Jul 10, 2017

Stephen Miller on Financial Crises, Capital Requirements, and the US Banking System

Higher and simpler capital requirements, rather than more regulation, may be the keys to a more market-disciplined banking system.
David Beckworth Senior Research Fellow , Stephen Matteo Miller Senior Research Fellow

Hosted by David Beckworth of the Mercatus Center, Macro Musings is a new podcast which pulls back the curtain on the important macroeconomic issues of the past, present, and future.

Stephen Matteo Miller is a Senior Research Fellow at the Mercatus Center at George Mason University. He joins Macro musings to discuss his work on the history of financial crises as well as the evolution of the U.S. banking system since the late 1800s. Specifically, Steph stresses the importance of capital requirements (how much capital or equity a bank holds relative to its liabilities) in combating financial crises.

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to macromusings@mercatus.gmu.edu.

David Beckworth: Steph, welcome to the show.

Steph Miller: Thanks for having me, David.

Beckworth: Am glad to have you one. There's a colleague, we've had some interesting conversations and I wanted to bring them on air, so here we are. Let's begin by asking, how did you get into this area? Why did you get interested in studying financial crisis?

Miller: I got into economics because I was initially interested in development and as I really started getting more into development, I realized I should have learned something like agricultural economics because I thought that was maybe a more important area. And I thought it was too late to retrofit because I didn't have any of either the biological background or anything like that. So I just thought, maybe I'll look at the financial side. And then that was right around the time we had the Asian crisis. And so that was for me, a shocking event as a young PhD student. And I thought, "Wow, what could you possibly do to maybe insure against a financial crisis?" And I know some people have proposed things like GDP, futures contracts or things like that, but I was thinking more about the sudden changes in financial flows because, or it's more than that.

Miller: I think Ragurham Rajan bank of India, official he, and now back at Chicago, he had mentioned that with the emerging market crisis, you would have an about face change in holdings. So it's not just a flow, it's a sudden reversal in stock in the stock holdings. And stocks are much larger than flow so that can be much more damaging. And in that sense I was thinking, "How could you insure against Asian Crisis risks?" So the first thing you would have to do, it was actually... I had taken a class in continuous time finance which... I only had two classes in finance, but that was the class that really gave me the idea because I remember we had to read this book by Hull on futures and options and it was just for... It wasn't required. It was for background reading it.

Miller: And in one of the chapters he talks about how the stochastic process that you could use as an underlying for a derivative contract for the purpose of hedging or ensuring could be as something as far removed as the weather in New Orleans. So I thought, "Wow, if you can model or price and hedge weather risks, then maybe you can do the same for an Asian Crisis risk." Now at the time I thought it's not something you could replicate. And in finance you talk a lot about replicating in a position or something with two other positions perhaps or so it wasn't clear to me that you could actually replicate a financial crisis as it were. But as I dug a little more into the literature, I found that there was one way to do it would be to think of a financial crisis as an exchange option.

Miller: And that's basically an option that allows you to switch from one risky position to another. So you're swapping out of one to another. The same as in the Black-Scholes framework is very similar. In the Black-Scholes you could switch between a safe and a risky. But here with the exchange option, you can more generally think about switching between two risky positions. So then I started thinking, "How would you measure it?" And so it seems that you could think of... You could measure the risk of an emerging financial market relative to the rest of the world. It's sort of like the capital asset pricing model or the alpha within the capital asset pricing. When you underperform, you sell and when you outperform you buy. And it worked very much like that. So it's basically the right to sell an underperforming countries is a put exchange option.

Beckworth: When you're talking about an insuring here, you're talking about the country insuring or investors insuring against?

Miller: Good question. So it's actually for investors. So at the time I was thinking, "People like Mark Mobius they might like their investments in Thailand, but they don't like what's happening to Thailand as a whole." So the idea was maybe they could insure some of that country risk as it were through a contract like that. Now it's probably going to be too expensive to be attractive to investors. So it may never work. That may be a continuing source of incomplete markets.

Beckworth: What about the credit default swaps on sovereign debt? Is that something similar?

Miller: Yes, it would be similar in a way. It's where you can hedge against default as it works. They can be useful products.

Beckworth: My understanding is credited defaults swaps for sovereigns though they're very thin markets, they aren't as deep as-

Miller: Yes, I'm less familiar with the CDS market, but-

Beckworth: And the GDP link bonds, which is I a fascinating idea. Which is kind of similar to what Scott Sumner has proposed not exactly the same, but he wants nominal GDP futures, but the GDP link bonds basically turns sovereign debt for emerging market and it's almost like equity, right?

Miller: Yes.

Beckworth: The value goes up during good times. Downtime goes down, obligations go down. But the difficulty is can you trust the statistics of the country?

Miller: Yes, exactly.

Beckworth: Because clearly if I'm Brazil, I'm going into deep recession and my statistics agency has an incentive to doctor the numbers and overstate the collapse. So I pay less to my creditors. So interesting ideas. I know Robert Shiller has been a big advocate of GDP linked bonds in a way to ensure against all these, the complete, the incomplete markets. So this is very fascinating stuff. This is kind of like taking finance to the limit. Right?

Miller: Right.

Beckworth: Get a complete world or Arrow-Debreu as a contingent contracts truly exist. So I love your big thinking. Very fascinating. So I can see why this became addicting to you and you write on that now. So with that framework, let's talk about the Great Recession and the financial crisis within that. So how do you see that unfolding? What caused the financial panic? And was it a big part of the story? Did the Great Recession become great because of the financial panic?

The Great Recession and its Causes

Miller: So just as I was finishing this research on the hedging against Asian Crisis, we had the other crisis, the more recent one. And that required that I learn a lot more about Marcus that I hadn't really thought about, which were, since I didn't have much of a finance background. So I started looking into the structured product market as a source of the risks. And I think it was two articles that I thought were very, they were actually non-econs, that sort of pointed me in the direction where I now stand. And that was a Frank Partnoy's Epilogue I think. Or when they re-released his book FIASCO, he wrote a... I think it was an epilogue to basically re-summarize what had happened because the book was written in 1997 I think after the structure product crash in 1995. So in the chapter he's going through how the more recent crisis was just like in '95 different products, but the same-

Beckworth: Securitization.

Miller: Structure products going bad. And the other article was Michael Louis's the end. It's different. It's a short read it and I think it got weaved into the book *The Big Short,* but that article itself was very good. I think the book was a little... It was different in focus but that short article, *The End* and Partnoy's pieces, they both said that most people focused on the supply, what went wrong on the supply side and you can think about greed or whatever you want to identify as the source of the problem, but they were really identifying the demand side. So you have to really understand why-

Beckworth: The demand for what?

Miller: …why was there such a demand for this structured products. And when I was looking into this-

Beckworth: Just real quick, when you say structured products, for listeners who don't understand, give an example of that. What would be a structured product?

Miller: So in 1995 the structure products that had difficulties were, I believe they were in many cases as Partnoy discusses, they were structured notes, which had maybe you'd take Philippine power bonds and I believe the payments were in Philippine currency or they would convert at the dollar exchange rate. So it worked as long as the exchange rate helped. But when it didn't, then the investors lost a lot. And so it turns out that if you read the financial crisis inquiry commission report, there's some discussion of how in the aftermath of that crisis around 2001 or 2000 the securitizers, the ones who were creating these structure products, they started thinking about, "That didn't work so well. What we did in the mid nineties after the so-called Tequila Crisis in Mexico, when a lot of that event caused a lot of these products I think to fail.

Miller: So then they started thinking about the securitizers. Can we broaden the underlying collateral list, diversify. And they came up with the multi sector CDOs and in that case I think they were putting things like airline leases and private equity funds and then you had or fees, I think-

Beckworth: Just to be clear, so you said CDO, that's collateralized debt obligations.

Miller: Yes.

Beckworth: Basically a debt instrument that's backed up by other debt obligations. So you're getting a piece... So if you buy a CDO, you're buying basically a debt instrument that gives you claim to a little piece of a number of income streams flowing from other debt instruments. Is that right?

Miller: Right.

Beckworth: Okay.

Miller: Or possibly private equity-

Beckworth: Or private equity.

Miller: Right. And so what happened around 2001 while we had the tech boom and bust, which wiped out the private equity fees and then the September 11th hurt the airline industry so that multi-sector, that diversification didn't help at all.

Beckworth: Interesting.

Miller: And then the securitizers, primarily the investment banks, but then as well a CITI Bank and a few of the other large banks like Bank of America, Wachovia, they started thinking about... And I think it was primarily the investment banks, but they were thinking, what's left? And they started targeting housing. And so I think that explains a lot of the change in the marketplace that there was this demand. I think you've talked a lot about it for safe assets as it were. These all turn out to have high ratings often but they had higher interest payments. So they were attractive to investors. So they began securitizing the housing backed CDOs and that would require them to first create mortgage backed securities, private label, mortgage backed securities. Those are different than what Freddie and Fannie would do.

Miller: Those are typically called agency mortgage backed securities and that would offer the funds to go out and find risky borrowers. So there are some papers in finance where they show that the causality really runs from the CDOs to mortgage backed securities to the lending, not the other way around. And I think, but a lot of the reform measures we have taken, especially like Dodd-Frank, they compartmentalize the problems. So you address risk in one sense in the housing lending.

Beckworth:         Okay, but not in the CDOs.

Miller: And then you also address risks in mortgage backed securities and then you also deal with collateralized debt obligations.

Beckworth: If I had a CDO back in 2005 and it was backed by real estate, I'm effectively getting all... I don't know how many but a large number of mortgages that are embedded in that layered down and I'm implicitly, again if I buy this debt instrument, I'm getting claimed to a little bit of each person's mortgage payment. The belief is it's diversified enough as it's relatively safe. What could possibly cause the whole housing market to crash.

