What Can Go Wrong in a Predictive Stress Test? - Finding Solutions in Best Practices

Given that a predictive stress test extrapolates based on history, it is important to be aware of the degree of extrapolation when choosing an applied shock. A quantitative measure that can inform the shock size is the standard deviation of the combined independent factors.

In this webinar we explore the relationship between the MSCI USA Index and the US Government Curve. We show how stressing one, two and three independent variables affects the outcome of our predictive scenario. Additionally, we provide insights on plausible shock magnitudes one may wish to employ using respective pricing histories of our selected assets. Finally, a set of useful diagnostics is offered and explained in detail.