Seminar: Matched Filter Approach to Portfolio Optimization

For NYU Community

Speaker: Prof. S. Unnikrishna Pillai



Portfolio optimization refers to selecting optimal capital distribution schemes for a set of financial assets such that the portfolio can outperform any other combination in terms of their measured risk and expected return.  Harry Markowitz, who pioneered the modern portfolio theory, devised a well-known portfolio optimization strategy whereby the portfolio's overall risk is minimized. This classic strategy provides a recipe for minimizing portfolio volatility; however it does not heed the overall expected return. This talk generalizes the Markowitz' strategy by simultaneously minimizing the overall risk of the portfolio while maximizing the expected portfolio return.  Interestingly, the solution to this portfolio optimization objective is realized by applying the classical matched filter design that is well known in electrical engineering.  In this context, the maximal expected utility strategy will also be reviewed in this talk where an investor's money utility is optimized. Simulation results will be presented for illustration purposes, with some additional remarks on quantifying "greed" using a risk-reward approach based on the famous "gambler's ruin problem."