Forecasting the Macroeconomy: Analysts versus Economists

• In Event: Financial Analysts II

Abstract

We investigate whether and to what extent aggregate earnings forecasts by sell-side analysts and forecasts of macroeconomic indicators by economists convey different information about the macroeconomy, and whether such differences have implications for forecast efficiency and the stock market. We find that the two sets of forecasts strongly covary over the 1984 to 2010 period, suggesting that they contain a non-trivial amount of common macroeconomic information. We also find that while real GDP growth forecasts have incremental predictive ability over aggregate earnings forecasts with respect to future aggregate earnings, the converse is not true. Additional tests suggest that analysts underreact to economists’ negative forecast revisions (i.e., aggregate earnings forecast errors are predictably more negative following economists’ downward forecast revisions), with the extent of underreaction more pronounced for more procyclical and larger firms. The stock market does not appear to see through such inefficiency—the market returns surrounding the first few “bellwether” firms’ earnings announcements are predictably more negative following quarters with more negative macroeconomic forecast revisions. Finally, we show that an “adjusted” aggregate earnings forecast that incorporates macroeconomic forecast revisions performs marginally better in terms of forecast accuracy and efficiency.

Authors

• Rebecca Hann, University of Maryland, College Park

• Maria Ogneva, University of Southern California

• Horacio Sapriza, Board of Governors of the Federal Reserve System