A word of caution on calculating market-based minimum capital risk requirements

This paper demonstrates that the use of GARCH-type models for the calculation of minimum capital risk requirements (MCRRs) may lead to the production of inaccurate and therefore inecient capital requirements. We show that this inaccuracy stems from the fact that GARCH models typically overstate the degree of persistence in return volatility. A simple modi®cation to the model is found to improve the accuracy of MCRR estimates in both backand out-of-sample tests. Given that internal risk management models are currently in widespread usage in some parts of the world (most notably the USA), and will soon be permitted for EC banks and investment ®rms, we believe that our paper should serve as a valuable caution to risk management practitioners who are using, or intend to use this popular class of models. Ó 2000 Elsevier Science B.V. All rights reserved. JEL classi®cation: C14; C15; G13