이익발표 후 잔류현상은 회계이익의 정보성에 대하여 투자자들이 과소반응하는 것과 관련되며, 이러한 과소반응의 정도는 투자자들의 비숙련성 또는 기업정보환경의 불투명성과 관련되는 것으로 보고되어 왔다. 투자자 속성과 관련하여 외국인투자자는 일반투자자와 비교하여 숙련성이 높으며, 외국인 소유지분의 비중이 높은 기업일수록 기업정보환경이 투명하다고 선행연구들은 보고하였다. 이러한 연구들에 기초하여 본 연구는 외국인 소유지분의 비중이 높은 기업일수록 보다 숙련된 투자자의 비중이 높아지며 기업정보환경이 투명하여 이익발표 후 잔류현상의 크기가 작을 것으로 예측하였다. 외국인 소유지분이 가능한 전체 표본에 대한 분석결과, 비기대이익의 크기에 따라 일정한 방향으로 이익발표 후 잔류현상이 나타났으며, 이러한 잔류현상은 외국인 소유지분의 크기와 음의 관계를 가지는 것으로 나타났다. 또한 이익발표 후 잔류현상의 크기에 영향을 미치는 것으로 보고된 기업규모, 거래비용, 성장성, 재무분석가의 수와 같은 기업특성변수들을 통제한 후에도 동일한 결과를 얻을 수 있었다. 회계이익의 지속성 차이 등을 고려한 추가분석에서도 외국인 소유지분 비중은 이익발표 후 잔류현상의 크기와 음의 관계를 가지는 것으로 나타났다. 이러한 연구결과는 외국인투자자들이 우리나라 자본시장의 효율성 향상에 기여한다는 것을 시사한다고 할 수 있다.
Previous studies have reported that the sign and magnitude of security returns in the post earnings announcement period are positively correlated with the sign and magnitude of the unexpected component of the earnings release. This predictability of stock returns after earnings announcement(i.e., the post-earnings- announcement drift) was first documented by Ball and Brown(1968). Ball and Brown(1968) reported that abnormal returns are predictable up to two months after annual earnings announcements and Foster et al.(1984) documented the same phenomenon up to 60 trading days after quarterly earnings announcements. This phenomenon implies that stock prices do not respond completely and immediately to information in the publicly announced earnings. Thus, this phenomenon is a stock market anomaly which is a deviation from the Efficient Market Hypothesis. While earlier studies attributed this phenomenon to the inadequacy of the CAPM as a model of asset pricing, or to research design problem that assumed information was available to the market before its public release, but none of those explanations could fully account for the observed drift. Rendleman et al.(1987) and Freeman and Tse(1989) provided some preliminary evidence that the drift in stock returns may represent misperceptions of time-series properties of earnings. In line with these conjecture, Bernard and Thomas(1990) provided evidence that the drift is related to the investors` misperception of quarterly earnings to be a seasonal random walk, although the actual process is a seasonally differenced first-order autoregressive process with a seasonal moving-average term. In other words, the drift represents biases in investors` assessment of the serial correlation of quarterly earnings changes. Based on Bernard and Thomas(1990), Bartov et al.(2000) conjectured that post-earnings-announcement drift will be most pronounced for stocks held primarily by unsophisticated investors who perceive the earnings process to be a seasonal random walk and will be least pronounced for stocks largely held by sophisticated investors who perceive correctly the time- series of earnings. They reported that the drift is negatively related to the institution holding variable which they used as a proxy for investor sophistication. This result indicates that the post-earnings-announcement drift is negatively related to the investor sophistication. In the above context of market underreaction to earnings information, information uncertainty was raised as a driving force of the post-earnings-announcement drift. Jiang et al.(2005) argued that with greater information uncertainty, investors trade more aggressively on their private signal, ignoring the public information such as earnings. Thus, the level of information uncertainty is positively correlated with the investors` underreaction to earnings information and the magnitude of the drift. The above discussions indicate that the level of investors` unsophistication and the information uncertainty are positively related to the magnitude of the drift. Also, prior studies reported that the foreign investors are sophisticated investors and that the firms with higher level of foreign investors show more transparent information environment. In this context, this study investigates whether the level of foreign investors affects the magnitude of the post-earnings-announcement drift. We conjecture that firms with higher level of foreign investors will exhibit lower magnitude of post-earnings- announcement drift. Empirical results are as follows. Firstly, the cumulative abnormal returns after the earnings announcement are monotonically increasing in the magnitude of the unexpected earnings, indicating the existence of the post-earnings- announcement drift in Korean stock market. Secondly, We regress the cumulative abnormal returns on the standardized unexpected earnings and the interaction variable of standardized unexpected earnings with the foreign investors` ownership. It turned out that the coefficient of the interaction variables is negative, implying that the foreign investors` ownership is negatively related to the magnitude of the post-earnings-announcement drift. Thirdly, we included other firm-specific variables that are reported to affect the magnitude of the post-earnings-announcement drift, such as firm size, growth, trading cost, and the number of analysts. With all these control variables, the main results appears to be same. Finally, we also controlled persistence-related variables. Prior studies reported that earnings persistence is lower in the cases of losses and earnings decreases than in the cases of profits and earnings increases. Also, earnings of fourth quarter were reported to have low persistence in comparison to earnings of other quarters. With all these additional controls, the main result that foreign ownership is negatively related to the drift does not change. There has been many concerns about the sharply-increasing foreign investors in Korean stock market. However, the results of our study imply that the increasing foreign investors have an effect of increasing the market efficiency in terms of reducing the underreaction to earnings information.