The Vasicek distribution derived in Bernoulli Mixture Models is not just a theoretical curiosity. It is the mathematical foundation of the Basel II Internal Ratings-Based (IRB) capital formula — the formula that determines minimum regulatory capital for credit portfolios in thousands of banks worldwide. This note makes that connection rigorous.
The key result is a two-step argument:
-
Conditional LLN (Proposition 11.17): In a Bernoulli mixture model with a common factor , the relative portfolio loss converges almost surely to the deterministic function conditional on each . This means idiosyncratic risk is fully diversified away in large portfolios — only systematic risk remains.
-
Quantile limit (Proposition 11.18): The -quantile of the portfolio loss converges to — the asymptotic loss function evaluated at the -quantile of the factor. For the one-factor Gaussian model this is exactly the Vasicek formula, and setting yields the Basel risk capital contribution per obligor.
The Basel IRB formula (QRM (11.37)). For obligor , the required regulatory risk capital is:
This lesson derives every term in (1) from first principles.
Learning objectives. By the end of this note you should be able to:
- state Propositions 11.17–11.18 and their key assumptions;
- explain how idiosyncratic risk diversifies in large portfolios while systematic risk does not;
- derive the Vasicek VaR formula for the homogeneous one-factor Gaussian model;
- trace each component of the Basel IRB formula back to the ASRF derivation;
- state the limitations of the IRB formula: ad-hoc , granularity, and one-factor restriction.
In this lesson
-
- The ASRF framework: formalising the asymptotics
-
- The one-factor Gaussian model and Vasicek VaR
-
- Derivation of the Basel IRB formula
-
- Limitations of the Basel IRB approach
Before you start
The full lesson, with worked examples and exercise solutions, is for members.
Read the full lessonNew to SelfMadeQuant.com? See membership options or browse the Quantitative Risk Management syllabus.