Foundation ERM Session 8: Extreme Events and Extreme Value Theory (EVT)

This presentation is based on a part of an academic course on Enterprise Risk Management (ERM) titled ‘Extreme Events and Extreme Value Theory (EVT)’ and covers topics such as: what do we mean by extreme events and why are they relevant in a financial context, Extreme Value Theory (EVT) and a case study on wind storm losses using EVT

1Session 8: Extreme Events and Extreme Value Theory (EVT)
2Session 8: Extreme Events and Extreme Value Theory (EVT)
3Extreme events
4Analysing fat-tailed behaviour
5Anchoring versus experience
6Example QQ-plot (versus Normal)
7Many (most?) investment return series are ‘fat-tailed’
8More periods give more scope for extreme events
9Skew(ness), kurtosis and Cornish-Fisher
10Flaws in Cornish Fisher (and hence skew/kurtosis)
11Joint fat-tailed behaviour
12What causes fat-tailed behaviour?
13Time-varying volatility
14Explains some market index fat tails, esp. on upside
15A longer term phenomenon too
16Crowded trades and selection effects
17Session 8: Extreme Events and Extreme Value Theory (EVT)
18Extreme Value Theory (EVT)
19In case study we will focus on
20Extreme value theory (EVT) results
21Block maxima results
22Generalised extreme value (GEV) distribution
23Limiting behaviour
24Main result for threshold exceedances
25Potential weaknesses
26Using EVT to Estimate VaRs
27Maximum likelihood (ML) estimation of tails
28Estimate the quantile that starts the tail
29Combined estimate
30Choosing the threshold
31Judging the start of the tail
32Session 8: Extreme Events and Extreme Value Theory (EVT)
33Estimated loss quantiles in SEK m
34Other analyses included in case study
35Application to ERM
36Important Information

Contents | Next | ERM Lecture Series

Desktop view | Switch to Mobile