POW! Empirical: Alternative and model-free risk/return analysis

POW! Empirical extends POW! Risk, by-passing the covariance matrix and the distributional assumptions typically associated with it, and directly accessing empirical data; where POW! Risk uses an "ex ante" model, POW! Empirical can be thought of as calculating "ex post".

With POW! Empirical you can

• estimate the same statistics as in POW! Risk, but on an empirical basis. For example, the 95% confidence Value at Risk will correspond to the actual 95th percentile over the period you have selected, and not that calculated from the normal distribution
• calculate draw-downs and run-ups, downside risk and other forms of semi-variance
• verify the level of skewness, kurtosis and autocorrelation
• create simulated portfolio returns over arbitrary periods, with or without bootstrapping