These are the broad findings of the restatements study:

• Over the 1997-2006 decade, restatements grew nearly eighteen-fold, from 90 in 1997 to 1,577 in 2006. However, the increase is largely driven by companies that do not trade on the major stock exchanges. Non-exchange-listed companies account for only 23% of all restatements in 1997, but increase to 62% by 2006.
• Restatement frequencies begin to accelerate in 2001 – well in advance of the passage of the Sarbanes-Oxley Act of 2002. Th is acceleration is likely due in part to the economic downturn about this time.
• The average market reaction to restatement announcements is negative throughout the study period. However, beginning in 2001, the magnitude of market reactions declines notably. This decline coincides with an increase in the number of restatements between 2001 and 2006.
• In particular years, restatement frequencies and market reactions are associated with several disparate factors. These include overall market returns and volatility, regulatory activities, and changes in the mix of underlying accounting issues. Regarding the shift in accounting issues:
  • Restatements attributed to fraud and those affecting revenues tend to have more negative market reactions. However, the percentages of both fraud and revenue restatements decline over the decade. Fraud is a factor in 29% of all 1997 restatements, but only 2% of 2006 restatements. The proportion of revenue restatements also decreases, from 41% in 1997 to 11% in 2006.
  • On the other hand, restatements related to accounting for non-operating expenses, non-recurring events, and reclassifications typically do not have discernibly negative market reactions. Together, these groups represent about 24% of all 1997 restatements, increasing to nearly half at the end of the study period.
• Across the decade, the average restating company increases in size, but remains similar to a comparison group of non-restating companies. Companies of differing sizes tend to restate different accounting issues, and several of the distinctions are consistent with expected variations in the activities of larger versus smaller companies.
• Finally, restating companies are typically unprofitable even before the restatement. In the year prior to announcing a restatement, more than half of restating companies report a net loss.