소유지배괴리도 계층별 회계투명성에 관한 분석

This study examines the association between the wedge and the accounting transparency among Korean listed firms belongs to Korean business group over the period 2001-2007. A Korean business group is a financial circle consisting of many firms related to various industries, and managed by controlling shareholders who usually possess voting rights in excess of cash flow rights. This difference between voting rights and cash flow rights gives controlling shareholders power to seek private benefits at the cost of outside shareholders. In turn, this incentive of the controlling shareholders may give rise to agency problems between controlling shareholders and outside shareholders. Extant empirical researches use firm-level wedge components and the findings on the effects of firm-level wedge on accounting transparency are mixed. This study furthers the inquiry into whether certain wedge-effects are more likely to impair accounting transparency than others by developing a framework that introduces the group-level wedge components. This study than examines whether the group-level wedge components, as distinct from the firm-level wedge components, impair accounting transparency among Korean business groups. This study uses accounting transparency proxied by conservatism, earnings management, earnings response coefficient and the cost of capital. This study introduces conservatism and earnings management as the proxies for the accounting information. It is expected that high wedge will be associated with low conservatism and high earnings management. The earnings response coefficient and the cost of capital as the proxies for the market response are used. When the external investors perceive that the wedge decrease the accounting transparency, it is expected that high wedge will be associated with low earnings response coefficient and high cost of capital. The regression results show that the effects of firm-level wedge on accounting transparency measures are inconsistent. However, the effects of group-level wedge on accounting transparency measures are consistent. The group-level wedge is positively related to earnings management and the cost of capital, and negatively related to conservatism and earnings response coefficient. In sum, these results show that group-level wedge decreases accounting transparency. Thus, this finding suggests that without properly controlling for group-level wedge, estimating the association between firm-level wedge and accounting transparency can lead to an erroneous inference regarding the effects of wedge on accounting transparency. To check these results for robustness, the association between the wedge and the accounting transparency is tested after controlling for the ownership of the controlling shareholders. The results are robust, which means that group-level wedge has a unique effect on accounting transparency that cannot be substituted by other determinants of accounting transparency such as the ownership of the controlling shareholders. The firm-level and group-level wedge are highly positively associated and it is difficult to separate the effects of each of the two wedge variables on the accounting transparency. This study examines VIF (Variance Inflation Factor) in the regression to investigate a multi-collinearity problem. The VIF in the regression is less than 2. This result suggests that the high correlation between the firm-level and group-level wedge does not necessarily cause a multi-collinearity problem. To mitigate concerns about autocorrelation, heteroscedasticity, and cross-sectional dependencies, this research conducts several sensitivity tests: Newey and West estimator, White's t in the ordinary least square regression, and Fama and Macbeth regression. The results show that unspecified autocorrelation, heteroscedasticity, and cross-sectional dependencies do not affect the main results. This is the first research to verify the association between group-level wedge and the accounting transparency (measured by conservatism, earnings management, earnings response coefficient, and the cost of capital) and the results of this research contribute the regulators, auditors, investors as well as academics.