A fuzzifying approach to portfolio optimization | |
| Author | Komgrit Leksakul |
| Call Number | AIT Thesis no. ISE-98-11 |
| Subject(s) | Fuzzy sets Portfolio management |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Advanced Technologies |
| Publisher | Asian Institute of Technology |
| Abstract | In this research, we first propose nonlinear programming formulation to determine the portfolio selection for multi period portfolio management, following the tradition portfolio theory. Since, this mathematical model is nonlinear programming so the solution approach for guaranteeing the global optimal solution is not available. In addition, this model is not suitable for using extensively in its original form to construct a large-scale portfolio. One of the most significant reasons behind this is the computational difficulty associated with solving a large-scale of our original model with a dense covariance matrix. Then, we propose to use the mean absolute deviation risk function, in a new linear mathematical model to replace the mean/variance risk function. In portfolio model, uncertainty normally exists, attributed by human perception of events. Such uncertain elements include budget, demand, and risk function. Then the corresponding fuzzy elements are fuzzified into fuzzy models, which are aggregated using Min operator and Max/Min operator. Using hypothetical data, we compare the total return obtained from the original models use of the fuzzy models. It is found that the total return from the former is substantially higher. We observe that the rigid requirements in the original models results in an unreali stic optimal solution, while the fuzzy ones seek to realize a desirable solution by relaxing some resource restrictions. This phenomenon is typical for decision-making behavior in portfo lio management. Furthermore, the Max/Min operator provides better solutions than Min operator, in term of total return and resources allocation. Consequently, we recommend the fuzzy portfol io model with mean absolute deviation risk fuzzified by the Max/Min operator for achieving a better solution. |
| Year | 1998 |
| Type | Thesis |
| School | School of Advanced Technologies |
| Department | Other Field of Studies (No Department) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Anulark Pinnoi; |
| Examination Committee(s) | Voratas Kachitvichyanukul;Fujiware, Okitsugu; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1998 |