Optimal Portfolio Allocation under Cumulative Prospect Theory

In this paper, a behavioral portfolio allocation model à la Cumulative Prospect Theory with loss control is numerically implemented. This optimisation problem is difficult to solve due to nonlinear objective function and nonlinear constrains. Moreover, the problem could be ill-posed. The structure of the optimal terminal wealth of this problem is a combination of three binary options whose settlements depend on the state of the market. The goal is to show the impact of each parameters (e.g. market parameters, investor's preferences parameters, initial gain or loss position) on the solution of this complex problem. Finally an application on real market data will compare the performances of the behavioral portfolio with the performances of an expected utility maximising portfolio.