The term "internal models" in the context of Basel IV refers to the methodologies and frameworks that banks use to calculate their risk-weighted assets (RWAs) and regulatory capital requirements for credit, market, and operational risks under the Basel framework. Basel IV, while not an official term, generally refers to the finalized Basel III reforms and enhancements agreed upon by the Basel Committee on Banking Supervision (BCBS) in December 2017, often called the "Basel III Endgame." These reforms, set to be implemented starting from January 2023 (with phased-in timelines in many jurisdictions), aim to reduce excessive variability in RWAs and improve the comparability of risk assessments across banks.

Internal Models Under Basel IV

Under the Basel framework, banks with sufficient sophistication and regulatory approval can use internal models to estimate risks and calculate capital requirements, as opposed to relying on standardized approaches. Basel IV introduces significant changes to the use of internal models, aiming to address concerns about their complexity, lack of transparency, and variability in outcomes.

Here are the key aspects of internal models under Basel IV for different risk types:

• Credit Risk - Internal Ratings-Based (IRB) Approach:


• Basel IV imposes stricter constraints on the use of internal models for credit risk under the IRB approach to reduce variability in RWAs.


• Key changes include:


• Removal of Advanced IRB (A-IRB) for certain exposures: Banks are no longer allowed to use A-IRB for exposures to large corporates, banks, and other financial institutions with total assets above a certain threshold (e.g., €50 billion). Instead, the Foundation IRB (F-IRB) or standardized approach must be used.


• Parameter floors: For portfolios where internal models are still allowed, Basel IV introduces input floors for key parameters like Probability of Default (PD) and Loss Given Default (LGD). For example, PD floors are set at 0.05% for corporate exposures.


• Output floor: Basel IV introduces an overall output floor of 72.5% of the RWAs calculated under the standardized approach. This means that even if a bank's internal model calculates lower RWAs, the final capital requirement cannot be less than 72.5% of the RWAs under the standardized approach.


• These changes aim to limit excessive optimism in internal models and enhance comparability across banks.


• Market Risk - Internal Model Approach (IMA):


• For market risk, Basel IV builds on the Fundamental Review of the Trading Book (FRTB), which was finalized as part of the Basel III reforms.


• Banks can use internal models for market risk, but the framework is more stringent:


• Expected Shortfall (ES): Replaces Value-at-Risk (VaR) as the primary risk measure, focusing on tail risk over a 97.5% confidence level.


• Liquidity horizons: Risk factors are assigned different liquidity horizons, reflecting how long it would take to unwind positions in stressed market conditions.


• Model approval and backtesting: Stricter criteria for model approval, including desk-level validation and backtesting requirements. Desks failing backtesting must revert to the standardized approach.


• Non-modellable risk factors (NMRFs): Risk factors that cannot be adequately modeled must be capitalized using a stress scenario approach, increasing capital requirements.


• The goal is to ensure that internal models for market risk are robust and capture tail risks more effectively.


• Operational Risk - Internal Models Prohibited:


• One of the most significant changes in Basel IV is the complete elimination of internal models (Advanced Measurement Approaches, AMA) for operational risk.


• Instead, all banks must use the Standardized Measurement Approach (SMA), which combines a business indicator component (based on income) with a historical loss component (based on a bank’s own loss experience).


• This change was driven by the high variability and subjectivity in internal models for operational risk, as well as challenges in validating these models.

Key Objectives of Changes to Internal Models in Basel IV

The reforms to internal models under Basel IV are designed to address several issues identified in the wake of the 2008 financial crisis and subsequent assessments of Basel III:

• Reduce variability in RWAs: Internal models often produced significantly different RWAs for similar exposures across banks, undermining comparability and confidence in capital ratios.


• Enhance credibility: By imposing floors and standardized approaches, Basel IV aims to ensure that capital requirements are not understated.


• Limit model risk: Over-reliance on complex models can lead to errors or manipulation, so Basel IV introduces constraints to balance model use with standardized methods.


• Improve transparency: Greater reliance on standardized approaches and mandatory disclosures makes it easier for regulators and stakeholders to understand and compare capital adequacy.

Challenges for Banks Using Internal Models

• Increased capital requirements: The output floor and parameter floors often result in higher RWAs, requiring banks to hold more capital.


• Operational burden: Revising or phasing out internal models (e.g., for operational risk) requires significant investment in systems, data, and processes to comply with new standardized approaches.


• Loss of competitive advantage: Banks that previously benefited from lower capital requirements through optimized internal models may face a level playing field with smaller or less sophisticated banks using standardized approaches.


• Regulatory scrutiny: Banks must meet stricter criteria for model approval and ongoing validation, increasing compliance costs.

Conclusion

Under Basel IV, internal models remain an important tool for sophisticated banks to calculate capital requirements, particularly for credit and market risk. However, their use is heavily constrained through parameter floors, output floors, and outright prohibitions (e.g., for operational risk). These changes reflect a shift toward standardization and conservatism in the Basel framework, aiming to strengthen the global banking system's resilience while addressing the limitations of internal models. Banks must adapt to these changes by recalibrating their models, enhancing data quality, and preparing for higher capital requirements. Implementation timelines and specific requirements may vary by jurisdiction, as national regulators (e.g., the European Union, U.S., or others) adopt Basel IV into local rules. For instance, the EU's CRR III/CRD VI and the U.S. Federal Reserve's proposals reflect tailored approaches to these global standards.

If you have a specific aspect of internal models or Basel IV (e.g., a particular risk type or jurisdiction) you'd like to explore further, let me know!