Robust trade-off portfolio selection (Q2218875)

The authors analyze the trade-off between two robust criteria in mean-variance portfolio optimization when there is parameter uncertainty: maximin utility and minimax regret. This approach is motivated by the observation that the portfolio chosen by either criterion may degenerate to one of the most uninformative portfolios. More precisely, while optimizing for the worst-case utility may yield an overly pessimistic portfolio, optimizing for the worst-case regret may result in a complete loss of robustness; it is demonstrated in the paper. Thus, seeking a trade-off between the two portfolios is sometimes necessary, not only desirable. On this way, it is proved that, under a widely used ellipsoidal uncertainty model, the entire optimal trade-off curve can be found via solving a series of semidefinite programs (SDPs). Then the model is extended to handle the union of finitely many ellipsoids, and it is shown that trade-off analysis under this quite general uncertainty model also reduces to a series of SDPs. For more general uncertainties, an iterative algorithm is proposed based on the cutting-set method.