Reading 37: Risk Management －LOS h

In factor push analysis, a factor or factors are pushed to an extreme to examine the impact on the portfolio. In scenario analysis, the effect on the portfolio from simultaneous changes in several factors is examined, which provides several different scenarios. In maximum loss optimization, the risk factors that have the greatest potential impact on the portfolio are identified. Once the factors are identified, procedures are put in place to limit their impact. In worst-case scenario analysis, all factors are pushed to their most damaging impact on the portfolio. Factor push analysis, maximum loss optimization, and worst-case scenario analysis are all forms of stressing models.

In stylized scenarios, one or more risk factors are changed to measure their impact on the portfolio. Some forms of stylized scenarios are similar to industry standards. The risk factors mentioned in the question are from those specified by the Derivatives Policy Group. In factor push analysis, a factor or factors are pushed to an extreme to examine the impact on the portfolio. In maximum loss optimization, the risk factors that have the greatest potential impact on the portfolio are identified. Once the factors are identified, procedures are put in place to limit their impact. In worst-case scenario analysis, all factors are pushed to their most damaging impact on the portfolio.

The primary purpose of stress testing is to model the effect of non-normal events that may not be reflected in the typical VAR calculation. Thus it is unlikely that stress testing would only reflect normal events. Stress testing is susceptible, however, to the analysts intentional and unintentional misspecification of the model, the failure to examine the by-products of major factor movements (how does a change in one factor affect the value of another), and the failure to include the simultaneous adverse movements of risk factors.