본 연구는 보고이익에 "전환"이라는 정보가 추가되었을 때 투자자들이 어떻게 반응하는지에 대한 증거를 제시하였다. 보고이익 ``전환``에 관한 정보는 투자자들에게 추가적인 효용증대나 효용감소를 가져올 수 있기 때문에 상당히 민감한 정보이다. 특히 신용상태의 평가에 상당한 중요한 고려요인으로 작용하는 것으로 알려져 있다. 또한 거래비용이론에 근거했을 때도 적자보고는 기업의 경영자에게 불리한 정보일 수밖에 없다. 또한 "0"을 기준으로 적자에서 흑자로, 흑자에서 적자로 보고이익이 전환될 때 prospect이론의 가치함수는 급격한 경사를 이루게 되어 투자자들의 효용이 급격하게 증가 또는 감소된다. 특히 대부분의 사람들은 손실을 꺼려하기 때문에 동일 금액에 대해서 손실환경에 속해 있을 때가 이익환경에 있을 때보다 더욱 큰 비효용을 인식하게 될 것이다. 따라서 전환정보는 투자자들에게 매우 민감한 정보이기 때문에 경영자들에게도 중요한 정보가 될 것이다. 본 연구의 목적은 보고이익 "전환"정보에 대하여 시장이 어떻게 반응하는지를 검증하고, 이익조정에 의한 흑자전환을 시장에서 어떻게 인식하고 있는지를 추가분석하는 것이다. 본 연구의 분석결과에 의하면 보고이익 전환은 추가적인 정보효과가 있는 것으로 나타났다. 즉, 보고이익이 흑자(적자)로 전환되었다는 정보는 주가에 단순한 비기대이익의 증가(감소) 이상의 영향을 미치고 있었다. 그러나 이익조정에 의해 흑자로 전환된 기업들은 시장에서 음(-)으로 반응하고 있었다. 본 연구는 보고이익전환에 대한 추가적인 시장반응을 검증함으로써 전환정보에 민감한 투자자들의 행동을 여러 가지 측면에서 이해할 수 있는 근거를 제시하였다는 데에 그 의의를 찾을 수 있을 것이다.

This study provides some empirical evidences on how investors react to the reversal of reported earnings signs. The reversal information will be a very sensitive to investors since such information may create additional increases or decreases in investors` benefits. Especially the reversal of reported earnings signs known as one of the most important factors determining firms` credit ratings. The reversal from positive to negative will be substantially disadvantageous to firms` managers under trade-expense theory. In addition, under prospect theory the value function has a steep slope around zero where earnings signs change from positive to negative or from negative to positive. Especially we expect that investors feel bigger disutility when firms report negative earnings than firms report positive earnings since most investors do not like to fall into loss status. In summary, the reversal of reported earnings signs will be very important information to investors since this could be very sensitive in an investors` value function. The purpose of this study is to investigate how the stock market responds to the reversal of reported earnings signs and also to analyze how the response will be different when the reversal is manipulated artificially through earnings management. The empirical results of this study reveal that the reversal of reported earnings signs provides incremental information over earnings information itself. In fact, the coefficient(& t-value) of interaction variable of reversal dummy from negative to positive and unexpected earnings(DNP*UE) and interaction variable of reversal dummy from positive to negative and unexpected earnings(DPN*UE) are 0.030(& 4.14) and 0.012(& 5.99), which are statistically significance at 1% and 5% level respectively.(See Table 4: Information Content of the Reversal of Reported Earnings Signs) Also, F-value for the interaction variables between unexpected earnings variable(UE) and the two reversal variables(DNP & DPN) is 28.20, which is statistically significant at 1% level. Such results support our hypothesis that earnings reversals will provide incremental information. This study estimates an additional model where a new variable(DG*UE) is added to test the information content of earnings` reversal from negative to positive when this reversal is created artificially through earnings management. The DG variable is defined here as "1" if net income minus discretionary accruals,(which may represent managed earnings, is negative. The results show that such reversal is recognized as bad news even it is from negative to positive. In fact, the coefficient(& t-value) of DG*UE -0.157(& -1.83), which is statistically significant at 10% level.(See Table 5 : Information Content of the Reversal from Negative to Positive Through Earnings Management) This study also tests whether the results will be different depending on the use of electronic reporting system, which was introduced in the year of 2001. The coefficients(& t-value) of two variables(DNP*UE and DPN*UE) for 2001 and thereafter are 0.007(& 2.58) and 0.058(& 3.69) respectively, which are stati- stically significant at 1% level. The explanatory power of this model is higher for 2001 and thereafter than for before 2001.(10.31% before 2001 and 19.64% for 2001 and thereafter). This result shows that investors could have utilized earnings information more easily after the adoption of electronic reporting system, and therefore the reversals information has been impounded into stock prices in more precise and efficient manner.(See Table 6: Information Content of Earnings Reversal in Each Period) In addition, this study examines the information content of the reversal at preliminary financial reports dates, when the information was released for the first time in the corresponding year. The results show that interaction variable between reversal dummy from negative to positive and unexpected earnings (DNP*UE) is not statistically significant, but interaction variable between reversal dummy from positive to negative and unexpected earnings(DPN*UE) is statistically significant at 1% level. F-value for the reversal additional variables (DPN*UE and DNP*UE) is 6.09, which is statistically significant at 1% level. However, the F-value only for the variable DPN*UE is not statistically significant.(See Table 7: Information Content of Earnings Reversal at Preliminary Financial Reports Dates) To investigate further the weak results for the interaction variable, DNP*UE, we add two additional dummy variables, reversal from negative to positive(DNP) and reversal from positive to negative(DPN), to the model. The results show that the added variable(DNP) is statistically significant at 1 % level, but still the interaction variable(DNP*UE) is not statistically significant. However, the F-value for DNP*UE and DNP is 12.05, which is statistically significant at 1% level. In summary, the results of this study show that the reversal of reported earnings signs has incremental information content. That is, the reversal infor- mation affects stock prices over unexpected earnings information. However, the artificial reversal created through earnings management results in negative effect even if earnings signs reverse from negative to positive. This study provides some empirical evidences on the earnings reversal infor- mation, which may be very useful to investors in practice. Therefore, it could provide a basis to understand behaviors of investors, who are very sensitive to the earnings reversal information. Also, the result shows that artificial earnings reversal from negative to positive through earnings management has negative effects. This results in an important implication for firms` managers, who are likely to manage their earnings as a way of escaping from their negative earnings status.