Creating portfolio risk and return models [52]

Go to: Summary | Previous | Next   
Bullet points include: Any sufficiently flexible model (i.e. one with sufficiently many model parameters) can be made to fit the past arbitrarily well Typically seek models that fit well but are also ‘parsimonious’, i.e. have relatively few free parameters There are various statistical methodologies that trade off parsimony versus goodness of fit e.g. Akaike Information Criterion

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile