研究目的:建立新的公司治理指數,並探討其與公司績效及過去20年台灣公司治理發展的關係。研究設計/方法:在113個公司治理變數中以逐步迴歸篩選出18個主要變數,分為董監事會席次、股權結構、及人事穩定度三大構面。分析2003至2019年1,485家企業。研究結果:新建的公司治理指數及其構面皆與公司績效呈正向關聯。研究限制/啟發:單一構面指標效果較綜合公司治理指數弱,顯示多面向衡量的重要。未來研究可延伸金融保險業及不同經濟情境下的分析。理論/實務/社會意涵:本研究協助投資者進一步理解台灣企業公司治理並為政府提供評等依據,對學術及實務界均具貢獻。創見/價值:研究提供全新公司治理指數,在不同期間能解釋公司績效,奠定未來研究基礎。
Purpose - This study aims to enhance corporate governance awareness in Taiwanese enterprises by assessing and reconstructing a comprehensive index. It delves into the correlation between governance and key performance metrics like ROA and Tobin's Q and tracks the two-decade evolution of corporate governance in Taiwan. Design/methodology/approach - Utilizing the TEJ database, we selected 113 corporate governance variables. Through stepwise regression, 18 pivotal variables emerged and were grouped into three dimensions: board of directors' seats, ownership structure, and personnel stability. Our study spans 2003 to 2019, analyzing 1,485 companies. Findings - The formulated corporate governance index, along with its three dimensions, shows a robust positive association with corporate performance. The index aptly signifies the potency and pertinence of governance to performance. Research limitations/implications - While individual governance indices for each dimension are slightly subdued compared to the holistic index, it highlights the need for multi-faceted corporate governance metrics. The research omits the financial insurance sector, advocating for extended exploration of governance across diverse industries and economic settings. Practical implications/Social implications - By leveraging the financial crisis and Taiwan's governance policy execution, this research differentiates specific analysis sub-periods. It not only deepens investors' grasp of Taiwanese corporate governance but also offers a robust benchmark for government evaluations, benefiting both the academic and practical realms. Originality/value - Introducing a pioneering corporate governance index for Taiwanese firms, this study can elucidate performance variations over distinct periods. It underscores the constructed index's validity, paving the way for future research.