International portfolio selection with Markov processes and liquidity constraints

This paper proposes an ex-post comparison of portfolio selection strategies applied to 64 stock exchange markets during the period 2008-2011. Assets are considered according to three different classes of liquidity constraints which differ for their average traded daily value. Moreover, in order to simplify the computational complexity of the choice we preselected assets among the global market considering different return characteristics: the joint Markovian behavior of the returns; their association with market stochastic bounds and their ex-ante reward-risk performance. For each liquidity class of assets we compare the ex-post performance of three portfolio selection strategies: one of them is based on the maximization of the Sharpe ratio; the other two are based on the maximization of a reward-risk performance measure that considers a Markovian evolution of the portfolio returns. The ex-post analysis shows that the reward-risk functionals allow better performance in terms of final wealth and demonstrates, with respect to liquidity classes, different features of portfolio composition in terms of assets nationality.