總體經濟因素與企業資本結構關聯性之研究—台灣上市電子公司為例

Abstract

資本結構長期受業界與學界的討論與研究，其對於公司營運的資金來源及財務體質亦影響深遠。過去研究多為著墨於各別公司變數對資本結構的影響，而總體經濟對於資本結構影響的探討則相對為較後期的研究。然而這些研究中，多只探討當期或前一期總體經濟因素對資本結構之影響，但本研究認為一方面總體經濟因素可能會有遞延影響資本結構的現象，另一方面，企業的經營者亦有可能以較前瞻的方式，針對未來的總體經濟狀況做出資本結構的調整。因此，本研究希望了解跨期性的總體經濟因子，是否確實對於資本結構有影響性，並是否可藉由加入跨期性的總體經濟因子，提升迴歸模型的解釋力。

著眼於台灣高科技產業本身的特質與台灣政府推動高科技產業的相關政策，都使台灣高科技公司的資本結構有其獨特性，故希望針對台灣電子工業的資本結構加以研究。因此，本研究以台灣上市電子公司為樣本，採跨期性的商業本票利率、狹義貨幣供給額變動率、股票市場本益比為總體經濟因子的代理變數，分別進行總負債比率和長期負債比率對這些總經因子變數與控制變數的季虛擬變數、公司層級因子進行迴歸，得出以下結論:(1)台灣上市電子公司的總負債比率季節波動一致性強，季虛擬變數即對總負債比率有明顯解釋能力。長期負債比率則無此現象。(2)加入單期的當期與前一期的總體經濟因子雖較僅有公司層級因子及季虛擬變數的模型解釋力有提升，但提升幅度有限。(3)加入跨期的總體經濟因子的模型，較僅有當期總體經濟因子模型及僅有公司層級因子及季虛擬變數的模型的解釋力皆有一定程度的提升。(4)上述研究結論在不同的總體經濟代理變數實證結果皆相同，具穩健性。(5)各總體經濟因子在各期對於資本結構的影響狀況不甚相同，本研究推測一原因為總體經濟對於資本結構影響的方式多元所致;另一可能為一控制效果，企業會綜合考量過去總體經濟狀況並前瞻未來總體經濟狀況，將資本結構不斷向最適資本結構調整，符合動態資本結構理論。

Capital structure has been studied for a long time, and has profound impact on companies’ operating capital and their financial conditions. The earlier studies have focused on how company-specific factors influence its capital structure and lately there have been some studies expressing how macroeconomic factors work. Past research shows that macroeconomics can influence capital structure through different angles. These angles include the value of collateral, agency problems, the public’s attitude toward companies’ investment, etc. However, these studies are limited to current data or one quarter lagged data of macroeconomic factors. In this paper’s perspective, a manager would consider past macroeconomic conditions but also take future outlook into account. The purpose of this paper is figuring out how the cross period macroeconomic factors, including lagged and leaded data, influence capital structure.

The sample data has been retrieved from listed electronic companies in Taiwan. The proxy variables of macroeconomics are interest rate of commercial paper, the exchange rate of money supply and PE ratio of the stock market. According to the empirical result, we discover that the quarter effect is very obvious for total debt ratio. Another conclusion found that the cross period macroeconomic data explanation describes the model better than the current and one quarter lagged data has in the past research. The final point expresses how the macroeconomic factor in different time periods can have different impact on capital structures. This evidence shows that it may be a controlled effect that adjusts capital structure to its optimal level.