Risk aggregation and Extreme Events [36]

Go to: Summary | Previous | Next   
Bullet points include: Practical ratings-based models generally employ Monte Carlo approaches But Monte Carlo merely a way of determining distributional form of some complex outcome and often very time consuming C.f. Analytical weighted Monte Carlo, see Kemp (2009) Analytical results often preferable (unless too ‘approximate’) Although not always how regulation is drafted Possible if Normally distributed latent variables driving default C.f. Gaussian copula / Large Homogeneous Portfolio

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile