ORDERED WEIGHTED AVERAGINGAGGREGATION METHOD FOR PORTFOLIO SELECTION

Portfolio management is a typical decision making problem under incomplete,sometimes unknown, information. This paper considers the portfolio selection problemsunder a general setting of uncertain states without probability. The investor's preferenceis based on his optimum degree about the nature, and his attitude can be described by anOrdered Weighted Averaging Aggregation function. We construct the OWA portfolio selec-tion model, which is a nonlinear programming problem. The problem can be equivalentlytransformed into a mixed integer linear programming. A numerical example is given andthe solutions imply that the investor's strategies depend not only on his optimum degreebut also on his preference weight vector. The general game-theoretical portfolio selectionmethod, max-min method and competitive ratio method axe all the special settings of thismodel.

ORDERED WEIGHTED AVERAGING AGGREGATION METHOD FOR PORTFOLIO SELECTION

Portfolio management is a typical decision making problem under incomplete, sometimes unknown, information. This paper considers the portfolio selection problems under a general setting of uncertain states without probability. The investor's preference is based on his optimum degree about the nature, and his attitude can be described by an Ordered Weighted Averaging Aggregation function. We construct the OWA portfolio selection model, which is a nonlinear programming problem. The problem can be equivalently transformed into a mixed integer linear programming. A numerical example is given and the solutions imply that the investor's strategies depend not only on his optimum degree but also on his preference weight vector. The general game-theoretical portfolio selection method, max-min method and competitive ratio method are all the special settings of this model.