Miller: It's actually two layers because you have the payments going in to the mortgage backed securities and those get redistributed to the CDO asset holders or the … And I think the biggest... There's a paper by Larry Cordell at the Philly Fed and two co-authors. That paper's really good in identifying what went wrong with those products. And it turns out I believe the two that were the worst among the CDOs, the ones that were the worst were the ones backed by home equity lines rather than your standard mortgages. And the synthetic CDOs, which were backed by credit default swaps, that was a later introduction or innovation,

Beckworth: Gary Gorton in his famous work Slapped by the Invisible hand. He makes the claim, which is interesting to me that many of the CDOs actually were okay. It was just some... The problem was there was some junk in there somewhere. Some subprime stuff that was troubling, but you didn't know for sure if you're a CDO had it. And if it did... so just more the uncertainty of not knowing where all the problems line. So everyone ran on the CDOs even though not all of them really had a problem. Is that a fair assessment?

Miller: So no. I think Gary Gorton was writing that right before the-

Beckworth: Earlier in the crisis.

Miller: So we didn't really know a lot about CDOs. And so I think that's why the Larry Cordell studies at all…is really important. Yeah. And there was also a Harvard undergraduate who had written a thesis on it that also inspired, I think Larry Cordell out to write this paper. Yeah. It looks like that was really the weapon of mass destruction as it were because they show that the write downs on... For the whole sample, they try to reconstruct every deal because that's just not available to anybody. And they find, I think the write downs were like 65%.

Beckworth: This paper you're saying indicates that there was a specific sector that caused the problems.

Miller: Yes.

Beckworth: It wasn't just uncertainty?

Miller: Right.

Beckworth: Okay.

Miller: Right. So it didn't take a large loss for instance, for there to be a complete wipe out of the collateral that backed the CDO.

Beckworth: Okay. So we had this run in the financial crisis and as a macro economist, I find that fascinating because there was a run on the banking system, the shadow banking system, or I might call the institutional level banking system. Even though you and I didn't see that at the commercial level, there was still this run at the wholesale institutional level banking system, which created, there was a panic. And if you look at all the money assets they use, like repo, commercial paper, things like that, they did collapse. There's the Center for Financial Stability in New York… division four, which includes all of these assets. And you do see this drop in the money supply. So in some ways it's very similar to the Great Depression. Great Depression was run on commercial banking, it will be a little bit different there. But the bank run then caused the collapse two, the bank run in 2007, 2008 caused the collapse …four but very similar mechanics. Right?

Miller: It depends. I know a lot of people say including John Cochran, say the crisis was a run, but in this case I tend to take the Fisher black view that banking crisis or solvency crises rather than liquidity. In a recent paper I've been looking at why banks actually may have held some of the very products that could have gone bust. And you see that, whether you look at short term funding or mortgage backed securities generally, or maybe it's the specific highly rated private labeled tranches that banks held, that turns out to be the one variable that does matter, the holdings of highly rated tranches. So this is another paper. It should be out soon. It hasn't come out yet.

Beckworth: What's the title?

Miller: *The Recourse Rule: Regulatory Arbitrage in the Crisis.* And in that paper I was building off both the Cordell Erdall insights about CDOs and another paper by Taylor Nadauld as coauthors. He's been involved in I think at least three studies where they look at different sectors of the financial system who were involved in securitizing. In this one paper with Erdall, not all than stilts they look at, the answer to the question, why did banks hold so many highly rated tranches?

Miller: And it seems that they identify what they call a securitizing bank. In other words, it's a bank that's involved in the process of making these CDOs and other types of debt products. And they have a reason for that. They also test other hypotheses, like was it greed or, or executive compensation? And they reject all the other hypotheses that could explain the holdings. But what they do find is the securitizing bank is primarily for skin in the game reasons. When you hold these things on your balance sheet, it's sort of an indication that you stand by-

Beckworth: The product you have created.

Miller: Right. But there could be even other reasons. For instance, you might be very familiar with the product. So if you know Goldman Sachs created these tranches, you might say, "Oh yeah, that looks like a good thing, so I'll hold them as well." So it's not just skin in the game reasons, but there are generally speaking reasons for a securitizing bank to hold more. But they look at a cross section right before the crisis and sort of a before and after.

Beckworth: So just to summarize though, your view of the great recession is there were structured products as CDOs, and some of them had really bad assets behind them.

Miller: Yes.

Beckworth: And there was just a few key ones it sounds like that really blew up and that ad spread to the rest of the financial system created the panic spread to the rest of the system.

Miller: Right.

Beckworth: Okay.

Miller: So it's the asset shock that may trigger the run in the sense that-

Beckworth: On good assets there's a fire sell of even good assets after the initial shock from the bad ones.

Miller: No, I think there are actually some authors out there, Boyce and Helwigan, I think they rule out the fire sale explanation.

Beckworth: But there's panic created and you end up selling other assets as well, don't you? Or is it-

Miller: Perhaps, but I would say that wasn't the feature of the crisis. It was probably a byproduct but not the key.

Beckworth: But by the end of the great recession though, stock prices were down across the broad asset prices went down.

Miller: I think the best way to explain why or where I'm getting at is with Citibank. This is in the Eriel, not all than Stelts paper, in the review of financial studies, they find that Citibank had about 10% of their entire portfolio allocated to structured products, CDOs as well as other private label securitized assets. And they said they found that they also had 6% equity. So it would take a 60% write down on the assets to wipe out the equity. While, as I mentioned in the Cordell study for CDOs, there was at least a 65% right down rate for CDOs and maybe higher in the ones that were originated closer to the time of the crisis. So that could explain exactly why the largest banks had their equity effectively wiped out.

Beckworth: Okay. And we're going to come back to that point and a little bit later in the show. There is the little amount of capital used to fund-

Miller: Yes.

Beckworth: That's 6%. Why was it 6% not say 10%? So you have the initial shock holding these CDOs, but then why wasn't there a cushion to deal with the blow? But let's move on. So we have this great recession, there was a banking panic, it was in the shadow banking system and we had Dodd-Frank come out of it as a result of all this. And I think a lot of your work and our colleagues here has shown some of the problems with Dodd-Frank, hasn't really addressed the main issues. In fact Larry Summers had a coauthored paper and Gary Gordon had a recent one where they argue the financial system really isn't that much safer post Dodd-Frank. So let's talk about this from historical perspective because my understanding is that the US banking system in general has had a lot of problems and particularly compared to Canada. Canada has had relatively few crisis even in 2008 but going way back, Canada's done better. So what is it about the US banking system historically that has set us up for so many panics?

The History of Banking Panics in the US

Miller: The work of Calomiris and Haber, the Fragile by Design book is really good as is a paper by Bordeaux, Reddish and Rockoff where they ask the question, why didn't Canada have a crisis in, and they just list a bunch of years when the US did.

Beckworth: Right. Every year we had and they didn't.

Miller: So according to their account, Canada since Confederation has never had a financial crisis a banking, sorry a banking crisis.

Beckworth: And that includes the great depression right?

Miller: Right.

Beckworth: And that's probably the most striking. We lost how many banks during the Great Depression? Is in the order of thousands, right?

Miller: Yeah, I think it was about 12 I forget. I think it was 12,000 but-

Beckworth: Okay. So we lost huge number of banks. We went through the Great Depression. They also went through the Great Depression, I mean in terms of our economic loss, but they didn't have the stress, which is just a mind blowing observation. So what, what's behind that?

Miller: In Canada, it seems that they never prevented banks from operating across the territory. So they would basically take in deposits from across the entire country and they would then originate loans and I guess by securities as well. But so-

Beckworth: They had branch banking?

Miller: They had branch banking and interstate banking.

Beckworth: And interstate.

Miller: And though that combination meant that the banks were always well diversified and they never had a shortage of funds because whereas in the US we had interstate banking restrictions from the start. But not only that, in some States, many States and I think Hugh Rockoff about it in his podcast where he was saying that there were a lot of unit banks, which means you have no branches, which means you're stuck in one location and you're subject to the economic conditions of that location. So if you're in rural Alabama or pick a rural farming area and there's a drought of some kind, then all the farmers might go bankrupt and the banks go down with them.

Beckworth: So their loans aren't very diversified. They all have all their loans in one industry. One local industry.

Miller: Right, right.

Beckworth: Okay.

Miller: So the combination of branching restrictions and interstate banking restrictions play a big role in why we've had so many banking crisis in the US.

Beckworth: Okay. Why question is why. So I understand the logic of that. Interstate branch banking would better diversify banks. So why didn't our forefathers... Why wasn't that made it into the laws?

Miller: So the Calor Meyerson and Haber book talks about that a lot, where it's really a political problem and you had a collusion between the rural populist politicians and the banking interests, the banking interest in one competition and I guess going back to Jefferson there was a real aversion to having interstate banking. So all of that gets factored into the political process and then we get laws and regulations that prevented interstate banking.

Miller: But if you go into a lot of other countries, you'll see the same bank with branches all over. So-

Beckworth: It's really odd.

Miller: Yeah it is odd. It's a very American problem. So we started to change that in the 1970s and Cal Meyerson, Haber also talked about this sort of the political collusion moved from a rural one to an urban one, so now it was urban populace joining forces, so to speak, or colluding with large banking interests. And you started to have a push for interstate banking that actually started in the 70s. And I think Crows talked about that in a 1999 paper. About how I think it was Maine started looking for out of state banks to maybe acquire some of their banks. And then you started to see States enter into similar agreements with each other. But then we had Riegle-Neil in 1994, which established interstate banking at the federal level.

Beckworth: So those would be like nationally chartered banks or-

Miller: There weren't so-called national banks prior to Riegle-Neil, but it-

Beckworth: But it allowed them now to go across state lines.

Miller: Yes they could operate-

Beckworth: The other ones you mentioned the state chartered banks that could go across state lines now with the-

Miller: No one was allowed to operate across state lines prior to the 70s.

Beckworth: But the deregulation at the state level who benefited from that? The state chartered banks or the national chartered banks?

Miller: Yeah, I think to the extent that you're now having better diversified banks because they're operating across economic regions-

Beckworth: And just to be clear, if you want to have a bank, you can charter it at the state level or at the federal level, right?

Miller: Yes, that's correct.

Beckworth: Okay. Let me ask this question then. So there are state bank regulators and there's federal bank regulators. So you have your state inspectors that come in and then you also have the federal reserve, obvious is the control over the currency, FDIC, who else am I leaving out?

Miller: Yeah, that's…

Beckworth: So is there confusion? Do they ever like have competing claims with like the federal reserve says, "Our regulators pull rank here, step aside FDIC." Is there ever this competition among bank regulators?

