Creating portfolio risk and return models [11]

Go to: Summary | Previous | Next   
Bullet points include: All rely on past data Models that explain a lot of past behaviour are by definition likely to be a good fit to the past Statistical models provide the best fit to past data Fundamental and econometric models in effect incorporate subjective views which it is to be hoped improve their likely future predictive ability Bigger differentiation is between historic and market-implied (aka market consistent) models C.f. ‘Real world’ versus risk neutral/market consistent economic models E.g. is “duration” a fundamental, econometric or statistical feature of a bond?

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile