Proprietary Benchmark · Q2 2026 Edition

The Ezelman CRR3 Impact Index

An anonymised, quarterly benchmark of Day-1 CRR3 RWA uplift across 50+ Ezelman mandates — broken down by asset class, portfolio type and jurisdiction. The data set the Big Four wish they had.

See the benchmark → Get full breakdown (members only)

+17.4%

Average Day-1 CRR3 RWA uplift · Q2 2026 · 50+ mandates · unhedged

The EBA's own impact study puts the industry at +9–10%. Our book shows +17.4% before management actions — and the gap is the story.

Breakdown by asset class

Day-1 RWA uplift on a constant-book basis, averaged across mandates. Peer-benchmarked to the EBA Q4 2025 impact assessment where available.

Asset class	Ezelman mandates avg.	EBA benchmark	Spread
Residential mortgages	+22.1%	+12.0%	+10.1 pp
Large corporates (IRB)	+18.6%	+9.5%	+9.1 pp
Specialised lending	+14.9%	+8.0%	+6.9 pp
SME corporates	+11.3%	+7.5%	+3.8 pp
Unrated corporates (SA)	+28.0%	+14.5%	+13.5 pp
Banks & FIs	+9.4%	+6.5%	+2.9 pp
Counterparty credit risk (SA-CCR)	+12.7%	+8.0%	+4.7 pp
Operational risk (BIC)	+4.8%	+3.0%	+1.8 pp
Output-floor bite (all-book)	+6.2%	+4.5%	+1.7 pp

Why Ezelman numbers are higher than the EBA

Selection bias — acknowledged. Banks hire us because their internal numbers are getting worse, not better. Our book over-indexes on uplift.
Pre-management-action numbers. We publish the number before repapering, portfolio re-segmentation, securitisation, and IRB permission refresh — the fixes that deliver 30–80 bps of CET1 back.
Conservative IRB re-calibration. We anchor on the upper bound of the PD floor literature until the bank's supervisor signals otherwise. Many advisory firms anchor on the lower bound. That is how you get a Pillar 2G surprise in year two.
Unrated corporate treatment. The SA risk weight for unrated corporates is where our book diverges most from the EBA — because the EBA's sample is skewed to G-SIBs with more rated names than European mid-tier banks actually have.

Methodology

How the index is built.

The CRR3 Impact Index is a rolling, anonymised aggregation of RWA uplift observed across Ezelman client mandates over the preceding four quarters. Every bank in the sample is a European credit institution; mandates are weighted by total-RWA exposure so the index is not distorted by a single large book.

We publish the number we see, not the number we market. When our numbers differ from the EBA's — as they do — we show the spread and explain it. We do not back-solve to a comfortable narrative.

The public edition of the Index (this page) shows averages. The full breakdown — by jurisdiction, by bank size tier, by IRB vs SA, with CET1 waterfall and management-action playbook — is a member-only asset, released quarterly to Ezelman Intelligence subscribers.

Next edition: Q3 2026, published end of September.