기대수익률의 추정에 의한 최적자산배분에 관한 연구 <평균-분산 모형과 평균-VAR 모형을 중심으로>

초록

In this paper we examine the optimal asset allocation both in the mean-variance and in the mean-VaR, and compare its differences using the simulation in Korean stock and bond market. we use monthly time series data set from August 2000 to July 2008. In the process, we also estimate expected return with three methods, that is, historical return, No view, specialists' view of major investment banks. Using the stock and bond portfolios, we maximizes mean-variance utility function or in mean-VaR utility function used by Alexander and Baptista(2002). Our simulation results show no significant difference between mean-variance and mean-VaR models in the optimal asset allocation. This paper indicates that mean-variance approach has usefulness in asset allocation though its limitation. Also, in terms of the optimal asset allocation we can't find significant risk-adjusted return on performance of asset allocation in using economists' market views of major investment banks.

키워드