Currently, the examining of the determinants and implications of holdings of cash are mainly focused on corporate characteristics, industry characteristics and macro-economic factors, there is still much left to be examined about the features of the decision maker. Given the great impact of split-share structure reform on the behaviors of controlling shareholder, which is the main decision maker, their preference of liquid assets might be affected in the meantime. This thesis is to find evidence to support this deduction which we first posed.We examine the cash holding behavior of listing firms to find any difference of controlling shareholders’preference of accumulating cash caused by split-share structure reform. We constrain the concept of controlling shareholder in the level of direct shareholders, which we name as superficial ownership structure. Our research provides evidence both for the effect of split-share structure reform and the impact of agency problem on the behavior of cash holding.This thesis chooses the largest shareholder as a representative of controlling shareholder to examine the distinction of the impact of controlling shareholder on the cash holding level before and after split-share structure reform. There are three ownership variables, the ownership ratio of the largest shareholder, the character of the largest shareholder, and a dummy variable which measures whether the largest shareholder can control the firm. There are also two variables set to describe split-share structure reform, one is a dummy variable which measures whether a firm has finished split-share structure reform, and the other is set to measure the different time for finishing split-share structure reform. The result shows that the preference of cash assets of controlling shareholders in non-state owned firms has weakened obviously, but that of the state-owned ones has not change enough to be noticed. Our research focuses on788firms listed on either the Shanghai or Shenzhen stock exchanges from2004to2011which have finished split-share structure reform until2011, and in the meantime meet other certain conditions we set previously. Therefore, our pooled sample contains6304sets of observations of firms. The feature of the sample is that it contains the same788firms over the eight years that we observe.The method to carry out empirical research is to conduct comparative study. We set two standards to classify the observations, one is to sort the observations into two groups in terms of the character of the largest shareholder, and the other is to sort them in terms of time of finishing split-share structure reform. We conducted comparative analysis between specific groups and multiple linear regression analysis in each group. To verify our conclusions, we designed a multi-angle robust inspection and the results showed the high reliability of our conclusions. |