The Bid-Ask Spread's Cost Components: Theory and Evidence

Posted: 2 Oct 1996

See all articles by Steven V. Mann

Steven V. Mann

University of South Carolina

Cynthia G. McDonald

University of Missouri at Columbia - Department of Finance

Pradipkumar Ramanlal

University of Central Florida - College of Business Administration

Date Written: September 1996

Abstract

We develop and test a model that provides improved estimates of the bid-ask spread's cost components: order processing, adverse selection, and inventory control. The model incorporates three unique features: (1) a dealer's response to inventory imbalances is not static but depends on the size of the imbalance and the dealer's aversion to inventory risk; (2) active inventory management by a dealer will result in a stationary stochastic process for inventory; and (3) inventory management will influence the adverse selection cost component. We estimate the spread's components using intraday data for NYSE/AMEX and NASDAQ stocks. We also examine the impact of our model's features on the cost estimates. The results suggest inventory costs are higher and order processing costs are lower than previously reported.

JEL Classification: G10, G12, D82

Suggested Citation

Mann, Steven V. and McDonald, Cynthia G. and Ramanlal, Pradipkumar, The Bid-Ask Spread's Cost Components: Theory and Evidence (September 1996). Available at SSRN: https://ssrn.com/abstract=7779

Steven V. Mann

University of South Carolina ( email )

Francis M. Hipp Building
Darla Moore School of Business
Columbia, SC 29208
United States
803-777-4929 (Phone)
830-777-6876 (Fax)

Cynthia G. McDonald (Contact Author)

University of Missouri at Columbia - Department of Finance ( email )

120 Middlebush Hall
Columbia, MO 65211
United States
(573) 882-0638 (Phone)
(573) 884-6296 (Fax)

Pradipkumar Ramanlal

University of Central Florida - College of Business Administration ( email )

PO Box 161400
Orlando, FL 32816
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
1,389
PlumX Metrics