動態資本結構調整、公司治理、與財務危機關聯性之實證分析

Abstract

過去文獻對於公司資本結構調整行為之研究不勝枚舉，然則大部分文獻在實證上皆只著重在探究財務健全公司的調整行為。過去學者已指出槓桿高低對於公司面臨的財務危機風險有重大影響，故相較於正常公司，資本結構之調整過程對於財務危機公司而言更是不可忽視。有鑑於此，本文以台灣上市櫃和曾經上市櫃公司為研究對象，試圖分別針對財務危機公司和財務健全公司以Ozkan(2001)所提出的動態資本結構調整模型佐以廣義動差法(Generalized Method of Moments；GMM)進行資本結構調整行為之比較，同時也將公司偏離最適槓桿比率的程度引入財務危機預警模型之建構，藉以瞭解公司偏離最適槓桿的程度和危機風險之關係。 本文實證結果主要有四點發現：1. 財務危機公司在總負債的調整上不僅不會朝最適槓桿比率調整，反而會加速偏離最適槓桿比率；健全公司則確實具有朝向最適槓桿比率調整之行為。2. 財務危機公司在危機發生前普遍存在長期負債使用過度之情況，故在長期負債的調整上，危機公司為了降低破產風險，會以償還過多長期負債為目標，因此在長期負債方面仍會具有朝最適槓桿比率調整之行為，但調整速度比健全公司相對較慢；在流動負債調整上，危機公司在自有資金不足及發行權益成本相對較高之情況下，會採用舉流動負債還長期負債之方式來調整過多的長期負債，因此流動負債的調整上將會加速偏離最適槓桿比率。3. 財務危機預警模型建構之結果顯示公司偏離最適槓桿程度和危機風險存在顯著正相關，代表公司債務之使用若超過最適水準將會提高公司所面臨的破產風險，此一結果符合靜態抵換理論(Static Trade-off Theory)之論述。4. 結合內部公司治理變數至財務危機預警模型建構之實證結果顯示加入公司治理變數可提升模型的解釋能力且能降低模型預測分類之誤差率。其中董監事持股率和大股東持股率和危機風險呈顯著負相關，代表當董事會的持股率以及股權集中程度越高，都將使董監事和大股東與公司利益目標越趨一致，進而提升監督意願，降低危機發生之風險。同時董監質押比則和危機風險為顯著正相關，越高的董監質押比就反應董監事對公司未來越不看好，且質押行為也會扭曲公司財務結構，進而提升公司財務危機之風險。

Most previous researches study the capital adjustment behaviors only for firms that are financial health. Scholars in the past have showed that level of leverage is highly correlated with distress risk. This means that the capital adjustment issue is much more important for distress firms. This paper uses Taiwan firms’ data and the GMM estimation technique to study the differences of dynamic capital structure adjustment behaviors between financial health and financial distress companies. We also derive the capital structure deviation level and combine it with the financial distress prediction model to study the effect of deviation to the distress risk. Our main findings are listed below. First, distress firms do not adjust their total debt to target level. On the contrary, they deviate from the optimal value with a fast speed. Not like the distress firms, financial health firms do adjust their total debt to target leverage. Second, overuse of long-term debt is a common characteristic for distress firms. For these firms, we discover a phenomenon that distress companies tend to use current debt to refund the long-term debt in order to eliminate the bankruptcy risk before the distress events. An appropriate explanation for this phenomenon is that firms which fall into distress usually don’t have enough own capital, and they can only take a higher cost of capital to issue new equity or bond securities. For that reason, the most possible way for these companies to eliminate long-term debt is through the issue of new current debt. This means that no mater firms are in danger or not, they will adjust their long-term debt levels to the optimal value. But for current debt, distress companies will chose to omit the optimal value and use more and more current debt for the purpose listing above. Third, the result of combining the deviation level with financial distress prediction model indicates that there exists a significant positive relationship between the bankrupt risk and the capital structure deviation level. This means that the overuse of leverage would raise the risk to bankrupt. This is consistent with the trade-off theory. Fourth, the result of building financial distress prediction model shows that the model with internal corporate governance variables has better ability to discriminate financial distress corporations form healthy ones. The ownerships for big stockholders, directors, and supervisors are negative correlated with the distress risk significantly, and there is a positive relationship between pledge ratio and distress risk with statistic significance.