Is American Banking Safe? You Might Not Like The Answer from Two Fed Veterans

As anybody who lived through the Global Financial Crisis of 2008 knows, banking can be hazardous. Failures can hit millions hard, wiping out life savings, tossing the economy into chaos, and messing with investments, spending, and overall growth.

Capital requirements are supposed to be crucial buffers shielding banks from catastrophes, rooted in centuries of financial evolution from Alexander Hamilton up through the New Deal regulatory regime and modern international agreements like the Basel Accords. But current regulators’ efforts to raise the capital ratios of big banks to safe levels are strongly opposed by most financiers, sparking debates on finding a balance between stability and financial risk, all amid intense political pressures.

In the following discussion with the Institute for New Economic Thinking, Walker Todd and Bill Bergman, who’ve been in the Federal Reserve trenches, give an insider’s perspective on the history of banking instability, the fallout of regulatory choices, and the ongoing battle to shield the public from financial risk — because it’s all of us who bear the brunt when things go awry.

Lynn Parramore: Capital requirements might seem like a puzzle today, with rules that are intricate and maybe a bit arcane. Can you break down how they’ve evolved in the U.S. and where key problems lie?

Walker Todd: The first question is: what is bank capital? The simplest answer is that it is – hopefully – the positive difference between the value of assets and the value of liabilities.

A question that economists should be debating and learning from with finance professors is: Do you want to use current market value, especially for assets? In practice, since 1938, the answer has been no, we’re going to use historic cost or book value for valuing bank assets. If you asked me where the root of the problem with capital requirements lies, I’d say it’s right there. It’s that decision and its consequences over time: the salami slicing, the regulatory tactics where the banks go to the Fed and say, oh, can’t you just give us another inch? And so on. Eventually, you get to where we are today where capital is almost a meaningless construct.

Bill Bergman: Walker’s mention of market versus accounting values brings to mind a joke that was flying around during the financial crisis of 2008/09: “To the left, nothing is right and to the right, nothing is left.” People didn’t trust the accounting numbers and they lost faith in the capital requirements for the banks.

Lynn Parramore: Let’s go back for a moment to an earlier crisis, to the lead-up to the Great Depression and the widespread bank failures in the U.S. Can you talk about the role of capital requirements in those events, as well as the role of the Federal Reserve?

Walker Todd: Remember that up until 1933, the U.S. was on a gold standard. More narrowly, it was on a gold reserve standard or gold exchange standard. Before 1933, banks were expected to have adequate capital to meet bank runs. Bank runs were not infrequent, but they were less frequent before 1933 than proponents of big government and big regulation of banks would have you believe. The largest banks, the equivalent of the J.P. Morgans of the day, would commonly have somewhere between 20% and 30% capital. A lot of that was attributable to the large gold reserve they held, which was either in their own vaults or at the Federal Reserve. The gold would count as capital if they hadn’t pledged it for something else.

Well, after 1933, the gold reserve went away and the capital requirements of most banks were reduced a great deal. About a third of the banking system was shut by early 1933. Another third was receiving vast infusions of, ultimately, investment money from the federal government, usually via the Reconstruction Finance Corporation (RFC). Only about a third of the banking system was able to survive without borrowing from the RFC or the Fed. So those banks had enough capital, but the other two-thirds were either gone from the scene or staying alive by borrowing heavily. Almost all in the middle group that borrowed from the RFC eventually made it through the program. They were required to pay back the RFC over a 10-year period. The RFC took warrants against bank shares, so that if you failed to pay it back at the end of the 10 years, RFC officials would wave the warrants around and say, you guys are out of here, we’re taking over the bank! They would appoint new management and buy enough shares with the warrants to make the other shareholders a minority interest.

Bill Bergman: Instilling confidence in financial institutions, particularly banks, raises crucial questions. Capital is a key component, but what do we know and when do we know it? As Walker mentioned, the accounting values are susceptible to manipulation by resorting to appeals to historic costs and other schemes that don’t give you the right answer for what the market believes — and the market can create its own reality. When market confidence wanes, we’ve learned that liquidity can dry up quickly for people who assert that they have a certain amount of capital.

Over time, we’ve enhanced and refined capital regulations, driven in part by the public support underpinning our banking system. In theory, we regulate capital in banks to prevent widespread issues that taxpayers would have to address, aiming for greater stability in the banking system. Unfortunately, it may work in the opposite direction. The regulation of capital arguably results in capital being set too low in the banking system because it combines with the safety net, and if the banks capture the regulators – which history teaches us they do — then banks get to the low number they want that lets them capture gains on the upside while socializing the losses.

