Dr. Darrell Duffie on Liquidity Strains at Year-End/Quarter-End and When Fed Reserves Will No Longer Be Ample
01:09:56 Share

The ECB is ready to do whatever it takes to preserve the euro. And believe me, it will be enough. Thank you. Let's close this door. I'm very pleased to welcome to Monetary Matters Professor Daryl Duffy, Professor of Finance and Economics at Stanford and the Stanford Graduate School of Business. Great to see you. Jack, thanks so much for having me on your podcast.

Daryl, we are recording at a pretty opportune time because you've done a lot of work on

spreads between two different interest rates, the SOFR rate and the IOR interest rates. And often the spikes and apparent stress in that interest rate spread makes itself known and makes itself visible at the end of a quarter. We're recording on Friday, December 27th, and it is getting very close to the end of the quarter. Normally, you wrote in a

paper that you've updated this year, but you originally published it in 2019, how the quarter end effect leads to a move in the spread of SOFR IOR between seven to nine basis points. Yet as we sit here today, according to the New York Federal Reserve's website, the spread between SOFR and IOR in regular terms, so I guess the average, is 20 basis points. And if you look at the 99th percentile, it is a 45 basis point increase.

spread. So how do you interpret that? Why do you think the spread is so wide and what does it indicate to you? Terrific, Jack. You packed a lot into one question. Just to make sure all of your listeners are on the same page, let me just unpack two crucial definitions that you used.

One is IOR. That's the interest rate that the Fed pays to banks for money that the banks leave at the Fed. And SOFR or S-O-F-R is the secured overnight financing rate. And that's the rate for financing of Treasury securities in the overnight repo market.

Now, those repos are about as safe as a deposit in the central bank, approximately. So the two interest rates should be virtually identical if markets are working perfectly. Of course, no markets are ever perfect. And the subject of our discussion today is how imperfect are the markets to the extent that those two rates, which are supposed to be the same

are actually different. And your question is related to what's going on at the end of each quarter and particularly at the end of the year that can drive those two rates especially far apart. That's the nub of the question. It happened at the end of the third quarter, big time, and it's happening again now, or at least the spread between the two has been moving SOFR up relative to IOR.

And that's because at the end of each quarter, and especially at the end of the year, foreign banks, that is non-US banks, most of them are measured for having enough capital by their regulators. And they do what's called window dressing. As the end of each quarter approaches, they want to look good for their capital requirements. So they reduce their balance sheets. They provide less funding in the repo market.

That means that all of those financings that were done in the repo market have to be done mostly by U.S. banks. And the U.S. banks are going to say, well, gosh, you know, we have balance sheets, too. And we're not as monitored as the foreign banks for quarter end, but we are monitored for other things. So we're going to have to charge a higher rate. It's their counterparties come to them asking for financing. They're going to say, well, it's quarter end. You know what happens? Quarter ends. We have to charge you more.

and it's been moving up at the end of each quarter on average, as you said, seven to nine basis points. But there are other things going on at the same time. One is the treasury market is getting bigger. That means more and more financing. Not only are there a lot more treasuries to finance, but the fraction of treasuries that get financed is also going up. And so there's just more and more pressure on the balance sheet space of the largest globally systemically important banks that are active in the repo market. Those are the household names

that you and I know and love like JP Morgan,

Citibank, Bank of America, Goldman Sachs, and Morgan Stanley. Those are the big five. And they cannot do an unlimited amount of financing for their counterparties because the Fed says, how come your balance sheets are so big relative to your capital? They know that's coming, so they don't open up their balance sheets to an unlimited extent. And so foreign banks, they are measured on the capital by the end of quarter, whereas the large U.S. banks...

They're measured all the time, so they don't do window dressing. And we are getting to that point where the funding needs are going on the US banks. So, Darrell, in terms of talking about why banks pull back from markets, I think almost the entirety of your answer put it on capital requirements. And basically, why banks would pull the market is because they are over leveraged relative to their capital.

instead of relative to their level of reserves, which is affected by quantitative easing or as we're doing now, quantitative tightening. So how much does the fact that bank reserves have been going down and may be approaching some potential lowest comfortable level of reserves, how much does that impact

what we're talking about. Well, that's the $64 trillion question. But before we leave the world of capital requirements and go into the world of ampleness of reserves, there is an exception even for the US banks. They are monitored at the end of each year for what's called GSIB scores, meaning separate from the leverage ratio capital requirement. There's a separate capital requirement that's very complicated, but includes a component which captures these repos.

So the size of their balance sheet at the end of the year really matters to the USG SIPs. They don't want to look big at the end of the year. So they also constrain their balance sheets like the foreign banks. So it's kind of a double witching hour with respect to capital requirements.

at the end of the year. And you can see from weeks ago, there were already elevated term repo rates, meaning financings that would run for multiple days across the end of the year. Those rates were already getting elevated as everybody was getting situated to get their financing needs met across the end of the year when everyone knew that balance sheet space would be tight. And those rates have been elevating for the last several weeks. I had one

contact at a hedge fund that told me his prime broker would not give them financing across the year end unless they signed up for what's called sponsored repo, which is a way to reduce the amount of balance sheet space used by the largest banks. We can get into that later if you like. So capital requirements are complex, especially for the U.S. banks at the end of the year. And we shouldn't just say that it doesn't really matter as much for the U.S. banks.

Now, coming back to the other part of your question related to the sufficiency of reserves, meaning how much central bank deposits does the Fed provide to banks in the US system? An amount that goes way up with quantitative easing when the Fed expands its balance sheet, and it goes down as it is right now with quantitative tightening as the Fed reduces its balance sheet, letting its government securities run off as they mature.

and not replacing them or not fully replacing them with new treasuries and agency government securities. So as that process continues and those government securities mature, the Fed's quantity of reserves provided to the banking system is also going down. And as that happens, it's just a question of assets and liabilities going down together for the Fed. The Fed's liabilities include these reserve balances that they provide to banks, which are deposits at the Fed.

Over time, there might not be enough reserve balances in the system to do what the system needs to do, which is not only financing that's done with reserves, but also meeting everyday needs for payments. Because these largest banks start each day with a certain amount of reserve balances, and through the course of the day, they have to pay out to others multiples of their opening of day balances. And how do they do that? They get funds from other banks during the course of the day.

And so they have to make sure that they start the day with enough balances so that they can pay out and meet their intraday liquidity requirements, meaning that the largest banks pretty much have to stay in positive balance through the day. If they don't, then their supervisors start to

and give them funny looks and send them texts saying, "What's happening? Why does it look like you don't have the ability to fund your own liquidity? It looks like you're relying on the Fed for liquidity. You're not supposed to do that." So the largest banks need a lot of reserves to make their intraday payments, especially when they also are being asked by counterparties to provide financing of securities using reserves for that. So there's two big demands.

For reserve balances, one is financing of securities in the repo market. The other is making outgoing payments. If they don't have enough incoming payments and opening of day balances to meet those two requirements, they start to get very stingy. And the SOFR rate, which is an average of repo rate, starts to go up.

When you say start to get stingy, I think there can be a compounding effect. You referenced this as self-fulfilling effect where I'm a bank, you're a bank. If I think you're going to be late in your payment, I'm going to slow down my payment to a third bank. And then that bank will be slower to pay and then it can exacerbate and it can feed on itself. And when we're talking about reserves, we're talking about commercial banks, banks

who have deposits with the Federal Reserve, kind of in the same way that a person or a company would have deposits with that bank. But everything you're talking about to a company or to a person would be totally invisible. This is just within the banking system. And then, okay, so Daryl,

Earlier, I think in your paper and just now, you referenced an answer to a question I've had for a really long time. I've known that before 2008, the amount of reserves in the system was tiny, about $40 billion, $50 billion, and that now it's $3.2 trillion. It peaked at over $4 trillion in the fall of 2021. I never really understood how that happened. How is it that

what we had a hundred fewer times you know reserves have gone up by a hundred times before the system worked fine and now we're having issues i never really under understood that but you referenced it in your paper which you know we can link to in the description of this that before the gfc great financial crisis banks liberally exploited daylight overdrafts of their federal reserve accounts to make their intraday payments

Earlier, you said that the banks are not supposed to rely on the Federal Reserve because they get in trouble if they do that. Were you referring to overdrafts? And tell us how overdrafts were the norm before the great financial crisis and now in these abundant or ample reserve regime, now banks are not allowed to do? Okay, so the answer is a little bit more subtle. Let me go back to the pre-financial crisis days when reserve balances, as you said, Jack, were really small, tiny compared to today.

but banks were not so concerned about going to the Fed if they needed extra cash during the day. They could overdraft their deposit account at the Fed, just like you and I could overdraft our banking account for a fee. In those days, there was no systemic risk regulation saying that every bank had to be self-sufficient for its liquidity. But after the financial crisis, Congress got very upset

about the government and the Fed supporting banks who didn't seem to have enough capital or liquidity to manage on their own. The spirit of the post-financial crisis regulations was that going forward, the banks should be able to meet their own needs without relying on the Fed. So a whole battery of regulatory requirements were introduced that required banks to maintain sufficient balances to meet their own liquidity needs without going to the Fed.

The consequence of that is the largest banks don't want to get a nasty note from their supervisors saying, are you really meeting these liquidity requirements? For example, would you really be able to resolve yourself if you were about to fail without coming to the Fed for extra liquidity? And banks are frightened about getting that kind of a note from their supervisors. So they don't want to go near liquidity.

a situation in which that might happen. So they have much higher demands for reserve balances today than they did in the past. And they're, in some sense, not penalized for having a lot of balances because another feature that distinguishes the pre-crisis world from the post-crisis world is that the Fed now pays interest on those reserve balances, whereas pre-crisis, the Fed didn't pay interest.

So the world is much different now. The Fed is now steering interest rates by the amount of interest it pays to the banks on their reserve balances. And banks are saying, well, OK, well, I'm getting fair market compensation for my balances at the Fed. I have no real reason to try to minimize those balances anymore. I'm getting a proper amount of interest on them. So I'm not worried if I have lots of balances. I want to also correct an impression that if you got into trouble and you didn't have enough liquidity,

that the Fed would stop you or want to stop you from coming to the Fed for additional liquidity. On the one hand, it says you must remain liquid on your own without coming to the Fed. And on the other hand, the Fed has introduced facilities, including the old discount window, but new facilities that would encourage banks to come to the Fed in an emergency if they really needed liquidity.

An example of that is the standing repo facility, which stands ready to provide liquidity to banks when they get into trouble and they really need liquidity, or there's just a shortage of reserves in the system. The unfortunate thing is with getting these two messages from the Fed, on the one hand, you need to be self-sufficient and we're monitoring that. And on the other hand, we want you to come to the Fed and get liquidity when it's really tight. The banks are going to err on the side of not going to the Fed.

because they don't want to demonstrate that they didn't meet their regulatory liquidity requirements. And so even when they're short of reserves, they're not going to go to the Fed, they're going to go somewhere else to get funding. And so are others. When the quarter ends, we're now seeing a lot of transactions at interest rates higher than what the Fed would be willing to offer to the banks. At the last quarter end in September,

about 600 billion of overnight financings were done in the repo market at interest rates well above the interest rate that the Fed was offering on the standing repo facility. How much above, Darrell? Because now we're 20 basis points above. Were they this high? Well, they don't say what the average was, but they show you the 75th percentile. And the 75th percentile was about 15 basis points higher than

than the interest rate that the Fed was offering for financing. So that means that there was 600 billion above that interest rate. A lot of that must have been at much higher rates. They also published the 99th percentile, and that was very high. I don't remember the exact number, maybe 40 or 50 basis points higher. But that's, of course, way out in the tail, and only very desperate dealers that really needed liquidity very badly. By the way, I don't want to lend the impression that

the sky is falling when this happens and banks are about to fail. It's not an urgent danger to the financial system, but it does interfere with the Fed's monetary policy and it causes consternation, especially with smaller dealers that depend on access to financing from these giant banks.

So banks who need financing at the end of the year, they could go to the Federal Reserve right now, but that would either be a bad look publicly or with the Federal Reserve. So they are tapping the repo market. So I guess reverse repo borrowers are selling treasuries and then buying them back the next day for a slightly higher price. And the lenders are doing the exact opposite on the other side of the transaction. So I'm looking at the 75th percentile right now. SOFR is 4.70%.

effective Fed funds rate is 4.33. So that would be a difference of 37 basis points. Although you're talking about interest on reserves, which is different than the effective Fed fund rate. I forget what the current effective Fed funds rate target is. 4.250 to 4.50. Okay. So we're above the middle of it where the Fed strives to target. So that's clearly an issue for those that are way up in the tail of

of those transactions are paying a lot more because the system is not working as well as it should on quarter ends. This is not exactly the quarter end. There's still another day, actually another two days that the market's open. And so we might see even higher rates across the next couple of days. Again, the sky is not falling. It's not evidence of a disaster looming, but it is evidence that even with enormous quantities of reserve balances,

financing can get difficult at the end of the year. So the sky is not falling, but this isn't a nothing burger. And I guess we have a 30 basis point differential between SOFR and IO interest on reserves at the 75th percentile. So it is increasing. And as you said, we've got two more days left before the end of the year. How much longer do you think the

before this stress begins to emerge in a more serious way. And the sky will start to fall, or the monetary equivalent of September 2019, where that's the monetary version of the sky falling when you have, I guess, SOFR blowing out by 300 basis points. And that severely inhibits the Federal Reserve's ability to have monetary policy, because if it's

target is X and the actual borrowing rate in the market is 300 basis points above X, that's a problem. Yeah. So I mean, quantitative tightening will only mitigate this problem. It's not going to cure it because we're seeing a quarter ends and year end as a capital requirement issue and reducing the pace of quantitative tightening will make reserves more available. But the availability of reserves is not

really what's causing the problems at quarter ends. It's the capital requirements. That said, there is going to be a problem with the ampleness of reserves sometime next year, most likely. For example, I went to a meeting early this month organized by the Bank Policy Institute

a symposium at which market participants talked about these issues, including is quantitative tightening going to have to come to an end? There was speculation in the room. It was Chatham House rules, so I can't tell you who said what, but there was speculation in the room that sometime around the end of the first quarter or between the end of the first quarter and the second quarter, it was a prediction that the Fed would reduce its pace of quantitative tightening. The last time that they were reducing the balance sheet

was in 2019 and they overshot and there wasn't enough reserves available to the market in the middle of September 2019. The 30 basis point spread that you were just talking about was dwarfed by a multi-hundred basis point spread that occurred on September. The Fed quickly reversed course, added a lot of reserves to the system. It happened again in March of 2020 when COVID hit. There wasn't enough reserve balances, rates shot up again.

It's been very quiet since then because the quantity of reserve balances has been huge. Right now, it's more than twice what it was in that nasty event of September 2019. So it's very interesting that we're already talking about stopping this quantitative tightening when reserve balances are twice as big as the last time

This became a problem. We'll see how all of this works. The standing repo facility, we'll see if it can actually do the job. It wasn't around in September 2019. So I'm not predicting major problems. I think the Fed will manage this.

And so I forget the medical term, but if you take one aspirin and you take it every day, in three months, you might have to take two aspirin. And there's some analogy to that perhaps with the level of reserves where what used to be an incredibly ample amount of reserves, the financial system gets used to it and needs more.

more and more reserves. So if the breaking points, what was the breaking point in September of 2019 in terms of a level of reserves and why might a much, much higher level that we're at now or slightly below that, you know, as we approach that with quantitative tightening, why might that be the level that can't be this time around? Okay. So it was around 1.5 trillion of reserves when it wasn't enough the last time. We're now at about 3.2 or so trillion, more than twice as much.

The effect that you just described of, you know, take one aspirin and then when it starts to not be enough, take two and so on. In the world of reserves, that's called the ratchet effect, which was first hypothesized, to my knowledge, by Raghu Rajan and Biral Acharya in a Jackson Hole paper. They hypothesized that once the Fed expands its balance sheet and provides an enormous quantity of reserves to the banks for quantitative easing reasons, the banks get somewhat addicted to

to having a lot of reserve balances. How do they do that? Well, it's hypothesized that when they have tons of reserves, they don't mind at all offering their customers immediate liquidity for things like commercial deposit accounts, transaction deposits, or credit lines that customers could pull on a moment's notice, as happened with SVB, for example. And the banks are not willing to offer these highly mobile

forms of liabilities unless they have lots of reserves. Once the Fed increases the quantity of reserves, the bank says, okay, we will deposit accounts more liberally. We'll provide credit lines more liberally because now we have enough reserves to manage the liquidity. Then once they get that business going and the Fed starts to tighten again, the banks are reluctant to

shut down the availability of those services like large transactions, deposits and credit lines. They don't want to reduce those anymore. So they have this kind of ratchet effect where they become semi-addicted to reserve balances. That's the hypothesis. And the courts deal out, I mean, there's a lot of controversy in central banking circles. I've had a number of central bankers ask me, is this really true? Can you prove this ratchet effect? And

There's lots of circumstantial evidence, but there's no smoking gun yet. The breaking point in September 2019, as you said, was $1.5 trillion. Now we're at $3.2 trillion. What is the lowest comfortable level of reserves and how do people estimate? I mean, I've heard that the way of estimating is just indexing it to GDP growth, which doesn't seem super scientific to me.

Yeah, well, I've seen a number of attempts to do this. One is to index it to GDP. Another is to index it to the total amount of deposits in the system or the total assets of banks. Another is to say, well, if 1.5 trillion was enough last time, then the system hasn't gotten that much bigger. It should be enough today.

I've given up on trying to estimate how many dollars of reserve balances is enough because there are so many structural changes in the financial system over time, particularly regulation, but also this ratchet effect. Regulation changes like money market fund reform and some of these liquidity requirements, the increased availability of the discount window.

the bank term funding program that was introduced by the Fed after SVB, all of these change quite significantly the amount of reserves that banks seem to feel that they need. And I've turned away from trying to estimate that number and toward signals, a canary in the coal mine that would tell us you're getting near. We don't know what the quantity of reserve balances is, but we're seeing signals that it's getting insufficient. One of those which

the Fed is using quite heavily is the interest rate spread that you mentioned at the beginning of our call. Either the spread between SOFR and interest on reserve balances or the spread between the effective federal funds rate and the interest rate on reserve balances. Those signals are helpful. I don't think they're enough

The reason being that the last time that those spreads got into bad territory, it happened instantly. It happened very quickly, just overnight, suddenly. So if you were looking at those spreads and saying, oh, it's looking okay, you might be surprised the next day to find that it's no longer even close to okay. Instead of looking only at those spreads, my collaborators, Adam Copeland, Yilan Yang, and I,

developed another warning signal, which seems to be exceptionally helpful in understanding when balances are getting low and for very natural logical reasons. I'll take a moment to explain that other signal if it's okay, Jack. Please. Okay. A few minutes ago, we talked about the stress that banks feel when others are paying them later in the day.

and they don't seem to have enough extra funding to lend because they're worried they're not going to get paid soon enough during the day to keep their balances above zero. So if you simply go to the active banks in the repo market and look at the time of day by which they've received the first half of their incoming payments, on average, that time of day is early afternoon. And when it starts moving to mid-afternoon, more than

100 minutes later than normal, you need to worry. That's a sign that these banks are getting paid sufficiently late in the day that they're going to start holding back funding to their counterparties. We found in our research that Signal is a canary in the coal mine. It does tell you when balances are getting too low. So if the Fed were to have a look at that every day,

and monitor how late in the day the largest repoactive banks are getting paid, they would have a pretty good idea of when balances are getting too low. As you said earlier, Jack, it's a feedback effect. If you start paying me late in the day, then I'm going to start paying the next guy late in the day.

The next bank will pay the other banks later in the day. It's a negative feedback loop that you don't want to see. Economists call it strategic complementarity. Once it starts going, it's a good sign that reserve balances are not sufficient.

for IOR spread is actually not a good predictor of if it's going to spike. So the fact that we have a spike now at the quarter end, it doesn't mean that it's going to spike to ridiculous levels. And we can show that a chart from your piece, just showing how at the beginning of September 2019, it was barely budging at all. And then boom,

boom, straight into the middle of September, it crept up. And then we'll show the next chart, which is again from the same piece that entire year, the payment time had been inching steadily higher. And in 2018, banks were actually being paid

not with no delay at all, or actually a negative delay. So I guess they were paying early. But by 2019, and in the middle of 2019, banks were delaying over an hour. And by September, the delay was up to almost 100 minutes. So I guess...

more than 90 minutes, which for a bank, JP Morgan paying Bank of America, I guess that is quite a big deal. So you're saying this is statistically a far better predictor right now. I want to get your general thoughts as well as a very established economist on these issues. You have access to this data when you're writing a paper on it, and I guess the Federal Reserve or the banks will give this data to you later.

After the fact, do you have access to this data right now? Does the Federal Reserve has access to this data right now? Or is this a thing where it, you know, we kind of have to be an archaeologist, but and it's very effective in interpreting it after it's happened. But, you know, the data can't be collected on time, like who has access to the data right now that we're referencing right now?

Okay, so I had the privilege of working with an economist at the New York Fed, Adam Copeland, and he is the member of our team that can actually touch these data and observe them bank by bank and construct this measure that I described in our paper, which is how late on average are the 10 most active dealer banks in the repo market getting paid. So I cannot myself look at those data. They're extremely confidential.

As you might imagine, they include every single payment in the Fedwire system, which is the backbone payment system by every bank in the system. So that would be, you know, the sort of data that nobody outside of the Fed should be able to look at. But once we get these aggregate measure, of course, we can publish this aggregate statistic in our paper. And in principle, economists at the Fed or regulators at the Fed or the FOMC,

could have a look at this number and say, "Gosh, it is getting a little bit high, but it's still not so high that we need to worry about it." As I mentioned earlier, reserve balances are not currently the problem with stress in this market. It's more capital requirements currently. And I expect that we won't see this canary in the coal mine chirping in the next few weeks or months. It'll be a while because there's still a lot of reserve balances.

But no reason not to have a look on a daily basis if you have access to the data. Okay. So you say right now, any indication of stress is mostly due to capital requirements, not reserves. And you're saying that for any time throughout the year, not just on a quarter end when things do get a little bit wacky. No, I mean, we can see from probably from this timing measure, but I don't have current access to it. And also from the interest rate spreads that you spoke of,

that there's a gradual signaling of reserve balances going down, but no signal yet that there's not enough reserves in the system. It's just not there yet. The spreads are getting smaller. SOFR to IOR is on a normal day, not a quarter end. It's in reasonable balance right now.

