Academic Research Insights: Does the Scope of the Sell-Side Analyst Industry Matter?

• Title: DOES THE SCOPE OF THE SELL-SIDE ANALYST INDUSTRY MATTER? AN EXAMINATION OF BIAS, AND INFORMATION CONTENT OF ANALYST REPORTS
• Authors:   KENNETH MERKELEY, RONI MICHAELY AND JOSEPH PACELLI
• Publication: THE JOURNAL OF FINANCE  • VOL. LXXII, NO.3 , JUNE 2017 (version here)

What are the research questions?

1. Do variations in aggregate measure of size and activity of sell-side analysts affect the quality of research produced by that industry?
2. Do those same variations in aggregate measures of size and activity of sell-side analysts affect optimism bias in the research produced?
3. Do those variations extend to sectors where information is easily obtained and modelled?
4. Do those same variations extend the circumstances whereby the decrease in analyst activity is dominated by knowledgeable or “star analysts?”

What are the Academic Insights?

As the first to document the attributes and influences from the collective group of sell-side analysts, this paper examines the relationship between analysts at the aggregate level and the capital markets over a 20 year period.  The authors construct a unique data set from 12.3 million analyst reports from IBES, available from 1990 to 2010, identifying not only analysts but brokerage houses and industry coverage variables.

1. YES.  After controlling for firm-level effects, the authors find that a decrease in analyst coverage results in a loss of aggregate accuracy in forecasting earnings and therefore the information value of the research.  Increases in forecast dispersion were also observed in the months following the drop in analyst coverage. A second tier effect: analysts use the reports of their peers, which affects the quality of their own research.  As the supply of available reports increases both within the covered industry,  the quality effects are observed to “spill over” to non-covered  industries.
2. YES.  Again, after controlling for firm-level effects, earnings forecasts exhibit an upward bias.
3. NOT SO MUCH.  The drop in aggregate forecast accuracy is less significant for industries where there is a greater analyst following and where earnings are more easily forecast (as defined in #2).
4. YES. MORE SO.  The drop in aggregate forecast accuracy and rise in aggregate optimism is more significant when the decline in the number of analysts is dominated by knowledgeable or influential analysts. Consequently, the remaining analysts were observed to exert less effort (produce fewer and less timely reports, fewer forecasts, less likely to initiate coverage of new firms) following the decline.

Why does it matter?

The question of the value added to capital markets from the aggregate group of sell-side analysts has only been addressed at the firm level.  Since analysts are generally considered to be a cost center, producing revenue for the firm only indirectly, it is an open question as to their contribution as a collective group.  The results presented in this article describe the collective group of sell-side analysts as a distinct form of competition separate from competition at the firm-level.  Further, the authors report that decreases in the overall level of analyst coverage at the aggregate level impacts the capital markets negatively.  At least 2 effects are documented:  changes in the accuracy of earnings forecasts and changes in forecasting bias (optimism).

The Most Important Chart from the Paper:

Elisabetta Basilico, Ph.D., CFA, (@ebasilico) is an independent investment consultant. With co-author Tommi Johnsen, PhD, she is writing an upcoming book on research backed  investing. You can learn more at http://academicinsightsoninvesting.com/