These results stand somewhat in contrast to the findings in Jorion (2002), who concludes that value-at-risk models are good predictors of future trading revenue variability.
(5.) See Hendricks and Hirtle (1997) for a discussion of the rationale behind the use of value-at-risk models for regulatory capital requirements and the choice of supervisory parameters specified in the capital standards.
"How Accurate Are Value-at-Risk Models at Commercial Banks?" JOURNAL OF FINANCE 57, no.
Banks whose value-at-risk models incorporate specific risk can use the specific risk estimates generated by their models.(12) Under the most recent announcement by the Basle Committee on Banking Supervision (1997), these model-based specific risk estimates are subject to a scaling factor of four until market practice evolves and banks can demonstrate that their models of specific risk adequately address both idiosyncratic risks and "event risks" that might not be captured in a value-at-risk model.(13) This provision holds out the prospect of harmonizing the specific risk capital requirements fully with the general market risk requirements as market practices with respect to positions subject to significant event risks become clearer.
More important, inaccurate value-at-risk models or models that do not produce consistent estimates over time will undercut the main benefit of a models-based capital requirement: the closer tie between capital requirements and true risk exposures.
The discussion of value-at-risk models in this paper might suggest that supervisory evaluation of banks' internal models is a daunting task, necessitating the hiring of large numbers of new staff with the same degree of technical and market expertise as the bank personnel responsible for developing and implementing the models.
Clearly, the use of value-at-risk models is increasing, but how, well do they perform in practice?
Although this article considers value-at-risk models only in the context of market risk, the methodology is fairly general and could in theory address any source of risk that leads to a decline in market values.
The two most important components of value-at-risk models are the length of time over which market risk is to be measured and the confidence level at which market risk is measured.
Thus, fewer historical data are available to aid in model calibration, and the modeling process itself is more complex, as the distribution of credit losses is quite asymmetric with a long right-hand tail.(27) Financial institutions have made considerable progress over the past two or three years in credit risk modeling, but it is fair to say that these models are at an earlier stage of development than the value-at-risk models used for market risk assessment.(28)
These factors have tended to be incorporated into value-at-risk models after the initial phases of model development.
Given the inherent limitations of value-at-risk models
, Rahl agreed with Sabatacakis that stress testing and scenario analysis are key to rounding out the picture of a portfolio's risk.