A Study on the Conservatism and Analysts' Optimistic Bias
Asian Tax Journal Vol. 11 No. 4 (2010), pp. 323-357
Abstract
Forecasts error(here after, FE) is calculated by subtracting earnings per share(here after, EPS)from analysts' earnings forecasts(here after, AEF), where AEF is said to be optimistically biased if FE is positive. In most previous studies, they believe that positive FE is driven from optimistic bias in AEF only, not considering the effect of accounting earnings. However, FE is likely to also have a positive value if economic losses are more asymmetrically reflected in EPS, based on accounting conservatism, than economic gains, even if the analysts' optimistic bias is not considered at all. Therefore, EPS should be considered when FE is calculated. Otherwise, it is likely to get the wrong conclusion that there exists analysts' optimistic bias in AEF although analysts' optimism in AEF does not exist. Accordingly, the purpose of our paper is to examine if conservatism that is reflected in EPS can cause the optimistic bias in AEF (namely, positive FE). To achieve the purpose, we divided the sample into three portfolios from NP1 to NP3, each of which has the similar magnitude of AEF but the magnitude of FE increases from NP1 to NP3. Therefore, because AEF of each portfolio is similar, it is likely that FE is related to accounting conservatism, if the observed conservatism in NP3 is higher than any other portfolio. To test empirically the hypothesis, we have used Ball·Shivakumar(2006) model with current cash flow(CFt) as the variable to distinguish economic gains and losses. The result shows that the larger conservatism is observed in NP3, leading to the conclusion that positive FE could be driven from accounting conservatism. And the above result is not changed even when we have used, as the variable to distinguish economic gains and losses, other variables such as RET, change in cash flow(ΔCFt), and industry －adjusted cash flow(INDCFt), respectively and other models such as Basu(1997) and Hwang·Lee·Nam·Park(2008).
Keywords
- Accounting earnings
- conservatism
- analysts' earnings forecasts
- forecast error
Related Articles
The Effect of Book-Tax Difference, Temporary Difference, and Changes in Net Deferred Tax Assets and Liabilities on Analysts' Earnings Forecasts
10(2) 225-260
Tax Avoidance and Corporate Transparency
20(2) 95-123
The Effect of Audit Quality on Quarterly Earnings Management
13(3) 9-51
The Annual Earnings Distribution's Characteristics in Small Cumulative Profits Firms at the End of the Third Quarter －Focused on the Income Smoothing－
12(1) 229-264
Firm Location, Analyst Following, and Bias in Analysts' Earnings Forecasts
11(4) 275-304