That’s the sad lesson in history.

Lynn Parramore: After the Great Depression, regulatory adjustments were implemented to prevent the recurrence of widespread economic crises. While these changes initially bolstered banking system stability, a few decades later there was a shift towards increased instability. What were the driving forces behind that shift?

Walker Todd: Going back a bit to the 1950s, after WWII, the banking system was pretty stable for a long time. Part of the reason was that Jesse Jones, the head of the RFC, had essentially replaced the heads and the directors in about a third of the U.S. banking system. He was appointing old-line, hard-nosed people who said, capital is capital, and we’re not going to pretend that liquidity is a substitute for solvency. That was the general attitude of the U.S. banking system all the way down to the 1960s.

Starting in the sixties, there was a heightened push for more housing finance in Washington. Banks were encouraged to make more loans than their capital supported. In particular, the term of mortgages began to expand. In 1933, you could not get a 30-year mortgage on a house anywhere. Then the Homeowners Loan Act was enacted in 1933, introducing a framework for amortizing mortgages with terms surpassing 10 years, commonly settling at 15 years.

Rolling into the fifties, sixties, you’re dealing with a structure on the mortgage lending side where everybody had 15-year mortgages. The goal was to facilitate homeownership for new buyers, achieved by extending mortgage terms to 20, 25, and eventually the standard 30 years.

If you’re carrying a mortgage loan for 30 years on the books of a bank, it is subject to a great deal of present valuation change — primarily due to interest rate changes over time. In theory, the bank should have enough capital to cover any losses — paper losses — they might incur from holding those mortgages. But if they failed to build up their capital, the usual consequence was their stock price would get marked down. In response to falling bank stock prices, the bankers would lobby Washington to give them more powers, like the right to branch into new areas and things of that sort. That’s how we got into the somewhat riskier world of banking after about 1970.

Bill Bergman: At that point, a significant phenomenon emerged—the acceleration of inflation, initially yielding positive outcomes, albeit temporarily, for the collateral value supporting mortgages. But at the same time, you also had the beginning of the end of low interest rates, and rising interest rates ended up deep-sixing half of our banking system. The savings and loan (S&L) industry, heavily invested in long-term mortgages but subject to limits on how high they could raise rates, bore the brunt of the impact.

That’s where we got the first real lesson about inadequate capital in the S&L industry that readers of economist Ed Kane’s work will be familiar with. Kane identified this blooming problem before anybody else did in his 1985 book, The Gathering Crisis in Federal Deposit Insurance. He pinpointed where the industry had been ravaged by the rising short-term interest rates and criticized the response of the Fed to catching up to the problem it had allowed to fester. Kane described how, in turn, many of those S&Ls, once they were effectively insolvent (not in an accounting way, but in an economic sense), had an incentive to gamble and take more risks. Kane was concerned about regulatory concessions on accounting principles — letting banks act as if they had more capital than they really had. Regulatory leniency, known as “forbearance,” allowed these deep-sixed banks to stay around, amplifying costs ultimately borne by taxpayers. The combination of lax capital regulations and a safety net exacerbated costs, exposing the general public to the financial risks of these institutions. That’s where we learned that everybody has a stake in this game. That’s what we care about. It’s not just the bankers and Wall Street, it’s all of us.

Lynn Parramore: Because when banks fail, regular people lose a lot.

Walker Todd: Right. Once we got into the seventies and the high inflation era, borrowing at the Federal Reserve began to take off in a way that we had not seen since before WWII. A couple of very large banks for the time failed. In 1973 C. Arnholt Smith’s U.S. National Bank was declared insolvent. The San Diego institution that had once had more than 60 branches and nearly $1 billion in assets was $400 million in debt, much of it from bad loans that U.S. National made to Smith’s other businesses. At the time, it was the biggest bank failure in U.S. history. Then in ’74, came Franklin National Bank of New York. I know that one because I walked into my first day on the job as a new lawyer at the Federal Reserve Bank of New York while Franklin was failing. My boss ordered me that day to go over to Franklin and help the examiners with the international loan portfolio.

Franklin failed, and in the late seventies, a bunch of savings banks in New York failed and the wave of S&L failures that characterized the eighties also kicked in. One question on the S&L side is, well, when did the S&L industry become insolvent in the aggregate? Most people would point to 1985, maybe. But Ed Kane would pound the table and say, no, this thing was effectively insolvent as early as 1974, because once the inflation took off, we would expose this gap between the current market value of the mortgages they held and the cost of funding the same mortgages. A gap emerged and the banks and S&Ls just couldn’t cover it.

Borrowing began to expand greatly. At the same time, capital levels maintained by the big banks began to decline a great deal. By 1982, the average capital of the large New York banks was probably in the neighborhood of no more than 2% to 3%. Old-line bankers would say this is ridiculous. The head of supervision at the New York Fed at the time, Ronald Gray, was very good on this and he was not above reading the riot act to bankers that he supervised about their low capital levels. My bosses told me that people like Walter Wriston, the head of Citibank, replied that we don’t really need capital in the banking system anymore. As long as we can borrow at the Fed, as long as we are liquid in the sense of having cash reserves plus unused borrowing capacity at the Fed, we don’t need capital. That’s where the battle began.

Lynn Parramore: Would you say that’s the moment when Too Big to Fail is getting institutionalized?

Walker Todd: Yes. The first institution that was claimed to be Too Big to Fail was Continental Illinois National Bank of Chicago, which failed in 1984. But the concept of Too Big to Fail is rooted in this inadequacy of bank capital because the Fed had to come up with an excuse for why they were lending to an obviously insolvent bank. The answer was that on historic cost or book value, they are not insolvent – and besides, as Wriston says, as long as we’ll lend to them they’re technically solvent in the sense that they can meet maturing debts as they mature.

Lynn Parramore: Can you talk more about your own personal experiences at the Fed, what the pressures looked like from the inside?

Bill Bergman. Sure. I was at the Federal Reserve Bank of Chicago, headquartered across the street from the Continental Bank that Walker mentioned. Continental failed before I started in 1990 after getting an MBA and another master’s in public policy at the University of Chicago. My favorite professor was Sam Peltzman, who introduced me to the work of economist George Stigler and the capture theory of regulation.

In theory, regulation should work for the public, for all of us. But to what extent do regulators end up being captured by those they regulate? Peltzman was a leader in this area and so was Stigler. Peltzman wrote a wonderful article back in ‘70 on capital investment in banking and its relationship to portfolio regulation. He argued that the banks were effectively substituting public capital for private capital. Banks were able to use the public purse to gather the resources on the downside if that case arose. That’s the concern Ed Kane brought up, as well as others like Senator Paul Douglas from Illinois, who was concerned about schemes to use public credit to privatize gains and socialize losses. I’m afraid that’s where we are today.

My first job at the Fed was in the economic research department, where, though not a formal economist, I summarized Midwest economic trends every six weeks. My insights stemmed from discussions with business figures, not just economic data—a distinction crucial in understanding what has evolved at the Fed and other agencies for bank regulation. One element is that mathematicians, over time, gained greater authority relative to people out there kicking the tires in the business world. After working in the economic research department for about eight years, my final position, starting in 1998, was in public policy analysis and regulatory policy analysis.

I still remember when they were talking about going into Basel II and Basel III. The economists and the mathematicians would get together with the supervision and regulation department and argue to the nth degree of sophistication about the right amount of tinkering we have to do to get to 8%, our basis for capital. Why was that the right number? I asked. No response. There was a great deal of high mathematics going into risk management, around 8%, but is it the right number? Maybe it’s set arbitrarily too low in the interest of the banks.

Later I wrote a paper when members of the Board of Governors, including Chairman Greenspan, advocated for the Fed to lead the regulatory apparatus for financial services holding companies. They emphasized the Fed’s concern about the safety net and the risks associated with the discount window and Fedwire, the Fed’s massive multi-trillion dollar payment system. They were arguing that access to Fedwire was a subsidy for banks. Yet every year the Fed was producing accounting statements asserting that it fully recovered the cost of providing Fedwire.

How could we deny subsidizing Fedwire while saying that we’re subsidizing it? I was invited to present my paper in Washington by a former Treasury counsel and attorney for the Reagan administration. That was the beginning of the end for me at the Fed. I was told to be a better team player. This relates to our discussion because the Fed’s capital requirements are driven by a need to adjust for the risk in the banking system, and part of that risk was the moral hazard we told Congress we were concerned about when arguing the Fed should be the lead regulatory authority.

