What type of economic news has the largest and most persistent effect on asset prices?:
How Economic News Moves Markets, by Leonardo Bartolini, Linda Goldberg and Adam Sacarny, FRBNY [pdf]: The Federal Reserve Bank of New York today released How Economic News Moves Markets,... in its series Current Issues in Economics and Finance.
Investigating how the issuance of new economic data influences asset prices in the stock, bond and foreign exchange markets, authors Leonardo Bartolini, Linda Goldberg and Adam Sacarny conclude that only a few announcements—the nonfarm payroll numbers, the GDP advance release and a private sector manufacturing report—have significant and persistent effects. Most of the other data releases examined, the authors find, generate only transitory or erratic responses.
Adopting the methodology used in other recent studies, Bartolini, Goldberg and Sacarny use high-frequency data on asset prices to track the effects of news about thirteen heavily watched economic indicators over the 1998-2007 period. Specifically, they seek to assess how stock prices, bond yields and exchange rates react to the “surprise component” in economic announcements—that is, the difference between the actual value announced for a particular indicator and the value that market participants had expected. The authors observe these reactions over two time horizons: within thirty minutes of the announcement and through 4 p.m. on the day of the announcement.
The analysis yields a number of lessons on the scale and persistence of asset price responses. In addition to identifying the announcements that have significant and longer-lasting effects, the authors determine that bond yields show the strongest response to economic announcements while stock prices show the weakest. As for the direction of these effects, news of higher-than-expected growth and inflation generally leads to a rise in bond yields and the exchange value of the dollar. While the size of news effects on asset prices tends to be consistent throughout the day, the immediate impact can be measured more precisely than the full-day impact.
I wonder if the same is true for polling data and outcomes in presidential elections, i.e. if these data releases move polls and votes more than others?
On the issue of predicting the outcome, see Bread, Peace, and the 2008 Election, by Lane Kenworthy from last February:
Douglas Hibbs’ “bread and peace” model has been extremely effective at predicting the outcomes of U.S. presidential elections. ...
The latest from Douglas Hibbs (June version) is here.