Prediction markets strong at forecasting US presidential elections, says new management insights
A case study of the 2004 U.S. Presidential election by researchers at Yale shows that prediction markets are proving to be a strong forecasting tool, one that may have an impact in calling the current presidential contest between Democrat Senator Barack Obama and Republican Senator John McCain, according to the Management Insights feature in the current issue of Management Science, the flagship journal of the Institute for Operations Research and the Management Sciences (INFORMS).
Management Insights, a regular feature of the journal, is a digest of important research in business, management, operations research, and management science. It appears in every issue of the monthly journal.
"Modeling a Presidential Prediction Market" is by M. Keith Chen, Jonathan E. Ingersoll, Jr., and Edward H. Kaplan of the Yale School of Management.
In their study, the authors relate that many firms are establishing internal prediction markets, while public prediction markets increasingly cover all manner of business, economic, and political events. Managers must decide whether to treat these markets seriously, especially when they price complex, interdependent events.
The authors' case study of the 2004 presidential election market suggests that they should. They explore the consistency of security prices associated with presidential election contracts that traded in the Intrade.com prediction market during the run up to the 2004 presidential election. In that prediction market, traders placed bets on various election outcomes such as "George Bush will win both Florida and Ohio" and "George Bush will be elected President of the United States."
The authors find that these prices were mutually consistent with the rules governing the Electoral College, and that traders appeared to quickly and efficiently assimilate new information as it unfolded over the campaign.
Turning to business, the authors suggest that prediction markets can be a valuable tool for managers who face decisions that may depend on the outcome of complex and interdependent events. Established prediction markets will likely do a good job assessing events in which interest is widespread. For events of narrower interest, establishing an internal prediction market may be an effective way to aggregate information within the firm.
The current issue of Management Insights is available at mansci.journal.informs.org/cgi/reprint/54/8/iv
Source: Institute for Operations Research and the Management Sciences