Later, I put together a catalog of all the references to credit readings from NRSROs [nationally recognized statistical rating organizations] in Federal Reserve operating and regulatory practices, including capital regulations. I recommended eliminating those references, fearing we were outsourcing risk management to the credit rating firms, potentially enabling them to mint money because you had to get such-and-such a credit rating for such-and-such a regulation. I proposed replacing credit ratings in capital regulations, as I believed reliance on inflated asset values could undermine the banking system’s actual capital. Although George Kaufman and others were sympathetic, regulatory authorities at the Fed instructed me to stop work on the proposal. Even at the time, George Kaufman said I was being too critical of the Board of Governors and getting in trouble. He told me to stop.

Walker and I share independent thinking that is motivated, I think, by a concern for public service. We both paid a price for that.

Walker Todd: In 1991, there was a banking bill called FDICIA, the Federal Deposit Insurance Corporation Improvements Act. The chief sponsor in the House was Henry B. Gonzalez, Chairman of the House Banking Committee. He was a great guy and had a good crackerjack staff. They were all in for the argument that the Federal Reserve should stop lending to insolvent banks. They essentially drafted a bill saying that.

Well, the Board of Governors and the Treasury went haywire over it and got it watered down to the extent that there could be a so-called “systemic risk override.” That meant that if the Board of Governors voted that there was a systemic risk out there and they had to lend anyway, and the Treasury Secretary would sign off on it, then they could go ahead and make the loan.

I had the misfortune of drafting an article for publication by the Cleveland Fed, which came out in ’93, about the emergency lending provisions of FDICIA. I explained how this came about. The sense of the provision was a section that said that the Fed was able to make emergency loans based on any collateral satisfactory to it – which could be just about anything. Changing that provision allowed securities firms to borrow directly from the Fed for the first time in history because the Fed discount window was originally set up to make loans only on types of collateral eligible for discount — and investment securities most definitely were not eligible for discount.

The Board staff turned out to be very unhappy with the submission of this article and kept intervening with my management to try to block publication. I was required to do 21 drafts over a year, which impeded my doing anything else. Eventually, it was published in mostly the same form in which I had originally submitted it, and at the last moment, a senior Board staffer called the research department at the Cleveland Fed to order them to stop the presses. But they couldn’t, because the president and research director were out of the office and couldn’t be reached. So it went out the door. That was it. I was given an unfavorable rating for the year and I left the next year.

Lynn Parramore: How do you view the regulatory landscape today? What concerns you?

Bill Bergman: Think about Silicon Valley Bank and the argument for the federal government and the FDIC to cover all of the deposits because of the broad systemic consequences. It’s kind of ironic when the leaders of our banking regulatory apparatus tell us that our banking system is sound. It’s sound because the government is behind it. You and I are behind it, the taxpayers and citizens. We are the public capital. The insiders make the upside but you and I are exposed to the potential downside.

Walker Todd: Starting with Basel III in the mid-nineties, the banks sold the Fed and the other regulators on the idea that capital should be adjusted for the perceived risk of assets. Not market value, mind you, just the perceived risk according to mathematical models. Gradually, supervision at the Fed was taken over by mathematically trained economists. There was a recent episode at the New York Fed, for example, where the head of bank supervision was an MIT-trained mathematician. He’s gone now, but that did happen. So those risk adjustments were blown up in the ’08 crisis here and the 2010 crisis in the Eurozone when the Europeans, in particular, discovered that the risk of sovereign debt varied from country to country. It wasn’t the same for Italy and Germany, even though both were members of the Eurozone. You might have a claim on Greece or Montenegro, but how do you know it’s any good? Yet they were giving full credit on their capital risk adjustment for risky assets. That drama is still unfolding.

Bill Bergman: It’s a hot topic now. There is a heated discussion about whether or not regulators need to be more aggressive in raising capital requirements. I’m afraid the proposals out there being debated are not nearly as fundamental, perhaps, as we deserve. Going forward, we can think about two possible better futures. One is simpler and higher capital requirements based on market values. That’s assuming that capital regulation is a good thing. I’m afraid that it’s ultimately corrupted no matter what happens, so maybe we should try another route, and that’s getting rid of the thing that causes the need for the requirements in the first place: having more spine in our Congress and in our regulators. They shouldn’t allow Too Big to Fail to be a threat to the rest of us. Why do we allow these people who are threatening us to enjoy subsidized public capital? Maybe consider a world without deposit insurance, without a central bank.

Walker Todd: I endorse what Bill said, and would add that the crucial thing is going back to market value accounting and maintaining adequate capital requirements. The best papers on this put the number somewhere between 20 and 30% for large banks.