The Empirical Test Of Beta Strategy

Since the 1960s,Sharp put forward the capital asset pricing model(CAPM)gradually.This model expresses the composition of assets in market for the first time,explaining the market using mathematical relationship between risk and return in investment.This marks the gradual formation of financial theory,which also means that modern finance analysis enters in a new period.Western empirical study shows that the capital asset pricing model does have a good ability to explain in the early stage,however,later empirical results show different results compared with previous ones.A series financial anomalies occur that cannot be explained by classical CAPM model.Based on the capital asset pricing model,some researchers focus on how to take action to get excess returns and some found that taking beta strategy which holds high beta portfolio and short low beta portfolio at the same time can obtain remarkable positive alpha in unconditional model,which means that beta investment strategy can get excess returns.This paper further improves the classical capital asset pricing model by introducing a lagged variable as information set which transfers a constant capital asset pricing model into a dynamic model.Moreover,it compares whether beta investment strategy in the condition model has better ability of explanation with unconditional model.Data is selected from 2000 to 2016 of Shanghai and Shenzhen stock exchange market about stock returns.Furthermore,according to the portfolio which are sorted based on beta from low to high,comparing the model with different unconditional variance,and try to further explain the reason about the difference between those two models.Studies have shown that under the condition model which using beta investment strategy,conditional model still can get excess returns.However,compare with unconditional model after contains lag information set,the relationship between excess returns and the market risk is more flatter,which means conditions model has a better explanatory power.Further research shows that the condition model of excess returns and market volatility has significantly positive correlation and market volatility can significantly explain the difference between conditional model and unconditional model.