Some aspects of active portfolio management

We propose a framework for considering active portfolio management, consistent with existing literature in the single period environment, but extended to the multi-period environment that most practicing active managers operate in. We derive several active portfolio management results including the Generalised Law of Active Portfolio Management (a generalisation that accounts for correlations and varying information coefficients and volatilities), and the Theory of Active Portfolio Management that suggests that mean/second-moment optimisation is the best active portfolio management approach in a multi-period setting, not mean/variance. In developing these results we uncover some inappropriate practices, and several simplistic practices that are theoretically supported by our framework.