A little less than a year ago, the US financial system was rocked by its first major banking drama since 2008. While the crisis was eventually contained, and only three lenders ended up collapsing, the experience re-ignited an ongoing conversation about the way we rescue troubled lenders. Not only did the Federal Reserve launch a new liquidity program called the Bank Term Funding Program as part of its support to the banking system in 2023, but regulators are now talking about changing existing facilities, including the Federal Home Loan Banks (FHLBs) and the discount window. For instance, Michael Hsu of the Office of the Comptroller of the Currency has proposed that banks be required to tap the discount window and "pre-position" collateral at the facility, just in case they one day need it. In this episode, we speak with Steven Kelly, associate director of research at the Yale University Program on Financial Stability, about the constellation of existing emergency facilities for banks, how they've evolved over time, and the changes that could be made to them now. This transcript has been lightly edited for clarity.
Key insights from the pod:
The impact of 2023 on banking regulation — 05:39
Can good banks fail? — 07:09
The role of the FHLBs last year — 08:54
Proposed changes to the discount window — 12:52
FHLBs vs. discount window — 15:58
The end of the BTFP — 18:06
Liquidity vs. solvency — 20:35
Deposit runs and social media — 21:49
Tension between monetary policy and banking regulation — 24:23
Carrots and sticks at the discount window — 28:06
What is the standing repo facility? — 30:49
What do banks actually do? — 32:42
The Basel III endgame proposals — 33:53
The future size and shape of US banks — 37:00
---
Tracy Alloway (00:17):
Hello and welcome to another episode of the Odd Lots podcast. I'm Tracy Alloway.
Joe Weisenthal (00:22):
And I'm Joe Weisenthal.
Tracy (00:23):
Joe, do you remember what we were doing this time last year?
Joe (00:28):
Nope, not really.
Tracy (00:30):
I should say around this time last year. But yes, I can't really even remember last week at this point. But in the middle of January, we recorded an episode on the discount window.
Joe (00:42):
Yeah, on the discount window. Yes, I do. Why were we talking about the discount window? Just for the fun of it?
Tracy (00:47):
No, because borrowing was going up.
Joe (00:50):
Right, yes. February, early March of last year, we did that episode about deposit rates and then that turned out to be very relevant. But even before that, you started writing about the rise in discount window borrowing.
Tracy (01:10):
Yes, that's right. I think I wrote something in December, I think the headline was like billions in discount window borrowing suggests all is not well with banks, which turned out to be fairly true come March because we did see a mini banking crisis, banking drama, whatever you want to call it — the collapse of three banks, including Silicon Valley Bank.
But since then, there has been this massive discussion about how to tweak all these emergency financing programs for banks. What should they actually look like? How do we want the lender of last resort system to operate? And what does it mean for banks if they have to hold back additional collateral in order to access these programs then hat does it mean for them from, you know, a capital or a revenue standpoint? So many things going on in this space right now. It's a little bit under the radar, but I think we should talk about it.
Joe (02:07):
Totally. It always is like there's this inherent challenge, right? And even with Silicon Valley Bank, even setting aside emergency borrowing. The thing I guess that people get anxious about is as soon as any financial institution makes moves to shore up liquidity, shore up finances, etc., that becomes a signal, right? To investors or depositors, whoever. It's like ‘Oh wait, why do they feel the need to shore up their finances?’ And then you become a target and you start worrying about the share price and equity, etc. And it feels like this is an inherent challenge for regulators, which is, of course, you want banks to be proactive. You want banks to have plenty of liquidity, you want them to have equity cushions. But if the act of doing so invites suspicions, then how do you get out of that puzzle?
Tracy (02:53):
Well, exactly. And I think this is most apparent with the discount window where people talk about the stigma of accessing the discount window. And, you know, when access it, it's supposed to be anonymous. The Fed doesn't publish who's actually tapping it until, I think, two years later. But you start to see rumors swirl.
In the case of last year, when we saw that discount window borrowing started to go up in December, I think that was the proximate time when people started to ask questions about, like ‘Well, wait a second. What’s going on with Signature? What's going on with Silicon Valley Bank?’ That might've contributed to some of the deposit withdrawals that we eventually saw.
So the super interesting thing now is that people are talking about reforming the discount window as well as some other facilities. We saw the Acting Comptroller of the Currency, Michael Hsu, also a former Odd Lots guest, talk about the idea of maybe we're going to require banks to tap the discount window every once in a while. Just so the stigma may be reduced, but also [so] they have the operational readiness to do it when they really need to.
Joe (03:59):
Can I say something? It's almost inappropriate. It's borderline
….
Tracy (04:03):
How are you going to make the discount window inappropriate? I'm curious now.
Joe (04:06):
So, you know what I think whenever I hear this idea of “If you just get everyone to do it, there's no stigma?’
