Universal portfolio generated by Hellinger distance

¹ Department of Mathematical and Actuarial Sciences, Lee Kong Chian Faculty of Engineering and Science, Universiti Tunku Abdul Rahman, Malaysia
² Department of Finance, Faculty of Accountancy and Management, Universiti Tunku Abdul Rahman, Malaysia

Abstract

There are numerous universal portfolios generated in different studies using various divergence functions to achieve one goal, which is to maximize wealth. To extend the exploration, through this study, we have generated a new universal portfolio using the Hellinger distance. We conducted a comprehensive performance evaluation of our newly developed universal portfolio against common strategies such as the buy-and-hold strategy and the constant rebalanced portfolio. This evaluation used diverse stock price data from the Bursa Malaysia local platform for trading stocks, bonds, and other securities. Our empirical findings demonstrate that the Hellinger distance-based universal portfolio outperformed the traditional strategies in two out of three datasets over a period of 535 trading days. Specifically, the Hellinger portfolio shows a significant improvement in wealth accumulation, with data set X achieving the highest wealth increase and exhibiting nearly 99% allocation to the most profitable stock. These results highlight the potential of the Hellinger distance-based approach in improving the performance of universal portfolios and offer a promising tool for investment decision-making in today’s complex and unpredictable financial markets.