Several studies have documented significant average underpricing for initial public offerings (IPOs) of common stock on the NASDAQ system - that is, the closing price on the first day of trading is significantly higher on average than the offer price. In our paper, we examine whether IPOs on the exchanges (NYSE and AMEX) display similar underpricing. In addition, we test whether differences in the initial and continued listing standards imposed by the exchanges and NASDAQ impact the level of underpricing of IPOs. It is widely acknowledged that the degree of underpricing of an IPO is related to the level of uncertainty surrounding the value of the firm - the higher the uncertainty, the greater the expected level of underpricing. Trading systems (i.e., exchanges and NASDAQ) impose different initial and continued listing standards on constituent firms. These listing standards include requirements related to public ownership, firm size, and the level of income over several years. By granting a firm a listing, a trading system provides a degree of certification that the firm has met its required initial standards and that it anticipates the firm will meet its continued listing standards in the future. This should reduce some of the ex ante uncertainty surrounding the firm's value and leads us to propose a certification hypothesis, namely, that the higher die listing standards imposed by a trading system, the lower the level of ex ante uncertainty associated with newly listing firms, and hence, the lower the level of underpricing. Traditionally, firms wishing to go public on the NYSE or AMEX first listed on the OTC (or NASDAQ) system and then transferred to the exchanges. From 1983 however, both NYSE and AMEX have allowed firms to list directly on the exchanges. In addition, in 1983 the NASDAQ National Market System (NASDAQ/NMS) completed its first full year of operation. NASDAQ/NMS imposes more stringent initial and continued listing standards than NASDAQ/non-NMS. Since 1985, an increasing number of firms have listed directly on the NASDAQ/NMS system. Together, the direct listings on the exchanges and NASDAQ/NMS provide an opportunity to examine directly the impact of listing standards on the underpricing of IPOs. We examine the level of underpricing for 55 NYSE, 41 AMEX, 158 NASDAQ/NMS, and 824 NASDAQ/non-NMS IPOs over the period 1983 to 1987. Significant levels of underpricing are found for all four trading systems, with the average levels being 4.82%, 2.16%, 5.56%, and 10.41% for the NYSE. AMEX, NASDAQ/NMS, and NASDAQ/non-NMS IPOs, respectively. Other factors such as firm size, the degree of ownership retention, the prestige of the lead underwriter (investment banker), and the quality of the auditor, also affect the perceived ex ante uncertainty of the IPO. Since these variables differ widely across our sample firms, we control for their effects on underpricing in a multiple regression framework. Our results provide some support for our certification hypothesis with the NYSE, AMEX and NASDAQ/NMS IPOs having significantly less underpricing than the NASDAQ/non-NMS IPOs even after controlling for these other proxies for ex ante uncertainty. However, we do not find any significant difference between the underpricing levels for the NYSE, AMEX, and NASDAQ/NMS offerings. We attribute this to the stricter listing standards applied by NASDAQ/NMS in order to meet the various state blue-sky laws. There are three major practical implications of our results. First, the underpricing phenomenon observed for NASDAQ IPOs is also present in initial offerings on the two major exchanges, NYSE and AMEX. Second, firms meeting the listing standards of either the exchanges or the NASDAQ/NMS, can reduce their level of underpricing by going public on one of these systems rather than on the NASDAQ/non-NMS system. Finally, for investors, our results suggest that higher levels of initial returns can be earned on NASDAQ/non-NMS IPOs.
