Emil Verner on Banking Crises, Credit Booms, and the Rise of Populism

Emil Verner is an associate professor of finance at MIT Sloan and is a research fellow at the National Bureau of Economic Research. Emil has written widely on financial stability, banking panics, and credit booms, and he joins David on Macro Musings to talk about these issues. Specifically, David and Emil also discuss the causes and policy implications of bank failures, the shortcomings of the Diamond-Dybvig model of bank runs, how financial crises spur the rise of populism, and much more.

Check out our new AI chatbot: the Macro Musebot!

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 [email protected].

Register now for Building a Better Fed Framework: The AIER Monetary Conference.

David Beckworth: Emil, welcome to the program.

Emil Verner: Thanks a lot for having me. Big fan of the show.

Beckworth: It's great to have you on, and you have quite the publication record already, in the top five journals. Of course, you're at a place where they probably expect the top five journals to be had. You're an associate professor at MIT— great program. And it's a lot of interesting research. What really caught my eye is this paper. It's made a big splash— *Failing Banks.* We're going to get to that, but [it’s a] really fascinating paper, and I'm hoping that it really changes the narrative, the dominant narrative that's out there, the Diamond–Dybvig model framing. We'll come back to that point in a bit. But tell us about your journey. How did you get into finance, to economics, [and] end up at MIT?

Verner: That's a good question. So, I think it was Napoleon who said that to understand someone, you have to know what was happening in the world when they were 20. And so, I was always interested in math and history, and economics seemed like a good way to combine those interests. When I graduated high school in 2008, it was like history was in the making, and I didn't really understand what was happening. But it was on the front pages that finance and banks [were] failing, and markets crashing was sort of center stage for understanding the world. And so, ever since then, I've been working on, thinking about, reading about macroeconomics and finance, and I ended up doing my PhD at Princeton, which was a great place to work on macro and finance. After that, I became an assistant professor at MIT and have continued pursuing those questions and those interests.

Beckworth: Okay, so, you cut your teeth on the Great Recession. It was your formative experience. So, it's fascinating to talk to people like you. I've had a lot of guests like you who— that's their experience, that's their baptism into macroeconomics/finance, and it's great to have another person on the show. It's interesting also, though, to see that there's a new generation who— their baptism was the pandemic and high inflation, which is a very different experience for those of us who went through the Great Recession. Nonetheless, you've written some really great papers that come out of this experience, and the big one that we're going to be talking about is *Failing Banks,* as well as another paper you've written on bank panics.

Beckworth: Before we do that, let's tee this up, Emil, and maybe ask a very basic question. Why do we care about banking panics, financial stability? I'm just going to throw a few reasons out there, but you're the expert here. You tell me if it makes sense and what you would add to the list. So, first off, I would just observe that people seem to care, just off the bat. That's more of an observation than a reason, but we saw what happened in 2023 with the banking turmoil with SVB, Silvergate, First Republic, [and] even Credit Suisse, the great financial crisis, as you just described, [from] 2007 to 2009, Dodd-Frank— a lot of regulation comes out of it. So, we really do seem to care about it, but I look at actual reasons. Something that you've written on, actually, is [that] the rise of populism tends to come out of financial crises, deep financial crises. So, I think that's an important point. Also, there tends to be slow growth, slow recoveries after a financial crisis. Are those the reasons we care? Are there other reasons? What would you add to that list?

Why Do We Care About Banking Panics and Financial Stability?

Verner: Yes, I think that's exactly right. Bank failures and crises have been recurrent challenges for market economies. Even the most advanced economies haven't been able to avoid or graduate from banking crises, and these crises are extremely disruptive. So, banks play an important and special role in the economy by intermediating capital between savers and borrowers and providing for a well-functioning payment system. When that function gets impaired, it's just really costly. I think that you were alluding to [how] some of the most severe macroeconomic crises come during banking crises. If we want to understand the really bad adverse economic outcomes, then banking crises are first and foremost on that list, and it affects the economy.

Verner: And when it affects the economy, it also affects broader society and our political system, as you alluded to. And I think, just from a research perspective, from a positive perspective, we want to understand these episodes, to understand what's the right economic model or framework to think about why major disruptions in the financial system can lead to real economic disruptions. And I think, ultimately, we want to understand these episodes in order to think about what [are] the right policies, both to fight crises when they happen, and also maybe, ideally, to prevent crises from happening in the first place, even though that's proven very challenging, historically.

Beckworth: Okay, well, we'll circle back to the policy implications later in the show. I definitely want to go there, because [there are] a lot of things that are happening right now, [like] Basel III, [and] there's push at the Federal Reserve to use a discount window more for liquidity, [and] some people want more capital funding. So, I want to return to that debate in just a bit, but let's go to your big paper. You have several big papers, but the one I want to start with is, *Failing Banks.* It's at the QJE. And maybe here, let's get things started by just laying out, what are the two big competing theories for why we have bank failures?

Breaking Down the Causes of Bank Failures and its Policy Implications

Verner: I think that whenever there's a major bank failure or crisis, you always hear this classic debate between liquidity versus solvency. So, on one extreme, you have the liquidity view that bank runs, even on maybe otherwise solvent institutions, can lead to failure. So, you mentioned Diamond–Dybvig before. The idea there is that a run can arise because everyone is expecting a run to happen, and it can become a self-fulfilling prophecy, and it can make the bank insolvent because the run is going to force the bank to take value-destroying actions like fire-selling assets.

