Nematrian
/

Stress Testing and Back-testing for UCITS [4]

Go to: Summary | Previous | Next   
Bullet points include: Risk depends on positions held and how they might behave in the future. Combines magnitude and likelihood. Risk is typically quantified using risk models. I.e. algorithms that take positional and other data and use them to derive risk measures that illuminate the risks being run. Usually ‘parameterised’, e.g. time-varying factors and covariance matrices. Magnitude relatively easy to compute (given accurate positional data). The real challenge is estimating likelihood. Back-testing tests (or calibrates?) chosen probability distribution. Stress testing places less emphasis on likelihood

NAVIGATION LINKS
Contents | Prev | Next | Library

Desktop view | Switch to Mobile