to distinguish regulatory classification from the bank’s intent to support, ideally offering a non-defaulting option alongside any deeper restructuring.
Finally, notably, documented criteria that clarify when to offer NPV-neutral concessions and when to escalate to the full NPE route allow risk, finance, and workout teams to act consistently, defensibly, and in a way that preserves both efficiency and trust.
06
Moratoria and factoring – two hotspots
Two areas stand out in the evolving default framework as particular sources of debate and operational challenge: moratoria and factoring. Both raise questions about how to balance regulatory consistency with practical realities, and both require careful governance to avoid unnecessary default flags.
Moratoria
The presence of a payment moratorium, even one mandated by legislation, does not in itself constitute a default. The assessment begins in the same place as any other case, i.e. whether the arrangement is a form of forbearance granted because the borrower is, or is likely to be, in financial difficulty. Only if that condition is met should the bank proceed to the NPV test at the original effective interest rate and consider the implications of Article 178(3)(d). The EBA’s position is that the existing framework should apply, without automatic exemptions. This means institutions must avoid blanket interpretations, instead judging each case on its merits and documenting clearly why a broad programmatic pause, such as sector wide administrative relief does or does not qualify as forbearance. While industry discussions have raised the possibility of carve-outs for general moratoria, these do not feature in the draft rules. In practice, consistent governance is essential e.g. a short, standardized memo template can record the borrower difficulty assessment, expected cash flow profile under the moratorium, and, where relevant, the NPV test outcome, ensuring results are auditable and proportionate even at scale.
Factoring
By contrast to Moratoria, the key issue in Factoring is precision. The regulatory direction is toward greater invoice level granularity to prevent “technical” defaults arising from administrative delays. The proposal introduces a 90-days-past-due tolerance applied at the invoice level, clarifying what types of arrangements qualify as factoring for this treatment. The aim is to ensure that delays in posting or disputes on individual invoices do not escalate into borrower-level defaults. For institutions, the main challenge lies in systems and data. Ageing and materiality checks must run for each invoice, purchased receivables must be tagged cleanly and linked to the correct obligor, and reporting should distinguish true credit-driven arrears from administrative noise. The benefit is a sharper default boundary i.e. fewer false positives, better alignment between back-office reality and regulatory reporting, and a more accurate reflection of borrower creditworthiness.
07
The NPV test in practice – what your engine must do
At the heart of the NPV test lies a simple but strict discipline which is the EIR set at origination is the one and only discount rate to be used. Every calculation, whether for the “before” or the “after” cash flow leg, must be anchored to this fixed reference point. From the test date (the moment the concession is granted, often at the end of a payment holiday), the bank must build the schedule of remaining contractual cash flows through to legal maturity. This forms the benchmark against which the restructured schedule is measured.
The restructured leg must capture, transparently and exactly, all the moving parts of the concession i.e. payment holidays, whether accrued interest capitalizes into principal, any rate step-ups, maturity extensions, fees, and deferrals. Alongside the numbers, institutions should record the rationale for forbearance i.e. the evidence that the borrower is, or is likely to be, in financial difficulty. Further to this, governance is completed by setting a clear threshold for NPV loss (no more than 1%, as required) and documenting procedures for overrides and the treatment of other unlikely-to-pay indicators. When these artefacts are locked down and applied consistently, the NPV result is reproducible, auditable, and defensible.
In practice, it is useful to view the NPV test from two complementary angles:
a. For regulatory classification, the capitalization view is decisive: here, the actual modified terms are taken as written, including the capitalization of interest during holidays, and the NPV of the restructured cash flows is compared with the original schedule, both discounted at the original EIR from the test date. This is the figure that drives Article 178(3)(d).
b. Alongside it, many banks maintain a timing-only view. In this version, the nominal cash flows are held constant but shifted by the length of the deferral, again discounted from the test date at the original EIR. While not a regulatory trigger, this secondary view is a valuable control and training tool. It isolates the pure time-value effect of deferrals and shows quickly whether a delay alone would breach the 1% materiality threshold.
To illustrate this more concretely, consider a €100,000 bullet loan with a five-year term, priced at a 5% EIR and repaid monthly. After the first year, the borrower is granted a 12-month holiday, with all subsequent cash flows pushed out by 24 months. On a timing-only basis, the NPV loss at the test date is:
This is nearly ten times the 1% cap, even though the nominal payments themselves are unchanged. Under the capitalization view, the impact may be even greater if interest accrues during the holiday and is rolled into the balance, enlarging the final bullet and pushing it further out in time thus reducing NPV unless offset. To stay within the threshold, banks can design NPV-neutral (or near-neutral) concessions. These may include modest rate step-ups, small fees, or micro-extensions calibrated so that the restructured leg, when discounted at the original EIR, remains within the 1% limit.
08
Process blueprint to stay compliant and proportionate
Building a compliant and proportionate process begins with the principle of NPV-neutrality. The most effective way to achieve this is through a standard “compensation menu” that frontline teams can apply consistently, temporary rate step-ups, micro-extensions of maturity, or small fees calibrated to offset the time value loss of deferrals or holidays. These tools should be embedded in a deal calculator that shows the ΔNPV in real time, allowing relationship managers to design borrower-specific solutions without breaching the 1% materiality threshold when the customer remains fundamentally sound.
Governance around the threshold itself is the next cornerstone. While regulation caps the loss at 1%, institutions may set a stricter internal limit. Clear policies should specify who has authority to approve a tighter threshold and under what circumstances other UTP overrides may be documented, for example, where collateral coverage is exceptionally strong. This ensures decisions are consistent, defensible, and free from ad-hoc exceptions.
Equally important is defining forbearance eligibility. Institutions need to spell out what constitutes “financial difficulty” in their portfolios and what evidence must be recorded. If the case does not meet this bar, it is not forbearance, the NPV test is not required, and Article 178(3)(d) does not apply. This gatekeeping step prevents misuse of the test as a generic repricing mechanism.
Alignment between the Definition of Default (DoD) and NPE cure rules should also be maintained. Cure horizons and “material payment” conditions must be synchronized so that borrowers experience a consistent path across risk, accounting, and reporting. As regulatory requirements evolve, both frameworks should be updated in parallel to avoid operational mismatches and confusing client communication.
Special cases require their own standard operating procedures. For moratoria, policies should predefine how to distinguish broad administrative relief from genuine forbearance, and how the NPV engine treats each path. For factoring, clarified invoice level rules need to be implemented so that administrative delays do not trigger “false” defaults; purchased receivables should be tagged cleanly, with ageing checks performed at invoice granularity.
Finally, the process must be operationalized in day-to-day classification. Beginning with a documented screen for financial difficulty, this is the gatekeeper for NPV testing. From the test date, build both the original remaining schedule and the restructured schedule, reflecting contract terms exactly (holidays, deferrals, capitalization, step-ups, fees, and maturity changes). Discount both legs at the original EIR, compute the ΔNPV%, and apply the decision rule:
If the loss exceeds the threshold, classify a default under Article 178(3)(d),
If not, document the outcome and check other UTP indicators to ensure the assessment is holistic.
Where default is recognized, apply cure rules and probation requirements, currently at least one year, with “material payment” conditions, while keeping DoD and NPE alignment front and center.
This classification must flow through booking and reporting. Defaulted exposures ordinarily migrate to IFRS 9 Stage 3, interest recognition switches to the net basis where relevant, and both COREP and FINREP returns must reflect the new status. At portfolio level, the defaults feed into IRB calibration and capital planning. As default definitions tighten, PDs can be expected to rise, while LGDs under A-IRB may fall if more cure prone exposures are captured; the net effect on RWAs is portfolio-specific and should be monitored through regular scenario analysis.
09
Model, data and controls – the things that will trip you up
The most common mistakes in implementing the NPV test stem from the basics, which to know where and how the discounting is applied. The rule is clear as the EIR set at origination must be used, never a current market rate or blended portfolio yield. Equally critical is the choice of starting point which is that the discounting begins at the modification date (often the end of any holiday), and both the “before” and “after” cash flow legs must be built forward from that point. A mismatch in either the rate or the start date can be enough to swing the result across the 1% threshold.
