文档介绍：The Market for Market-Making
Paul Schultz
University of Notre Dame
February, 2000.
Preliminary. Do not quote without permission.
I am grateful to John Affleck-Graves, Robert Battalio, Thomas Cosimano, Raymond Fishe, Marc
Huson, Aditya Kaul, Youngsoo Kim, Timothy Loughran and seminar participants at the University of
Alberta, the University of Notre Dame and the Securities and mission for helpful
suggestions. All errors are my own.
This paper can be downloaded from the
Social Science work Electronic Paper Collection:
http://papers./?abstract_id=213528
Abstract
Nasdaq market-makers can adopt new stocks or abandon old ones with almost no entry or exit costs.
Under these circumstances, the theory of contestable markets predicts that market-making
concentration in individual stocks should unrelated to the profits dealers earn trading those stocks.
However, I find that after adjusting for factors believed to determine spreads, market-maker
concentration is a highly significant determinant of trading costs. The problem is that entry does not
bring order flow. I show that dealers make markets in the stocks in which they receive order flow either
through purchase agreements or through internalization.
1. Introduction
On Nasdaq, there are a large number of potential dealers to trade any individual stock. The
cost to enter or exit market-making for a specific stock appears to be close to zero. Thus market-
making in individual Nasdaq stocks takes place in a close approximation to what Baumol, Panzar and
Willig (1982) define as a perfectly contestable market. In such a market, the possibility of hit-and-run
entry ensures that prices do not exceed the prices observed in a petitive market - even if
there are only a handful of market-makers. petition keeps prices in line and the number of
petitors is irrelevant.
In this paper I find that even after adjusting for variables typically used to proxy for inventory
holding costs, adverse s