Interview with Ron Feldman
Read the 2016 Annual Report Essay and see the Full Report
Published April 17, 2017 | April 2017 issue
“Too big to fail” is intrinsic to the Minneapolis Fed. For 40 years, our economists and senior officers have analyzed and advocated policies to reduce systemic risks associated with large, complex, and interconnected financial institutions. First Vice President Ron Feldman has been an architect of the Fed’s TBTF effort for nearly two decades. In the following conversation he discusses the evolution of TBTF at the Minneapolis Fed.
Origin of Minneapolis Fed TBTF concerns
Region: What was the original impetus for the Minneapolis Fed’s research and writing about too-big-to-fail banks?
Feldman: I was not here at the start, but I think the original impetus was around the issue of moral hazard associated with insurance. Moral hazard has been an issue for a long time for economists who have studied any kind of insurance—auto insurance, home insurance, health insurance. It’s the idea that when you provide insurance to someone, you’re protecting them from loss and, therefore, since they don’t bear the loss of the event, they’re going to be a little less diligent than they were before. So, for example, if your car is stolen, the auto insurance company has to buy you a new car, so you may not lock the door of your car as often as you used to.
That’s been known by insurance companies for a long time, and economists have been studying it as well. People here at the Minneapolis Fed looked at it in the context of insurance that the government was providing for financial institutions and their customers: deposit insurance. Deposit insurance means that when a bank fails, your deposit is covered and protected—below a certain amount, you’re going to get your money back.
My understanding from folks who were here at the time is that policymakers were also worried about the moral hazard from deposit insurance. Policymakers at the Board of Governors—including [Chairman] Arthur Burns—were concerned that because the insurance for deposits was so extensive, people who had deposits at a particular bank didn’t really care if the bank was in good shape or not. The worry is that if people who have money at the bank don’t care about the condition of the bank, the bank can take a lot of risks. More bank risk-taking can lead to bank failures, which can be costly for society as a whole.
I should say that while we’re talking about this in the context of the Minneapolis Fed, this was an issue that FDR talked about when deposit insurance was first created. So this was a really well-known issue.
People here, particularly Neil Wallace and John Kareken, decided to model moral hazard and deposit insurance more formally in the late ’70s, early ’80s, and showed that if you have 100 percent protection for the depositors, the bank has an incentive to take a lot of risk. So that was the impetus for their work. It wasn’t really about too-big-to-fail or large banks, per se, but it was about insurance for depositors in general.
Region: What policy advice flowed from this research?
Feldman: Well, again, this is not a new problem that economists just ginned up. It’s something that insurance companies have known about. When you buy an insurance policy, it doesn’t typically cover 100 percent of losses or costs. The policy typically has co-insurance. There’s some amount, that when you make a claim, you have to pay yourself. The idea is that if you’ve got some skin in the game, you’ll remain vigilant. There are other features of insurance policies, like deductibles, which also encourage the insured parties to bear part of the cost of their claim and thus take actions to avoid making the claim in the first place.
Economists here, John Boyd, Art Rolnick and others who were working on the issue, tried to borrow from what they saw in insurance markets and argued, “If there’s going to be deposit insurance, depositors shouldn’t be fully covered. There should always be a little bit of loss that the depositor has to bear.” In practical terms, if my family’s got $500 at the bank and the bank fails, maybe I would have 10 percent, or $50, at risk. So that was one policy prescription.
A second one was, if you’re worried about banks taking a lot of risks, don’t give them a lot of extra powers. This was during a period when people were considering—and actually implementing—expanded powers for banks. One of the policy prescriptions from the Minneapolis Fed was that before you give banks new powers that let them take on new risks, make sure that you’re not giving 100 percent deposit coverage. New powers with extensive government support will just induce more risk-taking.
Expanded powers at the time were, for example, allowing banks to pay more for deposits—for example, offering higher interest rates on saving accounts. But you might also think about this argument during periods when policymakers were allowing banks and bank holding companies greater ability to engage in insurance- or securities-related activities.
Region: So that was Kareken’s idea of, “Don’t put the cart before the horse.”
Feldman: Yes. Fix the issue with the deposit insurance before you give banks the “cart,” which would be new powers.
How did TBTF evolve from moral hazard concerns?
Region: The earlier focus was mainly on moral hazard generated by deposit insurance, but we now talk about TBTF—financial institutions that are considered too big or interconnected for the government to allow them to fail. What’s the link? How did TBTF concerns evolve from worries about excessive risk-taking created by insurance?
Feldman: One very important impetus for this change was the failure of Continental Illinois in 1984 and the generous support the federal government gave to Continental’s creditors. Continental was the nation’s seventh-largest bank at the time. Not only were Continental’s insured depositors protected, but the government also provided extensive protection to other creditors of the bank even though they had no formal, explicit insurance. Moreover, government officials declared that other nominally uninsured large-bank creditors would receive similar extraordinary government support if those large banks got into trouble. One congressman used the term “too big to fail” to describe Continental and these other large banks after Continental’s failure.
So it became clear that our concern about too much government insurance should not be focused exclusively on insured depositors. It made more sense to worry about uninsured creditors getting bailed out. After all, the federal government doesn’t even offer them explicit support. If they come to expect bailouts—because government bailed them out previously—the idea of imposing losses on anyone seems far-fetched. And then we would have a system with the 100 percent insurance and risk-taking—exactly what Kareken and Wallace warned about.
Further solidifying our focus on TBTF was a law put into effect in 1991: the Federal Deposit Insurance Corporation Improvement Act, or FDICIA. That law went a long way toward trying to reduce the amount of protection that the FDIC could provide uninsured depositors in routine cases. The idea was that the FDIC had to provide the least costly way of addressing a bank failure, and that made it very difficult for the FDIC to bail out uninsured creditors.
But FDICIA allowed an exception to this policy. In cases where there was a threat to financial stability, the government could in fact protect uninsured creditors at a bank, just as it had for Continental. It was more complicated in practice, but that is the gist.
Proponents of the exception argued that it would not be used because it was too onerous for policymakers to implement. To obtain the exception and provide broad protection, the government would have to get approval from many regulators in a very public way. Supporters of this exception thought that laborious process would discourage its use. We disagreed. We argued, in effect, “This exception is a really bad idea because it is going to be used to protect people at exactly the kinds of institutions where we don’t want bailouts because they’re so big.”
In our 1997 Annual Report essay, “Fixing FDICIA,” Art Rolnick and I recommended requiring a small amount of co-insurance for bank creditors when their bank fails. And we also said, when you put creditors at risk of loss, you can see from their behavior which banks they think are risky, and that can help you figure out which banks actually are risky and how much they should be paying for risk insurance.
Region: You also recommended greater financial disclosure, I believe.
Feldman: Right. Again, we were really trying to build off what we saw in private markets. If you want people to be able to judge the quality of their bank, ensure that the bank reveals more about itself and what it’s doing.
Region: Could you explain a bit more about pricing risk insurance premiums for banks?
Feldman: One way you can prevent banks from taking too much risk because of the insurance they get is to charge premiums that vary according to the risk of the insured. Again, we can see this in private insurance markets—if you’re someone who smokes, you pay a higher premium for health insurance than someone who doesn’t smoke. So if you’re a bank that’s risky, you should be paying a higher premium for the coverage you get than a bank that’s less risky.
How do you tell who’s risky and who’s not? A lot of people working for the government are looking at banks, and those analysts can give us their assessments. But we can also look to prices from markets when it comes to assessing the riskiness of a bank. Say people lent money to a certain bank, which is what depositors or bank bond holders are doing. If those people thought they would lose money if the bank failed, they would have charged higher interest for that loan than for a loan to a safer bank. The government can look at such prices as signals of the relative riskiness of banks.
Old and new in TBTF
Region: In 2004, you and Gary Stern published Too Big to Fail. How did that book evolve from some of the ideas you’ve just described, and what was new?
Feldman: Some things were similar and some things different. I think the most important thing that we tried to focus on—which sounds a little obvious now, but at the time it was new—was the rationale for protecting bank creditors in the first place. In some of this prior work, we didn’t even discuss that. We just assumed, “Oh, they get covered.”
What Gary and I tried to think through were the things government would do to make policymakers feel more comfortable in not protecting creditors.
The basic idea was, if you just pass a law—and this was the direction that people advocated at the time—saying, “We forbid coverage; we forbid bailouts,” that doesn’t work, as I suggested before. The reason governments provide bailouts isn’t just because they’re legally allowed to. It’s because they think the bailout will protect the economy from bigger harm.
