Strategic development and validation of internal models under CRD

CRD Internal Models: Supervisory Approval & Validation

The use of internal models to calculate risk-weighted assets requires supervisory approval from the ECB and national authorities.

  • 01Full compliance with CRD requirements for internal models
  • 02Significant capital relief through the IRB approach and advanced models
  • 03Solid model validation and governance in accordance with EBA standards
  • 04Successful regulatory approval procedures
11+Years of experience
120+Employees
540+Projects
ISO 27001certified

Model Risk Management under CRD VI/CRR III

CRR III has made internal model usage more restrictive: the output floor caps capital savings at 72.5% of the standardised approach, IRB approvals are now granted at exposure class level rather than institution level, and the revised ECB guide to internal models (July 2025) tightens requirements for governance, validation and risk parameter estimation. Banks must keep machine learning models explainable, integrate climate and ESG risks into their models, and provide stricter IT implementation evidence.

We support you in strategically navigating this complex approval landscape – from model adaptation to CRR III requirements through TRIM follow-up to optimising your capital efficiency under the new restrictions.

Our comprehensive range of model risk management services covers the entire model lifecycle – from initial conception through development and validation to regulatory approval and continuous monitoring.

2 service modules

Service scope

01

IRB Model Development and Validation

Development and validation of internal rating models for credit, market and operational risks in accordance with CRD requirements.

  • PD, LGD and EAD model development in accordance with IRB standards
  • Independent model validation and backtesting
  • Regulatory documentation in accordance with EBA Guidelines
  • Continuous model monitoring and calibration
02

Regulatory Approval and Governance

Support with regulatory approval procedures and the establishment of solid model governance structures.

  • Preparation and support for supervisory meetings
  • Building model risk management frameworks
  • Implementation of model governance processes
  • Change management and stakeholder communication

5 phases

Our Approach

We work with you to develop a comprehensive CRD Internal Models strategy that combines regulatory excellence with maximum capital efficiency.

  1. Analysis of your current model landscape and regulatory requirements

  2. Strategic model planning and business case development

  3. Model development, calibration and validation

  4. Regulatory documentation and approval procedures

  5. Implementation and continuous model monitoring

Your contact

Melanie Düring

Head of Risk Management

Developing internal models to CRD standards is one of the most complex and at the same time most valuable investments in modern risk management. Our clients benefit not only from significant capital relief, but also from markedly improved risk management capabilities and strategic decision-making foundations that create sustainable competitive advantages.
ADVISORI expert opinion

Our Model Risk Expertise

  • 01Deep expertise in CRD model regulation, EBA guidelines and ECB TRIM outcomes
  • 02Proven track record supporting numerous IRB approval procedures with BaFin and ECB
  • 03Interdisciplinary teams from risk management, statistics and regulatory affairs
  • 04End-to-end support from model development through validation to go-live

CRR III Output Floor: Action Required by 2028

The phased increase of the output floor from 50% (2025) to 72.5% (2030) progressively reduces the capital benefits of internal models. Banks must adapt their model strategy now to achieve maximum capital efficiency under the new restrictions.

8 QUESTIONS, BRIEFLY ANSWERED

Frequently asked questions about CRD Internal Models

What is the supervisory approval process for internal models under CRD?

Banks wishing to use internal models for calculating own funds requirements need approval from the competent supervisory authority. For significant institutions, the ECB is responsible; for less significant institutions, national authorities such as BaFin. The process involves an assessment of model quality, data foundations, IT implementation and governance structures. After initial approval, models are subject to ongoing monitoring and regular supervisory reviews.

What changes does CRR III bring for the use of internal models?

CRR III (effective 1 January 2025) introduced significant restrictions: the output floor caps capital savings from internal models at 72.5% of the standardised approach (fully phased in by 2033). IRB approvals are now granted at exposure class level rather than institution level, providing more flexibility for gradual adoption. At the same time, requirements for data quality, IT implementation and model validation have been considerably tightened.

What does the revised ECB guide to internal models 2025 contain?

The ECB guide published in July 2025 fundamentally updates supervisory expectations for internal models. Key changes include: tightened governance requirements with direct senior management accountability, new expectations for machine learning models regarding explainability, first-time integration of climate and ESG risks into credit risk models, stricter requirements for internal validation and internal audit in line with the EBA supervisory handbook, and expanded IT implementation evidence requirements.

What was the TRIM programme and how does it affect current model reviews?

The Targeted Review of Internal Models (TRIM) was a comprehensive ECB programme from 2016 to 2021 that assessed the consistency and appropriateness of internal models at significant institutions. TRIM findings have led to numerous supervisory measures that institutions must still implement. The insights from TRIM feed directly into the revised ECB guide to internal models and determine the focus areas of ongoing model monitoring.

How does the output floor impact capital planning with internal models?

The output floor sets a minimum for risk-weighted assets calculated with internal models: they may not fall below 72.5% of the standardised approach result. Implementation is phased in gradually until 2033 with EU transitional provisions. Banks must strategically adjust their capital planning and run parallel standardised and IRB calculations for their entire credit book. The strategic question is whether the IRB approach still delivers sufficient capital savings despite the output floor.

What requirements apply to model validation under CRD?

Model validation must be independent from model development and encompasses quantitative backtesting analyses, qualitative assessment of model methodology and ongoing monitoring of model performance. The 2025 ECB guide tightens requirements for validators and internal audit regarding independence and depth of review. Banks must report internal validation results to supervisory authorities at least annually and demonstrate the robustness of their IT systems and data quality.

What role do machine learning and AI play in internal models?

The revised ECB guide contains explicit expectations for the use of machine learning techniques in internal models for the first time. Banks must ensure ML-based models are sufficiently explainable and that their performance justifies the additional complexity. Supervisors expect transparent documentation of model logic, robust validation procedures for ML models, and evidence that algorithmic decisions remain traceable and fair.

How does ADVISORI support IRB approval and model authorisation?

ADVISORI guides institutions through the entire approval process: from the strategic decision for the IRB approach through model development and validation to the formal approval procedure with the ECB or BaFin. Our services include gap analysis of existing models against CRR III requirements, preparation of model documentation to ECB standards, support during supervisory reviews, and optimisation of capital efficiency considering the output floor.