上市公司董事長、總經理之性別如何影響企業的風險性及股利政策

Faccio, Marchica& Mura (2016)使用平均數差異檢定針對1999-2009年間歐洲企業CEO之性別如何影響公司之負債比率，得出負相關的結論；而Chen, Leung, & Goergen (2017)則針對美國S&P 1500指數1997-2011間企業進行研究，探討獨立董事之性別如何影響公司之股利發放政策。本篇論文結合兩者，同時針對董事長/總經理進行研究，並且加入八項實驗應變數，將領導者性別對公司的影響做一個更全面的討論，於此同時補足台灣內對此議題之研究缺口。本篇論文研究2006年至2017年間，共12年，台灣上市公司之董事長及總經理的性別，如何影響企業之負債比率、付息負債比率、現金比率、ROA波動度、Beta值、投資現金流量比率、股利發放率及營收成長率。這分別代表了公司之財務風險、營運風險、市場風險以及股利發放率。本研究採用平均數差異檢定實證董事長/總經理性別對八大實驗應變數的影響。並且使用複迴歸分析強健我們的結果。面對可能的內生性問題，本論文採用傾向分數配對以及差異中之差異法 (Difference In Difference)處理，並且得到一致的結果。最後本研究得出實證，在女性擔任董事長/總經理治理下的公司有較低的財務風險，較高的投資保守度以及較高的股利發放率。其中又以高投資保守度與高股利發放率結果最為強健。因此我們得出結論：女性董事長、總經理傾向使用市場投資人所謂的「鴿派風格」治理公司。。另外，將因果倒置後，我們也操作羅吉斯迴歸試圖探討具備哪些特性的東司有較高的傾向進行更換董事長、總經理性別由男至女的決定。在倒入30+公司治理變數進行探索性研究後，我們發現與配對樣本相比，公司在營業成本上升幅度較大、速動比率上升幅度較大時，較易出現更換女性董總的傾向。

關鍵字

性別轉換；公司治理；平均數差異T檢定；迴歸分析；風險；股利政策；鴿派治理；董事會；總經理

並列摘要

Faccio, Marchica& Mura (2016) extend the literature on how managerial traits relate to corporate choices by documenting that firms run by female CEOs have lower leverage ratio than otherwise similar firms run by male CEOs based on the data set of European companies in 1999-2009. Chen, Leung, & Goergen (2017), on the other hand, discover that companies with higher female independent director ratio tend to offer more dividend based on the dataset of S&P 1500 companies in 1997-2011. Our research combines the methodology of two of them by investigating how female chairman/president affect companies’ financial risk, operational risk, market risk and dividend payout ratio of Taiwanese companies. Our sample includes the data of all publicly traded companies in Taiwan from 1 January 2006 to 31 December 2017. Our methodology is as below: First, by the T-Test for Difference Between Means method, we discover that our hypothesis is proven. The companies with female CEO/Chairman tend to have higher currency ratio, lower investment outflow and higher dividend payout ratio compare to themselves before having the female chairman/president. Second, we use the Difference in Difference method to deal with the potential endogeneity issue. By apply the T test on our test group and comparison group which is selected by Propensity Score Matching method, we have the consistent result. Third, we do the robust test by using the regression analysis, trying to make our result more solid. The result of the regression rest proves our hypothesis once again, especially in two variables: Lower investment and higher dividend payout ratio. Our conclusion suggests that companies run by female are more “dovish.”