한국 코스닥 신규공모시장에서 수요예측제도의 정보생산기능 평가

We examine if the book-building procedure at the KOSDAQ market extracts private information from informed investors as argued by the dynamic information acquisition hypothesis a la Benveniste and Spindt (1989). Although, to test the hypothesis, it is desirable to have a direct measure of private information, most of previous studies have used price adjustment during book-building as a proxy for private information. In this study, using the unique data of the KOSDAQ equity issuing market, we construct direct measures of private information provided by institutional investors under the book-building procedure, which include the average bidding price and the oversubscription ratio. We collect each bid's information from book and measure limit prices by their quantity-weighted average and oversubscription at the issue price. We characterize these measures as private information provided during book-building, and analyze the effects of private information on issue price and whether the effects can be explained by the dynamic information acquisition hypothesis, namely a mechanism of information providing and compensation in KOSDAQ equity issuing market. We find that our measures of private information during book-building affect significantly the final IPO price. However, private information in bids has a vague and insignificant effect for the abnormal returns around the IPOs. In particular, during period 3 when Korean underwriters have a wide discretionary authority, public information, which is the KOSDA index return during book-building, affects significantly both the issue price adjustment during book-building and the abnormal returns around the IPOs. Consequently, our results do not support the dynamic information acquisition hypothesis at the KOSDAQ market.