Precision in periodicity is another area that frequently trips up models. If a loan is structured monthly, then every element such as holiday duration, deferral, and discounting exponents, must also be expressed in months, not rounded or approximated in years.
Capitalization logic is another blind spot where many concessions roll accrued interest into principal during holidays, and if the engine does not capture this correctly, the final bullet and all subsequent coupons will be miscalculated. Clean data handling is just as important and as such input fields should be strictly numeric, formula ranges locked down, and options like “mode” selections limited to controlled drop-down menus.
Finally, auditability cannot be an afterthought. Each test should generate a tamper-proof snapshot of the “before” and “after” schedules and the single EIR applied, since second line functions and auditors will inevitably ask to see it.
Getting the policy framework right is just as important as the technical details. Institutions should explicitly allow NPV neutral or near neutral compensation so that viable borrowers can be supported without triggering Article 178(3)(d). That means temporary rate step ups, micro extensions of maturity, or small fees can be used, provided they are consistent with conduct standards and borrower viability. The ≤1% threshold must be governed tightly where internal rules may set it below 1% but never above, with clear escalation channels for rare overrides. These exceptions should be limited to cases where other UTP indicators are absent and collateral coverage is demonstrably strong, and they must be documented rigorously.
On the systems side, the NPV engine itself should be reliable, consistent, and free from hidden complexity. At a minimum it must:
always discount at the original EIR,
(ii) use the test date as t₀, building both the original and restructured schedules from that point forward,
support monthly frequency natively, ensuring holidays and deferrals shift dates exactly; and
include a timing only view as a built-in sanity check, allowing teams to see immediately that, for example, a two-year pushout at 5% produces a ~9–10% PV hit before any compensation.
The framework should also include playbooks for edge cases. A moratoria playbook, for example, should spell out when a sector wide administrative pause does not constitute forbearance (and thus requires no NPV test) and when it does, together with practical guidance on scaling the NPV assessment efficiently across many accounts. For factoring, processes must reflect invoice level rules, applying a 90-days-past-due tolerance at invoice level so that administrative lags do not cascade into borrower defaults. Systems should tag purchased receivables cleanly and distinguish genuine credit arrears from process noise.
Taken together, these policy and system upgrades reduce avoidable breaches, keep the process proportionate, and make the institution’s decisions easier to explain and defend.
10
Capital and provisioning playbook – what management wants to see
Senior management needs more than raw test results, they want a single, connected view that links loan level concessions to capital, provisioning, and reporting outcomes. The first building block is an exposure level scorecard. For each type of concession, this should show whether the NPV test is likely to trigger default, how the exposure’s RWA evolves under both SA and IRB approaches, and the accounting journey under IFRS 9. That means tracking transitions from Stage 2 into Stage 3 and back again, identifying when interest recognition switches to a net basis, and clarifying what it takes for exposures to return to Stage 2, e.g. cash collections, the length of cure periods, and “material payment” criteria.
Complementing this is a portfolio level “what-if” pack. By assuming higher volumes of UTP driven defaults, banks can model the knock-on effect on PDs and risk weights, recalibrate A-IRB LGD in default paths as more cure prone exposures enter the default sample, and quantify sensitivity to expected loss shortfalls when provisions lag behind regulatory EL.
The third leg is mitigation analytics, where for common restructuring patterns such as payment deferrals, the analysis should quantify how much compensation is required to keep ΔNPV within the 1% threshold. For example, a 24-month pushout at a 5% EIR might demand a modest step-up in rate for a set period or a calibrated maturity extension. Bringing these three perspectives together gives decision makers a clear picture of the tradeoffs between borrower relief, capital consumption, and accounting impact.
From a capital standpoint, the effects materialize on two levels. At the exposure level, SA banks often see risk-weights rise sharply for defaulted unsecured positions, while IRB banks may experience pressure through expected loss and own funds deductions even if risk weights fall to zero (F-IRB) or become parameter driven (A-IRB). These burdens unwind on cure, but they tie up capital for as long as defaults remain open. At the portfolio level, a tighter default definition typically raises observed default rates and PDs, lifting risk weights. Under A-IRB, this can be partially offset if default datasets include more cure prone cases, nudging LGD down, but the net outcome is bank-specific.
Provisioning and profit and loss dynamics move in parallel. Defaults push exposures into Stage 3, driving up expected credit losses and placing near term pressure on P&L and CET1. Over time, cures and recoveries moderate these effects, but the timing can heavily influence reported results.
Operationally, default classification activates the full workout and NPE machinery, revaluations, enhanced monitoring, documentation, and reporting. For small, “technical” NPV breaches, this burden can be disproportionate, while repeated framework changes would compound the problem through costly model adjustments and revalidations. Reputationally, the default label constrains borrowers’ access to finance and can strain client relationships. Unless banks deliberately design NPV neutral concessions, they risk underutilizing viable forbearance and driving both bank and borrower toward worse end-states.
Finally, process alignment is critical. Any move to shorten DoD probation without synchronizing NPE cure rules would widen the gap between regulatory default and non-performing exposure treatment, creating confusion in operations and disclosures. Maintaining tight alignment across DoD and NPE definitions, mapping, and external reporting is essential to preserve both clarity and credibility, while ensuring that policy, systems, and communications evolve in lockstep.
11
Communication – reduce reputational risk
Reputation often hinges less on technical classifications than on how they are explained to clients. When a restructuring triggers what is, in essence, a technical default, the first touchpoint with the borrower should be plain, human, and transparent. The message must clarify why prudential and accounting rules can register a default even while the bank is providing support, what this means for credit reporting, and how the status can later be cured. Ideally, the borrower is offered two clearly framed choices. One is a non-defaulting, NPV-neutral option, such as a modest rate step up, a small fee, or a micro-extension, that offsets the time value loss and avoids the default label. The other is a defaulting option, which may offer greater short-term relief but carries the consequences of default classification. Setting out the timelines, costs, and cure conditions for both paths empowers the client to choose with eyes open, reducing surprise and preventing the damaging perception that “we asked for help and were branded in default.”
Strong communication must rest on robust processes. The first step is to tighten the NPV toolchain i.e. to standardize on the original EIR as the sole discount rate, build all cash flows from the test date, and model monthly contracts in months, including holidays and deferrals, so timing is exact. Every case should be run through two complementary views, the capitalization view, which determines the regulatory outcome, and a timing only control that makes the time value effect visible (for example, highlighting the ~9–10% present value loss of a two-year pushout at 5%).
Designing NPV neutral concessions is the next layer of protection. Tools such as modest step ups, small fees, or micro extensions can be calibrated within the calculator to keep ΔNPV within the ≤1% threshold, allowing viable borrowers to be supported without automatically triggering default. Governance must anchor this process by documenting the forbearance criteria, evidence of genuine financial difficulty, a strict ≤1% threshold policy, oversight of rare exceptions, and systematic checks for other unlikely-to-pay indicators.
Planning for capital and provisioning effects also helps avoid reputational strain. Running portfolio level “what-ifs” shows how different mixes of forbearance will affect expected loss shortfall sensitivity in F-IRB books and PD/LGD recalibrations under A-IRB. Operationally, a proportionate NPE playbook ensures that small, technical breaches do not translate into outsized workloads, while pre-agreed client scripts prevent unnecessary escalation of reputational damage when defaults are unavoidable.
Finally, maintaining up to date SOPs for special cases such as moratoria and factoring adds consistency. A standing memo should explain when a moratorium does not constitute forbearance (and therefore falls outside the NPV/default framework) and, when it does, how assessments should be scaled. For factoring, invoice level ageing rules with a 90-days-past-due tolerance should be applied to prevent administrative lags from generating false defaults, with receivables tagged cleanly in systems. These measures make outcomes more consistent, easier to defend, and materially less damaging to both customers and capital.
12
EBA NPV Test in Practice: Why Even Simple Deferrals Breach the 1% Cap
In order to illustrate the EBA rule and its impact on loan restructurings, an analysis was conducted to model a set of stylized cash-flow scenarios based on EBA’s requirements:
concession be assessed through an NPV test,
both the original and restructured cash flows must be discounted at the loan’s original effective interest rate (EIR),
the default classification triggered if the relative loss in NPV exceeds 1% even when the borrower ultimately repays the same nominal amounts.
