이익발표 후 잔류현상과 분기이익의 시계열속성과의 관계
Post Earnings Announcement Drift and Time-Series Properties of Quarterly Earnings
나종길(전남대학교)
33권 4호, 111~139쪽
초록
이익발표 후 잔류현상은 자본시장이 비기대이익에 대하여 즉각적이고 완전하게 반응하지 않고 비기대이익의 방향에 따라 일정기간 꾸준히 주가가 잔류하는 현상이다. 이는 곧 비기대이익이라는 이익정보에 근거하여 미래의 주가수익률을 예측할 수 있다는 점에서 효율적 자본시장가설에 어긋나는 자본시장이례현상이다. 이러한 잔류현상에 대한 가장 대표적인 설명은 자본시장이 회계이익의 시계열속성에 나타나는 과거 회계이익의 정보성을 충분히 이해하지 못한다는 것이다. 본 연구는 우리나라 자본시장에서도 이러한 이익발표 후 잔류현상이 존재하는가를 먼저 분석한 후 이러한 현상이 자본시장이 회계이익의 시계열속성을 충분히 이해하지 못하는 것과 관련되는가를 분석하였다. 분석결과 분기이익의 발표 후 일정기간동안 이익발표 후 잔류현상이 나타남을 알 수 있었다. 분기이익의 시계열속성을 분석한 결과 분기이익의 계절별이익차이는 과거 3개 분기의 계절별이익차이와는 양의 상관관계를, 4 번째 분기의 계절별이익차이와는 음의 상관관계를 가지는 것으로 나타났다. 또한 이러한 상관관계는 이익발표 후 60일 동안의 초과수익률로 측정한 잔상과 과거 4개 분기의 비기대이익과도 비슷한 패턴으로 나타났다. 이는 곧 자본시장이 분기이익의 시계열속성을 충분히 이해하지 못한다는 점을 시사한다고 할 수 있다. 자본시장이 반영하는 분기이익의 시계열속성을 분석한 결과 우리나라의 자본시장은 분기이익의 시계열속성을 거의 이해하지 못하는 것으로 나타났다. 본 연구는 우리나라의 자본시장에서 나타나는 이익발표 후 잔류현상의 존재와 그 패턴을 보고하였으며, 2000년도부터 보고되기 시작한 분기이익이 외국 선행연구에서 보고된 바와 같이 일정한 형태의 자기상관관계를 가지고 있음을 보고하였다는 의미를 가진다. 또한 자본시장의 참가자들이 분기이익의 이러한 시계열속성을 충분히 이해하지 못한다는 결과는 우리나라에서 분기이익이 공시된 기간이 짧은 것과 관련될 수 있다.
Abstract
Previous studies reported that the sign and magnitude of security returns are predictable, based on previously announced earnings. This ability of current earnings information to predict future abnormal returns are called ‘post- earnings-announcement drift.’ Ball and Brown(1968) first documented post-earnings announcement drift. That is, even after earnings are announced, abnormal returns continue to drift up for good news firms and down for bad news firms. Foster, Olsen, and Shevlin(1984) also reported the same phenomenon using quarterly earnings data. They reported that the drift exists even 120 trading days after earnings announcement and hedge returns of 6.31% over the 60 trading days could be obtained simply based on the announced earnings information. The predictability of stock returns based on current earnings information is a serious challenge to the efficient capital market hypothesis. The academic literature offers three possible explanation for this anomaly. 1) shifts in risks of firms with extreme unexpected earnings, 2) methodological problems in measuring the returns, 3) investors’ under-reaction to earnings information. While many studies suggested the first and second explanation for the drift, the third explanation as suggested by Bernard and Thomas(1990) has been accepted as the most plausible reason for the drift. Bernard and Thomas(1990) argued that one plausible reason of this phenomenon is the market inefficiency with regard to the serial correlation of the seasonally differenced quarterly earnings. However, Ball and Bartov(1996) argued that while market investors do know the serial correlation, they tend to underestimate the serial correlation. While continuing studies on this phenomenon are being executed in other countries, few studies reported the existence and the reason of this phenomenon in Korean stock market. Nah and Lee(2006) examined the post earnings announcement drift and related this phenomenon with the audit quality. Lee and Lee(2008) related the drift with firms’ information environment. However, no studies fully examined the pattern and the length of the drift and investigated the possibility of investors’ naivite with regard to time-series information of quarterly earnings. In this context, this study examines the existence of post-earnings-announcement drift. Also, by examining the serial correlation of quarterly earnings, this study investigates the market’s efficiency with regard to the time-series nature of quarterly earnings. Empirical results show that there exists a post-earnings-announcement drift in the Korean stock market and the magnitude of the drift is quite similar to the one in the American stock market. That is, there is a pronounced post-earnings- announcement drift, increasing monotonically in unexpected earnings. Most of the drift occurs during the first 60 trading days, and a long position in the highest unexpected earnings decile and a short position in the lowest unexpected earnings decile would have yielded an estimated abnormal return of 7.59% over the 60 days subsequent to the earnings announcement. Also, the results show that the autocorrelations in seasonally differenced quarterly earnings are positive with declining magnitudes at the first three lags, and strongly negative at the fourth. This autocorrelation structure of earnings suggests that quarterly earnings process is a seasonally differenced first-order autoregresssive process with a seasonal moving-average term, as suggested by Brow and Rozeff(1979,) Foster(1977), and Bernard and Thomas(1990). Further analysis indicated that the abnormal returns after the earnings announcement are correlated in a way which suggests that market investors do not understand the serial correlation of seasonally differenced quarterly earnings. While Ball and Bartov(1996) argued that the market investors in America are aware of the existence of the serial correlation, the results of this study implies that the market investors in Korean stock market do not understand the serial correlation. The difference in the market efficiency between America and Korea appears to stem from the short history of quarterly earnings in Korea which began to announce quarterly earnings in 2000. Future studies might examine the effect of firms’ information environment difference on the longevity of the post earnings announcement drift
- 발행기관:
- 한국회계학회
- 분류:
- 회계학