Stress testing / Liquidity and funding risk [25]

Go to: Summary | Previous | Next   
Bullet points include: Kwiatkowski and Rebonato suggest that risk managers should specify a set of conditional probabilities for a set of stress scenarios I.e. probability of X occurring given that Y has occurred More intuitive – think of a scenario such as Country X defaults and then easier to identify reliably the likelihood of other things happening given this scenario Impact of a given stress scenario then includes both direct impact plus additional impact associated with other events Weighted by probability of occurrence conditional on first event occurring, hence ‘Bayesian’ in nature Issue: specifying conditional probabilities may create inconsistencies

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile