본 연구는 경영자지분율과 회계정보의 가치관련성 간의 관계를 고찰하였다. 이전연구와는 달리 지분율의 구간을 나누고 각 구간에서 회계정보의 가치관련성이 다를 것이라는 가설을 설정하고 검증하였다. 1999년부터 2004년 간의 2,463 기업/년에 대해서 Ohlson 모형을 사용하여 분석한 결과를 요약하면 다음과 같다. 첫째, 지분율과 회계정보의 가치관련성을 고찰할 때 지분율의 크기에 따라 적절한 구간을 나누고 구간별로 서로 다른 관계를 가정하여 모형을 설정하면 모형의 설명력이 증가한다. 둘째, 지분율 10%와 24%를 기준으로 하여 세 개의 구간으로 구분하는 경우 모형의 설명력이 가장 크다. 셋째, 이익과 순자산의 가치관련성은 지분율구간에 따라 차이가 있다. 즉, 지분율 10%∼24%에서 이익과 순자산의 가치관련성이 10% 이하와 24% 이상에서의 그것보다 작다. 넷째, 재벌과 비재벌그룹으로 표본을 나누어 분석한 결과 비재벌그룹에서와는 달리 재벌그룹에 있어서는 구간별 회계정보의 가치관련성 차이가 나타나지 않았다.
This paper investigates the relations between management ownership and value-relevance of accounting information. Some prior studies argue that managers of low management ownership firms provide less useful accounting information because they have greater incentives to manage accounting numbers using their discretion to maximize their own wealth rather than firm values. They borrow their reasoning from Jensen and Meckling(1976). On the other hand, others argue that managers of high management ownership firms provide less useful accounting information due to entrenchment effect. They argue that managers of high management ownership with effective controlling power have less motivation to align themselves with value-maximizing decisions. These two conflicting arguments brought about many empirical studies the results of which still remain mixed. This paper focuses on the agency relationship between large and minority shareholders. Management ownership is defined as the sum of the largest shareholder`s ownership and that of special interest group. Coupled with entrenchment effects agency relationship proposes hypothesis that relevance of accounting information is different for differing levels of management ownership. Relevance of accounting information is measured by the explanatory power and magnitude of coefficients of Ohlson model. Samples consist of 2,463 firm-years that are listed in the Korean Stock Change or KOSDAQ for 1999-2004. Since how many break points we have and where they are throughout the management ownership levels are empirical question, we employ a grid search approach to explore the break points that produce the greatest explanatory power. With some constraints our search ends up with two break points, 10% and 24% which secure the largest explanatory power. Thus, we have three different intervals based on the management ownership; section1: management ownership 10% or lower; section2: between 10% and 24%; section3: 24% or higher. Empirical results indicate that Ohlson model`s explanatory power increases when taking into consideration varying relationship between value-relevance of accounting information and management ownership. Specifically, when we add interaction terms representing three different sections in the model, adjusted R2 increases significantly. When we classify samples into three groups with the criteria of 10% and 24%, we have the greatest explanatory power. That is, of thousands of possible schemes of classification 10% & 24% classification scheme shows the largest adjusted R2. The results also show that value-relevance of earnings(book value) measured by the coefficients of earnings(book value) in the Ohlson model is lowest for section2. In other words, accounting information of firms with management ownership between 10% and 24% turns out to be the least value-relevant. This finding supports our hypothesis that managers with ownership somewhere in the middle reveal opportunistic behavior taking advantage of minority stockholders. Additional analysis reveals that above results are driven by non-Chaebol group firms. We did not find high-low-high pattern of relevance of accounting information according to different management ownership levels for Chaebol firms. One possible explanation is that our measure of ownership is not appropriate for Chaebol group firms which have complicated ownership structure entangled by cross-ownership between affiliates. Comparison of discretionary accruals also supports that managers of high management ownership firms(section3) manage earnings to less degree than those of low management ownership firms(section3).