That on its own, as I said, is not a great signal because it can jump suddenly. But on average, over time, it does signal that there's probably enough reserves in the system. And you can look at these payment timing metrics as well.

and make a dashboard, look at all the numbers and come to a judgment about when it's time to stop quantitative tightening. And I'm pretty confident it's going to happen sometime next year. Why are you confident it's going to happen next year? Any stress we're seeing is due to capital requirements and not reserves. If you don't feel comfortable estimating lowest comfortable level reserves, and you're not seeing any stress now that's due to reserve scarcity, why are you saying that the Federal Reserve will stop quantitative tightening next year?

I'm taking the Fed at its own word. The FOMC, after September 2019, in its communications, particularly early this year, said explicitly, we are going to be more cautious this time than we were in 2019. The interpretation of that is they're not going to cut it as tight as they did at the time. At the time, they were trying to estimate what is the minimum level of ample reserves and try to come down pretty close to that level

Now they're not trying to do that. They want to leave a comfortable buffer to make sure that there's enough reserve balances. Now they can't be 100% sure and they should be looking at these metrics all the time, but I think they're going to pull up before they get quite as low in this case. There is one additional amber light, if you like, that's been blinking lately.

And that's another buffer that the Fed has, which is called the overnight repo facility. This is where money funds, mostly money funds, park their extra cash. In the past, as quantitative tightening has proceeded, that and also the availability of Treasury bills has been itself ample. That balance in the overnight repo facility has been coming down. And it's another liability of the Fed

As it comes down, that means reserves don't have to come down in lockstep with assets. I'm making this a little bit more complicated, but think of the assets of the Fed on the left-hand side of their balance sheet. And among all of their liabilities, there have been two very large components. One is the reserve balances that we've been discussing, and the other is this overnight repo facility. And that facility had been in excess of $2.5 trillion

more than a year ago, and it's been coming down. So that's a kind of a buffer as the Fed's balance sheet goes down. Either of the two liabilities, reserve balances or overnight repo facility, can accommodate that reduction of assets held by the Fed. Once the overnight repo facility is totally reduced to zero, all of the remaining quantitative tightening has to come out of reserve balances. And we're pretty much at that point now. There's very little left

in the overnight repo facility. So all the additional quantitative tightening will be coming, or almost all of it, will be coming out of reserve balances. And yeah, that reverse repo facility now is, I guess, a little bit over $100 billion is left in it. Yeah. I mean, if we want to fill out the complete picture for your listeners, there's one other significant liability, which is the government's own deposit account at the Fed, which is called the Treasury General Account. And you also want to keep an eye on that because the Fed can't control how fast the government spends money

particularly if there's a debt cap that starts to bind, then the Treasury General Account will have to wind down. As that winds down, that also provides more reserve balances. It's implicitly another form of buffer. And you don't know what's going to be. We had a close call last week with the debt cap ceiling being extended for a few months, but that means in a few more months, we may be faced with more volatility in the Treasury General Account.

And that will also impinge potentially on the availability of reserve balances either up or down. And other factors in your, I guess, regression of what impacts SOFR or spread is T-bill issuance, treasury bill issuance, treasury coupon issuance, and the redemption of that. So I guess has the large amount of, and I guess the T-bill issuance is

I guess, is correlated with a widening SOFR IRA spread, but it is less correlated than coupon issuance. So if the treasury funds itself with a one-month treasury bill instead of a 10-year treasury note or 10-year treasury bond, then that is less correlated to a widening of SOFR IRA spread than if it was a 10-year note.