Miller: I think there is at a certain level, but because they tend to specify who looks after what type of institution.

Beckworth: Okay so there is clear lines of demarcation?

Miller: Right. But I think during the crisis then that's when things got a little bit cloudy and-

Beckworth: Who has the final say?

Miller: Right. So I think the fed definitely took steps there because they were primarily involved with those banks that were in trouble.

Beckworth: So if I charter a state bank, the state regulators will come and look at it. Will the federal ones as well? Will the federal reserve come in and inspect me as well.

Miller: The fed would regulate the larger banks. Yeah. It really depends on who is your prime regulator I guess.

Beckworth: I'm a state bank. I can imagine at least the state regulator, and then imagine FDIC would also check my books out if I put that FDIC insured logo on the front of my bank. So I'm going to have at least... If I'm a bank, I'm going to at least have several different bank regulators.

Miller: Yes.

Beckworth: Okay. All right. So we have the problem historically then going back to the original question of limited banking. So you could have historically many places, just one office which made you very susceptible. Let me ask this other question. So I believe it was in the 70s that continental Illinois, a bank was bailed out. Is that right? Was that the 70s that happened or was that 80s.

Miller: 80s.

Beckworth: It was in the 80s… Did bailing out that bank and they bailed it out because they thought it was too important to fail. And since then there's been this argument running that that created some moral hazard. It created bad behavior. If a bank was big enough, they knew they would... Even if the government said no, no, they knew they'd be bailed. Is there some truth to that?

Miller: I'm not so sure. A lot of people, Almally, in her podcast and elsewhere has talked about the subsidy. I think Fisher Black is again somebody who inspired me in this area, he was saying that if there is a subsidy goes to the shareholders, not to the borrowers. So I think it may prop up the stock prices, but that's about it. And there is this, I found one study where if you measured it during the crisis that the subsidy if it exists, it was very small and very short lived. So he used an option pricing framework to measure them.

Beckworth: Here's a story that kind of is consistent with this moral hazard. At the end of the day, the government will bail you out if you're too big to fail. This is the whole critique of too big to fail, right? There's kind of that you'll take off more risks because you think there's a backstop. The government goes in and takes care of Bear Stearns, right? It makes an arrangement where JP Morgan takes over the assets. So there's kind of this, okay uncle Sam is going to step in when things get bad enough. And then sometime later we have the Lehman crisis. And my understanding is some people think that Lehman continue to take on risky assets during this period based on the understanding that it too would be treated like in worst case scenario like Bear Stearns. So is there any truth to that story?

Miller: Yeah, I can see that possibly being the case but the difference is that they were not commercial banks. So I think the fed wanted to maybe get involved with that problem, but really they shouldn't have been in the mix to begin with because they were securities firms and they should have failed because the investments they undertook went South. But there is this claim that I think Alan Meltzer mentioned it in congressional testimony and others have also mentioned it, that the bankers always come to the regulators and say, "If you don't do this, you're going to face a catastrophe." And so I think that's what officials have probably reacted to.

Beckworth: So what does the too big to fail then concern? Is it this fear that big banks will be bailed out and it creates bad behavior or is it something else?

Miller: Yeah, I think that's the standard story. I just-

Beckworth: But you discount that centered story.

Miller: Yeah. I think what's happened over time is we've gotten away from market discipline and towards regulator discretion. So if you look at the 19th century, Eugene White's written a good paper on the establishment of the Fed and how that changed the whole market discipline that we had. In the 19th century from 1865 till 1913 we had double liability for national banks and some States like California had triple liability. Other States had double and a few had single liability. But it seems that during that whole 50 year period, roughly 50 year period, there were very few bank failures. And what he found finds was during that period there were more voluntary liquidations because no shareholder wanted to be... They didn't want to pay out that double liability or triple-

Beckworth: So before things got really bad they liquidated, shut down the bank before it got to the point that they would have double, triple liability.

Miller: Exactly.

Beckworth: Okay.

Miller: So that seems to have worked quite well. So he puts a cost estimate on that. And it was something like $1 billion in $2,009. It was the total almost 50 year period. The cost of all banking crisis was about a billion. And then I think for the Great Depression, he uses Friedman and Schwartz's estimate, which was like 39 billion in $2,009. And then the SNL crisis was 200 billion in $2,009. And then the most recent one was 1.7 trillion or more depending on whose estimates you use. So I think what we've seen is we've gotten away from letting markets punish those who invest poorly. These are supposed to be the experts. They ones who know how to manage money or originate loans, whatever. And we've gotten away from allowing them to fail and we just keep allowing or trying to institute protection across the whole financial system.

The Contingent Liability Approach

Beckworth: And this leads us into discussion of capital requirements. But before we get there, I want to go back to this fascinating idea about double, triple liability. And I have a friend who's written a paper, he calls it contingent liability. But the idea is that shareholders are not going to lose just their investment, they're going to actually have to bail out a depositor. Something happens the bank has a bad outcome and depositors want their money back and the bank itself doesn't have it, the depositors can come to you, the investor. So that really makes you look the bank's balance sheet closely look at as business model, make sure it's doing things right before you put your money into it.

Beckworth: And as you mentioned, it also may incentivize you to liquidate before you get there. So I love that because what it does is it in simple econ 101 it aligns the incentives properly. It creates the right incentives which go into capital the idea that how you fund your activities. So bank can either borrow money and lend it out or it can use its own money to fund. And so my understanding is when you had this double, triple liability, there was more funding through capital, is that right? Or you had a bigger capital cushion which would absorb losses?

Miller: I think that's another issue that is a little bit cloudy, murky-

Beckworth: Not sure. Okay.

Miller: ... in terms of the history of capital and that's something I've been trying to get a better understanding

Beckworth: One of the claims of the contingent liability approaches, it would help banks on their own figure out the optimal amount of cash.

Miller: I think a paper by Ben Esty at Harvard showed that those subject to double liability had higher capital. But it wasn't much more. But we did have capital requirements and they were specified not the way we do it now as a ratio of equity to assets or equity to liabilities starting in 1865, I think it was all based on town size. So if you were a bank in a size of 200,000 then you had to have X dollars. And then if it was in a smaller town that would decline by half and then if it was... So they had ranges of think of them as a town size value capital requirements. And they were also, I think a barrier to entry of sorts because depending on how big your town was, you'd have to raise more capital in larger towns.

Beckworth: It seems reasonable because if there are unit bank restrictions, so if you only have one office and you're only in a local town, I can see why it makes some sense to approximate the capital holdings based on the population in your... Because you're not going to have a branch somewhere else and be liable for more so that seems reasonable.

Beckworth: But again that's going back to this point, and we need to move into this, again so one of the big issues during the Great Recession is some of these big investment banks and the big commercial banks, they would make loans the way they would fund them with the money to do the loans is they would borrow from other people. So they themselves were highly leveraged, highly indebted. Made them much more susceptible to shocks through any kind of slight disturbance in asset prices.

Beckworth: And so the argument has been since then, I think in your work too, Anat Admati,  John Cochran yourself, is that big banks, big financial firms need to fund their lending more through their own money, their own... So they have skin in the game. They create the right incentives and what a double and triple liability does at least in my mind, it creates the incentives. You're not going to make all these loans if you think you're going to get burned in the future and... You'll know the right amount of capital to fund with. Whereas now we're talking about let's increase bank's capital, but how do we know what the magic number is?

Miller: So that is the million dollar question, but so I think the way I would look at it is you do a cost benefit analysis, but even there it's really difficult. There's a one paper by begging now null and void, it's a working paper that they have put that optimal, they use an equilibrium model and find it. It should be like 15% I think. But it could in certain conditions the optimal ratio might be 10 or 20% and that's also been floated. That 15% number is also been floated a number of times.

Beckworth: I've heard 10, 15% like a knot and Annette Amati has said something like that.

Miller: Yeah, she said 20%.

Beckworth: But now we're talking about imposing it from the outside right?

Miller: Right.

Beckworth: As opposed to a bank on its own because of incentives that double liability reaching that number. So let's move to your suggestion. You've written several papers, let me mention some of them here that you've written and you've written one called *On Simpler Higher Capital Requirements,* also *A Primary On The Evolution and a Complexity of Bank Regulatory Capital Standards.* So talk us through that. How have our capital requirements changed? You've touched on some of them and what would be your ideal prescription moving forward?

Evolution of Capital Requirements and How to Improve Them

Miller: Right after the 1982 Latin American debt crisis, there was a push in Congress to do something about Capitol and that resulted in the International Lending and Supervision Act of 1983 which call for regulators led by the Fed to find a multilateral way to raise capital requirements, so that it wouldn't just affect us banks but banks abroad as well. So the end result was, first the US made a deal with the British regulators and then a Japan … Japanese officials signed on. That's discussed in Ethan Kapstein’s work about 20 years ago *Governing the Global Economy.*

Miller: And then when the Europeans saw that both Brits and Japanese have all signed on, then they decided to sign on. But in the process that gave us the Basel so-called Basel Accord, which call for raising capital, but then it added in all of these, you might call them exemptions or what it would do is it would set the capital requirement at 8% but it would reduce it for certain asset classes, like sovereign debt it would go to either zero or if it was short term or close to zero. If it was mortgage backed security it would be very low as well maybe 1.6% instead of 8%.

Beckworth: What's the principle here? The safer the asset-

Miller: If the regulator or the regulations say that a particular asset is safe, then it gets a discount, if you will, in terms of the capital that's required to back the investment in question.

Beckworth: Let's just be clear. If I'm a bank and I invested in loans or invest in other assets like treasuries, if I invest in a say safe asset, a treasury bond, then I don't have to worry as much about the capital I used to fund it. I can borrow more to buy a treasury bond than some other asset that's less safe.

Miller: Right. That's right.

Beckworth: Okay.

Miller: Yeah. So you would see treasuries and securities generally they would be discounted and loans would be, if you think of it that way, penalized… Commercial loans, especially they would have the highest capital requirement or commercial and industrial loans. And over time there were problems in the system and you see Basel officials making revisions and US regulators adopting those revisions. One of the revisions is the one that I was trying to look at in the Recourse Rule paper that I mentioned, because in 2001 people say the US never adopted Basel II, or maybe they did, but very late, like 2007 after the... Depending on how you define adopt. Let's just say they didn't adopt it. Then that's not quite true because in Basel II, there was a proposal to reduce the risk waiting which would reduce the capital requirement for a private label mortgage backed securities.

