Podcast: “It’s Hard to Regulate U.S. Banks!”

Unfortunately, more banks than you’d think share the risk factors that kicked off a run at SVB, according to a recent analysis from Gregor Matvos, a finance professor at Kellogg, and his colleagues Erica Xuewei Jiang, Tomasz Piskorski, and Amit Seru.

“There [are] plenty of other banks, 186 to be precise, that sure look like they could be subject to a run,” Matvos said.

On this episode of The Insightful Leader, Matvos shares his findings and what it would take to patch the holes in the U.S. financial system.

Podcast Transcript

Laura PAVIN: Hey everyone. It’s Laura Pavin.

If you look at any news headlines at all, you know that Silicon Valley Bank recently experienced a bank run—meaning that a whole bunch of depositors all tried to take back their money at the same time. It was the second biggest bank failure in U.S. history. The Biden administration recently stepped in to make depositors whole, but the whole incident has raised all sorts of questions about where to place blame for the bank’s failure and where the U.S. financial system goes from here.

So on this episode of The Insightful Leader, in addition to getting our arms around just what brought SVB to this point, we look at some new research from a professor right here at Kellogg, along with some of his colleagues. Because his analysis shines a light on the fact that, in some important ways, SVB is less of an outlier than you might think.

Insight editor in chief Jess Love spoke with Gregor Matvos, a finance professor at Kellogg, about it all. They started by talking about the role that treasury bonds played in the bank run at SVB.

Here’s Matvos.

…

Gregor MATVOS: Last year or so, the Federal Reserve increased interest rates, which led to a decline in the value of the holdings of banks, the asset side of banks, as we call it. So in other words, the loans they had made, the bonds they held, decreased in value.

LOVE: And why is that?

MATVOS: Well, think of it this way. You buy a government treasury bond that will repay over 20 years. Two years ago, interest rates are close to zero. Then the government starts issuing new bonds with interest rates at four percent. Obviously you’d rather have four percent than zero percent. That seems like a no-brainer. Well, if the four percent bond right now costs you $1 to buy, then the bond you bought in the past surely is not worth a dollar anymore. It’s worth less because it’s giving you less interest payments over time. In other words, it declines in value. And the issue banks have had is, first, not only are a lot of their securities somewhat longer maturity, in other words, they pay over time and therefore lose when interest rates go up, they also didn’t record these losses on the books. They pretended as though the value of these securities was the same as from the time when they issued them. In other words, if you made a loan a year ago or two years ago, you still sort of said, “Well, the loan suffered no losses; therefore, it’s worth the same.”

LOVE: Is that legal? Is that allowed?

MATVOS: Yeah, so recognizing losses in a mark-to-market way has always been a little bit fraught with banks. The recognition of losses tends to be slow. Um, sometimes that’s okay; sometimes that’s not okay. The interesting question here is the losses are quite substantial. So, depositors, particularly uninsured depositors, and we can get into detail of who those are, started getting worried about, “Wait a second. If all of us withdraw our funds, is there going to be enough to repay us?” And specifically imagine I’m one uninsured depositor, and I say, “Well, if all people like me pull out their money, is there gonna be enough money left for me?” And once people started worrying about there not being enough money and, well, through a good old bank run, there may in fact not be enough money left over and, well, it makes sense to run. And that’s partly what happened to Silicon Valley Bank. Now, what was especially interesting about Silicon Valley Bank is it’s depositors were a tight-knit community in which the news of potential demise spread even faster than it normally would.

LOVE: So was there a specific trigger or did somebody just out of the blue say, “Hey, I think there’s a problem,” and then go tell their 50 best friends?

MATVOS: Yeah, so it’s always very hard to put your finger on a trigger. I mean, it was clear that, sort of, banks were in slightly worse condition; there’s rumors that start swirling; and then enough depositors redeemed that Silicon Valley Bank said, “we had to sell 20 billion of our bonds at a loss.” I think that was actually a misstatement. They didn’t sell them at a loss, they sold them at market price. They just now finally had to acknowledge that the assets had lost value, which prior to that they had not done.

LOVE: Oh wait, this is just a clarification. When you say … did you just misspeak, or did they actually misspeak and say they had to sell at a loss?

MATVOS: I think they misspoke. I mean, in my view, just because you claimed something was worth 20 billion dollars and then you sold it for 18, [that] didn’t make it worth 20 billion dollars before. It was already worth 18 billion dollars; just on the books, you had it under the old value. And that’s part of what, sort of, when we’re gonna talk about what’s happening in the U.S. banking system right now is, well, there’s a bunch of assets [whose] values have not been corrected yet to reflect what their current market value is.

LOVE: Now, it seems like something that’s going to be important is this idea of uninsured deposits. And so this would be any deposit over $250,000. Is that right?

MATVOS: That’s correct. In the U.S., historically about half of all deposits are uninsured. Now, people tend to forget that and kind of close their eyes to this idea, because most of the time we don’t worry about banks … until we do. So uninsured deposits are not a problem until they are. And Silicon Valley Bank really stood out on that dimension.

LOVE: I find that just fascinating, this idea that, well, I guess we can get to that in a little bit, but I just wanna be like, “What were you all thinking?” Is the ideal scenario to just have 20,000 bank accounts? That seems like a complicated situation too.

MATVOS: I think you want to think a little bit about, who are the people who would have an uninsured deposit account, right? Over $250,000. Now it could be a depositor who is just forgetting that they have their salary kind of going into their account. They haven’t really paid attention, and more than $250,000 accumulates. Which, when it was a zero-interest-rate environment, probably doesn’t matter that much. You’re not getting paid anything in your deposit account; you’re not getting paid anything anywhere else. Might as well leave it there. Um, so let’s call this “sleepy individual investors,” if you will. Now, those quickly can readjust and maybe send money to two banks or put money in the treasuries, but it’s kind of a way of saving.

Another set of uninsured depositors would be something like charities or small businesses: You have to make payroll. It’ll be quite inconvenient to have your payroll spread across 50 banks and paying your workers across 50 bank accounts, and so on. So you have one account with one bank, and you run your payroll through there, and you probably get your loan through there, and all the other services. In some ways, banks offer these bundled services, and it kind of makes sense to stick [with it]. And it’s super convenient, and most of the time you really don’t have to worry about it, until you have to worry about it. So in short, you know, we shouldn’t be that surprised that there’s uninsured deposits, because it’s practical for small businesses, charities, and so on, you know, to keep more than $250,000 in one place. And most of the time, if you have a big first equity cushion … imagine that Silicon Valley Bank had been funded instead of 90 percent through deposits, 80 percent through deposits, and had 20 percent funding through equity. You can lose a lot of value and still not worry about deposits ever being impaired. But U.S. banks don’t do that. U.S. banks have more like 10 percent equity funding. And then if asset values do decline like they did, then all of a sudden we’re starting to eat into something.

LOVE: So I want to make sure we talk about your analysis. So you basically looked at what happened with Silicon Valley Bank and said, um, is this the kind of thing that could actually happen elsewhere? Are they a complete outlier or is this just emblematic of a broader system? So can you tell us a little bit about your analysis?

MATVOS: Yeah, so we looked at all banks in the U.S. Their data is public. And we just tried to look at a couple of things. We sort of said, well first, these hidden losses, are they big? Are they small? Is Silicon Valley an exception? And we found that no, for the average U.S. bank, we found something like losses that probably haven’t been mark-to-market on the order of 10 percent of their assets, which is substantial if you think that equity is about 10 percent of the assets.

So we said, okay, let’s look at these losses. And then we said, okay, so suppose Silicon Valley Bank, you know, defaulted because it was just … because of losses. How many other banks would default? Well, we find many. In other words, Silicon Valley Bank wasn’t a particularly big outlier on asset losses. It also wasn’t an outlier from the perspective of how much equity capital it had. Where it was really an outlier was what we call in some of our other work “uninsured leverage.” In other words, when it funded itself, it funded itself a lot with uninsured deposits. So a combination of losses and uninsured deposits is what triggered a run. If you just had losses and not uninsured deposits, you’d probably be fine. If you just had uninsured deposits and no losses—that was pretty much Silicon Valley Bank two years ago or one year ago—you’d be fine. Having both losses and many uninsured deposits, well, they were not fine.

LOVE: And this is something specific about Silicon Valley Bank because of the tech industry that they’re working with. Is that correct?

MATVOS: Well, turns out not quite.

So it was a pretty big outlier in uninsured deposits. But it turns out there’s enough banks who are somewhat outliers on the uninsured deposits, but also suffer big losses. And for those banks, runs would look very natural if you didn’t backstop uninsured deposits.

So the way we try to evaluate who would be at risk of a run is, we said, “Imagine that of these uninsured depositors, half of them wake up and freak out. Like, ‘Oh my God, I don’t know what’s gonna happen. Maybe I should pull out my money.’ So not all of them are running, but half of them sort of say, ‘Do we have a problem?’ And then you have to repay them. Is there enough money to repay the insured depositors at all?” Forget the rest of the uninsured depositors. Just the insured depositors who also have a claim on the bank, because we think the FDIC would step in at least then. And we find that there are about 190 banks for whom that would absolutely happen. In other words, if half the people ran, the bank would get shut down, and the other people who didn’t run would lose everything. So the incentives to run there would be enormous.

Now historically, or not historically, but if we looked at the letter of how deposit insurance works, the FDIC should generally only protect insured deposits. And that’s sort of why we took that cutoff. Uh, what it seems now, because of the issues in the banking sector, they will guarantee all deposits. So there is no reason to run because, well, the FDIC will stand behind all deposits, it seems.

LOVE: Right. So this is an ongoing situation, and as we are speaking, the Biden administration came out and basically said, you know what, we are going to guarantee all this.

MATVOS: All deposits. To prevent exactly the situations that we were a little bit concerned about. Yes.

LOVE: I see. So its specific makeup is pretty unique just because it’s catering to these large Silicon Valley companies. But when you combine it with your other measure of, uh, to come up with like a total calculation for the likelihood of a bank run, it actually is not as much of an outlier as we would want.

