특수관계자거래가 자기자본비용에 미치는 영향

We investigate the effect of related party transactions(hereafter ‘RPTs’) on the cost of equity. Our investigation is motivated by prior researches that suggest potential abuse of RPTs by controlling shareholders. If RPTs are used as means of self-dealing, the level of RPTs should be positively associated with the cost of equity. RPTs are potential means for controlling shareholders to expropriate outside shareholders via self-dealing(or tunneling). RPTs are not arm’s length in nature and may be more likely to expropriate wealth from outside shareholders. Several domestic papers report potential abuse of RPTs. RPTs are positively associated with discretionary accruals(DA) and analysts’ forecast errors. RPTs are negatively associated with the quality of accruals, earnings response coefficients(ERC), vlaue relevance of accounting earnings, and measures of firm performance. Bhattacharya et al.(2012) report evidence of both a direct path from earnings quality to the cost of equity, and an indirect path that is meadiated by information asymmetry, with the weight of the evidence favoring the direct path as the more important. If RPTs are potential means for self-dealing, we expect the level of RPTs to influence the relation between earnings quality(information asymmetry) and the cost of equity negatively. We examine the coefficient for interaction term(earnings quality×RPTs and information asymmetry×RPTs) using multiple regression. The dependent variable is the cost of equity. The independent variables include earnings quality(information asymmetry), RPTs and control variables. For a sample of non-financial listed firms(KOSPI) during 2000-2010, earnings quality shows a negative association with the cost of equity. And information asymmetry shows a positive association with the cost of equity. These relations are consistent with prior researches. We find statistically significant positive coefficient for the interaction term, earnings quality×RPTs. But we find statistically insignificant negative coefficient for the interaction term, information asymmetry×RPTs. We also find that RPTs(level) are not directly associated with the cost of equity in an additional multiple regression analysis where earnings quality and information asymmetry are controlled for. The results suggest that, after contrlling for the effect of earnings quality(information asymmetry) on the cost of equity, RPTs are not significantly associated with the cost of equity. Our results imply weak or insignificant associations between RPTs and the cost of equity. Our results also support that firm valuation is not systematically negatively affected by the level of RPTs. RPTs only weakly affect the relation between earnings quality and cost of equity. RPTs(measured as 1 or 0 dummy variable based on median observation) are marginally positively associated with the cost of equity in additon to the effect of earnings quality. However, RPTs do not affect the relation between information asymmetry and the cost of equity. When earnings quality and information asymmetry are controlled for simultaneously in a multiple regression, RPTs(measured as level variable) show statistically insignificant coefficient. Thus we conclude that RPTs do not systematically influence the relation between earnings quality(information asymmetry) and the cost of equity. Taken together, our findings for RPTs provide no support for the self-dealing hypothesis. The primary contribution of this paper is to document that RPTs do not affect the relation between earnings quality(information asymmetry) and the cost of equity. In this respect, the results of our study suggest that findings and implications of prior studies on RPTs need to be reconsidered.