FINM 33420

Statistical Inference for Risk Management

Autumn Quarter
Instructor: Jostein Paulsen

This course is a continuation of FINM 33410: Probability for Risk Management I. The main purpose is to provide statistical tools to estimate and test models introduced in that class. The emphasis is on maximum likelihood estimation, including its asymptotic properties, but some nonparametrics is also covered. Further topics are goodness-of-fit tests and model selection.

Furthermore, a presentation of copulas is given, together with some measures of correlation. A copula is basically a method to produce dependence between a group of random variables, and it is shown how to Monte Carlo simulate from a copula, as well as how to estimate unknown parameters. The course ends with an introduction to extreme value theory.

Prerequisite: FINM 33410