Mathematical models and a tabu search for the portfolio management problem in the Kuwait stock exchange

This article proposes two mathematical models to study and compare results of two cases of the portfolio selection problem in the Kuwait stock exchange (KSE) as an emerging market. The mathematical models attempt to balance the trade off between risk and return. Model-I maximises the expected return and maintains the risk to certain limits while Model-II minimises the portfolio correlation and constrains the expected return to a minimum acceptable level. Since, both models are turned out to be non-linear, a tailored tabu search algorithm is used to provide efficient solutions with reasonable amount of computational times. After testing the models using real data from KSE, the results indicated that Model-I is able to significantly beat the market using both quarterly and annual basis strategies.