Miller: They went from at least 4% to 1.6%, just like an agency mortgage backed security if it was originated by a private firm or like an investment bank or something. So that's why I wanted to see, did banks actually increase their holdings after that rule changed? And I find that they actually did, but it was... I mentioned securitizing banks. It was the largest securitizing banks that responded to the rule, which is actually what the OCC predicted. But this rule was intended to encourage securitization without encouraging risk-taking. But that was maybe true, but ex-post it looks like it encouraged both securitizing and the risk taking.

Beckworth: So what that rule did then is if you had a CDO that was built in such a way that had subprime in it, but it was still ready to AAA. You didn't have to fund that with much capital.

Miller: Right, right. I'm not the first to look at this, Engineering the Crisis by Friedman and Kraus.

Beckworth: Okay.

Miller: And in that book, they talk a lot about the recourse rule but they don't really have much data because that's really been the key problem. We don't have the data but on a lot of these holdings for individual institutions and the Eriel not all … paper that I mentioned, they also look at that hypothesis, but they don't have the data. They don't go back far enough-

Beckworth: This rule kicks in 2007 and-

Miller: 2001

Beckworth: 2001?

Miller: Right.

Beckworth: Okay.

Miller: Quarter four I think actually the day it was released, it was to be in effect. I think it was November 29th, 2001.

Beckworth: This is from Basel II, right?

Miller: Well, the changes were in proposals for the new Basel II which had not come out yet.

Beckworth: Okay.

Miller: So they weren't identical. But the Freeman and Kraus book that I just mentioned, they have a table where they compare Basel II with the Recourse Rule and they show that they're almost identical. They're identical with the highly rated tranches of private label.

Beckworth: So this rule insured encouraged the creation of more CDOs, some of them which had subprime back in to.

Miller: And I think in the way that Eriel not all … say it's that all of a sudden people have an incentive to hold more. And then not only that, you lower the capital requirement, so then-

Beckworth: You can just borrow money to fund them.

Miller: Right.

Beckworth: So you're holding stuff that may be questionable and you're levering up. You're getting more in debt along with it.

Miller: And risk retention went from holding equity to holding the highest rated tranches.

Beckworth: So what is your solution? So things have gotten really complicated. I think that's kind of what you're getting at right here.

Miller: Yes.

Beckworth: Things are getting complicated.

Miller: So then Basel III came out and we're debating and then you have liquidity coverage ratios and I think this is really a problem. The large banks don't have a problem because they can hire the compliance staff. But small banks have a problem with this. And I think that's what I've heard a lot of commentary about small banks because that's what they complain. One of the things they complain a lot about is the Basel style capital requirements. So my coauthor and I, Jim Barth, we actually just asked the question, what if you just went from a 4% leverage ratio, which was the 2014 regulatory value to 15% would the benefits outweigh the cost.

Miller: And we don't just look at the large banks, we look at all banks with at least 1 billion in total assets, bank holding companies with at least a billion in total assets. And we basically find that in most cases the benefits outweigh the costs of going from four to 15%. What we're doing here is we're asking, "So what are the benefits and what are the costs?" The benefits would be when you have higher capital, you have less run prone funding. Capital can be equity, it could also be longer term bonds like Coco's or long-term debt instruments. But the point is they're not prone to runs. And so if you have more or less run prone funding, more non run prone funding, then you lower the likelihood of a crisis. But the costs could be that when you require banks to have more capital, and I'm not talking about investment banks here, just the holding companies or the banking entities.

Miller: So if you have higher capital, that might mean that the argument is equity is expensive and therefore, this is what the critics of higher capital would say. And of course Admati and Helwig and others, they've all attack that. And I think they're right. I don't think it is, but what we do is we say, "Let's just say the critics are right." So we actually put... We look at what would be the costs. We're assuming the lenders would pass on the higher funding costs to the borrowers and that would result in fewer loans.

Miller: So even when we assume the costs are high, then you still find that the benefits outweigh the costs and we get in the latest version of the paper, it's not yet released. We just revised it. And that's to account for the fact that since we wrote our paper while the Minneapolis fed proposal came out and the federal reserve board also issued their own paper. So in the new version, which hopefully will be out soon, what we find is our optimal leverage ratio is about 19% but we're still only looking at, we're not saying we should set it at 19%, we're just saying that under the baseline benefits assumptions that we make, we get the optimal rate to be 19% and that's under the highest cost scenario. It would be even higher if you went to the lower cost scenarios.

Beckworth: You want to see banks go down this path.

Miller: Right.

Beckworth: Funding for capital. Now-

Miller: For one thing I think in the debate on higher capital, people lose sight of the fact that smaller banks tend to have higher capital anyway. They also tend to hold fewer securities therefore the ratio of risk weighted assets to total assets is higher for a small bank. So a lot of the studies when they translate their leverage ratios into regulatory capital ratios, they assume that ratios are much lower for the large banks. But when you look at the smaller banks that's a lot higher and therefore the difference between the optimal leverage ratio and the tier one rate to risk-weighted asset ratios is not so large. We get about, I think it was a 25% for-

Beckworth: So it would be terribly onerous for the smaller banks to go down this path to … more because they're already there.

