Thesis (M.Sc.)--Chulalongkorn University, 2009

This paper establishes empirical evidence on in which method between setting market view based on single-factor model as in the Black-Litterman traditional paper or setting market view based on multi-factor model should be used. The results are as followings: In general, the results show that the multi-factor model using SMB and HML factor derived based on Black-Litterman model extension of Krishnan and Mains (2005) can capture more information of stock excess return than that of the traditional model which leads to less volatility in predicted return and less portfolio weight change from time to time. However, the multi-factor model of dividend-price ratio and earnings-price ratio can not be considered as an improvement of Black-Litterman traditional model because the predicted return derived from the model is much volatile. This leads to a significantly shift in portfolio weight which violates Black and Litterman’s initial intention to reduce portfolio weight shift in each period