UN/BALANCED Episode 5: The U.S. Banking System Reacts to Rate Hikes

Welcome back! Apologies for the gap since episode 4. (We have a good reason, we promise.)

This month we take a look back at the ructions in the U.S. banking system since early March, what they mean, and how the financial system might look different going forward as a result. Along the way we cover everything from interest rates in China to Depression-era banking regulations to central bank digital currencies.

The first 15 minutes are free for everyone, but the full conversation—and full transcript—are for paying listeners. Please consider subscribing to get the full experience! As a reminder, all paid subscribers to The Overshoot are eligible for a substantial discount to paid subscriptions for UN/BALANCED.

Related links:

Thoughts on the Bank Bailouts — The Overshoot, March 13 2023

Banks Are Blowing Up While the Economy is Strong. Time to Worry? — The Overshoot, May 10 2023

Bank Regs Are Excess Profit Taxes — The Overshoot, July 20 2021

How a Regulatory Tweak Will Affect Banks, Money Funds, and the Financial System — Barron’s March 26 2021

Silicon Valley Bank bid report — FDIC

Review of the Federal Reserve’s Supervision and Regulation of Silicon Valley Bank — Michael Barr, April 28, 2023

Signature Bank bid report — FDIC

FDIC’s Supervision of Signature Bank — FDIC, April 28, 2023

First Republic Bank bid report — FDIC

Parachute Pants and Central Bank Money — Randy Quarles, June 28 2021

Economic Report of the President — Council of Economic Advisers, February 1970

Thanks again to White+ for the music and to George Drake Jr. for producing!

The full transcript is below, lightly edited for clarity.

Matt Klein: Hello, and welcome back to UN/BALANCED, a listener-supported podcast about the global economy and the financial system. I’m Matt Klein.

Michael Pettis: And I’m Michael Pettis. Today we’re going to discuss what’s been happening to the U.S. banking system as interest rates have risen and some banks have gotten into trouble. Some relatively large banks have blown up. The U.S. government has spent over $30 billion bailing out a few of the larger depositors, and a lot of surviving banks don’t seem to be in great shape.

Matt Klein: But first, I would like to make a brief special announcement. As you probably have noticed, there was a longer gap than normal between our previous episode and this one, and I’m happy to say that we have a very good reason, which is that my wife gave birth to our second child back in March, shortly after we recorded the previous episode that we released in April.

Everyone is doing great now, thankfully, but since the baby arrived a few weeks before her due date, we were not able to record as many episodes in advance as we’d hoped. We should be back to our regular monthly schedule from here on out. Thank you all for your understanding. Now back to the main event.

Michael, we both started our careers in the midst of major banking crises, and those experiences have played a large role in our thinking and in our historical interests.

Michael Pettis: Let me interrupt just so that people don’t think it was in the midst of the same major banking crisis. Your banking crisis happened much later than mine.

Matt Klein: Yes, that is a fair point! Nevertheless, I do think that experience has shaped our mutual interests. Looking at this latest episode, what I find interesting is that in some respects, it perfectly fits the template of how banks get into trouble. But at the same time, in other very important respects, it’s very different from what you and I lived through in the 2000s and the 1980s.

So maybe we can start by just going through kind of a quick, straightforward narrative of what actually happened with the beginning of this crisis with Signature Bank and Silicon Valley Bank back in March.

Why did these banks blow up when they did? All banks, what do they do? They borrow from one set of customers and they lend to some other set of customers, either by making loans or by buying bonds or something like that.

And the trick is to borrow from people who don’t want to get their money back anytime soon. And then you make loans to people who will pay you more because there’s some kind of risk involved. If you can manage that spread, then you make a nice profit and everyone’s happy.

The problem with Signature and Silicon Valley Bank is that they ended up borrowing overwhelmingly from the types of people that are most likely to leave at the first sign of trouble.

So you look at their call reports that they file with regulators and overwhelmingly they’re borrowing uninsured deposits from businesses. Not even the operating deposits that businesses use to make payroll, but the longer-term non-operating deposits.

Now, if you look at what regulators say when they classify the risks of depositors and how banks and bankers should hedge against those risks and what kind of assets they should hold against those deposits, these are considered the absolute riskiest flightiest characters. And yet, these banks were holding relatively long-term assets that could not be sold on short notice without taking some kind of a loss.

