44 Pages Posted: 18 Mar 2012
Date Written: March 15, 2012
We examine the role of macroeconomic announcements in shaping the earnings forecasts of equity market analysts. We find that earnings forecasts strongly respond to macroeconomic releases signaling changes in overall business conditions after controlling for analysts’ learning from firm- and industry-specific earnings surprises. We find this effect is industry-specific as analysts’ reliance on macroeconomic news is greater when they follow companies operating in more cyclical industries. In addition, medium-term forecasts respond much more strongly to macroeconomic news than forecasts for the current fiscal year. On average, macroeconomic surprises leads analysts to revise their current year earnings forecasts for cyclical firms by about 3 cents, and subsequent year’s forecasts by about 5 cents. This clearly suggests a strong stock price impact of macroeconomic news via the earnings channel. Macroeconomic reports convey earnings-relevant information, in particular, for longer-term forecasts. In fact, it is highly rewarding for analysts to incorporate macroeconomic information as accuracy of analysts’ earnings forecasts is higher when they put more weight on macroeconomic information in addition to firm-and industry-specific information. Our findings provide a rational explanation of herding suggesting that co-movement among equity analysts may be due to similar responses to public (macroeconomic and firm- and industry-specific) news. Moreover, our findings have important implications for stock valuation models whose performance depends largely on forecast accuracy as well as for recently developed mechanical earnings forecast models.
Keywords: earnings forecasts, equity analysts, forecast accuracy, macroeconomic news, business conditions
JEL Classification: E32, G12, G14, G17
Suggested Citation: Suggested Citation
Agarwal, Vikas and Hess, Dieter, Common Factors in Analysts’ Earnings Revisions: The Role of Changing Economic Conditions (March 15, 2012). Available at SSRN: https://ssrn.com/abstract=2024161 or http://dx.doi.org/10.2139/ssrn.2024161