Earnings predictability and bias in analysts' earnings forecasts.

Somnath Das, Carolyn B. Levine, K. Sivaramakrishnan

The Accounting Review · 1998 · 610 citations · 13 references

DOIFull text

Open access

Concepts

TL;DR

Higher demand for non‑public information exists for firms whose earnings are difficult to predict compared to those accurately forecastable with public data. The paper investigates how the predictive accuracy of past earnings and returns relates to analysts’ optimistic bias, hypothesizing that low‑predictability firms receive more optimistic forecasts. The authors assess whether the predictive accuracy of past earnings and return time‑series predicts the magnitude of analysts’ forecast bias. Results confirm that analysts issue more optimistic forecasts for firms with low earnings predictability.

Abstract

This paper examines cross-sectional differences in the optimistic behavior of financial analysts. Specifically, we investigate whether the predic- tive accuracy of past information (e.g., time-series of earnings, past returns, etc.) is associated with the magnitude of the bias in analysts' earnings fore- casts. We posit that there is higher demand for non-public information for firms whose earnings are difficult to accurately predict than for firms whose earnings can be accurately forecasted using public information. Assuming that opti- mism facilitates access to management's non-public information, we hypoth- esize that analysts will issue more optimistic forecasts for low predictability firms than for high predictability firms. Our results support this hypothesis.

References

13