Learning or Robust Control
Posted: 17 Dec 2007 Last revised: 14 Oct 2008
Date Written: 2007
Model mis-specification can cause substantial utility losses in portfolio planning. In this paper, we compare two approaches to cope with this problem, robust control and learning. We derive the optimal portfolio strategies and the utility losses due to model mis-specification. Surprisingly, neither learning nor robust control is uniformly superior to the naive approach where the investor simply ignores model risk. Furthermore, a comparison of the two approaches shows that learning takes some time to have an impact, so that short-term investors are in some (but not all) cases better off with robust control than with learning.
Keywords: Portfolio Selection, Model Uncertainty, Learning, Robust Control
JEL Classification: G11, G12
Suggested Citation: Suggested Citation