Liquidity Externalities and Adverse Selection: Evidence from Trading after Hours

This paper examines liquidity externalities by analyzing trading costs after hours. There is less than 1/20 as many trades per unit time after hours as during the trading day. The reduced trading activity results in substantially higher trading costs: quoted and effective spreads are three to four times larger than during the trading day. The higher spreads reflect greater adverse selection and order persistence, but not higher dealer profits. Because liquidity provision remains competitive after hours, the greater adverse selection and higher trading costs provide a direct measure of the magnitude of the liquidity externalities generated during the trading day. UNDERSTANDING NETWORK EFFECTS and liquidity externalities is one of the most important outstanding issues in market design (Madhavan (2000)) and market regulation (Macey and O’Hara (1999)). Liquidity externalities arise from bringing traders together in space and time to reduce search and trading costs. Bringing traders together creates liquidity externalities because the additional traders arriving in the marketplace reduce trading costs for all investors. To date, attention has focused almost exclusively on spatial network effects by examining the trading of securities in different markets at the same time (Lee (1993), Hasbrouck (1995), and others). In contrast, this paper studies the effects of temporal consolidation of trades by analyzing trading during and outside of exchange trading hours. A well-known liquidity externality arises from asymmetric information. Rational informed traders break up their orders across markets and over time. This provides incentives for liquidity traders to consolidate their trades geographically (Garbade and Silber (1979), Mendelson (1982, 1987), Pagano (1989a, 1989b)) and intertemporally (Admati and Pfleiderer (1988) and Foster and Viswanathan (1990)). It has been difficult to document the importance of these liquidity externalities, however. Unless there are barriers that restrict the flow of orders from one trading venue to another, in equilibrium, contemporaneous ∗Barclay and Hendershott are from the Simon School of Business, University of Rochester and Haas School of Business, University of California at Berkeley. We thank Tim McCormick for providing data and helpful comments. We also thank Rick Green (the editor); Joel Hasbrouck; Rich Lyons; an anonymous referee; and seminar participants at the Review of Financial Studies Conference on Investments in Imperfect Capital Markets, the University of California at Berkeley, the University of North Carolina, and the Economic Research group at Nasdaq for their helpful comments. Hendershott gratefully acknowledges support from the National Science Foundation. Any errors are our own.