What is the intermediary problem?

Many of the most contentious debates in our capital markets relate to how to manage the conflict of interest that information intermediaries are subject to by serving two potential masters. Do sell side analysts work for investors or try to preserve access to management? Are credit rating agencies selling a license to operate, given that the SEC and other regulators require use of rated securities, failing which the user pays some kind of compliance cost? Or are credit rating agencies in the business of providing timely information about potential default? Are proxy advisers working for institutional investors or are they selling ways to companies to improve their governance ratings? And there is much consternation about what ratings put out by ESG rating agencies capture.

On top of that, information gathering is a scale business meaning that the costs of generating value relevant data such as sell side ratings, proxy voting advice, credit ratings and ESG ratings are very high. Any kind of deep fundamental analysis is expensive. To illustrate, in my class, we spend 12 weeks analyzing the business model of one company. Two weeks are devoted to understanding revenues, two each to understanding labor and capacity, and one week each to materials, supply chain, capital structure, M&A (mergers and acquisitions), taxation, foreign currency exposure and a final week to bring everything together in a valuation. Clearly this model is deeply insightful but does not scale well. I am skeptical about claims that AI (artificial intelligence) will somehow make all this easy. Perhaps in the distant future, but not for a while, would be my naïve forecast. On the positive side, my students are smart, deeply motivated and have no ostensible conflict of interest as they are not getting paid in dollars but rather in learning by doing and in the satisfaction of understanding how their analysis routinely outperforms that produced by intermediaries with seven figure paychecks.

The huge costs of generating value relevant information led to two other structural problems. First, the market gets dominated by oligopolies as only two or three entities have the customer base left to defray these high costs. It’s no coincidence that we have three rating agencies (S&P, Moody’s, and Fitch), one dominant ESG rating agency (MSCI) and two main proxy advisors (ISS and Glass Lewis). Second, the high costs of producing value relevant information inevitably create incentives to cross-sell to the parties that are being monitored (the companies themselves) which in turns leads to the conflict-of-interest issues discussed above.

So, what to do?

We need a conversation around how investors and regulators can deal with these conflicts. Institutional investors, to some extent, are captive customers of these intermediaries as they need to comply with legal and fiduciary responsibilities imposed on them by law and custom. The usual ideas of fixing the potential “issuer pays” model in credit ratings and audit fees are not necessarily satisfactory as ESG ratings and proxy advice are paid for by investors, not issuers. To get the conversation started, I propose an investor funded wiki-based solution to the problem. 

A wiki-based solution:

We have 4000 odd stocks trading in the US. Let us start with say the S&P 1500, consisting of the most followed large, medium, and small companies. Create a wiki page for every company. As a pilot, start with a page for proxy advice for every company. The proxy statement for a December year end company comes out in say February and the votes are due before the annual meeting.

Let us randomly allocate 15 of these S&P 1500 stocks to each of the top 100 asset allocators, who apparently control an astonishing $57 trillion of assets. Have these teams produce an evidence-based rating on how an institution should vote on a proxy proposal. Make the data freely available to all institutional investors, who sign up to be part of the wiki. The top 100 allocators have the expertise and resources to do the work associated with understanding idiosyncratic governance issues for at least 15 firms assigned to them. The smaller institutional investor would technically “free ride” off the work of the top 100 but that’s completely fine.

Why? Unlike a trading signal where an investor would have incentives to hoard information, data aimed at improving governance can only help improve the value of the affected stocks for everyone invested in them. This is especially compelling for stocks that are part of a well followed index such as the S&P 500 as institutions are stuck with holding them till they fall off the index. Engaging with the laggards will surely release more value for everyone holding the index.

Having said that, most institutional investors have their own voting policies. For them, proxy advisor research helps facilitate a more efficient comparison of their voting policies to the thousands of proposals that are presented every proxy season. For the wiki research idea to be useful, each of the 100 asset managers in my experiment would probably have to follow a relatively consistent approach and format for analyzing and presenting the research.

What about anti-trust concerns?

The 15 stocks get randomly allocated every year by an uninterested party. These recommendations can be provided anonymously and for free disposal. Meaning that institutions are not required to follow these recommendations. ISS and Glass Lewis can continue to exist and institutions that want to buy their recommendations are free to do so. Ironically, no one seems to worry much about antitrust issues when we discuss the two or three information intermediaries in our capital markets. Why is antitrust a concern when investors share information that will release value for every investor?

Will the non-profit wiki model ever take off?

As one of the senior executives I showed this to remarked, “even sliced bread has to be sold.” But no one gave Wikipedia a chance over a paid for encyclopedia and they seem must done well. Did we underestimate the ability and willingness of informed citizens to contribute to addressing a social need? And there are several industries where the middleman has been disintermediated or at least replaced by a better intermediary that is less biased: travel agency business, Tesla selling cars without dealers, streaming services that have disintermediated video rental businesses, celebrities that reach their audience directly without agents and so on. Why is there relatively little discussion, among the left and the right, about reforming or replacing information intermediaries in our capital markets? The Wall Street Journal occasionally complains about the proxy duopoly without suggesting an alternative.

Do the large asset managers have an incentive to participate?

One objection is that the largest asset managers generally do not need the proxy advisor research and would presumably have little incentive to participate in the wiki research. Fair enough. Except it will release value for the smaller asset managers who cannot afford to invest as much the larger ones in research. And what could be a bigger ESG contribution than participating in a wiki project that benefits all shareholders?

I have, of course, not thought through every design feature of how such as wiki rating system would work. But I think the kernel of the idea that I flesh out perhaps has a shot at meaningfully addressing the conflict-of-interest problem that information intermediaries are subject to in our capital markets. Comments welcome.


This article was originally published on Forbes.com.