Home
Scholarly Works
Do Analysts' Forecasts Fully Reflect the...
Journal article

Do Analysts' Forecasts Fully Reflect the Information in Accruals?

Abstract

This study investigates whether financial analysts correctly weight cash flows, accruals and components of accruals in forecasting future earnings. This examination is in the spirit of Sloan (1996) who documents evidence that investors do not correctly distinguish between the cash flow and accrual components of earnings. We find that analysts do distinguish between accruals and cash flows although they generally underweight the information in both accruals and cash flows. More importantly, we find that analysts do not distinguish between discretionary and non-discretionary accruals even though discretionary accruals are less persistent than non-discretionary accruals. Our findings complement and extend the findings in recent studies on analyst forecast inefficiency with respect to the information in accrual and cash flow components of earnings using alternative research designs [Teoh and Wong (1998), Bradshaw, Richardson and Sloan (2000), Barth and Hutton (2000)]. Analysts are considered to play an important role as information intermediaries in educating investors about the future prospects of firms. They are trained in analyzing financial data and have industry expertise as well as detailed firm-specific knowledge through contacts with managers. Thus, one would expect analysts to correctly incorporate the information in earnings components specifically discretionary versus non-discretionary accruals. Our evidence complements evidence in recent studies that raises questions about the ability of analysts, on-average, to correctly incorporate the information in accruals and cash flows in forecasting future earnings. This in turn implies that other outsiders are also likely to find it difficult to undo earnings management via discretionary accruals and therefore provides a rationale for the existence of earnings management.

Authors

Ahmed AS; Zhou J; Nainar K

Journal

, , ,

Publisher

Elsevier

Publication Date

January 1, 2001

DOI

10.2139/ssrn.269213

ISSN

1556-5068
View published work (Non-McMaster Users)

Contact the Experts team