Dynamic portfolio selection under the predictability of asset return process
Yu Yun · Journal of Anhui University of Technology and Science · 2010
This paper presents a new method for dynamic portfolio selection based on the predictability of asset return process.By dividing the investment horizon equivalently,it introduces the return process of predictive variables into that of asset by modeling the risky asset weights in portfolio of each period as linear functions of the observation of predictive variables;it uses the expanded asset space to describe the asset return process through the matrix transformation,transfers the solution of conditional problem to that of unconditional problem,and applies the method of mean-variance analysis to obtain the coefficients of model,which maximizes the terminal wealth quadratic utility of each period;at last,it gives the algorithm of model and conducts an empirical study.Results show that predictive variables describe the asset return process by the expanded asset space;investor can rebalance portfolio according to the observation of predictive variables and makes.It optimal.It provides a method for the solution of conditional problem under dynamic setting.