Dispersed Macroeconomic Information: Announcements, Revisions & Stock Returns

Posted: 13 Oct 2008

See all articles by Thomas Gilbert

Thomas Gilbert

University of Washington - Department of Finance and Business Economics

Multiple version iconThere are 2 versions of this paper

Date Written: November 15, 2007

Abstract

I analyze the link between macroeconomic announcement surprises, intradaily returns on the S&P500 Index, and the subsequent revisions to the announced data. I show that announcement-day returns contain information about the future revisions of the released figures. This information is unrelated to the initial announcement surprises and predicts the future revisions: Prices increase when the subsequent revisions will be positive. This observation is strongest for real activity and investment variables such as Nonfarm Payroll, Industrial Production, and Factory Orders. I develop a rational expectations trading model where the final payoff is the sum of non-overlapping fractions of the economy that were previously observed as private signals. A preliminary public announcement does not convey new information to the market per se, but rather allows investors to deduce other investors' information. In turn, this allows them to assess the inaccuracy of the public signal and therefore estimate its future revision. In equilibrium, the risky asset's price changes in anticipation of the public signal's revision, even though the initial surprise may be in the opposite direction.

Suggested Citation

Gilbert, Thomas, Dispersed Macroeconomic Information: Announcements, Revisions & Stock Returns (November 15, 2007). AFFI/EUROFIDAI, Paris December 2008 Finance International Meeting AFFI - EUROFIDAI, Available at SSRN: https://ssrn.com/abstract=1282059

Thomas Gilbert (Contact Author)

University of Washington - Department of Finance and Business Economics ( email )

Box 353200
Seattle, WA 98195
United States
206-616-7184 (Phone)

HOME PAGE: http://faculty.washington.edu/gilbertt/

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
1,175
PlumX Metrics