Creating portfolio risk and return models [9]

Go to: Summary | Previous | Next   
Bullet points include: If risk model uses more than one factor model approach then usually hierarchy is: Fundamental factors Factor themselves may be identified using regression, likewise how to define factor (e.g. compare log size to straight size and see which fits the data better) Or using cluster analysis, see e.g. 3.8.4 of Kemp (2010) Industrial category (sector / group) treated usually as a 0 / 1indicator, except perhaps for  a clear conglomerate Econometric As industry categories may have been derived in part because of their perceived sensitivities to macro-factors there may be an element of circularity Statistical Identify any additional ‘blind’ factors (hope over time they become explainable?)

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile