Interest Rate Risk in the Banking Book (IRRBB) - EBA Guidelines and CSSF Expectations for Luxembourg Banks

Last updated: March 2026

Rates went up faster than most banks modelled. Between mid-2022 and late 2023, the ECB raised its deposit facility rate from -0.50% to 4.00%, a shift of 450 basis points in under 18 months. For Luxembourg banks that had built fixed-rate asset portfolios during the low-rate era, financed by short-term or variable-rate liabilities, the economic value of equity (EVE) moved sharply. Institutions that had assumed their non-maturing deposit base was stable, long-duration, and rate-insensitive found out how quickly those assumptions can break when depositors start moving balances to money market funds and term deposits offering 3% or more.

That rate cycle is exactly why IRRBB matters as a live risk rather than a modelling exercise. IRRBB sits in Pillar 2, not Pillar 1. There is no standardised capital formula. Instead, supervisors assess whether your interest rate risk management, measurement, and capitalisation are adequate, and if they are not, the consequence is a Pillar 2 add-on or a supervisory finding in the SREP. This guide covers the legal framework, the EBA guidelines (EBA/GL/2022/14), the six standard shock scenarios, the behavioural assumptions that drive the numbers, the supervisory outlier test, how CSSF and ECB teams assess IRRBB in Luxembourg, the reporting and disclosure requirements, CSRBB, and the errors that keep appearing in supervisory reviews.

Related reading: COREP Reporting Explained – the supervisory reporting framework that feeds the capital adequacy data underpinning your IRRBB assessment, and Pillar 3 Disclosure Requirements for Luxembourg Banks – which covers the IRRBB disclosure template (EU IRRBB1) as part of the broader Pillar 3 framework.

Legal Framework

IRRBB regulation in the EU operates across two axes: risk management requirements under the CRD, and disclosure requirements under the CRR. The distinction matters because they have different legal bases, different scopes, and different enforcement mechanisms.

Article 84 CRD – Risk Management

Article 84 of Directive 2013/36/EU (CRD) requires competent authorities to ensure that institutions implement systems to identify, evaluate, manage and mitigate the risk arising from potential changes in interest rates that affect an institution’s non-trading book activities. This is the legal anchor for the supervisory assessment of IRRBB. It also mandates the supervisory outlier test: competent authorities must apply at least one test comparing the institution’s EVE change under prescribed shock scenarios with a threshold expressed as a percentage of Tier 1 capital.

Article 84 CRD was amended by CRD V (Directive (EU) 2019/878) to strengthen the IRRBB framework, including the mandate for the EBA to develop regulatory technical standards on the standardised methodology and the supervisory outlier test.

Article 448 CRR – Disclosure

Article 448 of Regulation (EU) No 575/2013 (CRR), as amended by CRR2 (Regulation (EU) 2019/876), requires institutions to disclose their IRRBB exposures publicly. The disclosure covers the changes in EVE and NII under the prescribed shock scenarios, the key modelling assumptions (including NMD treatment and prepayment assumptions), and a qualitative description of the institution’s IRRBB management approach.

The EBA has developed implementing technical standards for the IRRBB Pillar 3 disclosure: template EU IRRBB1 (quantitative data on EVE and NII sensitivity under the six shock scenarios) and table EU IRRBBA (qualitative information on IRRBB management). These are part of the broader Pillar 3 ITS under Commission Implementing Regulation (EU) 2021/637.

EBA/GL/2022/14 – The Current Guidelines

EBA/GL/2022/14, the Guidelines on the management of interest rate risk and credit spread risk arising from non-trading book activities, is the current EBA guidance on IRRBB and CSRBB. These guidelines replaced EBA/GL/2018/02 (which itself had replaced EBA/GL/2015/08). The 2022 guidelines incorporated the revised BCBS IRRBB standards (April 2016) more fully into the EU framework and added CSRBB as a distinct risk to be managed alongside IRRBB.

The guidelines apply to all credit institutions and are addressed to both institutions and competent authorities. The IRRBB provisions of EBA/GL/2022/14 apply from 30 June 2023; the CSRBB provisions apply from 31 December 2023 (a six-month extension to allow institutions additional time to implement the new CSRBB framework). In Luxembourg, the CSSF has notified the EBA of its compliance with EBA/GL/2022/14, meaning these guidelines form part of the supervisory expectations that the CSSF and ECB joint supervisory teams apply when assessing Luxembourg institutions.

