Stress testing / Liquidity and funding risk [16]

Go to: Summary | Previous | Next   
Bullet points include: Could judge plausibility by whether stress lies within following domain of admissibility, where choose a confidence level, e.g. p=0.95, and set k2 so that n(k2)=p:  Defining ‘difference’ between distributions in this way is called using a ‘Mahalanobis’ measure But what should we use for V? Relying on V used in VaR model may endow it with too much credibility A more heuristic V?

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile