| Summary: | 碩士 === 中原大學 === 會計研究所 === 91 === This study discusses the relationship between earning forecast revisions,earning surprises and forecast errors, in an attempt to explore expectation management, earning persistence, persistence of meeting or beating earning forecast
(MBE), and the integration of meeting or beating earning expectation, expectation management and earning management. Moreover, we will explore whether investors regard the above mentioned.
This study concludes by multiple regressions that investors reward earning surprise, i.e. meeting or beating firms' management forecast. Modified Jones Model was used to estimate discretionary accrual components.
The empirical study concludes that:
A. Expectation Management: The empirical results are consistent with the expectation management hypothesis. A) The percentage of negative earning
surprises over the entire sample size is lower than the percentage of negative forecast error. In addition, the difference in percentage between negative earning surprises and negative forecast errors increases over time. B) The proportion of
firms with positive or zero forecasts that end with negative surprises is smaller than the proportion of firms with negative forecast errors that end with positive surprises, expectation management exists and the trend increases over time.
B. Rewards to meeting or beating firms' mandatory management forecast: Empirical study shows that with forecast errors under control, every 2.5% incremental increase of forecast error, CAR associated with a favorable earning surprise is
higher than that associated with paths associated with unfavorable earning surprise. This difference is not statistically significant.
Investors' reward to firms beating earning forecast is more than investors' penalty for failing to beat earning forecast. Also, the rewards are different for firms meeting earning forecast and firms beating earning forecast. Both hypotheses were supported and statistically significant.
C. Earning consistency: Empirical results supports that investors give higher rewards
to firms with positive earning, increased earning and not lowering earning forecast.
D. Meeting or beating earning forecast: Investors give higher rewards to firms frequently beating earning forecast.
E. Integration of meeting of beating earning forecast, expectation management and earning management: A) Expectation management: Empirical evidence supports the existence of expectation management. Investors give lower rewards to firms
meeting or beating earning forecast when they are aware of the possibility of firms expectation management to earning forecast. It is not supported due to its statistically insignificance. B) Earning management: Empirical evidence indicates that investors give lower rewards if they are aware that firms meet or beat earning forecast by earning management. Moreover, we calculated using random walk model and obtained results which are not significantly different.
|
|---|
