10.27.10
“Regulatory Spawn” and Proxy Advisors
On October 26th, 2010, a piece written by John Carney of CNBC was published and spread around the internet, claiming that the proxy advisors were similar to the credit ratings agencies in that their judgments were similarly biased (in the case of the proxy advisors “by a small clique . . . whose interest may conflict with clients”) and that the only reason these existed at all was because of a regulatory mandate, which required institutional investors to protect themselves by voting in accord with the advisors’ recommendations. Larry Ribstein of the U. of Illinois College of Law, in his regular blog, “Truth on the Market,” noted the article with approval: http://truthonthemarket.com/2010/10/26/proxy-advisors-as-regulatory-spawn/. I felt that the analogy was not only misplaced, but particularly insidious, because it denied the potential utility of both, and because it deliberately confused the very different business models of the two industries. I responded:
Larry,
Of course, we never had abuses of power by directors and chief executives before the regulators started empowering shareholders, did we?
The problem of the power of the proxy advisors is analogous to that of the credit rating agencies for only one reason: most institutional investors refuse to dedicate any resources to dealing with these issues, so they outsource it. If instead of regarding governance issues as a nuisance artificially created by regulators, they attempted to integrate consideration of them into their investment process, they would (a) be capable of making independent judgments rather than merely trying to cover their rear ends by relying solely on someone else’s opinion, and (b) they would then be in a position to make a fuss only about governance issues they themselves deemed substantive—but they would really make a fuss about those.
There is another important analogy with credit rating: as we saw in 2007 – 2008, it wasn’t that the debt ratings were unimportant, it was that they had been corrupted. And they were corrupted because bond investors relied upon somebody’s measure of risk, but refused to pay anything for it themselves. So the issuers paid instead. Similarly, one may argue that the voting recommendations can be important (yes, most of them are routine, but in that case, there’s no controversy), and that they key issue is only whether the system has been corrupted or not. The analogy to the credit rating system would only hold if the companies themselves were paying for the governance advisors’ recommendations. There have been accusations that ISS has some conflicts of interest due to the fact that they also have a consulting service for issuers, but they claim to be careful to avoid those conflicts, and the complaint Mr. Carney is making is that on the contrary, the recommendations are anti-management. (The other major proxy advisors have no such conflicts in any case.) For the analogy to hold, one must argue that the credit ratings agencies were an unnecessary nuisance because ratings they had issued on a lot of paper were too low, not that they were in fact far too high.
I agree that it would be better if more investors did some research on these issues and thought for themselves, rather than relying entirely upon third-party recommendations. But they would have to be prepared to spend a bit more on staff dedicated to considering these issues from an investment point of view. As it is, there is an almost total lack of coordination between corporate governance and investing at most institutions in the U.S. Then, when something that has been flagged for years goes very wrong, the portfolio managers who had ignored the risk run screaming to their lawyers, and a raft of after-the-fact litigation results. In such cases, any recoveries are paid for by the other shareholders. Is this really the better way to run this particular railroad?
Another contributor, Douglas Levene, responded to my post by arguing that I had avoided the central issue, that there was no economic incentive for institutions to pay attention to agency costs anyway, and they were only doing a minimal job of it because it had been mandated by regulators. Again, I think this is a dangerous misconception that needed to be met head on:
Douglas,
You’re assuming that they are correct in believing (if they really believe) that there is no economic incentive for monitoring agency costs. Actually, there are two: one, that more and more of their clients are demanding it; two, that poor governance is a risk factor, and correlates over the longer term with poor (often spectacularly poor) performance.
I know, there are some studies that claim to demonstrate otherwise, but they were, I believe, poorly constructed in two respects. First, they almost always looked for positive correlations between high governance scores and outperformance; governance is not a positive performance factor for the same reason that being law-abiding is not a sufficient reason in itself to hire someone—it’s an important requirement, not a sign of growth or performance potential. Second, most of the studies were event-driven. As with most risk factors, the event does not occur when the risks are first taken, and certainly not when they are ameliorated—they will occur somewhere down the road. As a veteran portfolio manager, I can assure you that plenty of investors who were told they might be supping with the devil, but thought they would have time to get out safely, ended up losing barrelfuls of money.
Meanwhile, there are plenty of other long-term studies which indicate the opposite, beginning with the landmark study of Gompers, Ishii, and Metrick (2003) which demonstrated an 8.5% annual performance spread from S&P constituents over the 1990s between the highest quintile on IRRC ratings, and the lowest. The sample was re-balanced each year, so that survivor bias should not have been a major factor. My own examination of the data indicated that about two-thirds of this return came from avoiding the low-rated components, which confirms my view that much of the benefit from governance monitoring is in minimizing risk.
There is another issue: many institutions do not pay for an in-house governance program primarily because their directors or trustees passionately hate the idea. They hate the idea because they are also CEOs and directors of other companies which may at some point be the object of governance concerns. It’s a conflict of interest problem as much as it is a problem of demonstrating and quantifying the economic incentives.
Yes: in an ideal efficient market, higher agency costs will translate rapidly into a higher cost of capital and underperformance. The problem is that this assumes (a) that the market is monitoring those agency costs itself, and (b) that real-life capital markets are efficient. I think we have had enough practical experience with the algorithms of efficient-market theorists in recent years to demonstrate that markets are far from perfect, or at least that when they do return to equilibrium it is sometimes with a very long lag and often a violent over-correction. As for the market being potentially a sufficient monitor of agency costs, this is contrary to the author’s hypothesis that monitoring these factors is a waste of time. If everyone is ignoring them, the invisible hand of the market will have no effect. It is true that the market could serve this function in theory, (which would create a free-rider problem), but this issue would not be solved by abolishing (or crippling) proxy advisors, nor by recommending that all investors vote blindly with management.