2017-07-012024-05-17https://scholars.lib.ntu.edu.tw/handle/123456789/682065摘要：由於董事成員們為股東們的最佳權益肩負起受託者的職責，董事會在制定主要公司未來策略走向、政策決定以及經理人薪酬上扮演者非常重要的監督角色。在這個子計畫裡，我們希冀對於董事會社會連結與公司財務以及投資政策的關聯性進行探討。 歷年的文獻主要針對董事會結構、董事成員特性(如董事獨立性、男女比例等)以及如何該特性如何影響公司決策上進行討論。但鮮少議題是針對董事會成員社會網絡之經濟上的價值與意涵以及其對於公司經理人決策的影響進行分析。因此本計畫想針對這個社會連結議題做幾個不同方面的聯結與探討。第一，本計畫希冀討論董事社會連結對於公司資本支出以及創新研發來出發，並進一步了解董事會成員其監督與諮詢功能對於公司制定投資決策的影響。第二，透過董事成員與公司高階經理人之間的社會網絡，本計畫將討論這種董事與經理人的內部連結機制如何影響公司的現金政策。第三，本計畫將延伸分析此內部連結對於公司盈餘操控及舞弊的關聯性。借由這些議題來了解內部連結對於股東利益之正反面效果。當然，了解董事會對於公司政策影響之研究勢必伴隨可能的內生性問題。而這個內生性問題將嚴重影響研究本身的正確性與可信度。舉例來說，我們所關注的公司政策本身可能影響了社會網絡的多寡。或者是，由於聲譽的考量，資本較大的公司容易吸引更多的董事。因此，董事社會網絡可能是反映出資本較大的公司有較多的董事會成員。亦或是，當公司有更多的分析師報導或機構投資人持股時，也可能吸引更多董事加入。而當這些因素同時影響公司決策時，將帶來董事社會網路與公司決策之間的虛假效應。所以本計畫將先嘗試著使用差異中之差異法(difference-in-differences)透過自然實驗方式(natural experiment)所帶給董事社會網絡的外生性衝擊來探討問題。利用差異中之差異法其橫斷面及時間序列的比較方式去減低省略之時間變量趨勢以及不可觀察因素之異質性所產生之內生性影響。再者，本研究將透過傾向分數配對(propensity score matching)去比較兩組在其他公司特性上極度相似而於社會連絡的變數上確有差異的公司來進行假說檢驗。最後，本研究也將利用兩階段最小平方法(2SLS)以及隨機配對的方式來探討計畫議題。本研究希望藉由上述的方法進一步對董事會網絡與公司政策之間的因果關係做一深入了解。希冀此研究對於政府對於規範公司董事會組織上與增進董事監督機制層面帶來一定的參考作用並藉此優化金融體系的運作避免大型公司舞弊所帶來之系統性財務風險。 <br> Abstract: Board of directors plays fiduciary role in major corporate strategical planning, decision making, and managerial compensation and represents the best interest of shareholders. In this subproject, we would like to focus on the effect of board networking on corporate financial and investment policies. Previous literature has involved with numerous analyses on the effect of structure and characteristics of board members on corporate decision makings. Nevertheless, less has to be understood is the economic value of the board directors’ network and how such directors’ networking affects managerial decision makings. In this subproject, we would like to explore our understanding on the effect of board network on corporate investment such as capital spending and innovation, the social relation between board members and corporate executives on firm cash holdings policy as well as earnings manipulation and fraud. However, to study the real effect of board connections on corporate policies would encounter with the endogeneity issue that biases our estimation and produces spurious relationship between board network and corporate policies. For example, it could be that board network reflects the size of board given that larger board tends to be associated with more connections. There is also a possibility that large firms are more likely to attract directors with more connections. Additionally, firms with greater analyst following and institutional investors are possibly to have more connected directors. Thus, our first approach to mitigate this empirical challenge is to employ the difference-in-differences method using natural experiments as exogenous shocks on board network. The cross-sectional and time-series comparisons in this difference-in-differences design help to alleviate the concerns of omitted time and unobserved heterogeneities between two different sets of firms. Our second approach is based on the propensity score matching that compares the outcome of interest between two types of otherwise identical firms except board connections. This propensity score matching approach mitigate the concerns of confounding observable factors that might drive the proposed relation. To further establish the robustness of the study, we would also apply the two-stage least squares method and fixed effect method to substantiate the empirical soundness of the study. By using the aforementioned identification strategies, we intend to establish the causal inference in order to better understand these important topics.董事網絡差異中之差異法自然實驗方式directorsnetworkdifference-in-differencesnatural experiment