Comparative alternative approaches on portfolio selection problem

In this study, Quadratic Programming, Goal Programming and Linear Programming all of which are Mathematical Programming methods applied to stock data, which companies take part in ISE 30 index between January 2000 and October 2006 period. We developed alternative models by financial theory. In a period of six years we analized that permanent twenty stock datas in ISE 30 index. The period between November 2006 and July 2007 is the test period in this study. The model is static. We compared the rate of exchange in yields of portfolio through all of models and come true in the test period. It is suggested that which alternative investment models are more profitable according to strategies of an investor who wants to invest by assestment the results found by the analysis of portfolio models.We realized highest income for some alternatives of goal programming in the result of analysis. Linear programming is agreeable for investor who accept a risk. Goal programming is agreeable for investor who has not limit of budget. Quadratic programming is also agreeable for investor who doesn?t accept a risk.Investors have to determine the model and objectives. If the objectives can?t be determined correctly, investors will be exposed to a loss. It is concluded that an investor should choice optimal mathematical programming method for himself to achieve purpose.