文档介绍：Market Microstructure With Uncertain Information Precision:
A New Framework
Simon Gervais1
Finance Department
Wharton School
University of Pennsylvania
Steinberg Hall – Dietrich Hall
Suite 2300
Philadelphia, PA 19104-6367
(215) 898-2370
e-mail: ******@
12 April 1997
1I would like to thank Gail Belonsky, Jeff Bohn, Raul Espejel, Robert Goldstein, Steve Hillegeist, Hayne
Leland, Richard Lyons, Jocelyn Martel, Stewart Mayhew, David Romer, Mark Rubinstein, and seminar
participants at l’Universit´e Laval, UC Berkeley, and the 1995 meetings of the Northern Finance Association
and the Financial Management Association for their advice ments. I would also like to especially
thank Matthew Spiegel and Brett Trueman for their numerous suggestions and ments. Financial
Support by the Conseil de recherches en sciences humaines du Canada (Award No. 752-92-0225) is gratefully
acknowledged.
Abstract
This paper provides the financial market microstructure literature with a relatively simple
framework to odate the possibility for the precision of information to be random. Indeed,
many researchers have discovered that the usual normal/exponential framework used in asymmetric
information models is intractable when the precision of information is random. This is why we
introduce a setting similar to that of Glosten and Milgrom (1985) in which shares of a risky asset
can only be traded one at a time.
We find that the bid-ask spread and insider profits both increase with a first-order stochastic
shift of the information precision, but decrease with a second order stochastic shift. On the other
hand, the effect of these stochastic shifts on trading volume is in general ambiguous, since such
shifts produce two counter-balancing effects. However, conditions are derived in which
effect is known. Also, the model’s conclusions are shown to be robust to settings that allow for
multiple trade sizes, insider risk aversion, elacticity of liquidi