An Empirical Examination of the Nasdaq/CHX Dual-Trading Experiment.

We analyze the effects of the SEC's experimental Nasdaq/CHX dual-trading program. The program, which began in 1987 and continues to the present, establishes an experiment in which the costs and benefits of competition between dealer and specialist market structures can be observed directly. Our primary finding is that the program led to significantly reduced mean quoted and percentage spreads for the dual-traded issues. Further, even though the CHX specialists quote lower spreads, they are not able to garner a significant number of trades from Nasdaq. Copyright 1999 by MIT Press.