1.The mean, standard deviation, skewness and kurtosis for each of the excess return variables are shown in Table 1. The skewness of RSOHU-RFR, SMB and HML are all greater than 0 which means that RSOHU-RFR, SMB and HML are all positively skewed. The skewness of RM-RFR is smaller than 0, which means that RM-RFR is negatively skewed. The kurtosis of RSOHU-RFR, RM-RFR, SMB and HML are all greater than 0, which means that the distributions of RSOHU-RFR, RM-RFR, SMB and HML are leptokurtic distributions. Based on the skewness and kurtosis, which are all not closed to zero, my research conclude that all the four variables do not appear normally distributed. Moreover, the Chi-test of RSOHU-RFR, SMB and HML are 56.2353, 42.5281 and 22313.4718, which are all greater than 23.6. The Chi-test of RSOHU-RFR, SMB and HML means that RSOHU-RFR, SMB and HML do not appear normally distributed. And the Chi-test of RM-RFR is 16.1203, which is smaller than 23.6. The Chi-test of RM-RFR means that RM appears normally distributed.
2. The correlation matrix for the four excess return variables is shown in Table 2. Most of the correlations are all positive, except the correlation between RSOHU-RFR and HML, which is -0.0357. Moreover, all the correlations are not very big. The absolute value of all the correlations are all smaller than 0.5.
3. The intercept of the CAPM model is 3.0111, which is greater than 0 and means there is an arbitrage opportunity. The coefficient of RM-RFR is 2.0544. Therefore, the relationship between RSOHU-RFR and RM-RFR is positive. When RM-RFR changes by 1 unit, the RSOHU-RFR will change 2.0544 units. Moreover, this coefficient of RM-RFR means that RSOHU-RFR is more volatile than market. The P-value of intercept is 0.0518, which is greater than 0.05. That means that the intercept is not significant. The P-value of RM-RFR is 3.1449E-09, which is much smaller than