On top of this, and this is what I think is kind of most interesting here, is that a lot of these deposits weren't even any earning any interest. Now, in 2021 or 2020, that doesn’t really matter, right? Because the interest rates you get on short-term money were effectively zero, regardless. So if you’re getting zero interest on a bank account versus zero interest on a Treasury bill, who cares?

But by the time you get to the end of 2022, the interest that you can get holding onto the safest possible short-term assets—short-term Treasury bills, or an overnight loan directly to the Federal Reserve—you can get yields of 4% or more. And yet these banks, for the most part, were still paying something on the order of zero. Maybe a little more than zero, but basically zero.

So the mystery to me, in part, is: why would these business customers even keep their money there in the first place? If you’re a corporate treasurer, it is not very hard to get a better yield for your shareholders and fulfill your fiduciary responsibilities. Nevertheless, that’s the situation.

Then you have a situation where both of these banks owned a bunch of bonds that were underwater in various ways, particularly Silicon Valley Bank. They had bought a lot of mortgage bonds during 2021, when mortgage bond yields were very low. Rates went up quite a bit since then, and what happened is that some depositors said, “well, this is going to be a problem.” And they start to leave the bank. The depositors see that there’s a loss, but more importantly, the bank sees that there’s a potential loss, so they try to raise capital to cover it, but then the depositors get very upset.

It turns out the deposit base for Silicon Valley Bank in particular was very concentrated among startups and venture capital firms. Those venture capital firms in particular could effectively advise their startups on how to behave and what to do. And they pulled their money very quickly. The post mortem says something like 40% of their deposits left in the course of a day, before they were shut down. Banks can’t respond to that.

So the interesting question here is, on the one hand, this is obviously very standard: the way banks fail is if someone stops lending to them, they fail. That’s just a very basic function of how banks are constantly borrowing. If they can’t roll over the debts, they blow up.

But there’s some interesting differences here in terms of what happened.

In particular, it’s not as if they took a lot of credit risk and then things blew up.

That’s actually not what happened at all. They were holding very safe assets, which if they’d held to maturity, they would’ve been paid back exactly what they thought they were going to get. Yet nevertheless, they blew up. That is obviously very different from a lot of the stereotypical examples.

What are you thinking of in terms of some of the historical examples that might spring to mind here? What makes this similar and what are some of the interesting differences?

Michael Pettis: Traditionally there are two reasons why banks can get into trouble. One is they become insolvent and the other is that they have horribly mismatched balance sheets.

In this case, there may have been an insolvency issue because the value of their assets dropped relative to the value of their liabilities. Typically, banks have insolvency issues because they make bad lending decisions. This was not a bad lending decision; this was a bad balance sheet management decision.

But they always run the risk of mismatches between liabilities and assets, because it’s the job of the banks to convert short-term savings into longer-term investment. Whenever depositors want their money back—whenever the liability side of the balance sheet contracts—that puts pressure on the asset side of the balance sheet.

And if the bank isn’t able to liquidate its assets very easily, which means at more or less the price at which it created them, then it gets into serious trouble.

To me, the really interesting story here is the stickiness of deposits—or the lack of stickiness of deposits. I had always learned, and I’ve always taught in my banking classes—maybe I’ll teach it less, or teach it with all sorts of restrictions—that there were certain types of deposits that bankers loved. Those were basically retail deposits. And the reason they love them was because it took an awful lot to get them to move.

Many banks are able to calculate what they call the “franchise value” of the deposits.

If you have a lot of institutional deposits, for example, if I’m a treasurer at a company, and I deposit $100 million in your bank, that $100 million is incredibly mobile. If another bank, if a JPMorgan calls me up and says they’ll give me five basis points higher yield, it’s worth my while to take my money out of your bank and put it into another bank and to get the five basis points of higher yield. By the same token, if a friend calls me up and says, “we’re hearing rumors about your bank”, it’s worth my while to move money out as quickly as I can.

But if I’m a retail depositor with $40,000 or $50,000 in the bank, it becomes much more complicated. First of all, I can’t move my money really easily. I have to take off a morning or an afternoon from work to go to the bank to move the money. This is the old days. It’s obviously not true today, but in the old days there was a cost of doing that, and my professor used to call it “shoe leather costs”. You have got to do a lot of walking around to get your money out of the bank and into another bank. So in that scenario, it would take quite a lot to get me to move my money.

I would need a huge difference in interest rates; another better or equally good bank would have to offer me a lot more money before I would take my money out of the bank and move it into that bank. 1% on, on $40,000 is $400. That’s $400 in a year. That’s not nothing, but if I think that the other bank is going to catch up within a month or so, it’s hardly worth the effort to move my money out. If I hear a rumor, I’m going to need a lot more than a rumor before I’m going to go through the hassle of going to the bank, taking time off from work, going to the bank, waiting in line, et cetera, et cetera.

One of the things that’s happened in recent years is that the frictional costs, or the “shoe leather cost”, of moving money around has collapsed. It’s almost zero.

So now it seems to me there isn’t a huge difference between a small depositor and a big depositor in the sense that either of them can move money very quickly, very efficiently. I don’t know if you can do it in a few minutes, but certainly you can do it in less than an hour at the first sign of trouble or at the first sign of differences in interest rates, et cetera, et cetera.

I’m not living in the States, I’m certainly not out in California, but it seemed to me that part of the problem may have been that this stickiness of deposits, which was a great asset for banks, has basically collapsed to zero, or close to zero, so there’s no longer any franchise value in deposits. They can move around at an incredibly rapid speed for extremely low cost, and therefore they’re much more volatile than they used to be. Would you say that that was one of the big problems with Silicon Valley Bank?

Matt Klein: I think there’s an element of that. Although I also think it’s worth stressing that both Signature Bank and Silicon Valley Bank didn’t really have the kind of franchise value and deposits that we think of with a normal retail bank because they didn’t really have retail customers. In that sense not much has changed. In retrospect, I think they arguably had a negative franchise value in their deposits because of the kinds of clients that they had.

But to your point, the normal default thing that makes sense is that if you’re a corporate treasurer and you’re managing large sums of cash, you would be very sensitive to differences in interest rates and risk.

What’s really interesting is that in both banks, that’s not what happened. And I think the reason is that they both tried to cater to specific industry sectors and offered them services that they couldn’t get elsewhere.

It’s kind of analogous to the situation you had in the 1950s or 1960s where the bank would not pay you interest, but you’d get a toaster, essentially. That’s kind of the thing that they offered, where you wouldn’t get interest—you’d be vastly underpaid compared to Treasuries or lending to the Fed or things like that—but there were other ancillary benefits.

In the case of Silicon Valley Bank, they offered very favorable mortgages to rich individuals. There’s an interesting question there of individual venture partners getting a very good deal on their mortgages and then pushing all of their venture funded companies to put cash there. That’s an interesting dynamic.

Signature Bank went in different direction. They really catered to the crypto industry. And they had all the systems set up that made it easier for various crypto businesses to pay each other in dollars internally through the bank. Again, that creates an interesting dynamic where, if the crypto industry tanks, as it did basically starting in the beginning of 2022, they’re pulling out their deposits for reasons totally unrelated to anything else happening.

With Silicon Valley Bank as well, you had a huge influx of deposits as various venture backed companies were raising a lot of cash. And then they stopped raising cash and they started burning that cash, and that started naturally pulling deposits back. These were underlying problems.

If you think about it from that perspective, the deposit franchise is negative, right? You have got yourself exposure to particular sets of financing that are not just volatile, but much more volatile than any other kind of industry, any other kind of deposit base you can think of.

And those deposits are correlated in a very unfortunate way, as it turns out, with interest rates, which is also problematic if you’re using those deposits to finance longer-term debts that then lose a lot of value as interest rates rise.

It’s kind of interesting to think “why would someone want to buy these businesses and assume the deposits?”

The normal thing is that when a bank fails and the government steps in, or not, there’s some kind of resolution and then there’s a takeover process. The Federal Deposit Insurance Corporation will make sure that insured depositors get seamless transition of service. And the feeling usually is “oh, well it’s good for a bank to do that because those deposits are valuable.” As you said, retail deposits are sticky. That still seems to me mostly, in the aggregate, mostly to be the case. And yet these deposits probably are not ones you’d want.

Michael Pettis: Not only are they ones that you probably wouldn’t want, but it seemed to me that Signature Bank was a particularly surprising event. During the big period of consolidation of the banks in the 1970s and 1980s, the classic argument was that small banks are overly exposed to a particular part of the economy, and so that increases the probability of default, right? If your economy, if your town does badly, the bank will collapse, and so you wanted to be diversified to protect yourself.

But here it seems they were highly concentrated, not just in one particular industry, but you know, for God’s sake, perhaps the most risky industry out there, right?

It’s incredible to think that no one considered the possibility that there may be a significant contraction in the crypto industry and how that would affect the balance sheet of the banks.