Research On Fuzzy Portfoilo Selection Models With Transaction Costs And Various Investment Constraints

Taking return of assets as fuzzy number and then studing portfolio problem based on possibilitic theory provides a way for charactizing fuzzy uncertainty in the financial market. Zhang has already given the optimal investment strategy by using the upper and low possibilitic characeristics he had proposed. Most of possibilitic portfoliot selection models have to assume distribution functions of possibility return are known, however, it is not always easy to specify them, at this case, studing how to invest based on interval thoery can be an effective way. In the same opinion, K.K.Lai et al.proposed the minimax semi-absolute deviation interval portfolio selection model. However, the model either proposed by Zhang or by K.K.Lai is established on a series of strict assumptions,such as not existing transaction costs or dividing stocks freely, as a result,the effectiveness and application of these two models is decoupled greatly. In order to make them much closer to the actual case, this paper studies fuzzy portfolio problems with transaction costs and various investment constrains respectively. Firstly, this paper sets up upper and low possibilistic portfolio selection models with transaction costs by regarding returns of assets as triangle fuzzy numbers,establishs upper and low possibilistic portfolio selection models with a variety of constraints by takeing returns of assets as trapezoidal fuzzy number, then gradually realizes the linearization of these two kinds of nonlinear problem by use of possibilistic theory,method of replacing absolute value and removing the non-linear conditions,and gives examples of each.Secondly, this paper establishs interval potfolio selection models with transaction costs and various investment constraints respectively, then gradully converses the interval programming problems into linear programming problems with parameter by using interval theory,the common means of putting muti-objective problem into single-objective one, replacing the absolute value and removing non-linear conditions,and gives examples of each.