February 26, 2019
Estimize is helping to level the playing field for many individual investors or small investment firms that lack access to Wall Street analysts.
Crowdsourced earnings forecasts are more representative of the market's expectation of earnings, especially as the size of the crowd increases.
New research by Gatton faculty Russell Jame, is the first of its kind to validate crowdsourcing financial forecasts.
Jame, along with co-authors Rick Johnston (University of Alabama at Birmingham), Stanimir Markov (Southern Methodist University) and Michael Wolfe, (Virginia Tech), had their article published in the Journal of Accounting Research.
The study, titled "The Value of Crowdsourced Earnings Forecasts," examines the legitimacy of online platform Estimize. Founded in 2011 by former hedge fund manager Leigh Drogen, Estimize gathers earnings and economic estimates from a diverse group of contributors including retail investors, corporate finance professionals, industry experts, amateur analysts and even students.
In their research, the authors suggest that Estimize earnings forecasts are incrementally useful in forecasting earnings, and more representative of the market's expectation of earnings, especially as the size of the crowd increases. When compared to traditional sell-side Institutional Brokers' Estimate System (IBES) forecasts 30 days prior to an earnings announcement, Estimize produces a more accurate consensus 60 percent of the time. That measure increases to 64 percent on the day prior to an earnings announcement.
“I think our findings relate to a growing trend of the democratization of information where consumers are relying less on expert advice and instead turning to their peers,” Jame said.
“For example, most individuals now turn to Wikipedia rather than Encyclopedia Britannica, Yelp rather than Michelin star guide, and Amazon review rather than Consumer Reports. Similarly, many investors are replacing or supplementing the earnings forecasts of professional sell-side analysts with crowdsourced earnings forecasts. Our results suggests this investors who are considering crowdsourced forecasts are accessing more accurate and less biased forecasts. In future research, I think it will be interesting to explore additional benefits of crowdsourced research. For example, does this competitive pressure help mitigate sell-side analysts’ conflicts of interest? Does the new information intermediary improve firm’s liquidity and ultimately lower the firm’s cost of capital?”
Additionally, Jame's research shows that Estimize forecasts are particularly useful because they are less biased than IBES forecasts and occur closer to the date of an earnings announcement. This allows the crowd to incorporate more public information, as well as information that may have been missed by the sell-side.
Jame and his co-authors conclude that Estimize is helping to level the playing field for many individual investors or small investment firms that lack access to Wall Street analysts. It provides a way for these investors to see what everyone is thinking about corporate earnings and in turn, to interpret any earnings surprise. As Estimize's contributor base continues to grow, evidence suggests that the usefulness of the platform will increase as well.
This article was originally published on UKNOW.
The Value of Crowdsourced Earnings Forecasts
Russell Jame, Assistant Professor of Finance and Garvice D. Kincaid Faculty Fellow, Gatton College of Business and Economics at the University of Kentucky
Rick Johnston, University of Alabama at Birmingham
Stanimir Markov, Southern Methodist University
Michael Wolfe, Virginia Tech
Journal of Accounting Research, Vol. 54, Issue 4, pages 1077-1110, September 2016