MATVOS: It’s still an outlier, but there’s plenty of other banks, 186 to be precise, that sure look like they could be subject to a run.

LOVE: So was that like 5 percent of the banks in America? Roughly?

MATVOS: So that’s about right. Yes, yes. And we were trying to be fairly conservative. So, for example, we assumed that if people ran on the bank, you could sell the assets at zero discount, which we know doesn’t happen. But we said, okay, you know, let’s try to give the banks a good shot.

LOVE: So, I wanna zoom out to the big picture here. This certainly seems like a disturbing finding. Um, and I wanna ask, like, basically, how this happened. So first of all, what was Silicon Valley Bank thinking? Did they not understand that this was a possibility?

MATVOS: I mean, it’s a great question. Uh, what could Silicon Valley Bank have done to prevent this? They could have bought interest rate insurance a couple of years ago. So had they done that, they would’ve been insured. Or they could have held very short maturity securities. They could have just invested everything in t-bills. But then they wouldn’t have made very much money. Uh, and in some ways, deposits are fairly cheap funding because they, well, people understand it will be backed up by the government eventually. So it’s a way to make a spread. And you make, I mean, in good times, you make a big spread, and you’re fairly profitable. And, well, then sometimes bad times come and you know, [things] go sideways. Uh, so yeah, Silicon Valley Bank could’ve prevented this, but it would’ve done so at, I think, quite an expense. Part of their issue specifically—uh, this is not more broadly for the banking sector—but part of the issue for them is they didn’t have tremendous lending opportunities.

For them it seemed like mainly what they were … their clients had just in some ways too much liquidity. They had to park it somewhere, and they parked it with them. And then they just invested in treasuries to earn something rather than completely zero.

Now for the rest of the banks, I think the issue really is at the end of the day, uh, banks are exposed to fluctuations in their asset values. You know, for example, when interest rates go up by 4 percentage points in a year, there’s very easy fixes to this. If banks are very well capitalized, we don’t have the issue that we’ve just seen in the banking sector. So, for example, in some related work, we looked at the capitalization of financial intermediaries that offer a lot of loans in the U.S. called shadow banks, but they can’t take deposits, and [we] asked, well, “how well funded are those with equity capital?” Because there, you know, you can’t rely on these cheap deposits. What are the debt holders and, you know, require for you to be somewhat safe. And it’s more like 20 percent capital instead of 10 for banks. So if you look at non-regulated financial institutions, in some ways, they have quite a lot of capital because they’ll have no fallbacks. And if U.S. banks had a lot of capital, we wouldn’t have this discussion right now.

LOVE: So are you envisioning a scenario where different banks that are involved in like different types of businesses or have different clients that they work with should actually be held to different capital levels?

MATVOS: No, I think they should all be held to a higher capital level!

LOVE: They’re not gonna be happy to hear this interview.

MATVOS: No, they won’t. Look, of course, we can go back to the drawing board, and we can keep on trying to rewrite regulations. And maybe we can come up with something sensible. But we’ve tried this many, I mean … the U.S. history is littered with trying to write better regulations for financial intermediaries. And then in 2008, we realize, ah, house prices matter. And now in 2023, we’re like, ah, interest rates matter, again, even though in the 1980s interest rates drove a third of the S&L industry—which is also banks—into default. I’d probably sleep a little bit better at night as a taxpayer if capital requirements were higher. Uh, which just provides you a cushion for, well, what if we get regulation wrong?

LOVE: Hmm. So do you have any other long-term takeaways that you see coming out of this particular incident.

MATVOS: That it’s, you know, hard to regulate banks. Um, I think banks are very interesting. I think they provide a lot of really useful services. We need banks in our economy. We are, you know, intermediation is one of the fundamental things we have to move money from people who have too much of it—we call them savers—to people who drive our economy—borrowers. That could be consumers; that could be firms. So we need that to operate well. Um, it’s not like the financial industry isn’t already very heavily regulated, you know. Medicine and finance are two of the most regulated industries in the U.S. Uh, the question is, can we come up with better regulation, and should we come up with sort of super fine regulation again? Or should we just say, like, there’s an easier fix, which is: You want to take deposits? You need a little bit more capital. By a little bit, I mean, five percentage points more, which banks consider quite a lot.

…

PAVIN: That’s it—from us, at least. As we record this, the Justice Department was reported to have started an investigation into the collapse of the bank. So the forensics on what happened at SVB are definitely still ongoing. But the overall cracks that Prof. Gregor Matvos revealed in his analysis do remain. Whether federal regulators do plan to act on his suggested fix—which is to require that banks secure more capital to fall back on in the event of a run—that remains a question.

[CREDITS]

PAVIN: This episode of The Insightful Leader was produced by me, Laura Pavin, and Jessica Love, Emily Stone, Fred Schmalz, Maja Kos, and Blake Goble. It was mixed by me. Special thanks to Gregor Matvos. Want more The Insightful Leader episodes? You can find us on iTunes, Spotify, or our website: insight.kellogg.northwestern.edu. We’ll be back in a couple weeks with another episode of The Insightful Leader Podcast.