內部人股票強制集保結束期間之股價行為研究－－以臺灣股票市場為例

The Stock Price Behavior in the Period Surrounding Lockup Expiration: The Case of Taiwan

本研究探討我國內部人股票強制集保結束前後期間的股價行為，由於內部人領回股票之作業規定，內部人實際領回股票的日期，大約是在集保結束日後第七日；因此，我們分別觀察強制集保結束前一日至後一日，以及結束後第七日至第九日等兩個窗口的股價行為。實證結果發現，兩個窗口都呈現股價下跌、成交量增加的現象，然而，第二個窗口的股價異常報酬率與成交量增加幅度都較第一個窗口大，顯示內部人實際賣出持股對股價的影響程度，要高於強制集保結束日的資訊效果。在橫斷面的分析上，上市上櫃後的股價漲幅、異常成交量、及內部人持股比率等三個變數，都與第二個窗口的異常報酬率有顯著的反向關係，研究結果支持股票負斜率需求曲線假說與內部人投資風險分散假說。另外，有創投資金參與的IPOs，在強制集保結束後，並沒有出現較大的股價跌幅，可能的原因是我國創投產業規模較小，其影響力也較不顯著。

This paper, using Taiwanese IPOs, examines the stock price behavior in the period surrounding lockup expiration. According to the regulations in the Taiwan stock markets, insiders of IPO firms have the obligation to deliver and lockup their shares in Taiwan Securities Central Depository Co., Ltd. (TSCD). Lockups serve several purposes. They reassure the markets that key employees will not sell out their shares in the period immediately after the IPO. In other words, these insiders/managers will continue exerting themselves for the operations of IPO firms. Lockups can mitigate the expropriation of minority shareholders by insiders when the information asymmetry is still sever in the aftermarket. In addition, the lockup agreements can support the stock price of newcomers by constraining the supply of IPO shares. When lockups expire, insiders are allowed to sell their shares in the secondary market, and the large increase in selling volume may dampen the stock price. Because the unlock date and the increase of shares supply are public information, the efficiency market hypothesis, arguing the current market price will fully reflect the available information, predicts that no slump should be observed on the unlock date. For the U.S. IPOs, there are similar lockups as the ones in Taiwan. Field and Hanka (2001), Ofek and Richardson (2000), Brav and Gompers (2003), Keasler (2000), Bradley, Jordon, and Yi (2001) have examined the lockup in the U.S. stock markets. They found the stock price decreases and trading volume increases after the unlock date. However, there are several differences between the U.S. lockups and Taiwanese lockups. First, the IPO lockup in the U.S. is an agreement between an issuing company and its underwriters. The typical lockup period lasts for 180 days, and the lockup covers most of shares that are not sold out in the IPO. On the contrary, in Taiwan the lockup is one of the listing requirements regulated either by “Taiwan Stock Exchange Corporation Criteria for Review of Securities Listings” or by “GreTai Stock Market (GTSM, Taiwanese OTC market) Criteria for Review of Securities Listings,” in which the lockup schedule and locked-up shares are specified. Secondly, owing to the facts that in the U.S. stock markets the stock price usually decreases, and trading volume increases after the unlock date, several websites or investment firms have provided the stock lockup information to public investors. In the Taiwan stock markets, however, neither official nor the press has provided any information concerning the lockup. Taiwanese investors are generally unaware of the unlock date and cannot figure out the number of shares insiders may sell out at the lockup expiration. Finally, insiders of the U.S. IPO firms are able to sell out their locked shares immediately after lockup expiration, but insiders of Taiwanese IPO firms cannot, for they normally get back locked-up shares about seven business days after lockup expiration; and they are able to sell out locked-up shares after having got their shares back. Differences in lockup arrangements between the U.S. markets and Taiwanese markets provide us a good opportunity to examine whether for Taiwanese IPOs the stock price will behave differently in the period surrounding lockup expiration. We therefore investigate the stock price behavior not only in the period surrounding lockup expiration (day -1 to day +1, window 1), which likely reflects information effects of lockup expiration, but also in the period surrounding insiders’ receipt of the locked-up shares (day +7 to day +9, window 2), which likely reflects the effects of insiders’ selling. To our knowledge, Chen et al. (2003) and Chen et al. (2004) are the only two articles which have examined lockup expiration of Taiwanese IPOs. They, however, focus their studies primarily on the stock price behavior in the period of window 1. Hence, these two studies are unable to capture (grasp) the effects of insiders’ selling. We examine the stock price behavior in window 1 and in window 2, and compare the information effect of the lockup expiration (window 1) and the effect of insiders’ selling (window 2). When lockups expire (window 1) we find an increase in average trading volume, and a negative, but not statistically prominent, abnormal return. On the other hand, when insiders get their locked-up shares back (window 2), we, however, find a 44% increase in average trading volume, and a statistically prominent three-day abnormal return of -0.72 percent. The stock price keeps going down after window 2. The empirical results support the hypothesis of downward-sloping demand curve for equity shares. Interestingly, the stock price goes up a bit in the period between window 1 and window 2, which is similar to what Chen et al. (2003) have documented. We find that the positive abnormal returns are mainly driven by firms whose insiders have filed for sale of shares, indicating that insiders may try to push up the stock price, so that they can sell their locked-up shares for a better price. In the cross-sectional analysis, we find that the negative abnormal return in window 2 is greater for firms with a greater post-IPO stock price increase, firms with a greater percent of insiders’ shares, and firms with a greater relative trading volume. Our results support the hypothesis of downward-sloping demand curves for equity share and of insiders’ portfolio diversification. Although Brav and Gompers (2003), Field and Hanka (2001), and Bradley, Jordon, and Yi (2001) find the stock price falling in lockup period is larger for venture-backed IPOs than for IPOs not backed by venture capital firms, we do not find such a pattern for Taiwanese IPOs. The underlying reasons could be as follows: First, the scale of venture capital firms in Taiwan is much smaller than that in the U.S. Second, venture capital firms in the U.S. will distribute IPO shares to their partners, which may likely sell out shares after lockup expiration, while venture capital firms in Taiwan are incorporated as company, which may likely continue holding IPO shares after lockup expiration. Finally, in Taiwan some venture capital firms are not insiders, and they may have already sold out their shares prior to lockup expiration.