I should also say, the premise of the whole book, which was also novel for the time, was that this TBTF issue hadn’t been fixed, that this exception process I talked about before was going to be used. There were going to be more Continentals. People thought it wasn’t going to be used and, again, we said, “No, it is going to be used. Banks are bigger than they’ve been before, they’re more complicated than they’ve been before, and because they’re bigger and more complicated, policymakers are more uncertain than they’ve ever been about what will happen if one of these institutions gets into trouble.”
Region: So the book was introducing the idea that policymakers need a credible commitment to not provide bailouts.
Feldman: Right. If you’re going to tell people that you’re not going to bail out a bank, you actually have to change something, the facts on the ground. We said there are two things you have to do. You have to make policymakers more knowledgeable about what would happen if a bank got in trouble, and you have to reduce the possibility that one bank getting in trouble would spread to the rest of the financial system and affect other banks.
Lessons from the Great Recession
Region: Unfortunately, just a few years after you published Too Big to Fail, the nation experienced the largest financial crisis and economic downturn since the Great Depression. What did the financial crisis and recession teach you?
Feldman: I think in large part, it validated what we’ve been working on here at the Minneapolis Fed for 40 years or more; it certainly validated the idea that we’ve got a big problem when a large bank gets into trouble.
It also validated the idea that policymakers act because they’re uncertain. Neel Kashkari, our current president, was involved in the response to the crisis at the Treasury Department. He’s told us all about how worried people in government really were. They didn’t know what was going to happen if some of these institutions went down, and so they had to act.
We also said, for example, that these banks need to have more protection against loss so that losses don’t spread from one to another. We need to think more carefully about what would happen if banks did get in trouble. That general idea was validated.
The crisis also made me rethink some of the old arguments I had engaged in on TBTF. I mentioned earlier that moral hazard had been a jumping-off point for our work on government support for bank creditors—again, the idea that if you’re insured, you take more risk. But before the crisis occurred, many policymakers were never convinced that moral hazard was much of an issue. They did not believe that banks take on more risk because bank creditors perceive they have government support.
And these same policymakers found moral hazard concerns expressed during the crisis to be downright dangerous. Those worried about moral hazard had the following concern during the crisis: “If you protect creditors now, that will make the moral hazard problem worse and lead to more risk-taking by banks in the future.”
While I understand this concern, it struck those accountable for the response to the crisis as off-base. Once you are in a crisis, you have to decide if you will really let the financial system and the rest of the economy collapse to ward off future risk-taking. That trade-off seems pretty clear to me, and I agree with our leaders who did provide support once the crisis got going. The time to try to fix TBTF is before a crisis, not during it.
But, and this is critical, it ended up that policymakers who did not have major moral hazard concerns before the crisis or during the crisis still supported many of the ideas that Gary and I advocated in our book. In that sense, it doesn’t really matter whether or not you believe that being insured is the cause of risk-taking at TBTF banks. If you have institutions that are so complicated, intertwined and big that when and if they get in trouble, they’ll bring everyone else down, then it doesn’t matter whether insurance led to more risk-taking.
The thing that really matters is that once these financial institutions get into trouble, if it spreads to everyone, then you’re going to have a bailout. The focus on moral hazard may not be critical to actually advancing useful reforms on the TBTF issue. Put another way, framing the TBTF discussion as one motivated by moral hazard obscured important areas of agreement—at least after the crisis—about how to actually fix the problem.
Preview of the Minneapolis Plan to End TBTF
Region: The Minneapolis Fed has led an initiative to end—to truly end—TBTF. It’s been a year-long process. You’re going to release the final report. Can you give us a quick preview? What are the key points?
Feldman: I think the main point of the report harkens back to some of the things we’ve just talked about—that the problem hasn’t been solved. That is, there are institutions that are too big to fail. That we need a credible way of actually imposing losses on the people who give money to the bank instead of on taxpayers.
And the specific solution we’re proposing is that instead of imposing losses on people who lend money to the bank—bond holders specifically, in this case—you ought to put the losses on equity holders. Equity holders are people who give money to a bank, who become part owners of the bank, but who don’t have any legal right to get their money back. We want more equity in banks. Historically, equity holders actually do take losses in the United States when a bank gets in trouble.
Region: And what are the other elements of the Minneapolis Plan?
