Robert Pozen, the chairman of MFS Investment Management and author of the newly published Too Big to Save, argues for smaller boards that meet more frequently and a professional class of directors who commit to serving fewer companies. Pozen has a varied background in public and private finance. The former vice chairman of Fidelity Investments served as Secretary of Economic Affairs for Massachusetts Governor Mitt Romney and is now a senior lecturer at Harvard Business School. What follows is an edited transcript of an interview with Pozen conducted by Mark Preisinger, the director of corporate governance at The Coca-Cola Co., at The Directorship Forum in November.
Do we, in your opinion, have a good rationale for when we bail out institutions?
The short answer is that we’ve bailed out too many institutions. If we had a good rationale, then we might decide that there are 10, 20 or even 30 financial institutions that are too big to fail. But we’ve recapitalized over 600 institutions. We need an articulated rationale for these bailouts. In my view, there are two good reasons for rescuing a troubled bank. First, if it is critical to the functioning of the payment system—the processing of checks and wires—and, second, if the insolvency of this institution would probably cause widespread failures in the entire financial system. That probably was true in the cases of Freddie Mac and Fannie Mae.
We also need a more disciplined process. Right now, if you yell systemic risk in a crowded room, we bail you out. The Treasury Secretary needs to write down on a piece of paper the specific reason for bailing out this institution, and make that public. Then we should have an independent body like the GAO do a review of every bailout after the fact. Through that process, we could start to develop some sense of whether these bailouts achieved their objectives.
It also seems that when we do these bailouts, Treasury is taking preferred stock with warrants. Does that make sense, particularly from a taxpayer perspective? Is that something we all should be concerned about?
This is one of the main themes of my book. I call it one-way capitalism. If taxpayers are going to bail out these institutions, we own the downside. However, we really don’t own much of the upside…If we’re going to bail out institutions in the future, the Treasury should take back a lot of warrants, or perhaps in some cases common stock. When JP Morgan redeemed the Treasury’s preferred stock, the Treasury realized almost $1 billion of profit on its warrants. But we as taxpayers should have received more than six times the amount of warrants and six times the amount of profits. We need those profits on the successful rescues to offset the losses that taxpayers are likely to incur on AIG and Bear Stearns.
How much blame do the boards have for what went wrong in the financial crisis?
I don’t think that the boards bear the most blame for the financial crisis. There are a lot of other groups and a lot of other factors that were more important. But on the specific question of the compensation system, a lot of boards dropped the ball. They approved a lot of bonuses based on one-year performance, and that performance soon evaporated. They didn’t require deferral of cash bonuses in many cases, and they often went along with guaranteed contracts and golden parachutes regardless of performance.
So, let’s say we get it right, and boards are composed the right way. Can they alone hold management accountable, particularly in financial services?
When we think about holding management accountable, we need to think about both the bondholders and the shareholders, as well as boards. One of the most unfortunate things about this financial crisis is how we’ve taken bondholders out of play. Financial institutions have issued $340 billion of guaranteed debt, in which the government is guaranteeing 100 percent of their debt. We have bailed out the Bear Stearns’ bondholders, some of the most sophisticated investors in the world. We’ve also directed AIG to pay out 100 cents on a dollar to very sophisticated investors on the other side of its credit default swaps.
If large bondholders never take a loss, they are going to stop being careful in choosing bonds; they are not going to push management to avoid excessive risks. This is moral hazard in the worst sense. If we want to hold bank executives accountable, we need to bring large bondholders back into action. That can best be done by requiring all large banks to issue subordinated debt, which would not be protected by the federal government if such a bank became insolvent.
The most important change in the shareholder area is one that’s already happened—stricter rules on when brokers may vote the shares held in the accounts of their customers. In most public companies, the “broker vote” involved 30 to 40 percent of the outstanding shares. Under prior rules, the brokers didn’t need instructions from their customers, and they tended to vote for management. Now, under the new rules, brokers can’t vote without their customers’ instructions, and therefore, they usually won’t vote. So that means a substantial shift of power from management to institutional investors, who normally vote their shares in every corporate election. Hopefully, institutional investors will use that new power intelligently.
What you’re suggesting relative to smaller boards and more focus on the business, is that applicable to boards across the board?
I would think that the new model should apply to very large and very complex companies. If you’re a director of such a company, you really should spend a lot more time on the board, and you should try to get on top of what the company is doing. In such a company, the board should be more professionalized.
There’s been a lot of discussion about the failure of the boards of mega-banks to adequately assess risk. How serious a problem is this?
The directors of many megabanks do not seem to have fully understood the risks being taken by these banks. In part, this may have happened because of undue reliance on internal risk models like VAR (variance at risk). VAR measures a bank’s risk exposure over a very short period, like a day or a week. Moreover, VAR covers risks only to a 98% degree of profitability. In other words, it does not deal with risks beyond the second standard deviation in a normal distribution curve.
However, the most important risks often materialize over much larger time periods, such as a year. And some of the most catastrophic risks have a probability of less than 2%, but they can destroy a company. So directors should not be lulled into complacency by risk models like VAR. Similarly, under Basel II, regulators currently allow large banks to set their own capital requirements based on their own internal assessment of the riskiness of their assets. This approach has at least three flaws. First, Basel II has a built-in conflict of interest – banks have an incentive to set their capital requirements at relatively low levels.
Second, the internal risk models of many banks turned out to be wrong – for example, some assumed U.S. housing prices would fall only once in 50 years. Third, these risk models are so complex that they cannot be understood by most bank directors, unless he or she happens to have a PhD in math from MIT.
Directors should insist on a clear delineation of all the assumptions underlying these risk models, and then push back hard on the validity of these assumptions. In the final analysis, quantitative models are not substitutes for common sense.