Tracy (04:13):
Oh, I know exactly what you're going to say...
Joe (04:15):
I tweeted this once. I always think about the scene in the movie Billy Madison, where the kid is really embarrassed because everyone can see that he peed his pants. And wait, who's that actor who’s in that? The comedian?
Tracy (04:29):
Adam Sandler.
Joe (04:30):
Adam Sandler. He splashes a bunch of water on his own pants. And so he’s like ‘Everyone around has water on their own pants, there's nothing embarrassing about it.’ And in my mind, I'm sorry, maybe I have juvenile humor, that is always where my mind goes. If you have everyone do it, no one can be embarrassed about doing it.
Tracy (04:48):
Okay. Well, on today's episode “Why Requiring Banks to Tap the Discount Window is the Equivalent of Splashing Your Pants With Water.”
Joe (04:57):
That's going to be the title of this episode.
Tracy (04:59):
Excellent. Well, we really do have the perfect guest. It's someone we've wanted to get on the podcast for a long time. And I don't think he's been on before though. We're going to be speaking with Steven Kelly, associate director of research at the Yale program on financial stability. So Steven, thank you so much for coming on Odd Lots.
Steven Kelly (05:15):
Happy to be here and talk about Billy Madison.
Tracy (05:19):
So, how much did the banking drama of last year change attitudes towards emergency lending facilities? I guess another way of asking the question is do we think that all these lender-of-last-resort type things actually stood up to the test last year?
Steven (05:39):
Yeah, so it's a tricky question. Obviously what comes after these crises is you go why didn't it work, right? It's there, the Fed can mint money, why doesn't it work? You know, why didn't SVB just post everything at the window? And we can get into the various reasons about that.
But basically it's hopeless to say ‘Oh, the discount window is going to work,’ once you have a name that's in the headlines. And that's sort of the pressure we put on it sometimes. What the discount window is great for is sort of a macro story. It's great for contagion, it's great for maybe a community bank that isn't facing the same kind of headline risks. It's never going to save that bank that's in the headlines.
And we can go into all the reasons about why, but so much of the franchise value just gets destroyed so fast. And when you're talking about replacing all your depositors, well, what is a bank but a collection of its depositors? So you can put all the money you want in the window, but it's never going to save that bank. That's just too much pressure on the window.
Joe (06:37):
Can good banks fail by taking on these countersignals to the market? So whether it's going to the Fed and trying to get additional liquidity, whether it's doing an equity sale at some point just to create that greater equity cushion. If the bank is fundamentally sound, can simply expressing concern be enough to bring it down? Because It does seem like people don't want to send those signals. Or ultimately if the bank is good, then these are good moves to take and they'll survive it.
Steven (07:09):
Yeah, the biggest thing is if you can't get capital. I mean, capital is what protects the deposit layer of the balance sheet. So if you can't get capital as a bank, you're out of business. So SVB came out March 8th last year with an 8-K that said, you kow, ‘We’re kind of looking at raising $2.25 billion, we have $500 million of commitment.’ That was enough to say, ‘Okay, they gave an inside look at the balance sheet and nobody wanted it.’
Joe (07:32):
So it wasn't that they were raising capital, it was that they announced that there was this gap.
Steven (07:36):
Yeah. If they had come out on March 8th and said ‘Warren Buffett is investing $2.5 billion, we would still have SVB today.
Joe (07:42):
Got it.
Tracy (07:43):
Maybe this is a good place to sort of back up and dive into what exactly happened with SVB. So there seems to have been a reluctance, or an inability, to tap the discount window soon enough. But we also know in retrospect, with the benefit of hindsight, that they were tapping another emergency… — well, I should be careful here. It's not really supposed to be an emergency lending facility. I'm talking about the FHLBs, the Federal Home Loan Banks. We used to call these, by the way, we used to joke in like 2009 that FHLB stood for ‘Free Hubris Loans for Banks’ or ‘Find Huge Lumps of Bucks.’
Joe (08:25):
I love ‘Find Huge Lumps of Bucks.’
Tracy (08:28):
But anyway, I mean, this is a facility, it was supposed to facilitate home ownership and help banks do mortgages for people, but it sort of transformed to become an alternate emergency lending facility — and we should definitely talk about why. But SVB was basically tapping that instead of the discount window. So walk us through what we saw from that particular bank, in terms of the choices they made to access different types of liquidity.
Steven (08:54):
Yeah, so we call the FHLBs the ‘FLUBS’, which they don't love, but we'll do it anyways. So I think it's not unique to SVB that they sort of relied on the FLHBs. You know, there was an SNL sketch after 2008 during the original stress test where they sort of did a bit where like the stress test was an actual exam that banks had to take. And Citigroup kept answering ‘government bailout’ to all the questions...
Tracy (09:20):
I never saw that!
