CONDITIONAL PRICE TRENDS IN THE AFTERMARKET FOR INITIAL PUBLIC OFFERINGS

Numerous studies have shown that, on average, initial public offerings (IPOs) are underpriced. However, while underpricing does exist on ave rage, there is widespread cross-sectional variation across specific is sues. Indeed, empirical results suggest that on the first day of tradi ng, less than 70% of IPOs close above their offering price, while at l east 25% close below the offering price.Using a sample of 2,096 IPOs o ver the period 1975 to 1991, we show that this initial mispricing is n ot restricted to the first day of trading. Our results provide strong evidence of significant price trends in the shortterm aftermarket that are in the same direction as the initial mispricing and that can last for up to three months after the offering date. Thus, after the first day of trading, initially underpriced IPOs continue to drift upward i n price and earn positive returns on 17 of the first 20 days following the first day of trading. This drift persists for approximately three months, resulting in an additional 6.4% return above that earned at t he offering. In contrast, prices of initially overpriced IPOs continue to drift downward after the offering, and these IPOs earn positive re turns on only three of the first 20 days following the offering. By th e end of the first month of trading, these overpriced IPOs have underp erformed matched firms of similar size by a further 4.4%, on average, from the price at the close of the first day of trading. Following the first three months of trading, there is little difference in the afte rmarket performance of both groups of IPOs and, consistent with Ritter (1991), both groups exhibit significant long-term underperformance. O ur results have important implications for investors who obtain an ini tial allocation of shares at the offering. Specifically, they show tha t the initial mark-et price relative to the offering price of an IPO c ontains important information about the short-term performance of the stock. This in turn, suggests a profitable trading strategy. IPOs whos e closing price on the first day of trading exceeds their offering pri ce (that is, that are underpriced), should be held for up to three mon ths in the aftermarket. This will yield, on average, an additional 6.4 % return above the 13.4% underpricing, resulting in a total three-mont h return of just under 20%, on average. In contrast, IPOs that close o n the first day of trading at a price below the offering price (that i s, overpriced IPOs) should be sold as soon as possible in the aftermar ket. While selling these IPOs immediately in the aftermarket cannot pr event the average overpricing loss of 2.7%, it avoids an additional lo ss of 4.4% if these IPOs are held until the end of the first month of trading. There are several possible explanations for our results. Firs t, it is possible that underpriced and overpriced IPOs have different risk characteristics. We verify this by showing that underpriced IPOs, on average, have higher systematic risk (beta) than overpriced IPOs o ver the first two years in the aftermarket. Adjusting returns for thes e risk differences, however, does not affect our results. Initially un derpriced IPOs earn significant positive risk-adjusted returns in the first three months following the offering, while overpriced IPOs have negative risk-adjusted performance. Second, the underperformance of in itially overpriced IPOs may result from underwriter stabilization acti vities. Consistent with other studies, our results do show evidence of stabilization activity in overpriced IPOs. We are, however, also able to document that these IPOs continue to drift downward in price and t hat they continue to underperform matched firms well after stabilizati on activity has ceased. Third, the market may underreact to the inform ation implied in the initial price of an IPO. This would be consistent with the evidence of underreaction to the information in seasoned equ ity offerings (Loughran and Ritter, 1995) and equity repurchases (Iken berry, Lakonishok, and Vermaelen, 1995). Our findings concerning signi ficant conditional trends in the short-term aftermarket for IPOs adds to the growing body of evidence in support of price momentum or drifts following important information events. Unlike most other studies, ho wever, our evidence shows that the direction of these price trends can be conditional on the direction of the information embodied in the ev ent.