Dymanic Portfolio Choice with Predictable Returns and Transastion Costs

Specialeforsvar ved Peter Lundemand Kristiansen

Titel:  Dynamic Portfolio Choice with Predictable Returns and Transaction Costs

Abstract:  Active investors main goal is to deliver the best risk-adjusted returns by choosing the optimal portfolio. In order to do so, they must base their choice on the current risks, expected returns, transaction costs, and the future evolution of the risks and returns. Standard single-period mean-variance portfolio optimization does not acknowledge that investors trade repeatedly over time and incur transaction costs in the process. These issues can be handled in a more sophisticated multi-period dynamic model presented in [Gârleanu and Pedersen, 2013] which we refer to as the model. This thesis derives the closed-form optimal dynamic portfolio policy and associated main theoretical results addressed in [Gârleanu and Pedersen, 2013]. The optimal dynamic policy is characterized by two principles: (1) aim in front of the target, and (2) trade partially toward the current aim. The aim portfolio is a weighted average of the current and future Markowitz portfolios and depends on the mean-reversion of the return predicting factors. Predictors with slower mean- version get more weight in the aim portfolio. Furthermore, we implement the model in Matlab and illustrate the performance on historical data for commodity futures and index futures by doing in-sample and out-of-sample tests. In the out-of-sample tests we find evidence that the model is too aggressive in building up large positions in response to the increasing factor levels. We show that it is possible to improve the net performance for the optimal strategy by adding constraints and thereby make the optimal portfolio less aggressive in building up large positions. Overall, for both in-sample and out-of-sample tests, we find that the optimal strategy, in the majority of the tests, has better net performance compared to benchmarks

Vejleder:  David G Skovmand
Censor:   Cathrine Jacobsen, ATP