Adverse Selection in Dealers' Choice of Interdealer Trading System
Peter C. Reiss
Stanford Graduate School of Business; National Bureau of Economic Research (NBER)
Ingrid M. Werner
The Ohio State University - Fisher College of Business
June 10, 1999
Dice Center Working Paper No. 99-7
London equity dealers routinely use two trading systems to trade with one another. In one, they trade directly by phone, using the best bid and ask quotes as a basis for negotiation. In the other, they submit and consume anonymous limit-orders in electronic limit-order books. In this paper, we use unique data to examine why dealers might prefer one system over another. We consider the predictions of two different theoretical models that emphazise the importance of adverse selection. Our evidence shows that direct, nonanonymous interdealer trades have subsequent price impacts several times that of the anonymous electronic trades. We interpret this as evidence that dealers typically resport to direct trade when they have information that is likely to affect future prices. Given the adverse selection of interdealer trades in the direct market, one should expect to see dealers charge more for these trades. Indeed, we do. Direct interdealer trades pay the best bid or ask, while limit order trades typically receive price improvement equal to one-third of the spread in the direct market. We conclude by asking how these differences in execution costs and subsequent price impact affect the initiating and posting dealer's profits. We find that both the poster and the initiator make significant positive profits from both types of interdealer trades.
Number of Pages in PDF File: 49
JEL Classification: G10, G20working papers series
Date posted: May 12, 2000
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo1 in 0.500 seconds