There has been much debate across the industry over the impact of Mifid II since its implementation. Asset managers, bankers and even some regulators have called for the rules to be rolled back, including the requirement for sell-side research payments to be unbundled from trading commissions. Critics stress that the decrease in research demand negatively affects the level of company coverage, especially for small and mid-cap firms. But the data reveals a different story.
In a new study conducted with the support of Columbia’s Chazen Institute for Global Business, we investigate the impact of unbundling on sell-side research. Examining analyst coverage of public firms in the US and the EU from 2014 to 2018, we find that while coverage of EU public firms has decreased in general (by 7.67 percent relative to the average coverage of these firms prior to Mifid II), the coverage quality has actually improved.
For example, the forecast errors of affected firms has decreased by 18.02 percent relative to the average forecast errors of these firms prior to Mifid II. We also find that individual forecasts have become more informative and capital markets have responded more significantly to analyst forecast revisions. And, contrary to industry concerns, the drop in analyst coverage has not come from small or mid-cap firms but rather is concentrated among large firms. So what story does this tell?
It shows that the unbundling requirement in Mifid II has enhanced analyst competition. To be clear, in response to the new regulation, asset managers internalized research costs, cutting budgets and becoming more selective in terms of what research to purchase. Catering to asset managers’ demand for quality, analysts were then driven to compete more intensively by producing superior research to maintain clients. This explains the simultaneous decrease in coverage quantity and the increase in coverage quality underscored in our research.
As further evidence for the competition mechanism, analysts who were squeezed out of the market tend to be those who produced less reliable research. For example, we ranked analysts by average forecast error and found that analysts who provided less dependable forecasts were more likely to drop out of the research market. At the same time, analysts who have remained in the labor market have produced more accurate research since the implementation of Mifid II.
When looking at the size of the firm in particular, small and mid-cap firms’ coverage remains almost unchanged. By contrast, large firms’ coverage has dropped on average by 10.74 percent. This finding speaks directly to the competition story because, prior to Mifid II, large firms typically had excessively high coverage levels and much of the research had low marginal values. In essence, these research services have been deemed unnecessary after Mifid II. Consequently, they have been competed out of the market, causing large firms’ coverage to drop more significantly.
Unbundling offers a unique opportunity to study the optimal way to pay for information. Putting a price on stand-alone research encourages analysts to compete by providing high-quality research. It also potentially restores analyst incentives because they are being evaluated by the service they actually provide, as opposed to conflating their performance with trading recommendations.
While it is clear that Mifid II has already had a profound impact on sell-side research – increasing coverage quality, if at the cost of quantity – some other important questions remain: do investors benefit from the new regulation, and if so, how? Will research production migrate more from the sell side to the buy side?
Time will tell how much Mifid II remains intact after its exhaustive review and, indeed, more time is needed to study its long-run impact on capital markets.
Yifeng Guo and Lira Mota are PhD candidates at Columbia Business School