주가표류현상은 준강형 시장효율성 가설과 배치되는 자본시장의 이상현상으로 이익공시 후 기간에서도 공시된 이익정보가 주가움직임과 체계적인 관련성을 갖는 것을 말한다. 본 연구는 비기대이익과 이익 공시 후 기간의 누적초과수익률 간 관계를 통해 국외 자본시장에서 검증되어 온 주가표류현상이 국내 자본시장에서도 발견되는지를 실증적으로 분석한다. 더 나아가 개별 기업의 정보불확실성과 이익의 질이 주가표류현상 수준에 미치는 영향을 살펴본다. 2003∼2005년 제조업에 속한 상장기업들을 대상으로 실증분석한 결과는 다음과 같이 요약된다. 첫째, 비기대이익이 커질수록 이익정보 공시 후 기간의 누적초과수익률이 증가한다는 것을 확인할 수 있었다. 둘째, 정보불확실성 대용치로 초과수익률변동성, 평균주식거래회전율, 기업상장연수, 재무분석가수를 사용하여 분석한 결과, 개별 기업의 정보불확실성과 이익정보에 대한 주가표류현상은 양의 관계를 갖는 것으로 나타났다. 단 기업상장연수를 대용치로 사용한 결과에서는 그 값이 유의하지 않았다. 마지막으로 이익의 질을 이익지속성, 초과이익지속성, 재량적 발생액의 정도, 유동발생의 예측오차로 측정하여 분석한 결과에서는 재량적 발생액의 정도를 제외한 분석에서 이익의 질과 주가표류현상 수준 간에 유의한 음의 관계를 발견할 수 있었다.

Post-earnings-announcement drift (PEAD) is a phenomenon which the price continues to drift up if the earnings surprise is positive and down if negative. Ball and Brown (1968) first reported that stock returns continue to drift in the direction of earnings surprises for several months after the earningsannouncements. Since then, a number of studies attempted to find the underlying reason for this phenomenon suggesting explanations such as mis-measurement of risk and investor irrationality. Nonetheless, these conjectures were not sufficient enough to provide an explanation for the existence of the PEAD. Thus, this phenomenon remains as one of the market anomaly that is in a conflict with semi- strong efficient market hypothesis. Although this phenomenon is an established anomaly in the US, there have been few researches attempting to explore the drift for other stock markets. This study aims to test for the presence of PEAD based on Korea Stock Exchange. Specifically, this study examines the relationship between earnings surprise and stock`s cumulative abnormal returns for several days following an quarterly earnings announcement. Considering there are only a few studies examining whether firm-specific characteristics affect the drift level, this paper also tests the relationship between them. We first examine if the drift is more evident when the firm`s information uncertainty is high. We measure the information uncertainty using several proxies including excess stock returns, average daily turnover, firm age and analyst coverage. If the firm`s information uncertainty is high, it is likely that investors put less weight on the company`s earnings release, and the price reaction to earnings announcements should be less than the rational level. Consequently, there should be a subsequent price adjustment following an earnings announcement as the initial price reaction does not fully reflect the information contained in the earnings announcement. Secondly, we test whether the firm with poor earnings quality shows higher drift level using four proxies for earnings quality. They are accounting earnings persistence, abnormal earnings persistence, absolute magnitude of discretionary accruals, and magnitude of estimation error in accruals. In this case, the reaction of investors might differ depending on the earnings quality. We predict that the investors underreact less to the firms with better earnings quality. For the empirical analyses, this paper uses 1,095 Korean listed firms (nonbanking firms with December fiscal years) for the years from 2003 to 2005, and the financial data were figured out from Kis-Value and Fn-DataGuide Databases. The empirical findings are as follows. First, the cumulative abnormal returns over the 60 trading days subsequent to quarterly earnings announcements are significantly increasing in earnings surprise. As we partitioned the sample firms into two groups according to the standardized unexpected earnings, cumulative abnormal returns for stocks announcing positive standardized unexpected earnings showed strong drift compared to the stocks of negative standardized unexpected earnings. Second, consistent with our predictions, the results suggest that the level of PEAD and firm-specific information uncertainty has positive relation when volatility of excess stock returns, average daily turnover, and analyst coverage are used as measures of information uncertainty. However, significant association is not found when the firm age is used as the proxy. As the empirical results are mainly consistent with our prediction, we conclude that firms with higher information uncertainty experience higher drift to their earnings surprises. Lastly, this paper conjectures that PEAD is negatively related with earnings quality. The test results appear to be supported by statistical significance when accounting earnings persistence, abnormal earnings persistence, and magnitude of estimation error in accruals are used as measures of earnings quality. However, the result was not significant when absolute magnitude of discretionary accruals are used as the proxy. The main contribution of this paper is that we demonstrate the existence of PEAD phenomenon in Korea market consistent with the US market. Empirical results represent that stock return do not fully impound the surprise in announced quarterly earnings immediately upon the earnings disclosure; stock returns continue to drift in the same direction as the earnings surprise announcement. Our results indirectly indicate that the inefficiency of the market exists in Korea market. Our paper also contributes to research on understanding the firm-specific characteristics affecting the drift level. We suggested information uncertainty and earnings quality as major two factors explaining the differences in drift level among the firms, and the results supported the prediction that firms with higher information uncertainty or poor earnings quality will show higher drift level. In this study, we empirically evaluated the prediction by including the interaction variable. In this way, we not only examine the mean effect but focus on the relation between information uncertainty (or earnings quality) and the magnitude of the drift level.