Verner: So, in this pure liquidity-based view, even if there's no fundamental problem with a bank or the banking system, runs themselves can make banks insolvent by first making them liquid, and so, we want to worry about runs. The other extreme is that bank failures are caused by fundamental weaknesses that drive a bank to insolvency. So, that could be the realization of credit risk, say, after a period of reckless lending. It could be realization of interest rate risk if interest rates rise very quickly and reduce the value of banks' long-term assets. It could even be things like fraud that we've seen repeatedly, historically, that make a bank insolvent.

Verner: Then, I think that it's important to emphasize that these two, the run versus the solvency views, aren't mutually exclusive. We have modern theories of bank failures and crises that argue that runs are more likely on banks when their fundamentals are weak. Nevertheless, there is this distinction about whether runs are, themselves, sufficient to create failures and crises or whether weak fundamentals and insolvency are really what's necessary to create a crisis and bank failures.

Beckworth: Okay, and from that, I guess, flows big policy implications, in terms of how you regulate the banking system and such. I often hear someone say, okay, let's say that it's a liquidity crisis. Eventually, it turns into a solvency crisis. You can't really always tease the two apart, and I guess that leads to a question. Is this really hard to identify and to wrestle with in the data?

Verner: I think it is challenging to identify and wrestle with precisely because, for example, as you said, the Diamond–Dybvig view as well, there's liquidity shocks that happen first that then lead to, for example, fire sales that make the system insolvent. But what we're really trying to get at in our research is, what is it that comes first? What is it that's the root cause? And can we try to identify— is it really problems, for example, on the asset side of bank balance sheets that lead to losses that then lead depositors or creditors to run because they're responding to those losses? Or can you have episodes where there's runs that lead to failures, even without major fundamental asset side losses?

Beckworth: Now, I want to throw another explanation out there. It isn't really an explanation or competing theory per se, but maybe it's a more fundamental story. I know you're familiar with this idea, but Charles Calomiris [and] I believe Stephen Haber— their book's called, Fragile by Design: The Political Origins of Banking Crises and Scarce Credit. They're basically telling a structural story that the US national banking system was flawed from the get-go. There wasn't branch banking. There were these reserve banks that accumulated reserves. This is very susceptible from the get-go. How would you place those structural stories and relate them to the two competing theories that we discussed that you just outlined?

Verner: Yes, I think that's a good question. I think that the structural perspective is very useful for understanding why some countries were more prone to bank failures and panics than others. So, the classic example is Canada versus the US. Canada had nationwide branch banking. Banks were relatively well diversified in their lending across the country and also in terms of the types of activities that they did, including investment banking activities that created some diversification in their revenues.

Verner: And the US system was thousands of small unit banks serving very local areas that were very under-diversified that would be exposed to a local downturn, say, in the local agricultural market or local industry would lead to losses on maybe just a few of a bank's customers that could then bring down the bank, make the bank insolvent, and then maybe lead to a run. And so, I think, in terms of the root cause of understanding why the US was prone to so many bank failures and panics in the 19th century— I think that that structural explanation gets you pretty far in terms of a cross-country perspective.

Verner: Nevertheless, though, I will say that even in countries with few large, diversified banks— like, say, Germany, for example— these countries haven't fully been able to avoid severe banking crises. So, even these types of systems can sometimes engage in rapid lending booms where everyone in the economy, including banks, get optimistic, and then down the road, that over-optimism is perhaps disappointed and there's losses, and even those big banks can become insolvent or close to insolvent, fail, [and] be exposed to runs. But, yes, I'm very sympathetic to the structural cross-country theory.

Beckworth: Just to park on Canada for a bit. So, historically, I think our listeners know that it did much better than the US banking system. Then, even during the Great Depression, it fared better? Is that correct? Then, in 2008, did it also do relatively well compared to the US?

Verner: Yes. So, in the Great Depression, there weren't the types of widespread runs or failures that we saw in the US. Canada had many fewer banks. They were well diversified. They branched across the nation. But nevertheless— actually, in one of my papers with Matt Baron and Wei Xiong, we looked at Canada specifically, and we argued that Canadian banks were also quite distressed in the 1930s, and that contributed to slow lending in the Great Depression. Canada also had a really bad downturn during the Great Depression. So, that's to say that this cross-country structuralist perspective helps. But even in systems where you have few large, diversified banks, crises haven't been completely avoidable.

Beckworth: It doesn't get you all the way there. You need these other theories, and we need the work that you're doing, that we're going to talk about. Just real quickly, though, before we get into what you did and what you found, what are the policy implications of these two competing views— the liquidity run view versus the solvency view?

Verner: Yes, that's a good question. That's ultimately why we're interested in this. I think that on the policy side, you can think about it in terms of ex ante and ex post during crises. So, ex ante, the liquidity view says that policies that are focused on making sure that banks have enough liquidity— So, that could be deposit insurance to prevent runs, central bank as a lender of last resort, other liquidity regulation— That's going to be important. The solvency view says, well, ultimately, banks fail because they become insolvent. And so, in order to make banking systems safer, they should be required to fund themselves with more equity capital— so, have higher equity ratios.

Verner: We also need to be worried about shocks that can make banks insolvent, like times when they grow rapidly in terms of their assets that can then lead to future losses down the road. So, that's on the ex ante perspective. Then, ex post, in the crisis, it also matters how you diagnose the crisis— if it's more about liquidity or more about solvency. If it's a liquidity problem, then, ultimately, liquidity type interventions, lending facilities from the central bank, should be sufficient. But if it's a solvency story, then, ultimately, liquidity is just going to paper over the cracks, and you need to focus on actually restoring bank solvency by recapitalizing them. Otherwise, banks aren't going to be able to service the economy and lend and promote recovery from the crisis.

Beckworth: Okay, we'll come back to this in much more detail in a bit. Let's jump into your paper. Again, the title is, *Failing Banks,* and I know you have this amazing data set. Walk us through the data set— how you used it and what you learned from it.

Exploring the Historical Banking Data

Verner: So, one of the exciting things about research and economic history, macro, and finance, now, is that the cost of collecting some of these historical data has fallen dramatically. And so, in this paper, which is joint with Sergio Correia and Stephan Luck— we've put together a new data set that covers most banks in the US from the Civil War, so, from 1863 until the present. For the historical sample from 1863 to 1941, we have this newly digitized data set that covers all national banks. So, it's important to remember that, in the US, there's a dual banking system. There are banks that are chartered at the state level and banks that are chartered at the national level, starting in the Civil War with the National Banking Act.

Verner: And we can see the balance sheets of all national banks, which are going to cover between 40% and 70% of commercial banking assets at the time. Then, we have a modern sample that's going to cover all commercial banks from the late 1950s until the present. This is more standard data from the Fed's Call Reports. So, what we have is over 38,000 banks, and because there are so many banks, and perhaps also because of some of these structural issues with the undiversified unit banking nature of the US banking system, there are lots of bank failures— over 5,000 bank failures. We also see lots of information about these banks in failure.

Verner: So, we see, for example, what the receivers that came into the failing banks thought about the quality of the assets of the failing banks, why they thought the bank failed, what the recovery rates were on the assets, and how much depositors got back of the money that they had deposited in failing banks. And so, this is going to let us open up the hood and say much more about, what are the characteristics of failing banks, and what are the commonalities in failing banks, as well as differences across this very long sample where we have periods where there's very little policy intervention through deposit insurance or the Federal Reserve. We have the period before those two institutions are in place. Then, we have the modern period where we have deposit insurance and the Fed, which, in principle, can be there to prevent some of these liquidity-based failures.

Beckworth: I love that you have the sample that spans these big institutional changes— the deposit insurance and the Fed, as you said— so you can kind of control for that. That's what's really great about this sample. So, what were your key findings? What are your takeaways? What did you learn?

*Failing Banks*: Key Findings and Takeaways

Verner: We started, basically, by just saying that if there are a few basic facts about failing banks, what should they be? Do they look random or are there markers that we can see in the run up to failure? And we came up with three basic facts in failing banks that are quite robust throughout this entire period. The first is that failure happens gradually in the form of deteriorating solvency. So, failing banks— both in the national banking era 120 years ago and in the modern sample— they undergo this period of rising loan losses on the asset side, typically in the five years before failure, and this is going to erode their capital base [and] lead to losses.

Verner: So, typically, failing banks have this gradual period of declining solvency that comes from the realization of losses on the asset side. That's the most common feature of failing banks in terms of their asset side. Once, then, that banks are in failure, we see relatively high losses on their assets in failure and pretty low recovery rates on these assets. For example, in the period before the FDIC, so, before 1934, failing banks had recovery rates on their assets of something below 60%— in the mid-50s percent or so. This suggests that often, by the time that banks failed, their assets were very troubled, and this is also what the receivers that came into failing banks argued or saw in these banks.

Verner: The second fact that we found is that, on the liabilities side, you see that failing banks tend to rely increasingly on what we call non-core funding. So, this is non-core retail deposit funding. So, often, as losses rise, failing banks have to rely more and more on other types of funding like time deposits or broker deposits or market-based funding from other banks. The reason that this matters is that these types of funding are generally more expensive and more sensitive to risk. And so, what that means is that on the liabilities side, the cost of the funding that banks have is also rising as their asset side is deteriorating.

Verner: Then, the third fact about failing banks that we find in our data is that one of the reasons banks often get into trouble— not the only one, but one of the common ones— is if they grow very quickly. It turns out that if you want to predict which banks are most likely to fail in five, six years, then the banks that are growing really quickly right now are the ones that are most likely to fail. Of course, some are actually growing quickly because they have good business models and have some edge over their competitors, but very often, this rapid asset growth is a sign, often, of a lot of risk-taking that's not appropriately priced by these banks, and so that helps understand often why these banks then end up into trouble down the road, is because of this rapid lending growth.

Beckworth: So, these three facts, these three findings— they all point to what theory being the better one to explain when we have banking panics or banking stress. Which ones work the best?

Verner: What we argue, what we see in the data, is that bank failures are actually quite predictable based on these weak fundamentals, and so, that suggests that this idea that you can have runs on otherwise healthy banks that are going to turn them, first, illiquid and then insolvent, but if the run hadn't happened the bank would have been fine— those types of failures seem to be quite uncommon. In fact, we argue that they're extremely rare, even before deposit insurance.

Verner: So, what we see is that before deposit insurance, there are lots of failures that involve runs. That is, lots of failures where banks have a lot of deposit outflows before failure, but these failures are predictable in the sense that these are the banks that often have the weakest fundamentals. And so, what that's telling you is that, at a minimum, very poor fundamentals is a necessary condition for bank failures in most failures in US history, even before the Fed and even before federal deposit insurance was there, in a period when it could have been plausible that there were non-fundamental runs that caused bank failures.

Beckworth: Okay, so, that's pretty striking, and it also kind of goes against the popular narrative. It seems like it's easy for commentators, policymakers to instantly go to the liquidity story, the bank run story, and given the span of time that you cover it, it probably seems really shocking, for example, for the Great Depression or for 1907. Maybe [during] some of the banking panics, I believe 1893, 1896, 1870s, there were some other financial stresses then, but what you're telling me [is that] in spite of those well-known stories we tell, you find very little evidence, if any, for the bank run Diamond-Dybvig story.

Verner: That's right, yes. So, what we argue is that runs, on otherwise healthy banks, are not a common cause of failure. They might happen. We're not saying that runs— sometimes depositors panicking because they see a long line at the bank, like in Mary Poppins— We're not saying that doesn't happen, but in terms of being an important cause of failure, we argue that these are very uncommon. So, it seems that banks, for the most part, can manage pure liquidity problems. For example, if they have good assets by borrowing against those assets from other banks. And what's necessary for failure is to be insolvent or very close to insolvent. In other words, liquidity problems alone are not sufficient or necessary for failure.

Verner: So, I think what you see is often that the immediate cause or the immediate characteristic of many banks that failed, especially before deposit insurance, was that there was a run. Then, I think that there's a natural inclination to say, well, the bank failed because of the run. In a way, yes, the bank failed today because of the run, as opposed to next week or in three months or in six months. But ultimately, the root cause of the failure is that the bank had typically very poor assets and was essentially insolvent. And if it hadn't had those fundamental problems, it would have been able to manage the run. So, it's about what's the root cause of failure and of crisis versus what's the immediate cause of failure?

Beckworth: So, part of the challenge, then, is what we see is the run phase. We see it make the news. We don't see the slow deterioration of the balance sheet that's leading up to that, the fundamentals, getting to the place where this run happens. That doesn't make news. I think you said that five years leading up to that [is when] we see some change take place?

Verner: Yes, exactly. So, typically, in failing banks, on average, their fundamentals start to deteriorate about three, four, five years before. In the modern period, where we can measure capitalizations or equity-to-assets more precisely, the typical failing bank has a reduction in its equity-to-assets of 8 to 10 percentage points in the 5 years before failure. So, it essentially goes from being reasonably capitalized to insolvent, and it looks similar in the historical data even though there's some more challenges in terms of the accounting for equity back then. But the root cause or the root problem is those losses. The runs don't come out of the blue and the runs are often a symptom or a reaction to those deeper problems.

Beckworth: Well, I want to come back to the Diamond-Dybvig model and talk about its history a little bit more and its implications and also some more recent cases— how to interpret it in light of your findings. But before we do that, Emil, I want to jump into another paper of yours, because it complements this paper that we just talked about. You have a QJE 2021 paper titled, *[Banking] Crises Without Panics.* Tell us how the findings in this paper complement what we just talked about.

*Banking Crises Without Panics*

Verner: The *Failing Banks* paper, I see as a micro level paper just focused on the US, using micro level data on banks from the US going back to the 1860s. And this paper, *Banking Crises Without Panics,* which is joint with Baron and Xiong, is a more macro cross-country perspective. So, what we did in that paper is we collected new data on bank equity returns, bank losses, and also new information on narrative accounts of when there were banking panics or panic runs. And we did this for 46 countries going back to 1870 again. And in that paper, we tried to think about, from a macroeconomic perspective, what are the consequences of different types of crises?

Verner: So, in history, we've seen that there are some episodes of banking panic runs without widespread bank failures or bank losses, especially in the US— for example, say, the panic of 1884. There were runs, but there weren’t widespread bank failures or bank losses. Then, we also look at, what about the converse case— situations where there were widespread bank failures and bank losses but, say, without panics? For example, in Japan, in the early 1990s, there weren't panics until 1997, often because policy forestalls these panic runs. And what we found, basically, is that large asset losses— solvency problems that come from asset losses— are necessary for banking crises to have severe macroeconomic consequences.

Verner: So, this fact that many of us know, that banking crises are associated with severe macroeconomic downturns— what's necessary for that is for the banking sector to essentially have experienced large losses and be severely impaired in terms of its capitalization. Liquidity problems and panic runs then look like they can be important amplification mechanisms. So, it looks like they can make things work by making the crisis more acute. But often, these panics happen after the realization of large asset losses. So, for example, if you look in bank stock returns, the panics don't just happen out of the blue. Panics often happen after bank stocks have already fallen by 40%, 50% or so. So, we interpret this as panics being the final phase of the crisis when creditors [and] depositors become especially concerned about the solvency of the banking system. Then, they run on the banks. But at that stage, the banking system is often already very impaired.

Beckworth: So, you have this knockout one-two punch here with these two papers, making a very consistent argument across both of them. I just want to bring up a little earlier history in the US, the Antebellum Period. I've had George Selgin on the show before, and he's written a lot about this, and we’ve talked about it before— how it used to be the view that the free banking system, Antebellum Period was just viewed as a complete basket case, wildcat banking. But then, when people like Hugh Rockoff, Arthur Rolnick, Warren Weber— they went out and said that it is actually more nuanced. States like New York, that had a better free banking system set up, actually did fairly well. So, they begin to undermine that story. Then, there was work being done, I guess, in the '80s and the '70s, on the Great Depression, kind of getting to what you did. Folks like Barrie Wigmore, Elmus Wicker— they found that there weren't bank runs per se, but [that there are] fundamentals, like you said, that then led to subsequent bank runs.

Beckworth: I bring all of this up because the Diamond-Dybvig model was written in the early '80s when many of these views— this understanding of the Antebellum banking system, the Great Depression— was still very much one of runs, of liquidity issues, and the public having a panic and running on the bank. And so, I guess, just to circle back to this, given what you have found, what has been your response? Because it seems like this view is still fairly popular, and what you're finding really undermines the application, I guess, or the widespread relevance of the Diamond-Dybvig model. So, what has been your engagement, your response to these findings?

Responding to the Diamond-Dybvig Model of Bank Runs

Verner: That's a good question. So, I think that on the one hand, as you said, many people who've looked at the data, and especially economic historians, would agree with our assessment. So, you mentioned Elmus Wicker. You also mentioned Charles Calomiris before. Their analysis of, for example, the failures during the Great Depression, is very consistent with the story that we're telling, that it was the banks that were insolvent that ended up failing, and the panics happened in places where there were concerns about solvency. We're sort of building on that with a broader and longer sample but contributing to that tradition. At the same time, we have also come across some resistance.

Verner: We want to be nuanced in how we talk about these findings. We're not saying that runs don't matter, for example. Runs still matter for how banks fail, and you can have insolvent banks that are able to survive for some time, and the run will be the proximate reason why the bank fails. And if you have lots of runs, then that can be very costly or very disruptive, because it can lead to contagion. But nevertheless, we do want to reframe the discussion a bit or be a part of reframing the discussion to focus exactly, as I've said before, on the runs being, often, the response and a symptom of these deeper problems. And we were talking about policy before as well— I think that also has some important policy implications about how we think about bank regulation and crisis intervention. And I think, even today, especially, for example, after the failures in 2023, there's a lot of focus on liquidity regulation and banks having access to the discount window in a more flexible manner, and I think that this research suggests that we should be more focused on basic solvency issues.

Beckworth: Okay, well, that's interesting. I had Anat Admati on the show not too long ago, and I was expressing my enthusiasm for the increased use of collateral at the discount window. She's like, "I want none of that." She said, "Best case scenario, it's costly. Worst case scenario, it just makes things worse. It does no good. It's just a waste of resources," liquidity regulations. So, she wants to go all in on, of course, the capital funding side, and I understand where she's coming from. She wants banks to fund with more capital. I don't think that she's someone who'd go 100% equity banking, but she's saying that we definitely have room to grow with more of a capital cushion.

Beckworth: My colleague at Mercatus is Tom Hoenig, who used to be the Kansas City Federal Reserve Bank president. He was also the FDIC Vice Chair for a while, because he was a bank regulator before he became president at the Kansas City Fed. And he's real big on banks’ need to fund with more capital, and he’s also highly critical of the Basel regulations, because he thinks that it's a way for big banks to game and actually avoid really funding with capital. He does not like the risk weighted capital ratios. He thinks that it's a workaround to that. I guess, coming back to these present cases that you just mentioned, like March 2023, is it fair to say that they were viewed more from a Diamond-Dybvig perspective, when, in fact, they maybe they should have been viewed more from a solvency one? I guess that's what I'm hearing you say. Is that fair?

Verner: I think so. I certainly— being on Twitter in March 2023, you had many people saying, “This is great. This is a classic example of Diamond-Dybvig, and we can use this to teach our students about Diamond-Dybvig.” And I think that I even heard Doug Diamond on a podcast say, “No, the failure of SVB was actually not a Diamond-Dybvig run. SVB was a really poorly managed bank. They got the risk management wrong. They had really large losses on their long-term hold-to-maturity portfolio that basically meant that they were insolvent.” And so, if we want to think about, especially in this last series of failures, asset losses, in this case, from interest rate risk, not from credit risk, but from interest rate risk, was a key reason why some of these banks failed.

Verner: Subsequently, around the failure of SVB, there's research that has then looked at the runs that happened. And actually, lots of other banks were exposed to pretty large deposit outflows around this time, and the banks that weren't insolvent— this is along the lines of the argument that we're making— they can manage these types of runs by borrowing from the FHLBs, by borrowing from the discount window. And so, in other words, if you're not insolvent or have very weak fundamentals, then a run, in itself, is usually not going to be sufficient to cause a failure of the bank. And so, that's why we should be focusing on the essential root cause of these failures, which is the solvency side.

Beckworth: Let's go back to one other recent example, and that would be the great financial crisis where you cut your teeth, your baptism by fire into financial crises. So, the story there is, or at least the popular story, is that it was a run on shadow banking or institutional money assets suddenly disappeared as depositors, big banks had left. Wall Street left, they panicked, [the] repo market— all of that good stuff there. So, would you tell a more fundamental solvency story as opposed to the run story for that period as well?

Applying the Bank Solvency Story to the Great Financial Crisis

Verner: I think that's right. I think that often these runs were happening in response to really large losses on these mortgage-related securities that these banks had on their asset side and often had too much exposure to. And this was gradually being realized, for example, by banking analysts starting in the fall of 2007. In this paper, *Banking Crises Without Panics,* we looked at the US case, specifically, and you see that even before, for example, the failure of Bear Stearns, aggregate US bank stocks had already fallen, I think, by about 60% or so. So, this was consistent with this idea that you start with asset losses, that pushes a bank or banking system close to insolvency, and then, depositors and other creditors are going to run on that. But the run is happening as a response to those initial losses rather than being the cause of those losses. Then, the question becomes, well, how much of an amplification mechanism— how much worse do runs make it? That's where it becomes much more difficult, but it's important to start with, well, what is the deeper root cause?

Beckworth: So, maybe, in another timeline, if the fundamentals had been better, we wouldn't have had the big run on Lehman, on money markets, and the financial crisis would have been much milder, less severe.

Verner: I think that’s right, but I think what these papers, and also some of my research on credit booms and the consequences of risky credit booms, shows is that it's not that these shocks just happen out of the blue. Often, it's because of a consequence of rapid lending growth that finances high leverage along with large increases in valuations, for example, in real estate or real estate-related assets. Then, once those expectations aren't met, then, all of a sudden, the whole boom unravels, and losses have to be realized. And so, the boom itself matters for the subsequent crisis and sows the seeds of the crisis. The crisis doesn't just sort of come out of the blue because of some unexpected shocks.

Beckworth: So, in other words, we probably could not have had a counterfactual history very easily during that period.

Verner: Yes, we would have had to start in 2002, 2003 rather than 2007, 2008. Though, on the margin, I think that there were things that could have been done in terms of having banks conserve more capital in the run-up to 2007, 2008.

Beckworth: And I know you've written on this elsewhere, as you just mentioned— the importance of the run-up and credit with the non-tradable sector, right? That's key, like housing, things like that. That's where you really see the most damage created from these credit booms, correct?

The Impact of Credit Booms

Verner: Yes, that's exactly right. So, in another set of papers, I've looked at trying to understand credit booms, what causes them, why they often but not always lead to growth slowdowns, and financial crises. This research here uses a lot of new disaggregated data on credit to different sectors of the economy around the world, and there are two key markers of especially risky credit booms. By risky, I mean credit booms that are more likely to end in growth slowdowns and banking crises.

Verner: The first is credit booms that finance demand in the economy rather than productive capacity, such as, for example, household credit booms when there's rapid lending to households. Those are much more likely to end in growth slowdowns and crises. The reason is that they're often based not on improving fundamentals, but based on increased availability of credit— what people sometimes call an expansion in credit supply or easy credit conditions. And so, when you have those types of booms, that can stimulate demand in the short term, but in the long term, it can lead to these over-indebtedness problems, for example, in the household sector.

Verner: The second characteristic on the corporate side is credit booms to fragile borrowers, especially that are reliant on real estate collateral, like firms in the non-tradable sector, which tend to be smaller, [and] tend to use real estate-based collateral more. These can also be more risky or often more likely to be risky "bad booms" because of these collateral feedbacks that lead to lots of increase in leverage [and] lots of borrowing against rising real estate prices. But ultimately, if these booms aren't based on good fundamentals, improvements in productivity, then they're subsequently followed by debt overhang, falling asset prices, defaults, banking sector losses, and banking crises.

Verner: And that really contrasts with [how] there are some credit booms. They're less common, but there are some credit booms that are based more on good fundamentals and improvements in productivity. Historically, that's often happened in the tradable sector, especially manufacturing, where a lot of countries have grown by having rapid productivity growth in the manufacturing sector. In those types of episodes, when there's been rapid credit growth to the more tradable sector, those are more likely to end badly. We argue that that's because these are based on better fundamentals and productivity and not based on increased availability of credit or expanded credit supply that's not based on those good fundamentals.

Beckworth: So, then, you would attribute the slow recovery from the Great Recession, the Great Financial Crisis, to this very story, that the non-tradable sector households had all of this debt, and they had to slowly repair their balance sheets, [and] it takes time. Is that fair?

Verner: Yes, that's exactly right, and that's the experience that's been repeated by many countries throughout these credit booms, whether it's in Scandinavia and the Nordic countries in the 1980s, or Mexico up to its crisis in 1994, and even some of these historical episodes that we were talking about earlier.

Beckworth: And it might even be China today. It looks like they have a big real estate sector that's highly leveraged.

Verner: I think that's right, yes. I don't understand China as well, and I think that China has some of its own peculiarities. But it's certainly another example of a major real estate-based credit boom that's led to a lot of misallocation of credit, both in terms of geographically and to the real estate sector. And that takes a long time to unwind and, it takes a long time to work off that debt overhang. That's, I think, what we're seeing in China.

Beckworth: Okay, and a big part of the story is the banking sector that funds those credit flows, the, maybe, excessive credit growth. So, let's circle back to the banks just for a little bit and spend some time on the policy implications, prescriptions. So, you outlined earlier the two broad contours that you could take if you think it's a Diamond-Dybvig world, bank runs on solvent institutions, and more deposit insurance, lender of last resort facilities by the Fed.

Beckworth: I guess, in the limit, you could take Mervyn King's “pawnbroker for all seasons,” or— the central banks there are the first and last aid to the markets. Then, the other extreme— maybe two corner solutions, if you would call it that, would be just fund, fund, fund with capital. Where would you land? You’ve got these two corner solutions, pawnbroker for all seasons, it's all the central bank, it's all government intervention, or banks need to fund with more capital. I imagine that the truth is somewhere in the middle between those two, but how do you lay that out and how do you land that?

What Are the Necessary Policy Prescriptions?

Verner: Yes, I think that's right. I think, not to be too boring on your podcast, but I think somewhere in the middle seems right to me. The pawnbroker for all seasons, to me, the challenge that I have with that— and I'm not necessarily against some of these proposals that try to have banks pre-position more collateral at the Fed to be able to more quickly respond to funding pressures. But again, it doesn't really address often the root cause of bank failures and crises, which we argue is solvency, and there are some costs like we were getting at before. The more liquidity requirements that you do, for example, do actually require banks to then do less lending.

Verner: Then, on the other side, the narrow banking, or 100% equity-funded banking— that would obviously make crises and failures much less likely within that system. But again, I think it goes too far. I don't think that we need to go all the way there. Bank liabilities, deposits, are a valuable product that banks create. People like them. They serve a useful role in the economy. So, I think, somewhere, [there is] a more intermediate step where we just ask banks not to have 10% capital, but a number that's a fair bit higher. Non-bank financial intermediaries have roughly 30% equity capital financing, and I think it would be nice to have the banking system move in that direction as well.

Beckworth: Okay. One last question about Diamond-Dybvig. I know I've been beating up on it a lot here. So, that's me, listeners. That's not Emil. It's David Beckworth who's been giving out the attention to Diamond-Dybvig, but one last question on it. Again, to me, at least, it seems like it receives a lot of attention. You talk to someone in March 2023, and you ask them what happened, they'll tell a story, and they'll say, “See Diamond-Dybvig, 1983.” They'll default to that. I guess my question to you is, why is it so persistent? Why is it so popular? Why has it become the default story that people often tell? Now, obviously, you don't, and your colleagues [don’t], and as you mentioned, a lot of economic historians do not. But it seems to be, for a lot of people, the default narrative that's embraced when there is stress in the banking system.

Why is Diamond-Dybvig So Popular?

Verner: It's a very good question. I think, on the one hand, if a car is racing down the highway, and it crashes and hits a tree, people might say, “Well, the car crashed and there was a really bad accident, because the car hit the tree.” But really, the root cause was, well, the car was speeding way too fast. So, sometimes we sort of focus too much on the narrow immediate, because that seems easy to identify.

Verner: Historically, I think there's also the perspective that, in a way, Diamond-Dybvig kind of exonerates the banks and the mistakes that the banking sector has made in terms of rapid lending growth, poor decisions, [and] too much risk taking, by saying, "Look, there's factors that are outside of our control, like these runs that happen when depositors panic." And that's not necessarily because the banks made poor lending decisions. And so, I think that that narrative has always been there and been helpful to try to, to some extent, exonerate banks and also to provide more protections for banks, even though I think it's, in terms of understanding especially bank failures— it's quite ahistorical, even in the period before deposit insurance.

Verner: One fact that I actually didn't mention that I just wanted to get in is that in the bank failures we look at, most of them have the cause of failure or the cause of death as classified by the receiver from the OCC that went into the failing banks, and the most typical causes of failure are losses on assets or economic downturn or even fraud. Runs account for less than 2% of bank failures, even though, of the cause of failures assigned by this OCC, are receivers, even though many of these failures involve runs. In other words, not attributing a failure to Diamond-Dybvig or to a run requires taking a step back and really understanding, what was the deeper reason for why a bank failed? In other words, we have to ask people to become better at doing causal analysis and understanding cause and effect, and we know that that's very difficult, in general.

Beckworth: Yes, good luck with that, with the popular media and the public. But that's why you're doing the good work you are, fighting the good fight. You're helping us be more careful and, hopefully, nuanced when we do have future banking turmoil. Alright, Emil, I want to turn to some other research you've done that's related, in the time we have left, and that ties back to this concern about financial stability that I mentioned earlier, and that is the implications for populism.

Beckworth: So, when President Trump was first elected, and during this time, there was a lot of talk about, “Oh, populism is on the rise,” and some people attributed it to different things. But one of the causes that was pointed to was the financial crisis, 2007-2009. It really left behind the scarring, the problems that you outlined, households having to heal their balance sheet. I think, clearly, there were other factors as well. You could look at maybe globalization having some effect, even though I think that, on balance, globalization is a great thing. But clearly, there are dislocations happening, but a lot of people said that the Great Financial Crisis was the catalyst that really sparked the rise of populism, not only in the US, but in Europe. And you have a paper on this, and we've talked about another paper on the show, too, that you cite in yours. 

Beckworth: It was a 2016 paper that was titled, *Going to Extremes: Politics After Financial Crises, 1870-2014* in the European Economic Review that makes this case that after looking at, I believe, 800 general elections, that when you have deep financial crises— not just a garden variety of recessions, but deep financial crises— it tends to lead to the rise of the far right [and] populism. I always found that very convincing, so I was excited to see your article that makes a similar case for Hungary. So, tell us about your paper and the takeaways that you see for the current environment we're in.

*Financial Crisis, Creditor-Debtor Conflict, and Populism*

Verner: We were motivated by exactly that observation in the paper that you mentioned, that financial crises often seem to create environments that lead to the rise of populist political parties or movements. And so, I think it's useful to just define populism, because you see that term thrown around a lot now. And so, the way political scientists would define populism is a thin-centered ideology that divides society into two antagonistic, homogeneous groups— the good people versus the corrupt elite. Populists come in and claim to represent the people, the good people and the general will, and populism is thus sort of characterized by anti-elitism and anti-pluralism.

Verner: You can then start to see, well, why does a financial crisis or a debt crisis actually create a fertile environment for that type of political movement? It is— well, in a debt crisis, there is a basic cleavage between creditors and debtors about how the burden of adjustment to the crisis should be distributed. Often, in crises, there's lots of lending growth before the crisis that then leads to high debt burdens. The question is, well, who should actually pay for the crisis once the crisis comes? Should the debtors have to repay in full, or should there be some sort of haircuts or loss sharing when the world didn't turn out as well as was expected during the boom before the crisis? And populist parties are really well positioned to take advantage of this cleavage, because it dovetails with their aim to speak for people, usually debtors, against the elites, bankers, finance, capitalism, all of that. And so, we looked at this in the context of the case of Hungary. So, I've actually written a few papers about the case of Hungary. Hungary had sort of a very severe household debt crisis during the global financial crisis.

Verner: So, what happened in Hungary was, before 2008, two-thirds of the mortgages that were originated were denominated in foreign currency, mostly Swiss franc, and this was actually quite widespread in the non-euro areas in Europe, from Iceland to Poland, Hungary, and that meant that households had these really big foreign currency exposures. And in the crisis, safe haven currencies like the Swiss franc appreciated a lot and emerging market currencies like the Hungarian forint depreciated a lot. And so, that meant that people who had taken out these foreign currency loans saw their installments rise by 50% or more. So, you took out a mortgage, say, for $100,000. All of a sudden, you owe $160,000 and your monthly debt has become much more difficult to service.

Verner: And so, Hungary was an interesting case, because they had this very severe household debt crisis, and at the same time, there was this rise in far-right populist movements. The main one that really came out of nowhere was a party called Jobbik, and Jobbik basically came in and promised, to debtors, debt reliefs by intervening in these private contracts, these mortgages, and saying that there should be write downs, maturity extensions, and other forms of forgiveness, and this should be paid for by the banks, many of whom were international banks that were coming into these new markets [and] competing for customers.

Verner: And this party benefited a lot from the crisis and saw a big increase in their vote share, and we see this by looking at individual survey data, by looking at local ZIP codes where there were lots of these foreign currency loans. That's also where there were lots of these people who voted for this far-right populist party. And this position then became adopted by the right-wing movement in Hungary, led by Viktor Orban, which ultimately led to some of these policies that actually did provide some debt relief for foreign currency debtors.

Verner: So, the reason I like the Hungary case, even though it's an extreme case, is it shows a very clear-cut case of a situation where debt, and how to deal with debt, was a major political issue that created a major divide in society, and there was an opportunity for a populist party to come in and insert itself in that cleavage and gain support in a way that more establishment centrist parties that don't feel comfortable with, for example, intervening in private contracts, weren’t able to do.

Verner: And so, to your earlier question of, well, what explains the rise in populism that we've seen over the past few decades or so? I think that the financial crisis was really an accelerant on top of some of the long-term trends that we had before. So, I think it's rising inequality [and] trade globalization [that] certainly contributed to the rise in populism as well as immigration. Cultural factors also mattered. So, the distance between ruling elites and ordinary voters seemed to get bigger.

Verner: That created the opportunity for new political movements that argued that they were more in touch with the concerns of ordinary voters, and the crisis sort of reinforced those dynamics and created an environment for a backlash against financial globalization against banks, especially international banks, and Hungary was sort of a microcosm of that and somewhere where it happened a bit earlier and in a more extreme way than in other places as well.

Beckworth: Well, that's a great story and a great paper. We'll provide a link to it in the show notes, along with your other papers that we've talked about. So, circling back to what we said earlier about these credit booms that are excessive— In the case of Hungary or the world, the Great Financial Crisis— what do you recommend policymakers do, if anything? Say, we're past the point. We didn't stem the excessive credit creation in the first place, but say we're in 2008, and we're struggling, and we want to avoid the rise of populism. What can we do? Can we adopt state contingent debt contracts, more supportive fiscal policies? Any suggestion for the policymakers listening today, if we find ourselves in that place again?

How Do We Stem the Tide of Populism in the Future?

Verner: Yes, that's a good question. What should the policy playbook be right before a crisis? I think, ex ante, you want to focus on times of rapid lending growth and focus on what credit is financing. Is it financing [the] productive parts of the economy where there's strong productivity growth, or is it more financing consumption and sectors where there's lots of collateral-based borrowing that can sort of get out of whack with asset price growth? If it's too late to do that, then what becomes important is focusing on the solvency both of borrowers and of the financial system. Policies that prevent banks from paying out a lot of capital through dividends right before crises, and policies that are focused on recapitalizing the banking sector during crises— those are key to making sure that the banking sector can resume servicing the economy.

Verner: But at the same time, it's important to not just focus on the banking system, but to also focus on borrowers and to think about different ways to help borrowers repair their balance sheets. So, this can be through debt relief programs, especially programs that target giving borrowers liquidity in the short term, and that can help restore demand in the economy. And I think having an approach that focuses on the banks and the borrowers seems, to me, to also be less likely to lead to major backlash than one that's mainly focused on the banking system. And that was part of the debates that we saw after 2008, was that there was a lot of concern that policies were focused too much on banks and not enough on indebted homeowners.

Beckworth: Okay. Well, with that, our time is up. Our guest today has been Emil Verner. Emil, thank you so much for coming on the program.

Verner: Thanks for having me. It was great fun.