臺大管理論叢 NTU Management Review VOL.30 NO.1

91 NTU Management Review Vol. 30 No. 1 Apr. 2020 individual CEOs, by looking into the proportion of independent directors and shareholding ratio of institutional investors, and explores the differences in behavior between professional CEOs and family member CEOs, in regards to salary reduction and performance recovery. 2. Design The duration of the sample lies between 1999 and 2004. The subjects are Taiwan Stock Exchange listed and over-the-counter listed companies. The raw data comes from the Taiwan Economic Journal and annual reports. A total of 3,107 observations are employed, containing 2,092 family businesses across 19 industries. This study defines a CEO salary reduction (“pay cut”) as an event where total annual compensation is less than that in the previous period. This study also defines the following variables: performance recovery index (ΔPerformance: ΔROA, ΔRET), relative industry performance index (mROA, mRET), the proportion of independent directors (IND), shareholding ratio of institutional investors (INS), company size (SIZE), debt ratio (LEV), market-to-book ratio (MB), industry type (Industry), deviation between control and cash flow rights (SEP), manager shareholding ratio (ManagerHold), abnormal remuneration received by CEOs (AbnormalPay), and a dummy variable that represents the research year (Year). Model 1: Probability( Paycuts it ) = ∝ 0 + ∝ 1 mROA it -1 + ∝ 2 mRET it -1 + ∝ 3 IND it + ∝ 4 INS it + ∝ 5 SIZE it + ∝ 6 LEV it + ∝ 7 MB it + ∝ 8 Industry it + ∝ 9 SEP it + ∝ 10 Manager Hold it + ∝ 11 AbnormalPay it + ∝ 12 Year + ε it Model 2: ∆ Performance it, it +1 = ß 0 + ß 1 Paycuts it + ß 2 SIZE it + ß 3 LEV it + ß 4 MB it + ß 5 Industry it + ß 6 SEP it + ß 7 ManagerHold it + ß 8 Year + ν it