We study the aftermarket for 72 initial public offerings (IPOs) using comprehensive trade and quote-change data from every market maker for the first three days of trading. Underwriters quote higher bid prices than other market makers for issues that commence trading at or below the offer price. Underwriters repurchase large quantities of stock in the aftermarket without risk by overselling the issue by the amount of the overallotment option. If the IPO is hot, the overallotment option is exercised. If not, the short position is covered with aftermarket selling. We discuss several reasons for underwriter support.

On May 26 and 27, 1994, several national newspapers reported the findings of W. Christie and P. Schultz (1994) who cannot reject the hypothesis that marketmakers of active NASDAQ stocks implicitly colluded to maintain spreads of at least $0.25 by avoiding odd-eighth quotes. On May 27, dealers in Amgen, Cisco Systems, and Microsoft sharply increased their use of odd-eighth quotes, and mean inside and effective spreads fell nearly 50 percent. This pattern was repeated for Apple Computer the following trading day. Using individual dealer quotes for Apple and Microsoft, the authors find that virtually all dealers moved in unison to adopt odd-eighth quotes.

The NASDAQ multiple dealer market is designed to produce narrow bid‐ask spreads through the competition for order flow among individual dealers. However, we find that odd‐eighth quotes are virtually nonexistent for 70 of 100 actively traded NASDAQ securities, including Apple Computer and Lotus Development. The lack of odd‐eighth quotes cannot be explained by the negotiation hypothesis of Harris (1991) , trading activity, or other variables thought to impact spreads. This result implies that the inside spread for a large number of NASDAQ stocks is at least $0.25 and raises the question of whether NASDAQ dealers implicitly collude to maintain wide spreads.

SOES bandits are individual investors who use Nasdaq's Small Order Execution System (SOES) for day trading. Their average profit per trade is small, but they trade dozens or hundreds of times per week. Bandits usually establish a position before most market-makers have updated their quotes, and lay off the position at favorable prices through Instinet or SelectNet. It is noteworthy that they trade profitably with market-makers despite having less information. Bandits keep the profits and bear the losses from their trades. Thus they have greater incentives to trade well than the employees of market-making firms.

Nasdaq's Small Order Execution System (SOES) allows orders to be submitted by computer, thereby assuring rapid execution at quoted prices. We examine trading in the 20 largest Nasdaq stocks around the time of a rule change that reduced the largest SOES trades from 1000 to 500 shares. We show that SOES trades contain information about short-term price movements and that SOES trading declined dramatically with the rule change. However, quoted and effective spreads were unaffected by the rule change.

The relative merits of dealer versus auction markets have been a subject of significant and sometimes contentious debate. On January 20, 1997, the Securities and Exchange Commission began implementing reforms that would permit the public to compete directly with Nasdaq dealers by submitting binding limit orders. Additionally, superior quotes placed by Nasdaq dealers in private trading venues began to be displayed in the Nasdaq market. We measure the impact of these new rules on various measures of performance, including trading costs and depths. Our results indicate that quoted and effective spreads fell dramatically without adversely affecting market quality.

On May 26 and 27, 1994, several national newspapers reported the findings of Christie and Schultz (1994) who cannot reject the hypothesis that market makers of active NASDAQ stocks implicitly colluded to maintain spreads of at least $.25 by avoiding odd-eighth quotes. On May 27, dealers in Amgen, Cisco Systems, and Microsoft sharply increased their use of odd-eighth quotes, and mean inside and effective spreads fell nearly 50%. This pattern was repeated for Apple Computer the following trading day. Using individual dealer quotes for Apple and Microsoft, we find that virtually all dealers moved in unison to adopt odd-eighth quotes.

On May 26 and 27, 1994 several national newspapers reported the findings of Christie and Schultz (1994) who cannot reject the hypothesis that market makers of active NASDAQ stocks implicitly colluded to maintain spreads of at least $0.25 by avoiding odd‐eighth quotes. On May 27, dealers in Amgen, Cisco Systems, and Microsoft sharply increased their use of odd‐eighth quotes, and mean inside and effective spreads fell nearly 50 percent. This pattern was repeated for Apple Computer the following trading day. Using individual dealer quotes for Apple and Microsoft, we find that virtually all dealers moved in unison to adopt odd‐eighth quotes.

The relative merits of dealer versus auction markets have been a subject of significant and sometimes contentious debate. On January 20, 1997, the Securities and Exchange Commission began implementing reforms that would permit the public to compete directly with Nasdaq dealers by submitting binding limit orders. Additionally, superior quotes placed by Nasdaq dealers in private trading venues began to be displayed in the Nasdaq market. We measure the impact of these new rules on various measures of performance, including trading costs and depths. Our results indicate that quoted and effective spreads fell dramatically without adversely affecting market quality.