本研究以股權所有權與公司績效作為研究標的，實施期間為2006年至2016年，使用資料為天下雜誌「2000大調查」中三大產業裡公司規模為最大之企業，分別為鴻海、大聯大及國泰金。首先本研究以單根檢定檢驗研究資料是否呈現定態的時間序列，所運用之單根檢定包含PP、KPSS與NP單根檢定法。結果發現股權所有權及公司績效資料皆為I(1)數列。 接著利用Enders and Granger(1998)門檻自我迴歸模型(TAR)以及動差門檻自我迴歸模型(M-TAR)進行門檻共整合檢定，再進一步利用Enders and Granger (1998)及Enders and Siklos(2001)門檻誤差修正模型(TECM)，研究股權所有權與公司績效之間是否存在著長短期非線性因果關係。結果發現當以公司績效為被解釋變數時，不同產業的董事持股比率對公司績效間均有長期因果關係存在；另外當以持股比率為被解釋變數時，公司績效對大多數董事及大股東持股比率皆不存在有長期因果關係，但以鴻海董事及大聯大大股東持股比率為例外。 This paper’s purpose is investigate the asymmetric causal relationship between ownership and corporate performance.The research material is from「CommonWealth」magazine from March 2006 to Decemer 2016.At first, our research employs PP, KPSS and NP unit root test to make sure what type the time series is. After, make use of threshold error- correction model (TECM) studied by Enders and Siklos (2001) to confirm the long-run equilibrium between ownership and corporate performance. This paper finds that no long-run causal relationship exists between ownership and corporate performance except 鴻海board’s shareholding ratio and 大聯大majority shareholder’s shareholding ratio are dependent variable when ownership is dependent variable. This paper also finds that long-run causal relationship exists between ownership and corporate performance when board’s shareholding ratio is variable and corporate performance is dependent variable.
