Optimal Portfolio Selection and Index Tracking for the Shipping Equity and Freight Rate Markets

Abstract

This paper reproduces the performance of two international shipping stock indexes and two physical shipping indexes by investing only in stock portfolios that our algorithms determine. In our analysis, we use daily stock data and address the index-tracking problem with the differential evolution and genetic algorithms. Our proposed portfolios are constructed by a subset of stocks from the two shipping indexes or the Dow Jones Composite Average and/or the FTSE 100 indexes. To test the performance of our heuristic, we examine three different scenarios: buy-and-hold, quarterly, and monthly rebalancing, and we account for transaction costs where necessary. Finally, to eliminate any data-snooping bias from our results, a reality check is performed on all competing portfolios. Our contribution to the shipping finance literature is twofold: for the first time, the index tracking problem in the shipping industry is addressed by employing the differential evolution and genetic algorithms; second, investors who do not have fully access to the two shipping stock indexes or the two physical shipping freight rate indexes, can replicate their performance by investing in the proposed portfolios.

Keywords: Index Tracking; Shipping, Differential Evolution and Genetic Algorithms