“…To solve the bias problems that may be caused by the omission of variables in the equation of the model, as much as possible, this study controls the effects of these variables. Ten control variables are selected: stock returns (Ret) (Wang G et al, 2021), liquid ratio (Lr) (Zhang et al, 2015), enterprise total profit growth ratio (Np) (Huo and Zhang, 2017), enterprise age (Age) (Carnahan et al, 2010), enterprise liabilities to assets (L/A) ratio (Lev) (Díaz-Fernández et al, 2015), enterprise total asset turnover ratio (Tato) (Edward and Marciano, 2019), the enterprise growth rate of net profit (Profit) (Gao et al, 2021), enterprise operating growth ratio (Revenue) (Fu and Shen, 2020), equity nature (State) (Li, 2011), and Tobin's q (Tq) (Lee et al, 2021) as control variables, and selects assets scale (Size) (Goll et al, 2008) and return on equity (Roe) (He et al, 2020) as replacement variables for the robustness test. Ret, Lr, Np, Lev, Tato, Profit, Revenue, Tq, Roe are collected from the financial annual reports of the enterprises.…”