Eight times a year, the 12 members of the Federal Open Market Committee (FOMC) meet to discuss current monetary policy and the risks of pursuing price stability and sustainable economic growth. Though the members may have diverse views, the committee has to decide on a single forecast for different key economic variables, such as for what next year’s inflation rate will be.
New economics professor, Xuguang Sheng, is particularly interested in why—and how—professional forecasters disagree based on similar information. “If you look at economic forecasts for things like GDP and inflation rate, you’ll see that these professional forecasters have access to the same information, the same econometric models,” explains Sheng. “You expect that they’ll give similar forecasts. But this is not the case. The question is why these forecasters disagree.”
This is a question that Sheng has explored for the past few years. In a co-authored article published in the Journal of Econometrics, Sheng helps explain that there are two primary sources of forecast disagreement: differences in the forecasters’ prior beliefs and differences in their interpretation of the same public information. The paper won the Heinz König Young Scholar Award from the Centre for European Economic Research.
The paper explores why economic experts, such as the members of the FOMC, have different opinions, and whether there is a relationship between these different opinions. It further explores whether this potential relationship could have negative implications for future economic development. Sheng and his co-authors propose a method for gauging the underlying uncertainty surrounding a point forecast. For example, this month, the Congressional Budget Office predicted that consumer price inflation rate in 2011 will be 1.2%. If the current uncertainty is similar to that seen, on average, in the past, their analysis suggests a probability of about 70% that inflation could be between 0% and 2.4% in 2011. This broader interval could help yield more accurate economic forecasts.
Besides his research on forecasting, Sheng has several other projects in the works, including a few that deal with more theoretical concepts that pique his interest. For one, he is currently working on how to optimally combine probability values in a panel data context. In another, he is interested in developing new tests for panel unit root and co-integration—which is a frontier topic in econometrics.
Additionally, Sheng is writing a paper that proposes a new method to measure forecast uncertainty for financial analysts’ earnings forecasts--a method that has never been used in accounting literature, according to Sheng. “Some methods in economics are standard, but never reach the field of accounting. We want to get this message out in a big way.”