/


Creating portfolio risk and return models [13]

Go to: Summary | Previous | Next   
Bullet points include: Using traditional fundamental and/or econometric factor modelling isn’t the only way of seeking explanation or meaning: PCA in effect focuses just on magnitude of contribution to variance Trace of covariance matrix (i.e. sum of the variances of each security in the universe) equals the sum of its eigenvalues (N.B. is equal weighting best?) The most important PCA components may therefore just be (larger magnitude) random noiseUsually when asked to explain how something works, we expect the answers (i.e. ‘drivers’) to be ‘causative’ or ‘informative’, like extracting radio signals from background noise Is it possible instead to focus on meaningfulness? C.f. also next session

NAVIGATION LINKS
Contents | Prev | Next | ERM Lecture Series


Desktop view | Switch to Mobile