Regulatory capital determination and Its implications for internal ratings-based credit risk model development and validation

Focusing on the interconnections between the Basel regulatory capital formula and several well-specified statistical models, this working paper seeks to understand some of the important issues embedded in the Basel Accord. These include: Where does this formula come from? What risks does it try to capture? Why does the Basel Accord stipulate that the formula be implemented on a basis of homogeneous segments for retail exposures or similar risk ratings of wholesale obligors? Is there any desirable property on the number of loans for a segment (or obligor group)? Why is LGD treated as a constant as opposed to a random variable? When covering expected loss – and determined independently – how is the loss reserve related to the minimum regulatory capital? Answers to these questions have some important implications for Basel model development and validation.