Abstract
In this paper, we propose an online data-driven sliding window approach to solve a log-optimal portfolio problem. In contrast to many of the existing papers, this approach leads to a trading strategy with time-varying portfolio weights rather than fixed constant weights. We show, by conducting various empirical studies, that the approach with proper choice of window size possesses a superior trading performance to the classical log-optimal portfolio and mean-variance portfolio in the sense of having a higher cumulative rate of returns.