Joe (09:21):
Yeah, I don't remember this either.
Steven (09:21):
That's sort of what happened with the FHLBs, particularly prior to March and prior to, you know, the supervisory pressure we've seen since where if you ask a bank ‘Hey, what do you do if you really get pinched?’ They go ‘We'll go to the FLUBS, we have a great relationship with the FHLBs.’
I mean, that's another piece, that these FHLBs are a lot more commercial in nature than the Fed. But that was sort of the contingency funding plan writ large across the system. And, you know, it sort of works if you need a billion or $5 billion if you're SVB. It doesn't work if you need $40 billion because the FHLBs take government collateral, they take mortgage collateral. If you need to start posting commercial and industrial loans, something like that, or corporate bonds, you’ve got to go to the Fed. And if you're not set up at the Fed because it's annoying to do that, it's too expensive. You don't want to leave collateral there, whatever the reason, you run out of time.
Joe (10:10):
So some banks just aren't set up with the Fed? [Is it] too annoying?
Steven (10:14):
Exactly.
Tracy (10:14):
That’s the operational readiness argument for making them splash water on their pants slash go to the discount window.
Steven (10:23):
Right, so SVB literally couldn't get collateral to the Fed in time before the run was out. Again, the SVB story was over, so it didn't matter. But it's useful to be ready to go at the Fed because they can take, effectively, your whole balance sheet. Basically any asset a bank will have, you can put to the window. I mean, that's why it exists. And the FHLBs, because of this sort of housing origin and whatever other reasons, just don't have that range.
Tracy (10:49):
The same thing happened at Signature, by the way. And I think there was a really good speech by a Fed official, I can't remember who it was, but they were talking about how it had basically been five years since Signature tapped the discount window. And when it came time to tap again, things were blowing up. You need to access emergency liquidity. The Signature staff didn't really understand the rules around collateral eligibility and what they had to do in order to actually go to the window and borrow money. So I can see the argument for why you would want people to practice it.
Joe (11:25):
Hey, what happened to Signature? Who took over their assets again? I'm making a bit of a joke there. I'm aware.
Steven (11:35):
I'll have to check my Bloomberg. I think there's some sleepy bank no one's ever heard of.
Joe (11:40):
NYCB. Okay, so that sort of blew my mind at the time. Here are these institutions and I guess that weekend, obviously, I guess the Fed or the Treasury determined that there was some sort of emergency aspect of it. And we all know that there was this sort of rescue and they opened up this new program, the BTFP, which we'll talk about.
But it sort of blew my mind that you could have this crisis. And part of it is, well, what time is it? What time is the window actually open? Can we reopen the window? That sort of blew my mind. So what happened immediately? Or maybe not immediately, but in the wake of all of this? We've sort of talked about the runup. What changes did we see regulatory-wise in the wake of the SVB and Signature disaster?
Steven (12:25):
So the biggest change, which is long overdue, is you’ve got to post more collateral at the window. It's this term ‘prepositioning’ that we're starting to hear more and more of. And, you know, you’ve got to practice. And we're hearing more from regulators that they would like a little more practice in this, you know, and this is probably the direction that supervision needs to go. Where like, ‘Hey, if you're not practicing, we're going to dock you. If you can't show that you can show up at the window and get liquidity when you need it, we're going to dock you from a supervisory perspective.’
But really, there's been a lot of pressure to send more collateral to the Fed. There is something like $3 trillion of collateral at the Fed — we don't get daily updates on this — but it's in that vicinity and we know it's been growing. And we're also hearing from regulators that there's maybe some reforms to be had and some new liquidity measures to take
Tracy (13:26):
Is pre-positioning a synonym basically for encumbrance? So the idea that, I have a certain number of assets on my balance sheet, I'm going to have to use more of those or set more of those aside at the central bank in order to satisfy these new requirements. And therefore, they're going to be less available to me to do stuff, to do creative and hopefully revenue-generating things with them like repo them out, or something like that.
Steven (13:52):
So yes, that's part of the story. And the other thing I think about is banks have a lot of loans. So, those are just sitting there. If you move those to the Fed, you get over the hurdles of moving them. It's not like you're repoing out your loans necessarily as a community bank. Just house them at the Fed, don't leave them at the FHLBs, send them to the Fed. And so that's going to be a piece of it.
The other thing that may come here is some sort of carrot with that stick. So the hardest thing about the discount window is that it really can't be any cheaper. And that's the stigma, is that it's really expensive. So if you have a deposit which is yielding zero, and you've got to go to the window, the Fed can say ‘Oh look, it’s just the top of the Fed funds rate.’ But that's still, you know, right now it's 500bps more than deposits.
Joe (14:38):
Oh, I see. So even if it’s just like, you're just borrowing, even if it's just at the Fed funds rates, like really banks are borrowing from their depositors — that's way more, or whatever it is now, it's not zero anymore, but whatever the typical deposit is.
Steven (14:51):
Right, so it's incredibly expensive. It would be great if you were Coca-Cola and you could go to the discount window, but you can't. So the way to de-stigmatize is to offer some sort of carrot. And you can say ‘Hey, if you pre-position collateral, we'll give you some credit towards your LCR or we'll give you some credit towards these other things. So you can self-insure less and do more profitable things with your balance sheet,’ that's maybe a viable route. And that's why prepositioning is sort of coming into vogue.
Tracy (15:16):
Wait, can you talk a little bit more about the rates available? I guess specifically the discount window versus the FHLBs? Because this is something I never quite understood. So if you go to the FHLB, the FLUB, I think they actually, they do something where they do look at unrealized gains and losses on your securities in order to see whether or not you're a viable entity to be lending money to. So if you are super stressed, they might not actually lend you money. But on the other hand, it seems like everyone kept going to them. And I'm assuming it's because the rate of borrowing is more attractive than the discount rate?
Steven (15:58):
Yeah, so that's a big piece of it. And it's the confidentiality. So the valuation thing you're thinking of is only on available for sale. So that's part of the story , you can still hide the losses and be held to maturity to some degree. But the pricing is definitely advantageous. And nobody can write that story that you wrote last February, Tracy, where you're saying it looks like there's…
Tracy (16:17):
Ahem. December, please. Give me my extra two months.
Steven (16:21):
I think you told these banks to borrow so you can write the story. Nobody can write that story because it's not public data. We don't get a weekly balance sheet from the Fed.
Joe (16:30):
Why isn't it public?
Tracy (16:31):
Because you don't want to run on the banks!
Steven (16:32):
Because the FHLBs are a private entity. They're cooperative, basically put together by the banks. And so that's a piece of the stigma and the funding is really good because this is a GSE, it's a government-sponsored enterprise. So when you have a crisis and there's a flight into government money market funds, what can they buy? They can buy Treasuries and they can buy FHLB debt.
And so you really have cheap issuance. And the FHLB pays out their earnings to members. They don't pay it out based on, you know, the biggest bank gets the biggest or everybody gets an equal share. Whoever borrows is who gets the earnings back. So it's literally a rebate to anybody who borrows. And so what we see is the FHLBs are always competitive with the Fed, often cheaper. And, you know, that drives part of the story too. They just have this built-in discount.
Tracy (17:24):
So one other thing that's happened recently is the Fed has basically said they're going to end the BTFP program I think in March, which was when it was supposed to end. And I was kind of amazed at some of the arbitrage stories that came out a little while ago. The idea [was] that banks could basically get free money from the Fed because of the way the rates were set on the BTFP versus other financing sources. How big of an issue was that? How much did that play into the decision to end it? And then I guess lastly, given what we're seeing now with one particular New York-based bank and the troubles there, is there a possibility that the BTFP gets extended?
Steven (18:06):
So I would say it's unlikely absent a wider crisis. The arbitrage story isn't likely why it ended. It ended because things, you know, NYCB notwithstanding, things have been calmer. There's not really the ‘unusual and exigent’ circumstances the Fed looks for.
I think it's probably why we found out in January that it was going to end in March and why they announced the rate change. So the way the rate thing worked is the BTFP is for one year, the Fed charged one-year OIS plus 10 basis points. So that was sort of the penalty built in. But what they're lending is reserves and if you are a bank, you can leave those reserves at the Fed, not do anything with them, and earn interest on reserves. So this is sort of a new post-2008 thing that actually weighs into the expenses of the Fed. So when OIS plus 10bps drops below IOR, you could just run that trade infinitely as long as you have the collateral and just sort of harvest the carry basically.
Joe (18:59):
Just for listeners and for myself, remind me again what exactly the BTFP stipulated? It was rolled out as part of the SVB emergency. I guess there were all these concerns about all these losses on the hold-to-market book, but remind me of what the actual design of that program was?
Steven (19:18):
So the biggest thing about the BTFP is that it took collateral at par value.
Joe (19:23):
Right, par. And so there were a lot of these Treasuries, in particular in [SVB’s case, they] were way off par because rates had shotup.
Steven (19:29):
Right. And so the critique at the time was ‘Oh my gosh, this is not how central banking works. You can't just lend at whatever value, blah, blah, blah.’ And that was pretty overblown, because the BTFP still charges a market rate. So just like we were talking about before, they're not charging the deposit rate, they're charging 500bps at the time.
So all it did was term out a bank's losses. Because, you know, a mark-to-market loss on a held-to-maturity security from interest rates is representative of your funding cost over time to hold that security. Banks typically don't pay that actual rate, right? If they're paying cheaper deposits. And that's why we ignore the accounting.
But once you take that security to the Fed, if you take a 30-year Treasury to the Fed for 30 years, just keep rolling that discount window loan, you're going to pay the market rate. And so those losses, you're going to realize them over time. So it solved the liquidity problem, but it didn't ignore these losses. Banks still had to deal with them.
Tracy (20:20):
You just alluded to something that I wanted to ask you. This is sort of a provocative question, I know it's going to get you going. But, liquidity versus solvency? That's not even a question, that's just a statement. What's the difference? Does it matter?
Steven (20:35):
So this is a common ‘versus’ framing for basically synonyms in banking. Like, every time a bank fails, it's either one political party — it's definitely the executive who ran the bank — [they say] ‘We just had a liquidity problem, we just needed more liquidity from the Fed. People freaked out.’
The question they can't answer is, why your bank, right? We don't have a panic that takes down JP Morgan, right? The idea that something can be a liquidity issue alone doesn't exist. These banks aren't chosen at random. And every bank at the very end looks like a liquidity issue because the last thing they do is either fail to make a payment or look like they're about to fail to make a payment and the regulators show up.
Joe (21:17):
Well, can you have like a pure self-fulfilling prophecy? Like couldn't someone start a false rumor or misunderstand social media? And I remember in the wake of SVB, it was like ‘Oh, social media caused the bank run. Or all of these people on a ski trip in Aspen,’ I think was one of the stories. They were all like WhatsApping with each other and that's what caused the bank run. Is that a real thing where a bank could go down? You say, ‘Well, yeah, but why are they targeting you?’ But maybe everyone just on the WhatsApp group says ‘This is the bank that's in trouble.’
Steven (21:49):
They could, but we just don't see that as really happening. You know, SVB was running on negative accounting equity for months and investors had discounted it. It's not just the SVB case. There has yet to be a case study where Twitter can come out or, you know, Bill Ackman can just like take down Goldman Sachs. Because it goes back to what we were saying about Warren Buffett or the capital to raise — if the franchise is strong and you have contingent capital, you don't have to worry. If you don't have contingent capital, the capital structure breaks down.
Joe (22:20):
This is the thing, Tracy, that I still to this day don't quite understand. Which is that there was no question that they were running negative equity and it was right there in probably the 10-Q or some SEC filing. But, you know, all of these startups supposedly love the bank. They target it, they understood they had these special products so that founders could get mortgages by posting their RSUs as collateral, which other banks didn't. I don't understand why they couldn't have monetized that franchise value, which now seems to be gone.
Steven (22:52):
So they did for so long. I mean, that's how they could afford running at negative accounting equity. It's like Amazon, right? How long did they take to turn a profit? But you had long-term viability.
And the thing with SVB is like, okay, you can have the ‘mostest,’ ‘loyalist’ depositors in the world, like every bank thinks they have, and SVB probably did. But what happened, you know, what Tracy was observing with her article, and what was happening before the run is they had to spend their money. The deposit balance at SVB was dependent on new IPOs that just weren't happening. So these venture capitalists are as loyal as can be, but they're just spending down their cash balances and so the balance sheet unwinds.
Tracy (23:33):
How much of the banking fragility that we've seen over the past year is basically an interest rate story? So, you know, setting aside IPOs, which dried up when interest rates increased, you also just have the losses on the bond portfolio. And to me, this is kind of a non-issue, but it's also kind of a fundamental tension in the banking system, which is that you've built all the rules around the idea that the best type of collateral is either cash or government bonds. Which is fine when government bonds are really boring and not very volatile and there's not a lot happening. But when inflation starts to go up and the primary tool the central bank has to manage that is to affect the price of bonds, then we seem to have this tension enter the system.
Steven (24:23):
For sure, that's how you end up with the BTFP. Which, you know, sort of fits in this long trend, particularly at the Fed, of like ‘What is a Treasury and how much do we want to monetize it? How often do we want to be intervening? What different lending facilities do we have to set up? Who do we let?’
So it does sort of sit within that post-2008 tension of we're really building the system on top of these safe assets and we kind of have to keep them money-like. But yeah, that is the key vulnerability. But also there are so many banks who have that same vulnerability as SVB that didn't fail.
So that's sort of the built-in macro vulnerability. But the interest rate risk, you know, was also in tech, innovation, and in crypto. And so that's why we've seen banks like Schwab, banks like Bank of America, huge unrealized losses [there], but less concerned about the franchise.
Tracy (25:14):
Yeah, when I say “non-issue,” it seems like there's a tension, but also I find it hard to believe that the banking system is going to come down because banks have bought too many US Treasuries. That doesn't seem realistic.
Steven (25:29):
Right. This was always the nuclear option that the Fed had. The second you're worried about JPMorgan going down because of too many Treasuries, the Fed's going to cut rates and just recapitalize the whole system. So it was also sitting in this tension of the Fed was tightening and didn't want to ease up on tightening. So that made it a harder dance too.
Joe (26:01):
You wrote about this a little bit at the time, and I think even before SVB, what happens if we get into this situation in which the Fed is trying to put out a financial fire at the same time that it's trying to fight inflation, which was certainly the case in March 2023, because at that point the hiking cycle had not yet reached its peak. And so, you know, they re-expanded the balance sheet. Some Twitter people thought that was QE or it wasn't.
But talk to us about that tension, you know, it’s like you mentioned, let's say JP Morgan was getting into trouble, but that could happen at a time of high inflation. And as Tracy mentioned, it could happen at some banks through the Treasury channel. It literally did. How do central bankers think about this tension? Or how do you resolve that?
Steven (26:47):
Well, central bankers might not. I mean, at the absolute limit, this is why crisis interventions are capital injections and guarantees and fiscal. It's sort of all these things at once and, again going back to discount, why it's not always enough and why it doesn't solve every crisis.
But that's exactly why you have to see quote-unquote ‘innovative’ things like valuing collateral at par. And the BTFP and the arbitrage and sort of the terming out losses is a little unique because all the losses really built up just in Treasuries at the time.
You know, if you're thinking about something like commercial real estate — to pick a random asset class — the credit risk is endogenous to the Fed, right? So it's not a case where the loss is necessarily materialized over time. The Fed can come along and write basically a put option on credit risk in a way that even valuing collateral at par, it sort of couldn't with the BTFP.
Tracy (27:40):
So I mentioned at the beginning that we are seeing various attempts to tweak and in some cases change significantly the way these various facilities are used. If you were Michael Hsu at the OCC or if you were Michael Barr, the vice chair of supervision at the Fed, if you were the ultimate Michael, basically, how would you be arranging this constellation of facilities?
Steven (28:06):
Yeah, there’s a few things. I mean one, we can talk about the standing repo facility, but I might pop a blood vessel if we do that...
Joe (28:14):
I want to see that. We'll get to that.
Steven (28:17):
The biggest thing is you have to have some carrot at the window. So it used to be, even in recent history, that there was a premium to go to the discount window, right? You want banks to like be evaluated by the market when they're getting funding and you don't want haircuts at the discount window being your effective collateral requirements.
So there is a reason to not run everything out of the Fed, right? They're not asset managers. But at the same time, you have this tension where you want them to come when the time is right. And so you have to have [somewhere[ for the banks to go because the Fed can't go any lower on price. It used to be a hundred basis point premium. We've seen the Fed lower this in crisis, they lowered it to zero over Fed funds in Covid and it's been there since.
So it's clear that they are keeping it, you know, they’re keeping the discount window rate at top of Fed funds. So they really can't go any lower because then, you know, we're in the BTFP problem where they're taking in less than than you can pay on interest on reserves. So you have to have some regulatory carrot and stick basically for the discount window. So that's a big piece.
The second thing is getting the FHLBs out of the lender of last resort game. We sort of got a unique political moment in that the FHFA put out a report a few months back kind of saying this thing, right? You know, my sense is Sandra Thompson, the head of the FHFA cares about affordable housing, right? She doesn't want to be in this world of, like, bankers just lending to each other and it goes to a trillion dollars in a crisis.
And it's just sort of so far from where those institutions started, it's so far from the goal of housing. Like, get them out of the lender of last resort game. Don't let them pay dividends based on who borrows. Pay dividends based on who does affordable housing or something along those lines. And basically write up a bunch of term sheets for all these different potential 13(3) facilities that you're going to have to roll out.
Because the other piece of this is, going back to your rates question, Joe, all we talk about now is central bank intervention. Anytime a market blows up, it's ‘Where's the ECB? Where's the BOJ? Buy equities now, and bail out this bank.’ Especially since 2008. And where was all this before? Well, they just cut rates when we didn't have to worry about the zero lower bound. They just cut rates and that was sort of the Greenspan playbook, right? Just let some financial froth come out and then clean up the mess with rates. So that's sort of the other reason that we're talking about this more and more and more. We're worried about the zero lower bound.
Joe (30:34):
Talk to us about the standing repo facility. It seems like a good idea, you know, just always be there and get some…
Tracy (30:40)
Whoa, his head just exploded!
Joe (30:41)
Oh shoot! Oh shoot! It was nice knowing you, Steven.
Steven (30:44):
It's okay.
Joe (30:47):
What's the downside? It always seemed like a good idea.
Steven (30:49):
So the standing repo facility is, first of all, basically the discount window for Treasuries and agencies. The nice thing about it is that it adds primary dealers. You hear the Fed talk about it and they like want to add all these banks to it, but it's just the discount window. You can bring Treasuries to the discount window. So that whole piece of it, of like ‘let's get banks involved,’ it would really only be valuable if you as a bank at the depository subsidiary had collateral in the tri-party repo market. Because The Fed runs this program out of the tri-party repo market. It's not like the discount window in that sense.
So that part's sort of goofy. It's nice to have it for the primary dealers, but there are two problems we have with it. One, and Zoltan [Pozsar] has talked about this on this podcast at length, which is you're relying on the primary dealer's balance sheet to sort of on-lend it to everybody, repurpose the liquidity for every hedge fund that needs it in a time of crisis. And that just doesn't work because balance sheets get pinched. And then the alternative is like, okay, you let every hedge fund come directly to the Fed. And there are political and legal issues with that.
The other thing is, again, it's not like it takes a matched book, right? The standing repo facility is not going to take your Treasury and your future. So if you look at something like the basis trade and the risks we have around the basis trade, I take very little comfort in the standing repo facility even though some folks do.
Because when the basis blows out, you basically have the cash price fall and the futures tighten, right? And that's what we saw in March 2020. And all the standing repo facility can do is replace your repo funding at that new market value. So it goes back to this issue of like where you value the collateral? It's not going to recognize ‘Oh, you have a future and you have a Treasury. So I'm going to lend, you know, at the collective value of that portfolio.’ It's going to lend at the cash value of your dash-to-cash Treasury that everybody's trying to get rid of. And so you're just going to get caught in that spiral.
Joe (32:37):
What does a bank do? No, seriously, what is the main product that a bank offers?
Steven (32:42):
Deposits.
Joe (32:44):
Can you explain it? And like when I think [of it], it's like ‘Oh, I want to go to the bank, I want a loan.’
Tracy (32:46):
If a bank offers mostly deposits, why are they all so bad at actually matching the benchmark interest rate?
Steven (32:52):
Because deposits are a service. That's exactly why.
Tracy (32:56):
I disagree. I get nothing from my bank.
Joe (32:57):
This is my argument. We should be paying the bank for deposits. I love easy online banking services and free ATM withdrawals. Why aren't I paying them?
Steven (33:08):
Well, not only that. It's the ledger, Tracy. It's the ledger of the whole economy. You cannot make a payment that isn't a deposit transfer happening somewhere at the back end. And that is a service that banks offer. And that is exactly why the franchise value can erode so quickly. Because if you say ‘Oh, we have all the liquidity we need at the discount window because we're highly capitalized,’ and failed banks always have great capital ratios, right? So they can take their collateral to the window, haircut it, whatever, but you have no deposit franchise left and the deposit franchise is what was allowing you to borrow at zero and lend it at three so that's your whole franchise value.
Tracy (33:43):
I realize we've made it through this entire conversation without even touching the Basel [III] endgame proposals. Should we do it?
Joe (33:49):
Yeah, let's go for it.
Tracy (33:51):
All right, Basel.
Steven (33:53):
I mean it's almost not worth it because it's not going to look anything like it does now. I mean, I don't even know…
Joe (33:59):
What does that mean?
Steven (33:59):
Well, there’s a lot of places in it that look like easy fixes. You know, there are weird charges that show up for climate financing or like random distortions that happen in housing. So the Fed's going to look very responsive. It's going to look like it changed a lot of things.
The other thing is they've sort of signaled that they want more consensus than they had [signaled] putting out the proposal. And they had two dissents putting out the proposal — they had Mickey Bowman, who they're never going to get. She hates everything the Fed has done on the regulatory front in the last year. And it's Chris Waller, who has really talked about the operational piece, which is a little bit distortive.
So the proposal sort of looks at charging for operational risk based on like the size of a business. So the size of an asset management business would cause you to need to hold more capital. But those businesses tend to be very stabilizing. Look at what we've seen happen to Morgan Stanley. It's a diversifying business, it's sort of an all-seasons business. So I think we'll probably see a lot of changes on the operational risk charges as well.
Joe (34:56):
Prior to SVB, I believe there were a lot of fights around the regulatory limits and whether stress tests about banks that weren't the mega too big to fail banks, but weren't necessarily like the little tiny community banks out in the middle of nowhere. And I think SVB and some of these others sort of fell in that middle and, in a way, probably harmed themselves. Because in retrospect, they probably just would've been better off taking a little bit of hit to profitability for sort of tighter regulatory requirements. What is happening with regulation for some of these more mid-sized banks?
Steven (35:31):
Well, I mean the goal is to bring them all to sort of recognize them as big banks which, you know...
Joe (35:38):
Why not do that?
Steven (35:39):
There are maybe legal reasons and blah blah blah. But I'm really not convinced. Fundamentally, if we take a step back from the capital regulation debate, we're talking about changing ratios from like 12% to 13.5%. I get why a bank is annoyed, I get why there’s all these interest groups involved. But from a systemic perspective, that's just not that interesting. That's not going to be the difference between 2008 and not. And it's also not going to be the difference between a profitable banking system that beats Europe and China and not.
Tracy (36:08):
I mean, it is true that SVB had a carve-out as a smaller bank and there is a discussion about whether or not those carve-outs should exist. But just backing up for a second, big picture, I feel like in the US, we have yet to decide what we want the banking system to actually look like. So there's this sort of It’s aWonderful Life vision where you have all these local banks, community banks even — in New York — and they know you and they build up that relationship and you get those benefits.
But on the other hand, our experience of last year is that maybe there is a benefit to being extremely large and efficient and having a funding advantage and things like that. And it feels to me like the regulators, politicians, basically everyone involved in this equation, has yet to figure out exactly what they want.
Steven (37:00):
Yeah, and it's a hard thing to talk about because you can't go out as Jay Powell and be like ‘I think we should have less banks.’ Because you'll have fewer banks by Friday, right? It's a hard thing to talk about.
And they've pushed back on this idea of a barbell banking system, which is sort of the midsized ones get hollowed out, they either downsize or upsize and you're left with community banks and bigger banks. And that is sort of the verdict of 2023, is you would say, ‘Okay, big banks did well, small banks did well — let's just get rid of the midsized banks.’
But you know, there are small banks that are very dependent on the local economy. I will say, big picture, you cannot be a niche bank that is also under the pressure of financial markets. You can't be focused on Silicon Valley and also need to raise equity and have attentive headlines. If you're a community bank, you can probably run on negative equity longer than a mid-sized bank that has to go to market and things like that.
Tracy (37:57):
Alright, Steven Kelly. Thank you so much for coming on Odd Lots and letting us trigger you for basically 40 minutes. Really appreciate it, that was great.
Joe (38:05)
That was a lot of fun. Thanks, Steven.
Steven (38:06):
Thanks, guys.
Tracy (38:20):
So Joe, I really enjoyed that conversation. I have a feeling it's going to be a very relevant one in 2024 as we start to see more movement on these various issues, including maybe reforming the discount window, whatever the Basel endgame actually ends up looking like.
There are a few interesting things that I would pick out there. So one of them was Steven's emphasis of how important the actual banking franchise, the deposit franchise, is to funding. And, you know, if the franchise starts to go, that's when you do get the deposit issues and then you can't actually raise capital. And I think some of that did get lost in the conversation around SVB and Signature and First Republic where it was more like ‘Oh, these banks kind of got unlucky. They bought too many bonds or whatever.’
Joe (39:09):
No, that really connects some dots and crystallizes a lot of things. And I had forgotten with SVB that prior to the run that did happen on the bank, there was the deposit shrinkage that was simply a result of Silicon Valley financial conditions at the time, which is that there was no IPO window for a while there and there was no new fundraising. So these startups and stuff did not have fresh cash coming in and they were in survival mode and, you know, they're like spending down their money all the time.
So there was this sort of like natural, it was not a run, it wasn't even about the Treasuries, it was not about the report on the Substack in January of that year that’s like [by] Byrne Hobart, the author of The Diff newsletter. He’s like, ‘By the way, Silicon Valley Bank is insolvent, you guys should check this out.’ And people ignored it for about four weeks. It was not about that, it was just about the fact that the deposits were going down.
Tracy (40:03):
However, Joe, I remain reluctant to pay my bank a fee. I don't want to.
Joe (40:09):
No, I mean, I like having free banking and I like having free access to ATMs and a website and a nice app and stuff like that. But it does really make sense and sort of crystallized this point, which is that that is the only sub-market rate borrowing in the world, right? Like basically, for the banks.
And the reason they can get submarket rates because they also throw in this service for you. But there was that chart we had at our recent Odd Lots trivia night that Josh Younger showed, which is the Fed funds rate and then ‘What is this rate below it?’ And there really is only one rate in the world that's going to ever be below the Fed funds rate. And that's the special rate that banks can borrow from their own customers.
Tracy (40:52):
Deposits, yeah. You did mention, I think earlier in the intro that around this time last year, so in addition to Bill Nelson on the discount window episode that we did in January, I think in February, probably, we spoke to Joe Abate over at Barclays about exactly this issue. So deposit rates, the beta to benchmark interest rates. So, yeah, I think we were pretty on the ball.
Joe (41:16):
We were pretty on the ball and talking to Steven put a bunch of things together. A lot of light bulbs went off. It's like, I get why this is the case or I get why that's not really an ultimate fix, etc. So I really enjoyed that conversation.
Tracy (41:31):
Okay. On that self-congratulatory note…
Joe (41:34):
And the FLUBs, that's a good one! I'm going to start calling it that. That's so much easier to say than FHLBs.
Tracy (41:39):
Shall we leave it there?
Joe (41:39):
Let's leave it there.
You can follow Steven Kelly at
@StevenKelly49
.