Quantile portfolio optimization under risk measure constraints (Q2441473)

The authors study the class of distortion risk measures which satisfy the properties of coherence, as well as the law invariance and comonotonicity. This class of measures is used to quantify the risk of the agent. At first, the dynamic market model is presented and the general form of the portfolio optimization problem is formulated. A distinctive characteristic of this formulation, versus the classics framework to study the problem of utility maximization, is that the decision variable is the quantile function of the terminal payoff. The main results start with the formulation of a general model of a portfolio selection with budget and risk constraints; the proof of the main theorem requires several technical results based in the analysis of the problem with the Lagrange multipliers. These results allow for constructing iteratively two sequences of parameters that converge to the Lagrange multipliers required to satisfy the budget and risk constraints simultaneously.