股票市場對強制性財務預測誤差預期之研究

Evidence of Stock Returns Foreseeing Mandatory Management Forecasts Biases

本研究以 1999 至2001 年間有發布財務預測的上市及上櫃公司，探討股票市場是否對強制性財務預測誤差存有預期。實證結果顯示管理當局宣告財務預測時帶給市場新訊息，且市場對於預測誤差有所預期，惟強制性原始財務預測偏向保守預測。進一步分析原始財務預測平均而言趨向保守原因，結果顯示管理當局具備良好的預測能力，保守的預測與預測動機有關，暗示管理當局不願完全揭露其不偏估計。由於發佈動機為外顯因素，市場係洞察財務預測動機，藉直接觀察強制性揭露情境而能對財務預測誤差有所預期，未必表示股價所反應之資訊集合大於管理當局之資訊集合。

Documented market reactions to mandatory management earnings forecast releases suggest that these forecasts provide relevant information to other market participants. While management forecasts enrich the information set impounded into stock prices, non-management provided information also influence investors’ beliefs. More specifically, efficient market hypothesis posits that stock prices are determined by an information set much larger than the manager’ information set. Thus, the market might be able to deduce the quality of management forecast from other information. This is an important issue in the debates over mandatory disclosure of management earnings forecasts. Outspoken opponents of mandatory disclosure often denounce the inherent uncertainty of forecast information. They claim that large forecast errors misled investors. Based on efficient market hypothesis, present study argues that stock returns foresee management forecast errors realized subsequently. Consistent with the prediction, the empirical evidence shows that pre-announcement and announcement periods stock returns do foresee subsequent management forecast errors. On average, mandatory management forecasts are conservative. What is reflected in pre-announcement period returns may be due to intentionally biased reporting. Either that investors have a richer information set than management do or that management did not incorporate all their information in their earnings forecasts is consistent with the above empirical finding. To explore the possibility of mis-presentation, this study distinguishes forecast ability from motivational factors that both give rise to management forecast errors. Further analysis that controls differential forecast ability suggests that mangers are good predictors but may not present their forecasts in good faith. Expected forecast bias estimated from motivational factors is positively related to pre-announcement and announcement periods stock returns. Therefore, it is likely that the market forms bias expectation from circumstantial events leading to management’s incentives to mis-present their forecasts. The study contributes to the management forecast research by decomposing forecast errors resulting from forecast ability-related factors and from motivational factors. There are two policy implications of empirical findings. First, it is shown that mandatory management forecasts provide useful while not fully revealing information to the market. Released forecasts might enhance market efficiency. Second, the market is efficient to expected biases. Since investors are price-protected, some large forecast errors are not as harmful as they appear to the operation of our market. In sum, mandatory disclosure requirement is a successful enactment despite of highly political disputes.