Book a Demo
The EVE supervisory outlier test (SOT) is the supervisory tripwire for interest rate risk in the banking book (IRRBB): it flags a bank to its supervisor when the decline in its economic value of equity (EVE), under the worst of six prescribed interest rate shock scenarios, exceeds 15% of Tier 1 capital. In the EU the trigger sits in Article 98(5) of the Capital Requirements Directive (CRD), and the scenarios and calculation are specified by Commission Delegated Regulation (EU) 2024/856, in force since 14 May 2024.
Why does the supervisory outlier test exist?
IRRBB has no Pillar 1 capital charge. The Basel Committee decided in its 2016 standard (BCBS 368) that the risk depends too heavily on behavioral assumptions, such as deposit stability and prepayment, for a single formula: a standardized charge would over-capitalize some banks and under-capitalize others. In place of a charge, the Committee built what it called "enhanced Pillar 2", resting on three mechanisms: comparable public disclosure, a standardized framework supervisors can impose, and a quantitative outlier test.
The outlier test is the quantitative tripwire of that design. Basel Principle 12 requires supervisors to publish criteria for identifying outlier banks and to apply, at minimum, this test: a bank is an outlier at the latest when its maximum ΔEVE across the six prescribed scenarios exceeds 15% of Tier 1 capital. That tightened the 2004 test, which used 20% of Tier 1 plus Tier 2 capital under a simple ±200bp shock.
The 15% of Tier 1 figure is a minimum stringency, not a fixed line. Supervisors may set tougher thresholds, add further tests and treat other banks as outliers on broader grounds. The EU used exactly that option by adding a second, earnings-based test, the NII supervisory outlier test, which flags a decline in one-year net interest income of more than 5% of Tier 1 capital under the two parallel shocks. How the two lenses relate, and why a bank can look safe on one and exposed on the other, is covered in EVE vs NII explained.
How is the EVE SOT written into EU law?
The EU wrote the outlier logic directly into Level 1 legislation. Article 98(5) CRD requires competent authorities to exercise supervisory powers at least where an institution's economic value of equity declines by more than 15% of Tier 1 capital as a result of a sudden and unexpected change in interest rates under any of six supervisory shock scenarios. The scenarios and the common modeling and parametric assumptions were mandated to the European Banking Authority (EBA), whose draft technical standard became CDR (EU) 2024/856, in force since 14 May 2024.
The purpose of the delegated regulation is comparability. Every EU bank computes the same test the same way, so that outlier status means the same thing in Dublin and Milan. Where Basel expresses the test as a supervisory expectation, the EU has made it directly applicable law in every member state, applying to all institutions in CRD scope rather than only the large internationally active banks of the Basel scope.
What are the six shock scenarios?
The test uses six instantaneous shocks to the risk-free curve, defined in the Basel standard and reproduced in Article 1 of CDR 2024/856, applied separately to each material currency:
- Parallel up
- Parallel down
- Steepener (short rates down, long rates up)
- Flattener (short rates up, long rates down)
- Short rates up
- Short rates down
Each shape is built from three currency-specific parameters: a parallel, a short and a long shock size. Annex I of CDR 2024/856 lists them for 28 currencies; for the euro they are 200bp parallel, 250bp short and 100bp long. For currencies not in the annex, banks calibrate the shocks themselves using the method in Article 2 of the regulation.
The Basel Committee recalibrated the shock sizes in July 2024 (BCBS d578), with currency-specific factors and the six shapes unchanged, for implementation by 1 January 2026. In the EU, however, the shock table sits in CDR 2024/856, which has not been amended, so EU SOT figures still use the original calibration; see our IRRBB outlook for 2026–2028.
The bank computes ΔEVE under each of the six scenarios, and the test compares the worst outcome against 15% of Tier 1. One bad scenario is enough.
How is ΔEVE calculated for the outlier test?
A comparability test is worthless if every bank can choose its own assumptions, so Article 4 of CDR 2024/856 closes the main degrees of freedom:
| Element | Prescribed treatment |
|---|---|
| Balance sheet | Run-off: existing positions mature and are not replaced |
| Own equity | CET1 instruments and other perpetual own funds without call dates are excluded |
| Behavioral and automatic options | Must be reflected, with key behavioral assumptions adjusted to each scenario |
| Commercial margins | Included or excluded in line with the bank's internal approach; exclusion requires a transparent, consistent method and notification to the supervisor |
| Discounting | A general risk-free curve per currency (for example OIS), without credit or liquidity spreads |
| Post-shock rate floor | Starts at –150bp for immediate maturity and rises by 3bp per year to 0% at 50 years; if observed rates are already lower, the observed rate applies |
Where the regulation is silent, banks apply the assumptions of their internal measurement system or of the standardized approach. That is how the most consequential constraint enters: under paragraph 111 of the EBA IRRBB guidelines (EBA/GL/2022/14), the behavioral repricing of retail and non-financial wholesale non-maturity deposits (NMDs), together with operational deposits, is capped at a weighted average of 5 years, per currency. Because NMD assumptions dominate EVE results, the cap matters more than any other single parameter; how banks actually model these deposits is the subject of Non-Maturity Deposits Explained.
The post-shock floor is a calibrated compromise. Down-shocks cannot assume absurdly negative rates, but rates are not floored at zero either, reflecting the fact that several European currencies have traded below zero.
How to Build an IRRBB Framework That Holds Up Under Scrutiny
How are results across currencies aggregated?
Results are computed per currency and then aggregated per scenario with deliberate conservatism: losses are added in full, while gains are weighted at only 50%, on the logic that a gain in one currency is an imperfect hedge for a loss in another. One EU-specific refinement applies: for ERM II currencies with a formally agreed fluctuation band narrower than the standard ±15%, gains are weighted at 80%, acknowledging the tighter link between those curves and the euro. In practice this means the Danish krone.
What happens when a bank breaches the 15% threshold?
A breached supervisory outlier test does not create an automatic capital charge. It obliges the competent authority to exercise supervisory powers, drawing on the Article 104 CRD toolkit: requiring the bank to reduce the risk, imposing additional own funds through Pillar 2 requirements (P2R), restricting business, or requiring changes to models and assumptions.
There is an escape valve written into Level 1: the authority is not obliged to act if it concludes, on the basis of its supervisory review, that the institution's IRRBB management is adequate and it is not excessively exposed. This "hard trigger, soft consequence" design is inherited from Basel Principle 12, where outlier status is explicitly a trigger for a supervisory conversation, not a sanction.
In ECB practice, a breach typically leads to closer SREP scrutiny of the IRRBB element, a documented supervisory dialogue and, where the underlying risk management is found wanting, consequences in P2R or Pillar 2 guidance (P2G). Outlier numbers have been falling: the EBA's January 2026 report on its IRRBB heatmap found fewer outliers on both ΔEVE and ΔNII as banks adjust to the new rate environment.
How often is the SOT reported, and can outsiders run it?
EU banks report their SOT results quarterly through the dedicated IRRBB supervisory reporting templates, filed since the 30 September 2024 reference date, so a breach is visible to the supervisor in near-real time rather than at an annual review. The EBA's 2026 reporting simplification consultation closed on 10 July 2026, with final standards due by end-2026 (see EBA CP/2026/07 and IRRBB reporting).
From the outside, EU Pillar 3 disclosure (template EU IRRBB1) publishes ΔEVE under all six supervisory scenarios together with Tier 1 capital, so any reader can run the outlier arithmetic: compare the worst-scenario ΔEVE against 15% of Tier 1. The sign pattern across scenarios is informative too. A bank that loses EVE in the up-shocks has assets repricing later than its liabilities, which is the classic long-duration position.
Key takeaways
- The EVE supervisory outlier test flags a bank when its worst ΔEVE across six prescribed shock scenarios exceeds 15% of Tier 1 capital.
- In the EU the test is law: Article 98(5) CRD sets the trigger, and CDR (EU) 2024/856 (in force since 14 May 2024) prescribes the scenarios, shock sizes and calculation.
- EU shock sizes remain the original calibration in Annex I of CDR 2024/856 (EUR 200/250/100bp); the Basel 2024 recalibration has not yet been transposed.
- The calculation is tightly constrained: run-off balance sheet, own equity excluded, risk-free discounting and a post-shock floor from –150bp; the 5-year NMD cap comes from EBA/GL/2022/14.
- A breach triggers supervisory powers, not an automatic capital charge, but in practice it means closer SREP scrutiny and possible P2R or P2G consequences.
For the complete picture, from EVE and NII sensitivities and the six supervisory shock scenarios to outlier tests and how Basel, the EU, the UK and the US differ, see Mirai RiskTech's guide IRRBB regulation across Basel, the EU, the UK and the US: a complete reference.
