/

Stress Testing and Back-testing for UCITS Funds


This presentation explores stress testing and back testing for UCITS funds, including proposed regulatory guidelines, back testing issues (such as periodicity, look-back bias, selection effects and market-implied risk measurement). It also explores the different interpretations that can be placed on the term 'stress testing'.

Slides
1Stress Testing and Back-testing for UCITS Funds
2Agenda
3CESR Risk Management Principles - February 2009
4Basic principles
5Visualising likelihoods
6Back-testing
71. Back-testing is closely linked to calibration
82. Return behaviour depends on periodicity
93. Focus on entire distributional form or on just one quantile?
104. Does the model fit well 'period by period'?
115. Look-back bias, in sample and out-of-sample testing (1)
125. Look-back bias, in sample and out-of-sample testing (2)
13Selection effects: (1) Noise versus Meaning
14Selection effects: (2) Example data
15Selection effects: (3) Principal components analysis
16Selection effects: (4) Including 'meaning'
17Selection effects: (5) Can be very important!
18Market consistent risk measurement
19Back-testing in other regulatory frameworks
20Stress Testing
21Stress Testing - Additional drivers
22Different meanings given to the term 'Stress Testing'
23Reverse stress testing
24Stress testing - other observations
25Stress testing - statistical techniques
26Summary
27Important Information



NAVIGATION LINKS
Contents | Next | Library


Desktop view | Switch to Mobile