民國八十年後我國新上市公司須強制公告財務預測(強制上市當年揭露於公開說明書)，而在此之前則可自行決定是否揭露盈餘預測。此一規範形成新上市公司在上市時，有強制性預測、自願性預測、與未預測三種情境。針對此一規範，本文檢測此三種情境對折價幅度與蜜月期長度的影響。實証結果發現：（1）強制性財務預測的新上市公司，新股折價幅度小於非強制性或自願性揭露的樣本公司，表示強制性財測規範確有助於降低資訊不對稱。（2）自願揭露者的新股折價幅度大於未揭露者，此種結果類似於隨意說說(cheap talk)情境，原因或為研究期間我國訴訟成本相當低。此一結論與資本市場成熟的國家的訊息傳遞(signaling)結果，並不相同。（3）強制性財務預測的新上市公司，蜜月期小於非強制性或自願性揭露的樣本公司，自願揭露者的蜜月期大於未揭露者。考量我國股市漲停跌幅限制所造成影響之敏感性分析，亦支持前述之結果。 In 1991, Taiwan Stock Exchange mandated all initial public offering (IPO) firms to include their management forecasts in the prospectuses. This study explores the association between the 1991 disclosure rules and the IPO anomalies including positive abnormal returns accompanying the IPOs and the subsequent price reversals. Specifically, we partition our IPO observations during the test period based on the existence and the type of management forecasts. Our results support the notion that disclosure regulations help mitigating both information asymmetry and investors' overreaction. First, the post-1991 IPOs are significantly less under-priced. Second, consistent with the cheap talk hypothesis, the pre-1991 IPOs with voluntary earnings forecasts appear to be subject to greater under-pricing as opposed to the observations without management forecasts. Third, the magnitude of subsequent price reversal appears to be greater (less) for the pre- (post-) 1991 IPOs. Fourth, the duration of the under-pricing period for the post-1991 (pre-1991) IPOs appears to be longer (shorter). On the other hand, the duration of the under-pricing period for the pre-1991 IPOs with (with no) voluntary disclosures appears to be longer (shorter).