Portfolio Selection Model Theory And Its Applications

This paper describes the modern theory of portfolio selection process of d-evelopment, analysis the different methods and models and the differences bet-ween them. We introduce the model of the mean-variance and mean-VaR. w-hen the singular-r covariance matrix is positive or definite, we discuss the gen-eral expression of analytical solutions of the mean-variance model, by comp-aring the definite singular covariance matrix and positive covariance matrix, w e discuss the problems of the mean-variance efficient frontier and the effective combination. In addition, using the generalized inverse matrix method, Jiang c-hun Fu, Daiyong Long (2008) study the efficient frontier and efficient analytic-al solution of the combination, but some of the conclusions did not complete. we not only appropriately amendment them and give a integrated discussion. with covariance matrix linear subspace generated by all relations, it gives the e-xact combination of cutting-edge expression.With the Markowitz's mean-variance model, further study the model of mean-VaR, when the covariance matrix is singular, we discuss the optimal sol-ution of mean-VaR, and comparison with the classical mean-variance, the-n we found that only appropriately choise the confidence level, in this case th-e optimal solution of mean-VaR model exists.Finally, under the of normal distribution, comparing with the model of mea n-variance and mean-VaR model's risk measure criteria, we acquire effecttive p-ortfolios and the efficient portfolio frontier.