19 Pages Posted: 22 Aug 2010
Date Written: July 19, 2010
Financial crisis risk is now firmly in the spotlight after a turbulent past quarter century bore witness to a number of market events previously thought to be once-in-a-lifetime occurrences. The apparent inability of risk models to deal with real world financial markets produced a whole line of popular books decrying unrealistic assumptions behind the extended CAPM-type models, especially the bell curve assumption. We closely examine the old paradigm and present evidence which clearly shows that this focus on the distributional axioms is misguided and that the main deficiency of the old paradigm lies elsewhere, namely in equating risk with presently observed volatility. We present a new approach to forecasting risks and thoroughly examine its performance and implications over a number of crisis events in the past 25 years, while comparing it to traditional methods. Our approach, which is based on the assessment of risk-taking behavior of market participants in the endogenous risk framework, dominates alternative methods of risk estimation, while still allowing finance professionals to utilize familiar risk metrics like Tracking Error and Value-at-Risk.
Suggested Citation: Suggested Citation