We present the results of the structural break test in Table 3, which shows that structural changes occur around 1997 and 2007, bothof which are the beginning of important financial crisis events. It should be noted that there are two crisis event after July 1997, that is,the Asian crisis and the dot-com crisis. Both of them are typical crisis events that have significant impacts on the stock market. The firstbreak point in July 1997 shown by the BP test implies that these two crisis events can be viewed together and start from July 1997, ina statistical sense. After April 2007, the global financial crisis began and it was also followed by the European sovereign debt crisis.Therefore, we extend the model by adjusting the dynamic changes in parameter m and θ to measure the impact of the investorsentiment index during different periods. Specifically, considering the structural changes occurred in July 1997 and April 2007, weincorporate dummy variables for both “intercept” and “slope” to adjust the structural changes. The inclusion of “intercept” and “slope”dummies allows us to investigate the influence of investor sentiment on the long-run correlation more comprehensively. For thispurpose, we incorporate dummy variables to adjust the structural changes of parameter m and θ. In regard to formulae (11) and (12), weset D1 equal to 1 when t is in the period between July 1, 1997, to April 2, 2007, and 0 otherwise. In addition, we set D2 ¼ 1 when t is laterthan April 2, 2007, and 0 otherwise.