Abstract:
We believe that the traditional approach to risk management of using a single methodology for risk estimation and end of horizon risk measures based on a normal distribution, fails to capture the real risk of a portfolio. We believe that risk, by its very nature, needs to be estimated in a multi-dimensional manner, using parameters which are asymmetric, and capture the intra-horizon risk of an asset. We propose a risk structure which we believe will be superior in this regard. We further propose a framework to disentangle the return of a portfolio due to skill versus luck. We believe this is a critical factor in deciding if the future returns of an investment strategy are likely to persist.