The Six Standard Supervisory Shock Scenarios

The IRRBB framework prescribes six interest rate shock scenarios for EVE measurement and two for NII. These scenarios are defined in Annex III of EBA/GL/2022/14, drawing directly from the BCBS IRRBB standards.

The Scenarios

The six EVE scenarios are:

Parallel shock up: the entire yield curve shifts upward uniformly.
Parallel shock down: the entire yield curve shifts downward uniformly.
Steepener shock: a combination of a downward shift at the short end and an upward shift at the long end, applied simultaneously to steepen the yield curve.
Flattener shock: a combination of an upward shift at the short end and a downward shift at the long end, applied simultaneously to flatten the yield curve.
Short rates shock up: a shock concentrated at the short end of the curve, diminishing toward longer tenors.
Short rates shock down: the inverse, concentrated at the short end downward.

For NII measurement, only the two parallel shock scenarios (up and down) are prescribed under the standard framework.

The shock magnitudes are currency-specific, calibrated from historical interest rate data. For the euro specifically, the calibrated shock sizes in Annex III of EBA/GL/2022/14 are: parallel shift of 200 basis points, short-end shock of 250 basis points, and long-end shock of 100 basis points. The steepener and flattener scenarios combine the short-end and long-end shocks with scalar adjustment functions that shape the full term structure. Other currencies have different calibrations; for instance, the USD calibration uses a parallel shift of 200bp, a short-end shock of 300bp, and a long-end shock of 150bp. Institutions with material exposures in multiple currencies must run the scenarios for each currency using the currency-specific calibrations.

EVE and NII: What They Measure

EVE (Economic Value of Equity) is the difference between the present value of all banking book assets and the present value of all banking book liabilities, including off-balance sheet items. A change in EVE under a shock scenario measures how much the economic net worth of the banking book would shift if rates moved as prescribed. EVE captures the long-term, full-horizon effect of interest rate changes.

NII (Net Interest Income) measures the projected change in the institution’s net interest income over a forward-looking earnings horizon, typically 12 months, under a given rate scenario. NII captures the short-term earnings impact. An institution can have a stable EVE but volatile NII, or vice versa, depending on the repricing structure and the maturity profile of its book.

The relationship between EVE and NII is not symmetrical. I have seen institutions where the EVE impact under the parallel-up scenario was small (because assets and liabilities repriced in rough alignment over longer horizons) but the NII impact was large (because the short-term liability repricing outpaced asset repricing in the first 12 months). The supervisory framework requires both measures precisely because they capture different time horizons of the same underlying risk.

Behavioural Assumptions: Where IRRBB Gets Hard

The EVE and NII numbers that feed the outlier test and the Pillar 3 disclosure are only as good as the behavioural assumptions underlying them. Understanding these assumptions before looking at the outlier test results is important, because the assumptions are what produce the numbers the outlier test evaluates. For products with contractual cash flows (fixed-rate loans with known maturities, fixed-rate bonds), the modelling is straightforward. For products where the cash flow timing depends on customer behaviour, the modelling is the risk.

Non-Maturing Deposits (NMDs)

NMDs are liabilities with no contractual maturity date. The depositor can withdraw at any time. Despite the contractual position, historical experience shows that a portion of NMD balances is stable: depositors do not move their money even when rates change. This stable portion is called “core deposits,” and the fundamental modelling question is: how much of your NMD base is core, and what effective maturity should you assign to it?

The BCBS framework provides caps on the core proportion and average maturity by NMD category:

Retail transactional deposits: core proportion cap of 90%, average maturity cap of 5 years.
Retail non-transactional deposits: core proportion cap of 70%, average maturity cap of 4.5 years.
Wholesale deposits: core proportion cap of 50%, average maturity cap of 4 years.

These are the BCBS caps. In the EU, EBA/GL/2022/14 applies a uniform maximum average repricing maturity of 5 years across all NMD categories (with limited exceptions for institutions using the standardised approach), which differs from the BCBS’s differentiated 5/4.5/4-year structure. EU practitioners operate under the EBA cap. Within these caps, institutions use their own internal models to estimate the actual core proportion and effective maturity, based on historical deposit volume analysis over at least 10 years. Banks must distinguish between stable and non-stable NMDs using observed volume changes, then identify the core proportion within the stable balance as the amount unlikely to reprice even under significant rate changes.

This is where the assumptions get tested against reality. The 2022-2023 rate cycle showed that some institutions’ NMD models had been calibrated during a period of persistently low or negative rates, during which deposit outflows were minimal. When rates moved to 4%, depositor behaviour changed. Retail clients who had never moved a savings balance in a decade suddenly transferred to term deposits or money market funds. Institutions that had modelled 80% of their retail non-transactional NMDs as core, with a 4-year average maturity, found those assumptions aggressive in the new environment.

Supervisors now scrutinise NMD assumptions more closely than any other element of the IRRBB model. If your internal model assigns a core proportion near the regulatory cap, expect to be challenged on why your institution’s depositor base is assumed to be more stable than the cap implies is prudent.

Prepayment Optionality

Fixed-rate loans with prepayment options (mortgage loans being the primary example) introduce optionality risk. When rates fall, borrowers prepay and refinance at lower rates, shortening the effective duration of the asset side. When rates rise, prepayments slow, extending duration. The modelling challenge is estimating prepayment speeds under each shock scenario, which requires historical prepayment data, an understanding of contractual prepayment penalties (which vary by jurisdiction and product), and an assessment of borrower refinancing behaviour.

In Luxembourg, where residential mortgage products typically have relatively long fixed-rate periods (10-20 years) and prepayment penalties governed by the consumer credit legislation, the prepayment modelling dynamic is different from markets with 30-year freely prepayable mortgages. But optionality risk still exists, and supervisors expect it to be quantified.

Pipeline Risk

Pipeline risk arises from commitments to lend at a fixed rate that have been offered but not yet drawn. If rates move between the offer date and the drawdown date, the institution is exposed. This is relevant for mortgage pipelines in particular. EBA/GL/2022/14 expects institutions to include pipeline risk in their IRRBB measurement where it is material. Many institutions model pipeline risk separately from the main IRRBB calculation and then aggregate the results, which is acceptable provided the methodology is documented and the aggregation is consistent.

The Supervisory Outlier Test (SOT)

The supervisory outlier test is the threshold mechanism that identifies institutions with potentially excessive IRRBB. The assumptions described above feed directly into the EVE calculation that the outlier test evaluates. Under the BCBS framework as implemented in the EU, a supervisor must implement at least one outlier test comparing the institution’s maximum change in EVE, across the six prescribed shock scenarios, with 15% of its Tier 1 capital.

If an institution’s worst-case EVE decline exceeds 15% of Tier 1 capital, it is flagged as a potential outlier and subject to enhanced supervisory review. This does not automatically trigger a capital add-on, but it does trigger a formal supervisory assessment of whether the institution’s IRRBB exposure is commensurate with its risk appetite, risk management framework, and available capital.

Alongside the EVE outlier test, an NII supervisory outlier test exists. Commission Delegated Regulation (EU) 2024/856 defines the NII SOT: the change in NII under the parallel shock scenarios, expressed as a percentage of Tier 1 capital, is compared against a threshold of -5% of Tier 1 capital. The EBA’s original draft RTS (EBA/RTS/2022/10) had proposed a lower threshold of -2.5%, but this was revised upward to -5% during the Commission adoption process (the EBA published its opinion on this adjustment as EBA/Op/2023/03). The NII SOT captures earnings risk over a one-year horizon, complementing the EVE SOT’s focus on long-term economic value. Both tests feed into the supervisor’s overall assessment of whether the institution’s IRRBB exposure warrants enhanced review.

Supervisors may also apply additional outlier tests using different capital measures, including CET1 capital or the margin by which regulatory capital exceeds minimum requirements. Any additional test must be at least as stringent as the 15% of Tier 1 threshold for the EVE test.

In practice, I observe that most Luxembourg banks run below the 15% threshold under normal conditions. The banks that approach or breach it tend to have concentrated fixed-rate mortgage books financed with short-duration deposits, or significant portfolios of long-dated government bonds acquired during the low-rate period. The rate cycle since 2022 has reduced some of these legacy positions, but residual duration mismatches remain across the sector.

CSSF and ECB Supervisory Assessment

IRRBB is assessed as part of the SREP (Supervisory Review and Evaluation Process). The treatment differs depending on whether the institution is a significant institution (SI) directly supervised by the ECB, or a less significant institution (LSI) supervised by the CSSF.

Significant Institutions – ECB/JST Assessment

For Luxembourg’s significant institutions, the ECB joint supervisory team (JST) assesses IRRBB within the SREP’s “assessment of risks to capital” block. The JST reviews:

The institution’s internal measurement system (IMS) for IRRBB: methodology, scope, assumptions, and model governance.
The EVE and NII sensitivity results under the six standard scenarios and any institution-specific stress scenarios.
The supervisory outlier test outcome: whether the 15% Tier 1 threshold is approached or breached.
The quality and stability of behavioural assumptions, particularly for NMDs and prepayment models.
The ALCO governance framework: frequency of IRRBB reporting to the board, limit framework, and escalation procedures.
The ICAAP treatment of IRRBB: whether the institution allocates adequate internal capital to IRRBB.

If the JST concludes that IRRBB management or capitalisation is inadequate, the consequence flows through the SREP score. A poor IRRBB assessment contributes to a lower overall risk score, which can result in a higher Pillar 2 Requirement (P2R) or Pillar 2 Guidance (P2G).

P2R vs P2G for IRRBB

The distinction between P2R and P2G matters for IRRBB. P2R (Pillar 2 Requirement) is a legally binding additional own funds requirement under Article 104(1)(a) CRD. It must be met at all times and a breach triggers automatic distribution restrictions. P2G (Pillar 2 Guidance) is a supervisory expectation that the institution holds own funds above the overall capital requirement (OCR), including P2R and the combined buffer. A failure to meet P2G does not trigger automatic distribution restrictions under Article 141 CRD, but it does trigger a supervisory dialogue and potential escalation.

IRRBB typically flows into the P2R via the SREP risk-to-capital assessment, not as a named IRRBB add-on but as part of the overall P2R calibration. Some supervisors also address IRRBB through P2G where the concern is about stress sensitivity rather than current-state capitalisation. The ECB has been clear that IRRBB can contribute to both P2R and P2G depending on the nature and severity of the risk.

Less Significant Institutions – CSSF Assessment

For LSIs, the CSSF conducts its own SREP assessment, following the EBA SREP guidelines (EBA/GL/2018/03, consolidated version). A revised version of the SREP guidelines is in EBA consultation but has not been finalised as of March 2026; EBA/GL/2018/03 (consolidated) remains the current reference. The CSSF’s assessment methodology for IRRBB is substantively aligned with the ECB approach, but the intensity of engagement differs. LSIs may not receive the same depth of on-site IRRBB review as SIs, but they are subject to the same regulatory framework and the same outlier test threshold.

CSSF Regulation No. 15-02, which transposes certain CRD Pillar 2 provisions into Luxembourg law, provides the domestic legal basis for the CSSF’s SREP activities, including the assessment of IRRBB.

CSSF Circular 12/552 and Governance Expectations

Circular CSSF 12/552 (as amended, most recently by Circular CSSF 24/860) is the CSSF’s principal circular on central administration, internal governance and risk management for credit institutions. While it covers all material risks rather than IRRBB specifically, it establishes the governance framework within which IRRBB management must sit: board-level risk oversight, the risk management function’s independence and authority, the ICAAP process, and the internal control framework.

For IRRBB specifically, the CSSF expects Luxembourg credit institutions to comply with EBA/GL/2022/14 in full, including the requirements on IRRBB governance, ALCO structures, limit frameworks, and stress testing. The CSSF’s off-site supervision function monitors IRRBB exposures through the supervisory outlier test data, through Pillar 3 IRRBB disclosures (for institutions that produce them), and through IRRBB-specific data requests during the SREP cycle. Where the CSSF identifies that an institution’s IRRBB framework falls short of the EBA guidelines, the supervisory response mirrors the ECB approach: a governance finding in the SREP, and potentially a P2R or P2G adjustment to address the risk.

In my experience, the CSSF’s IRRBB focus for Luxembourg LSIs concentrates on three areas: NMD modelling quality (particularly for institutions with large retail deposit bases), the adequacy of internal capital allocation to IRRBB in the ICAAP, and whether the institution’s ALCO is genuinely reviewing and challenging IRRBB results or simply receiving them. These are the areas where the gap between EBA guideline requirements and actual practice tends to be widest in the LSI population.

Reporting and Disclosure

IRRBB has both a public disclosure channel and a supervisory reporting channel. Practitioners regularly conflate the two. They are different processes with different audiences, formats, and frequencies.

Pillar 3 IRRBB Disclosure (Public)

Article 448 CRR requires institutions to disclose their IRRBB exposures as part of their Pillar 3 report. The disclosure uses two components:

Template EU IRRBB1 presents the quantitative EVE and NII sensitivity results under the six prescribed shock scenarios (for EVE) and the two parallel scenarios (for NII). For each scenario and each currency with material exposure, the institution reports the change in EVE and NII. The template also includes columns for internally modelled NII sensitivities where the institution uses its own scenarios beyond the prescribed ones.

Table EU IRRBBA provides the qualitative context: how the institution defines and measures IRRBB, the key modelling and parametric assumptions (including NMD treatment), the hedging approach, and the average and longest repricing maturity assigned to NMDs. This last data point is important because it makes the NMD behavioural assumption directly visible to market participants and supervisors.

Pillar 3 IRRBB disclosure follows the standard Pillar 3 frequency: semi-annual for large institutions, annual for SNCIs. The audience is public: counterparties, investors, analysts, rating agencies.

Supervisory Reporting (To the NCA) – The J Template Set

Separately from the Pillar 3 public disclosure, institutions submit standardised IRRBB supervisory reporting to their national competent authority. The EBA ITS on supervisory reporting includes a dedicated IRRBB template set: the J templates (Annex XXVIII/XXIX of the ITS), reported quarterly. These are the supervisory equivalent of what the C templates are for credit risk and liquidity. The J template set covers:

J 01.00: EVE and NII supervisory outlier test results, plus market value changes. This is the headline template that supervisors review first.
J 02.00 to J 04.00: Breakdown of IRRBB sensitivity estimates, providing granular detail on the drivers of the EVE and NII results.
J 05.00 to J 07.00: Repricing cash flow profiles, showing the maturity and repricing structure of the banking book across time buckets.
J 08.00 to J 09.00: Relevant parameters on behavioural modelling, including the NMD core proportion, repricing maturity, and prepayment speed assumptions the institution uses.
J 10.00 to J 11.00: Qualitative information on IRRBB management, governance, and risk appetite.

The J templates are a distinct reporting obligation from the COREP own funds templates (C series) or the liquidity templates (C 72.00 for LCR). While they sit within the broader supervisory reporting ITS, they are produced by the ALM or market risk function rather than the standard regulatory reporting team. This ownership split is a practical reality: the data that feeds J 01.00 through J 09.00 comes from the IRRBB measurement system, not from the general ledger or the COREP data warehouse. Institutions need a clear handoff process between the ALM function (which owns the IRRBB model and produces the sensitivity results) and the regulatory reporting function (which is responsible for submission logistics and data quality controls).

The distinction between the J templates (supervisory reporting to the NCA, quarterly, not public) and the Pillar 3 templates EU IRRBB1/EU IRRBBA (public disclosure, semi-annual or annual) is a point of confusion for practitioners. Both cover similar ground (EVE/NII sensitivities, behavioural assumptions), but they serve different audiences, use different formats, and follow different frequencies. The data must be consistent between them: a material discrepancy between your J 01.00 SOT results and your EU IRRBB1 Pillar 3 disclosure will attract supervisory questions.

Internal Reporting

EBA/GL/2022/14 requires institutions to have a robust internal reporting framework for IRRBB. This means regular reporting to the ALCO (or equivalent risk committee) and to the board, covering at minimum: EVE and NII sensitivities under the prescribed scenarios and any internal stress scenarios; limit utilisation; the supervisory outlier test result; the key behavioural assumptions and any changes to them; and the basis risk, optionality risk, and credit spread risk positions.

The frequency of internal IRRBB reporting should be at least quarterly, with more frequent reporting (monthly or even weekly) for institutions with material or complex IRRBB exposures. The guidelines expect the ALCO to review and challenge the results, not merely receive them.

Credit Spread Risk in the Banking Book (CSRBB)

CSRBB is the second risk covered by EBA/GL/2022/14 alongside IRRBB, reflecting the EBA’s view that credit spread risk in the banking book warrants a distinct management and governance framework rather than being subsumed under either IRRBB or credit risk. CSRBB is the risk to the economic value and net interest income arising from changes in credit spreads in the market, for banking book positions. It is related to, but distinct from, credit risk (the risk of default) and from IRRBB (the risk from changes in the risk-free rate).

CSRBB materialises when the market credit spread on an instrument widens or narrows, affecting its market value or the cost of replacing it. For a bank holding a portfolio of corporate bonds in the banking book, a widening of credit spreads reduces the economic value even if the risk-free rate is unchanged and no defaults occur.

The guidelines require institutions to have a CSRBB management framework that is conceptually separate from IRRBB, with its own identification, measurement, and monitoring processes. In practice, many Luxembourg banks are still developing their CSRBB frameworks. The measurement challenge is significant: isolating the credit spread component from the overall yield change requires a clean decomposition into risk-free rate, credit spread, and liquidity premium, which is model-dependent and data-intensive.

Supervisors assess CSRBB as part of the SREP IRRBB review. The expectation is not that institutions have a fully standardised CSRBB model on day one, but that they can demonstrate they have identified where CSRBB sits in their banking book, have a governance framework to monitor it, and are working toward quantitative measurement. A bank that has not addressed CSRBB at all will attract a supervisory finding.

Common Errors and Supervisory Findings

Based on EBA, ECB and CSSF supervisory communications, and on what I see discussed in risk management networks across Luxembourg, the following issues appear most frequently:

Overly optimistic NMD assumptions. The most common finding. Institutions modelling core NMD proportions at or near the regulatory cap, with average maturities at or near the maximum, based on historical data from the low-rate period. Supervisors expect back-testing of NMD assumptions against the 2022-2023 rate cycle and recalibration where the data shows the assumptions were too favourable.

Governance gaps. ALCO meetings that receive IRRBB reports as information items but do not review or challenge the assumptions. Limit frameworks where IRRBB limits have never been breached because the limits were set too wide. Board members who cannot explain the institution’s IRRBB risk appetite in practical terms. The guidelines expect active governance, not passive receipt of reports.

Inadequate documentation of behavioural assumptions. The NMD model exists, but the documentation does not explain why a particular core proportion was chosen, what data was used, how the model was validated, or under what conditions it would be recalibrated. Supervisors expect model documentation at a standard comparable to IRB credit risk models: methodology, data, validation, limitations, and a clear owner.

Failure to run reverse stress tests. EBA/GL/2022/14 expects institutions to consider which interest rate scenarios would most severely affect their capital and earnings. Reverse stress testing asks: what rate scenario would cause the EVE decline to breach the 15% outlier threshold, or cause NII to fall below a level that threatens the institution’s viability? Institutions that only run the six standard scenarios without exploring tail scenarios are not meeting the spirit of the guidelines.

No CSRBB framework. Some institutions have not begun to address CSRBB as a distinct risk. Supervisors are moving from “we expect you to start thinking about this” to “show us what you have.” An institution with no CSRBB identification, measurement, or governance framework is behind the supervisory expectation.

Reconciliation gaps between internal models and Pillar 3 disclosures. The EVE and NII figures reported in template EU IRRBB1 should be consistent with the results presented internally to the ALCO and the board. Differences arise when the internal model uses different assumptions or perimeters than the Pillar 3 template. These differences must be explained in table EU IRRBBA. Unexplained gaps attract questions.

What Is Coming

The IRRBB regulatory framework is not static. Several developments are relevant:

The EBA is developing regulatory technical standards (RTS) under Articles 84 and 98(5a) CRD that will specify the standardised methodology for IRRBB measurement and the detailed supervisory outlier test calibration. These RTS have been subject to an extended timeline. When finalised, they will formalise elements of the framework that currently sit in guidelines rather than binding regulation. Institutions using internal models should monitor the RTS consultations for any implications on model governance, validation, or the scope of the standardised methodology as a backstop.

The EBA is also revising the SREP guidelines. A consultation paper was published proposing a comprehensive update to EBA/GL/2018/03, incorporating CRD VI/CRR3 changes, DORA, ESG integration, and refined IRRBB assessment methodology. When finalised, the revised SREP guidelines will replace the current consolidated version. Institutions should monitor this for changes to how IRRBB assessment feeds into the SREP scoring framework.

The question of whether IRRBB should move from Pillar 2 to Pillar 1 (a standardised minimum capital requirement for IRRBB) remains open at the BCBS level. The BCBS IRRBB standards (April 2016) kept IRRBB in Pillar 2 but established the standardised methodology as a potential Pillar 1 tool. The EU has not moved toward Pillar 1 treatment, and there is no current legislative proposal to do so, but the discussion is not closed.

The interaction between IRRBB and the Fundamental Review of the Trading Book (FRTB) at the trading book boundary is a supervisory focus area. The CRR3 trading book boundary rules (Articles 104 and 104a CRR, as revised) tighten the criteria for boundary classification. Instruments reclassified from the banking book to the trading book exit the IRRBB perimeter and enter the FRTB capital calculation. Institutions should assess this boundary impact as part of their CRR3 implementation.

Frequently Asked Questions

Is there a Pillar 1 capital requirement for IRRBB?

No. IRRBB remains in Pillar 2 in the EU. Supervisors assess IRRBB adequacy through the SREP and can impose additional capital requirements (P2R) or expectations (P2G) if they determine that an institution’s IRRBB exposure is not adequately covered by existing capital.

What are the supervisory outlier test thresholds?

The EVE outlier test compares the institution’s maximum EVE decline, across the six prescribed shock scenarios, with 15% of its Tier 1 capital. The NII outlier test, as defined in Commission Delegated Regulation (EU) 2024/856, uses a threshold of -5% of Tier 1 capital. Breaching either threshold flags the institution as a potential outlier for enhanced supervisory review.

Which EBA guidelines apply to IRRBB now?

EBA/GL/2022/14 is the current guideline. It replaced EBA/GL/2018/02, which in turn replaced EBA/GL/2015/08. The 2022 guidelines cover both IRRBB and CSRBB. Any reference to the 2015 or 2018 guidelines is outdated.

What are the J templates for IRRBB supervisory reporting?

The J template set (J 01.00 through J 11.00) is the standardised IRRBB supervisory reporting framework, reported quarterly to the NCA. J 01.00 covers the EVE/NII SOT results, J 02.00-J 04.00 break down sensitivities, J 05.00-J 07.00 cover repricing cash flows, J 08.00-J 09.00 cover behavioural modelling parameters, and J 10.00-J 11.00 cover qualitative information. These are distinct from the Pillar 3 disclosure templates (EU IRRBB1, EU IRRBBA), which are public and follow a different frequency.

How often must IRRBB be reported internally?

EBA/GL/2022/14 requires at least quarterly internal reporting of IRRBB metrics to the ALCO and the board. Institutions with material or complex IRRBB exposures should report more frequently (monthly or weekly). The Pillar 3 IRRBB disclosure follows the standard Pillar 3 frequency: semi-annual for large institutions, annual for SNCIs.

Can NMD core proportions exceed the regulatory caps?

No. The BCBS framework specifies hard caps on core proportions: 90% for retail transactional, 70% for retail non-transactional, and 50% for wholesale deposits. Average maturity caps under the BCBS are 5, 4.5, and 4 years respectively. In the EU, EBA/GL/2022/14 applies a uniform 5-year maximum average repricing maturity across all NMD categories. Internal models must produce estimates within these caps.

Key Takeaways

IRRBB sits in Pillar 2: there is no standardised Pillar 1 capital charge. Supervisors assess adequacy through the SREP and can impose P2R or P2G if they find IRRBB management or capitalisation insufficient.
The current EBA framework is EBA/GL/2022/14 (IRRBB provisions from 30 June 2023, CSRBB from 31 December 2023), replacing EBA/GL/2018/02. Six standard shock scenarios are prescribed for EVE with EUR-specific calibrations (parallel 200bp, short 250bp, long 100bp), and two for NII.
Behavioural assumptions drive the IRRBB numbers. BCBS NMD core proportion caps apply (retail transactional 90%, retail non-transactional 70%, wholesale 50%); EBA/GL/2022/14 applies a uniform 5-year maximum average repricing maturity across all categories. Models calibrated in the low-rate era need recalibration against the 2022-2023 rate cycle.
The EVE supervisory outlier test flags institutions where the maximum EVE decline exceeds 15% of Tier 1 capital; the NII SOT uses a threshold of -5% of Tier 1 capital (Commission Delegated Regulation (EU) 2024/856). The CSSF applies these tests under its Pillar 2 framework (CSSF Regulation 15-02), with governance expectations anchored in Circular CSSF 12/552 (as amended).
IRRBB supervisory reporting uses the J template set (J 01.00 through J 11.00), reported quarterly to the NCA. Pillar 3 disclosure (EU IRRBB1, EU IRRBBA) is a separate public obligation following standard Pillar 3 frequency. Both must be consistent.
Common supervisory findings: overly optimistic NMD assumptions, governance gaps (passive ALCO), missing reverse stress tests, no CSRBB framework, and reconciliation gaps between internal models and Pillar 3 disclosures.

Pillar 3 Disclosure Requirements for Luxembourg Banks – covers the full Pillar 3 framework including the IRRBB disclosure template EU IRRBB1 and table EU IRRBBA as part of the Article 448 CRR requirements.
COREP Reporting Explained – the supervisory capital reporting framework. The capital adequacy data in COREP underpins the Tier 1 capital figure used in the supervisory outlier test denominator.
FINREP Reporting Explained – financial reporting templates. The banking book asset and liability balances in FINREP are the starting point for the EVE calculation.
MREL Reporting Requirements – MREL and IRRBB interact through the capital stack: a P2R add-on for IRRBB increases the MREL target for institutions in resolution scope.
COREP Reporting Errors – common data quality issues in COREP that can cascade into IRRBB measurement errors where COREP data feeds the IRRBB model.

Sources and References

Directive 2013/36/EU (CRD), Article 84 (IRRBB management requirements): https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32013L0036
Directive (EU) 2019/878 (CRD V), amending CRD including Article 84 on IRRBB: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019L0878
Regulation (EU) No 575/2013 (CRR), Article 448 (IRRBB disclosure): https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32013R0575
EBA/GL/2022/14 – Guidelines on the management of interest rate risk and credit spread risk arising from non-trading book activities (current IRRBB/CSRBB guidelines): https://www.eba.europa.eu/regulation-and-policy/supervisory-review-and-evaluation-srep-and-pillar-2/guidelines-on-the-management-of-interest-rate-risk-arising-from-non-trading-activities
Commission Delegated Regulation (EU) 2024/856 – RTS on supervisory outlier tests for IRRBB (EVE and NII SOT definitions and thresholds, finalised from EBA/RTS/2022/10): https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R0856
Commission Implementing Regulation (EU) 2021/637 (Pillar 3 ITS, including IRRBB templates EU IRRBB1 and EU IRRBBA): https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32021R0637
EBA/GL/2018/03 (consolidated) – Guidelines on SREP and supervisory stress testing (current; revised version in consultation): https://www.eba.europa.eu
BCBS IRRBB Standards (April 2016) – Interest rate risk in the banking book: https://www.bis.org/bcbs/publ/d368.htm
CSSF Regulation No. 15-02 (Pillar 2 transposition, Luxembourg): https://www.cssf.lu
Circular CSSF 12/552 (as amended by Circular CSSF 24/860) – Central administration, internal governance and risk management: https://www.cssf.lu/en/Document/circular-cssf-12-552/

Disclaimer: The information on RegReportingDesk.com is for educational and informational purposes only. It does not constitute legal, regulatory, tax, or compliance advice. Always consult your compliance officer, legal counsel, or the relevant supervisory authority for guidance specific to your institution.