Asset allocation using reliability method

This paper proposes a numerical method for the asset allocation problem based on the conventional Advanced First-Order Second Moment (AFOSM) reliability analysis. The proposed method separates the space of decision problems from the space of uncertain returns. By this separation, an uncertain asset allocation problem can be converted into two recursive optimization problems. One is defined in the space of random returns and the other in the space of decision variables. The proposed method can accept general objective functions like nonlinear utility functions, nonlinear loss functions or nonlinear transaction costs. An adaptive algorithm is designed to find the solution of the recursive models. The convergence of the algorithm is also proven. Numerical illustrations are included in support of the theory and Monte Carlo simulation is used to evaluate the quality of solutions.