Feldman: There are three elements overall. The first is requiring banks to hold more equity, as we’ve just discussed.
The second is a tax on what we’re calling shadow banks—financial firms like asset managers, finance companies and hedge funds, many of which do not face bank-like supervision and regulation. The idea here is, if you make it harder to get bailouts in the banking sector, some of the banking sector’s riskier activities might flow to nonbank financial firms and, as a result, we could see cases where our higher equity proposal just leads to unmanaged risks elsewhere.
To avoid that outcome, we will tax shadow banks’ risk-taking. We do that by taxing their leverage—that is, the funding of their activities through borrowing. More leverage is associated with more spillovers into other firms and the economy when the financial firm that’s borrowing gets into trouble. If these firms are going to fund their activities with debt, they’re going to pay a tax, and that tax is going to make the cost of funding a shadow bank the same as funding a regular bank.
That latter point is key. Making the funding costs the same would discourage the flow of risky activity away from regular banks to shadow banks. In an ideal world, we would just raise capital requirements for shadow banks to equal those of regular banks. But there are so many types of shadow banks with so many disparate capital regimes that this option did not seem viable. The tax was a way to get to the same place without that complexity.
And the third part of the Plan is for community banks. We think that currently, supervision and regulation of community banks is not focused on real risk-taking, but instead is focused on activities for which supervision is costly but provides relatively little benefit. And we think there has not been an effort to really fix that because no one wants to relax regulations on banks, even small banks, if big banks still are too big to fail.
So, once we fix the TBTF problem, we want to go in and focus on where smaller bank risk-taking can truly lead to problems, like if they grow too fast or have excessive exposure to one type of asset, such as commercial real estate. If we focus on those issues, we can then try to remove other regulations focused on activities that are relatively less risky, for which supervision has less payoff.
Connection between the Minneapolis Plan and previous Minneapolis Fed TBTF work
Region: What is the connection between the Minneapolis Plan and this long history of research and analysis at the Minneapolis Fed?
Feldman: I’ve alluded to it a little bit, but I think the main connection is, we don’t think the current system is going to work because it’s not credible. We talked earlier about how the book with Gary focused on this idea of credible commitment, that we’re not just saying we’re not going to bail out individual depositors or creditors of banks. We’re actually going to take steps on the ground to reduce the incentive to provide bailouts.
The current regime says there are some people who have lent money to banks—these are the special bond holders I mentioned earlier—and when banks get into trouble, we’re going to impose losses on these bond holders. That hasn’t happened historically, and we don’t think it’s going to happen in the future, so that strategy isn’t credible. Why not? Because if you impose losses on these bond holders, other debt holders get really nervous. If they can, they get their money out. Specifically, lenders will respond by charging much more to lend to the bank or selling the debt they have at low prices. These reactions all create more instability.
What is credible is imposing losses on equity holders. So the essence of the Plan is actually closely linked to the history we have of trying to make sure that this isn’t just talk; we’re actually going to change the facts on the ground.
Will the Minneapolis Plan end TBTF?
Region: It’s been 40 years, nearly to the month, since Kareken and Wallace wrote their first paper on this problem. Are you optimistic that the Minneapolis Plan will finally solve TBTF?
Feldman: I’m optimistic that if you implement the Plan, it will solve TBTF. So the question is, am I optimistic that the officials who would have to implement the Plan—our elected officials in Congress and the administration—would put such a plan in place?
I think Neel Kashkari has made a good point that our collective memory associated with financial crises wanes over time. The history of financial regulation in the United States is cyclical: We tighten bank regulation, trying to make sure that we don’t do bailouts, but then, over time, people worry that those rules make it harder for the economy to grow, and they forget that those rules are in place because we had a big financial crisis, and that crisis really hurt the economy. So then rules are loosened. And we end up with banks that can bring down the financial system when a big shock hits.
So I’m optimistic that we can get reform if we can act while people remember just how devastating the financial crisis was, how many people lost their homes, how many people lost their jobs, how many people felt that it was fundamentally unfair that creditors of these banks were protected and individuals who were hurt by the Great Recession weren’t.
I just hope that everyone who would consider not implementing the Plan would do the cost-benefit analysis and think, “This might seem costly now, but the next time we have a financial crisis, if you are part of the group that didn’t act, that’s really on you. You bear the responsibility because you could have prevented it, but didn’t.”