| Summary: | 碩士 === 銘傳大學 === 管理學院高階經理碩士學程 === 93 === This study was based on 1998 - 2003 sample data from ten stock-listed department stores in Taiwan. To avoid bias caused by a single equation model, this analysis was constructed by a simultaneous-equation model with three equations. The board structure, financial policy, and corporate were major parameters for three equations.
Because most of the sample variables are time series data, first we used the method of Augmented Dickey-Fuller (ADF) to check whether the data are stationary, and Akaike Information Criterion (AIC) to find out the best lags of variables. The stepwise regression and the method of variance inflation (VIF) then were applied to select the suitable and no multicollinearity variables in the equations.
The empirical results employed by three-stage least equations (3SLS) show as the followings.
1. The relationship between board structure and corporate value are: (1) the most positive significant factor of board structure to corporate value is overall board ownership; (2) the most negative significant factors are overall board size, insider board size, institution board size, outsider board ownership and board ownership hypothecation; (3) insider board ownership plays a negative significant role in explaining corporate value in the same time period, but positive in one time lag of three months.
2. The relationship between board structure and financial policies are: (1) the most positive significant factors of board structure to investment policy are outsider board ownership and board ownership hypothecation, and the most negative is board ownership; (2) the most positive significant factors of board structure to debt policy are insider board ownership and chairman as general manager, and the negative factors are insider board size, institution board size, outsider board ownership and institution board ownership; (3) board structure is not a significant factor to dividend policy.
|
|---|
