Basel III Endgame: Navigating the CRR3 Output Floor
By Jonas Adam Mohamed Osman, known as Yonas Osman
CEO of Quantica Risk
60 Rue François Ier, 75008 Paris, Francequanticariskmodel.comHow the CRR3 output floor reshapes capital planning for EU banks, why internal models lose ground, and what treasury teams must do now.
Executive answer
The CRR3 output floor, phasing in from 2025 and reaching 72.5% of standardised risk-weighted assets by 2030, caps the capital benefit banks can draw from internal models. For institutions with historically favourable IRB models — particularly in low-default portfolios such as large corporates, banks, and specialised lending — the floor binds hardest, pushing required capital upward even where internal models suggest lower risk. Boards and CFOs need multi-year capital trajectories that reflect the floor, not just current model outputs.
Practically, this means recalculating RWA under both the internal model approach and the standardised approach every reporting period, then applying the higher of the floored and unfloored figures. Banks with diversified low-risk books often see the floor bind well before 2030 on a binding-constraint basis, so capital issuance plans, dividend policy, and portfolio mix decisions should be stress-tested against the full glide path now rather than reactively in later transition years.
Why the output floor changes capital incentives
Before CRR3, banks with sophisticated IRB models could hold materially less capital than standardised peers for the same exposure, creating a strong incentive to invest in model granularity. The output floor compresses that gap by design, since it is calculated at an aggregate level across credit, market, operational, and CVA risk rather than per exposure. This changes the return on investment for model development: incremental model refinement that once reduced RWA materially may now have little effect once the floor binds, redirecting resource toward data quality and standardised-approach optimisation instead.
The floor also interacts with sovereign and institution exposures where internal models traditionally produced near-zero risk weights. The EBA's monitoring exercises have shown that globally systemic and large universal banks are most affected, while smaller retail-focused banks with naturally higher standardised risk weights see less relative impact. This divergence matters for competitive dynamics: banks previously advantaged by internal models must now compete on funding cost and balance-sheet efficiency rather than capital arbitrage alone.
Managing the transitional arrangements
CRR3 includes transitional provisions — notably a temporary cap on RWA increases for residential mortgage and unrated corporate exposures — designed to smooth the glide path through 2032. Banks need granular exposure-level tracking to determine which transitional treatments apply and how they unwind, since the benefit of these provisions erodes mechanically each year. Treasury and finance functions should build the floor's full trajectory, including the eventual loss of transitional relief, into their three- to five-year capital plans rather than treating 2025 figures as representative.
Supervisors, including the ECB under its SSM mandate, are increasingly asking banks to demonstrate how the output floor interacts with Pillar 2 add-ons and the combined buffer requirement, since double-counting risk between the floor and supervisory add-ons is a live concern. Banks that model the floor in isolation from their broader capital stack risk understating the true binding constraint, particularly where MREL and leverage ratio requirements also tighten over the same horizon.
Strategic responses for banks and investors
Leading banks are responding with a mix of balance-sheet actions: originate-to-distribute strategies for capital-intensive low-margin assets, selective exposure reduction in sovereign and bank books where floor impact is highest, and repricing of long-dated corporate lending to reflect the true floored cost of capital. Some institutions are also revisiting legal entity structure, since the floor applies at consolidated and in some cases solo level, creating incentives to rationalise subsidiaries with duplicative capital requirements.
Investors and rating agencies are now asking banks to disclose floored and unfloored capital ratios side by side, and markets are beginning to price the gap as a signal of model conservatism versus aggression. Management teams that proactively communicate their floor trajectory, rather than waiting for disclosure requirements to force transparency, tend to see less volatility in credit spreads when the binding constraint eventually materialises in reported ratios.
Conclusion
The output floor is not a one-off compliance exercise but a structural reweighting of how capital is allocated across a bank's book, and its effects compound with every other Basel III endgame element — the revised standardised approach for credit risk, the new operational risk framework, and CVA risk capital. Banks that model it as an isolated add-on rather than an integrated constraint will misjudge both their capital headroom and their competitive positioning over the next decade.
Jonas Adam Mohamed Osman, known as Yonas Osman, CEO of Quantica Risk, approaches the output floor as a capital allocation problem first and a modelling problem second, helping banks separate genuine risk reduction opportunities from capital effects that the floor will neutralise regardless of model quality.
Primary sources
More in Banking & Market Risk
- FRTB Implementation: Lessons for Trading Desk Risk
By Jonas Adam Mohamed Osman · 8 min read
- IFRS 9 ECL Overlays: Discipline Beyond the Model
By Jonas Adam Mohamed Osman · 8 min read
- IRRBB Supervisory Outlier Test: What Banks Must Prove
By Jonas Adam Mohamed Osman · 7 min read