Essays on earnings warnings

This thesis consists of three essays on earnings warnings. Essay 1 examines whether maintaining a strong disclosure reputation explains firms' voluntary behavior of issuing an earnings warning in the face of an earnings disappointment. Using "Analyst Following", "Analyst Forecast Dispersion" (with a negative sign), and "Management Horizon" as disclosure reputation proxies, I find that firms with higher existing disclosure reputation are more likely to warn after controlling for the litigation motivation and other firms characteristics.; Essay 2 re-examines market reaction to earnings warnings. Prior research documents that the market reacts more negatively to warning firms than to those that anticipate negative earnings news but do not warn. This finding is inconsistent with Ross (1989), who argues that in an efficient market, the timing of uncertainty resolution is irrelevant to stock prices. The issue is even more intriguing because warnings are voluntary. I examine four potential explanations.; I find no support for the risk-revision and market overreaction explanations and some support for the omitted-other-accounting-information and correction-of-previous-overvaluation explanations. Despite the evidence, the differential market reaction remains. More interestingly, using the four-factor buy-and-hold abnormal return method, I find that in the post-event period, warning firms experience normal returns while non-warning firms' stock performance drifts down. Assuming that the long-term return test method is valid, this finding suggests that warning firms have fully given their bad news in the event-quarter while for the non-warning firms, the bad news arrives gradually and eventually.; Essay 3 investigates the intertemporal liquidity changes around earnings warning and earnings announcement as well as in the post-event period. In theory, a more transparent firm disclosure policy improves stock liquidity and lowers the cost of capital, but the evidence is mixed. Earning warnings provide another avenue to examine this issue.; I find that warning firms have higher liquidity than non-warning firms in the pre-event period, consistent with the disclosure reputation argument for earnings warnings. However, I find no statistical difference in future liquidity changes between the two groups, implying that the warning decision has no noticeable impact on a firm's future stock liquidity.