Forecasting has occupied minds of people for a very long time. In the vein of recent economic turmoil, criticism on the capability and the usefulness of the forecasting industry gained new momentum. Research on biases in financial analysts’ forecasts produced some remarkable studies in the past decades and covered a wide array of potential disruptions. Among them, researchers find the significance of concepts from behavioral economics such as over- and underreaction to novel information and (anti-)herding – a state when people pay more attention to reaction of others rather than to the initial information that causes this reaction in the first place.
Investigating the impact of the economic cycle on the forecast accuracy.
In a comparative study that covers Earnings-per-Share (EPS) forecasts for companies from the U.S. and Germany from almost 25 years, Dr. Jan Peter Schmütsch looks at the nexus of these biases. Specifically, his study “EPS Forecasts: Herding, Optimism, and the Business Cycle” investigates following questions:
- Are financial analysts too optimistic with regards to EPS performance of American and German companies?
- Does the business cycle shape the degree of analysts’ optimism?
- Is there evidence for the herding bias among analysts on both sides of the Atlantic?
The DJIA and DAX indexes serve as the basis for the study.
To shed light on these questions, the study uses a sample of 31 companies from the Dow Jones Industrial Average (DJIA) Index that are covered by 184 analysts over a period of 24 years. In addition, 42 companies of the DAX Index covered by 113 institutions over 17 years are examined. Using different types of statistical and accounting measures and methods, including the (anti-)herding detection method, the study reveals ample evidence for excessive optimism among Earnings-per-Share forecasters for both U.S. and German subsamples (DAX-companies experiencing larger forecast errors). Furthermore, the study shows that in periods of high uncertainty forecasters release overly optimistic estimates. The recession periods show the largest absolute negative deviations and therefore the strongest optimism. Interestingly, the seasonality aspect also appears to shape forecast errors, which can be attributed to the dispersion of new information that is incorporated into later forecasts and diminished uncertainty.
The results of the study confirm widespread optimism.
With regards to herding behavior of analysts, the study finds that except for DAX forecasters from expansionary periods, anti-herding prevails for both indexes and is even more pronounced for longer forecast horizons at any level. Moreover, the bias significantly affects the accuracy of forecasting results for both the U.S. and the German subsamples.
All in all, the study finds that overoptimistic prognoses are more pronounced in times of recessions, with a tendency of financial analysts to anti-herd throughout all forecasting periods. Given these findings, investors should look at analysts’ outlooks with more skepticism during contractionary periods.
Please refer to the original paper for further details: