Optimal Convergence Trading for Cross-listed Stocks

Abstract

With the intention of maximizing an investor's terminal utility, we construct a non-threshold ased trading model within which the optimal trading weights for daily rebalancing are derived analytically via stochastic optimal control. Having released the constraint that the cointegrating vector is equal to one, we propose a more practical trading strategy that applies to a much wider range of categories of cointegrated assets. We explore extensive out-of-the-sample experiments on the cross-listed stock portfolios, facilitating comparative studies among Chinese and European, UK and US stock markets. We further test the time-delay arbitrage of the strategy using the cross-listed stocks by employing two parallel trading mechanisms, respectively equity-based contracts for difference (CFD) and real shares trading. Our empirical results illustrate that the time-delay arbitrage of the cross-listed stocks strategy based on the analytical solution of weights yields relatively stable and better performance than that of the home market index. Our research is instructive for the practitioner's trading decision in cross-listed stocks and other kinds of convergence investment.