Slow Motion Earnings Revisions on Wall Street

Response Times Range from 9 Minutes to 5 Days

Speed matters on Wall Street, but information processing delays can range from 9 minutes to nearly five days when news triggers a revised earnings estimate for a publicly traded company. Research from the University of Maryland’s Robert H. Smith School of Business explains the range in revision processing times and confirms the value of companies that collect information from individual analysts and distribute consensus estimates to investors.

“Three broad groups of variables explain the activation delays,” says Smith School accounting and information assurance professor Musa Subasi, co-author of the paper featured in the Journal of Financial Economics.

Using data from the Thomson Reuters Institutional Brokers’ Estimate System, the study examines nearly 1 million forecast revisions covering U.S. publicly listed companies between 2003 and 2013. Time stamps on each forecast revision allow the authors to measure the gap between its announcement by the analyst and activation on Thomson Reuters products, and cross-reference the delay with other variables.

One major factor that influences revision processing times is investor demand. Larger firms with more investors tend to get higher priority treatment. “There’s more pressure to process these firms first,” Subasi says. “It’s based on a ranking in the queue.”

The second factor is processing difficulty. Estimates can be based on generally accepted accounting principles (GAAP earnings) or actual earnings that exclude nonrecurring items like layoffs and restructuring. “How big is the surprise?” Subasi says. “When the difference between actual earnings and GAAP earnings is large, analysts need more time to reach a consensus.”

Major one-time events like stock splits, mergers and acquisitions also add complexity and correlate with longer activation times for revised earnings estimates.

The third factor relates to the limitations of human resources. Analysts and investors get stretched to capacity when several companies report major events at the same time. “Activation delay increases on days with larger numbers of concurrent announcements,” Subasi says. “You end up with a bottleneck.”

Overall, he says, the Thomson Reuters Institutional Brokers’ Estimate System holds up to scrutiny. “They’re following the manual,” Subasi says. “The system is very transparent.”

Read more:
Determinants and Consequences of Information Processing Delay: Evidence from Thomson Reuters’ Institutional Brokers’ Estimate System is featured in the Journal of Financial Economics. Authors include Ferhat Akbas from the University of Kansas, Stanimir Markov from Southern Methodist University, Musa Subasi from the University of Maryland, and Eric Weisbrod from the University of Miami.