Miller: I'm sure they would complain or at least some would complain, but I don't think it's impossible.

Beckworth: Will if the alternative is less... If doing this means less regulation? If that was a tradeoff, I'm sure this is real and let's use that as a segue into a bill that was just passed in the house, The Choice Act. The Choice Act has, among other things, I may call an exit ramp, they have this proposal that if you do fund with 10% capital, is that the right number?

The Choice Act

Miller: Yes. At the holding company.

Beckworth: At the holding company. Okay, so if you fund with 10% capital and some of the regulations, some of the compliance costs from Dodd-Frank disappear, is that right?

Miller: Yes. So it's an off ramp-

Beckworth: Off ramp.

Miller: ... so to speak. And vice chairman Hoenig at the FDIC he's been floating an alternative proposal because in his view I think he doesn't think many of the large banks would actually take the off ramp. They like the compliance costs and if that's correct then there are sort of public choice reasons for that because maybe all these regulations serve as an entry barrier. So it's like a moat that no one can cross unless you have really... If you have that infrastructure.

Beckworth: In other words, so the small medium sized banks they do keep adequate capital, the 10%. The big banks never change. Big investments never change. Moreover they keep out new entrants into their … of large banks.

Miller: That's right.

Beckworth: Okay.

Miller: So Hoenig has therefore proposed an alternative because he's concerned that they won't bite and opt for the higher capital, which is still the problem-

Beckworth: Right. Because they're too big to fail. They could still cause systemic problems for the rest of us. So how does this proposal fix that? What does he do?

Miller: All right. He thinks of separating... You have a parent holding company and then you have these intermediate holding companies, one for traditional banking and one for nontraditional banking, which would be everything from securitizing and et cetera. So you sort of separate what you might call traditional banking from nontraditional banking. Everyone gets a 10% leverage ratio. So that's even below the level, the optimal value. We find-

Beckworth: And when we say that was … you say leverage ratio, that means how much capital you have to fund with, right?

Miller: Right. Equity to assets.

Beckworth: All right.

Miller: Although I think maybe we should be looking at equity of the liabilities instead total. Let's call them short term debt. Whether it's deposits or other short term securities. I think in that case what's, in a sense there are some attractive features of the Hoenig proposal because he actually gets to an issue that I think has been lost in the discussion, which is should you have higher capital or capital adequacy? Should it be at the subsidiary level or at the holding company level? And I think it should be probably at the holding company level. Merton, Miller, Fisher Black, a lot of people in the past I've talked about establishing capital requirements at the subsidiary level-

Beckworth: To just be clear, what are the differences between those levels?

Miller: In the US we use these holding companies we have for at least or close to a hundred years, if not more. And that was one way to... There were a number of reasons but one reason for it was to get around the branching restrictions. That's why they created this organizational form. And what it does is it allows you to buy multiple banks and they all wind up being under a holding company. So, and Paul Kupiec is another person who's been talking about why you want to have capital at the subsidiary level because it's not clear if the holding company will actually bail out a failing subsidiary. So then you still have a problem.

Beckworth: So the subsidiary you mean the actual bank?

Miller: Yes, the commercial bank within the parent holding company.

Beckworth: Okay. All right.

Miller: Right.

Beckworth: Is there any hope that the Choice Act going forward will reflect some of these concerns because I know this bill just passed the house? It's going to be changed probably in the Senate. Any hope that in the Senate that they'll address some of these issues?

Miller: I'm not sure. In a sense, I feel a little more comfortable with regulators specifying the capital requirements than putting it in legislation. Because if it's the wrong value, it's probably harder to change the legislation than it is to change the regulations. So just in terms of the big picture, I think the regulators have a better sense but there's still, it's hard to debate that because there are a lot of people who think what Basel is doing is correct. But I think the Hoenig proposal has some merits in the sense of you're really talking up bank subsidiary capital requirements as well as holding company. But I think that that should be a key focus of if we really want to stop too big to fail or whatever you want to call it, the bailing out of banks, I think subsidiary level capital requirements would work better. We probably want them to be market value that's even more troubling. Anat Admati has been talking about that as well.

Beckworth: So this this Hoenig rule would break down those barriers. So-

Miller: You separate the nontraditional banking from traditional banking.

Beckworth: Okay. So they don't have the choice to continue doing compliance costs and regulations. They have to.

Miller: Oh no, I'm sorry. He does call for eliminating much of the regulation that they currently face.

Beckworth: But in return they have a high capital.

Miller: Right.

Beckworth: Okay. Whereas the Choice Act still gives them the option to maintain the regulatory burden of Dodd-Frank if they don't fund with high capital.

Miller: That's right.

Beckworth: Okay. All right. Very interesting. All right, our time is up, but this is a fascinating conversation. I could continue on for some time. Our guest today has been Steph Miller. Steph, thanks so much for stopping by.

Miller: Thank you, David.

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate