Grading on a curve in the equity ratings business? ? ?

Wednesday, a major national broker-dealer announced its intention to change its equity research ratings system, starting on June 2 of this year.

The amazing thing about this new system is that it mandates a distribution of ratings within each "coverage cluster" (defined as a single analyst or multiple analysts sharing a common region, sector, or classification). Specifically, the guidelines will mandate that at least 20% of the coverage cluster will be ranked "underperform."

While there may be a perception among the investing public that analysts are reluctant to give negative ratings to the equities of covered companies, this new direction by a large firm closely resembles "grading on a curve." It basically says that in every classroom, the top students (up to a mandated limit) will get high grades, and the bottom students will get low grades.

Consumers of such research must therefore realize that a highly rated company is not highly rated in absolute terms but rather in relation to the other companies in a "coverage cluster," and likewise with negatively-rated companies.

The head of research stated, "The rationale for imposing distribution limits is simple: we want to ensure that our analyst distributions correlate more closely with historical return statistics of stocks." But this is ludicrous! By the very definitions of their ratings, the analysts are assessing the future performance potential of the company in question. They should not be forced to conform their prediction of future performance to a "historical" distribution of some group of companies in the past!

This is like saying that candidates for leadership positions will be judged on their own merits to a certain degree, but that the leadership positions must also be filled in a way that reflects the demographic percentages of the general populace. Why an investment firm would decide to do that with stock ratings is astounding.

It is even more astounding when investors consider the dramatically reduced analyst coverage at many big firms in the wake of their settlement with regulators in 2003 (in which three sell-side firms were accused of issuing fraudulent research reports, and seven others with issuing research reports which "contained exaggerated or unwarranted claims"). If stocks are now being rated based on their relative positions within a "coverage cluster," investors should be aware that many companies they might assume would be in that coverage cluster are not even covered, which further calls into question the concept of grading on a curve.

We have recently shared some of the guidelines that many decades of investment experience have shown to be valid guidelines for selecting businesses with superior prospects for capital investment. These guidelines are not relative guidelines but are based upon crucial, objective standards such as return on invested capital and compound annual earnings growth. We would advise investors that grading on a curve is a bad practice when it comes to deploying their own capital.


Subscribe to receive new posts from the Taylor Frigon Advisor via email -- click here.