As illustrated in Figure 2, a simple deferral or holiday, without any change to principal or coupon, can create an economic loss that breaches the cap purely due to the time value of money.
What Figure 2 illustrates
The first panel demonstrates this with a €100 loan, five-year maturity, monthly payments, and a 5% EIR. After a 12-month holiday, all remaining payments are pushed out by an additional 24 months. While the nominal repayment profile is unchanged, discounting at the original EIR reduces the present value of the shifted schedule to €90.50 compared to €100 for the original, producing a ΔNPV of 9.5%. This is nearly ten times the 1% threshold—showing how timing alone can trigger default classification.
The second panel extends the analysis across a range of EIRs (3%, 5%, 7%) and deferral periods (6–36 months). The “Uncompensated ΔNPV %” column highlights the loss when no adjustment is made i.e. the longer the deferral and the higher the EIR, the greater the present-value hit. For example, a 36-month deferral at 7% produces a loss of nearly 19%.
The results also quantify the compensation required to keep the ΔNPV within the 1% cap. This can be structured in different ways e.g. as a coupon step-up (e.g., +260 bps for a 5% / 24-month case), as an upfront fee (around €8.50 per €100), or as a one-time fee at maturity (about €11.50 per €100). Each option neutralizes the time-value effect and keeps the concession within the regulatory boundary.
The analysis makes clear why the EBA retained the 1% rule as even modest deferrals breach the cap unless paired with compensating features. Banks therefore need to industrialize their approach by embedding NPV calculators in frontline tools, standardizing compensation menus, and documenting outcomes to ensure decisions are transparent, auditable, and defensible.
13
Conclusion — embrace the constraint
By retaining the 1% cap, the EBA has signaled unambiguously that the NPV test will remain a cornerstone of default classification. The expectation is not that banks look for ways around the rule, but that they build their processes around it. That means industrializing the test which entails deploying an auditable NPV engine that always applies to the original EIR from the test date, embedding it in front line tools, and calibrating product features so that viable restructurings can be structured without breaching the threshold. This technical discipline must be coupled with tight governance i.e. clear criteria for what qualifies as forbearance, documented evidence of borrower difficulty, a firm ≤1% threshold policy with controlled overrides, and full alignment of cure rules, reporting, and client communications. The result is consistency, across the loan file, regulatory reporting (COREP/FINREP), and the message delivered to customers.
From the industry’s standpoint, the cap is often viewed as demanding. Banks cite operational burden, the risk of technical defaults, and the possibility of discouraging forbearance. Yet EBA’s rationale is deliberate as they aim to harmonize, increase comparability, and consistency with accounting logic, while preventing arbitrage around materiality thresholds. In short, the constraint is not a flaw but a feature.
Institutions that embrace this constraint stand to gain. By standardizing compensation menus, automating the NPV test, running regular capital and provisioning “what-ifs”, and maintaining proportionate NPE workflows, banks can safeguard capital, minimize reputational risk, and deliver better borrower outcomes. The strategic play is not to wait for the bar to shift, but to design within it, ensuring that decisions are transparent, repeatable, and defensible.
14
Contact
Author
This article has been drafted by zeb consulting. Any text or images included in this article has been created based on the public available information as well as external and internal expert opinions. The article is indented for general information only and is not intended to be, and should not be relied upon as, legal, regulatory, financial, investment, tax, or other professional advisory. zeb does not warrant that this article is objective or complete, and neither zeb nor its employees shall bear any liability arising from or relating to the content in this article.
zeb Consulting is as a leading strategy, management and IT consultancy which has been offering transformation expertise along the entire value chain in the financial services sector inEurope since 1992. Zeb has five offices in Germany – Frankfurt,Berlin, Hamburg, Munich and Münster (HQ) – as well as 10 international locations. Our clients include European large-cap and private banks, regional banks, insurers as well as all kinds of financial intermediaries. Several times already, our company has been classed and acknowledged as “best consultancy” for the financial sector in industry rankings.
Share:
