ave accounting scandals (e.g., Enron, WorldCom and Global Crossing) and the 2002 Sarbanes-Oxley Act (SOX) changed management-analyst earnings dynamics? In their December 2006 paper entitled “Mechanisms to Meet/Beat Analyst Earnings Expectations in the Pre- and Post-Sarbanes-Oxley Eras”, Eli Bartov and Daniel Cohen examine whether companies have changed behaviors post-SOX with respect to accrual earnings management, real (transaction-based) earnings management and earnings expectations management. Using earnings forecast and financial data for thousands of companies during 1987-2005, they conclude that:
- Across the entire sample, companies meet or beat earnings expectations 65.3% of the time, but this frequency declines significantly in the Post-Scandal (see timeline below) period. There is a substantial decrease in the frequency of just barely meeting or beating analyst expectations between the Late Pre-Scandal Period (26.8%) and the Post-Scandal Period (21.6%).
- The use of expectations earnings management (based on the proportion of downward estimate revisions that result in positive surprises) declines in the Post-Scandal period relative to the preceding seven years, as does the use of accrual earnings management. There is no change, however, in the use of real earnings management.
- Results are robust to varying economic conditions.
The following figure, taken from the paper, shows the timeline set by the authors to define periods of potentially different company earnings management behaviors.
The paper also provides a survey of research on earnings management.
In summary, earnings guidance/forecasts are probably more accurate (to be precise, less intentionally biased) post-SOX than they were pre-SOX.
