To read the full version of this content please select one of the options below:

Estimating Value at Risk: A Subjective Approach

KEVIN DOWD (Professor of financial risk management in the Centre for Risk & Insurance Studies at the University of Nottingham)

Journal of Risk Finance

ISSN: 1526-5943

Article publication date: 1 March 2000


This article outlines a subjective approach to estimating value at risk (VaR) and its related confidence intervals based on priors of the profit/loss distribution and its parameters. In the tradition of Bayesian statistics, this pro‐duces probability density functions for VaR that allow for subjective uncertainty. The author shows that imple‐menting this approach can be intuitive, straightforward, and applicable to any parametric VaR. One of the more difficult issues in this area is how to assess the precision of estimates: VaR estimation is usually straightforward, but estimating a confidence interval for a VaR estimate is not. This article suggests that, by inferring VaR from prior beliefs, rather than thinking of VaR as dependent on an “objective” P/L distribution, interpreting estimated confidence intervals is less problematic


DOWD, K. (2000), "Estimating Value at Risk: A Subjective Approach", Journal of Risk Finance, Vol. 1 No. 4, pp. 43-46.




Copyright © 2000, MCB UP Limited