Yeah. Let me walk through the mechanics on that. Treasury bills are another very safe, very short-term asset like repos and like reserve balances. You can think of a bank, for example, which has a portfolio decision. How much do I put into bills? How much do I put into reserve balances at the Fed? And how much do I put into overnight repo? As T-bills become very plentiful, the interest rate offered on T-bills goes up and I'm going to move money

out of my reserve balances account if I can and put it into treasury bills because I get a higher interest rate. And so that's going to push the, by, you know, it's a choice I can make in order to get market clearing. That means the interest rate on reserve balances, if the interest rate on reserve balances is constant, the only thing that can adjust is the repo rate. And that will go up as a substitute for treasury bills. So repos and treasury bills tend to move together.

The supply of treasury bills matters a lot. The other element of that is as you fund more and more with treasury coupon issues, you're also putting direct pressure on the repo market because every time the treasury auctions a lot of coupons, the treasury notes particularly, but also the bonds, they circulate in the repo market. They demand financing in the repo market and there's more pressure on the largest dealers to provide financing on treasury securities. And that puts upward pressure on the repo rate.

And another element of that is that on the day of a treasury issuance, the day after the auction, when the treasuries are actually issued to the largest dealers, they've got to pay the Treasury Department first thing in the morning for those treasury securities that reduces their reserve balances. Those reserve balances completely disappear from the system because they go into the Treasury General Account. So they're no longer available to the banking system.

And further, even more exacerbating this, the Treasury Department needs to be paid first thing in the morning. So if you're one of these largest banks and you're thinking about when shall I pay my counterparties today? Shall I delay a payment until later in the day to relieve the pressure on my reserve balances?

Well, you don't have a choice if it's the Treasury Department and you're paying for the Treasury securities that you just won in the auction the day before. You have to pay for those first thing in the morning. So with 60 to $75 billion large issuances that we're seeing lately, that's a big effect in the Treasury repo market and drives repo rates up. Notably, we discovered in our research.

And how, Daryl, does this apply to a world in which over the past few years, the Treasury has funded itself with a higher percentage of Treasury bills, shorter term issues than normal. And we are moving towards administration where the incoming Treasury secretary or likely Scott Besant has been quite critical of the Treasury.

Biden Yellen Treasury funding itself with Treasury bills and has hinted that they may issue a lot more coupons going forward. Yeah. So, I mean, that causes a lot of question marks. The new Treasury secretary also been said to have influenced the choice of the new head of the Council of Economic Advisors, who wrote a research paper when he was at Hudson Yard

with Nouriel Roubini, suggesting that the government had been using its Treasury bill issuance policy as a way to do stealth quantitative easing to stimulate the economy. Now, I actually think that's very unlikely that the administration would do that. First of all, the communication to the Treasury's debt management office would be extremely tricky.

But secondly, the Debt Management Office just doesn't do that. It doesn't even think about monetary policy. It's trying to finance the US government's debt at the lowest possible interest rate. It's not conducting monetary policy. But nevertheless, if the new Treasury Department Secretary decides that fewer Treasury bills will be issued, that will put more financing into the coupons, that will raise the coupon sizes.

But it will also reduce the supply of treasury bills. And as I said, that would relieve pressure on repo markets because if there's very few treasury bills available, their interest rates will go down. And that means substitution out of the repo market will relieve pressure in the repo market on rates and cause repo rates to go down. But isn't coupon issuance more correlated with

than bill issuance with IOR spread, right? Yes, on the days of issuance, absolutely, because it's the coupons that put pressure on the issuance date, the date that the Treasury Department actually provides them into the market. That puts a lot more upper pressure on repo rates for coupon issues than for Treasury bills. But on a day-to-day basis, on a normal day, if there's a lot of Treasury bills in the market, that means Treasury bill rates will be higher

And that tends to raise repo rates relative to IOR. Okay. And raising repo rates, sorry, raising treasury bill rates and raising SOFR rates

has not been a problem because money has come out of the reverse repo facility because the reverse repo rate is a fixed rate. But actually, it has gone down in the floor and it went down by 30 basis points instead of 25 basis points. So it now is aligned with the lower end of the Federal Reserve's target range instead of five basis points above that range. Why do you think the Federal Reserve did that?

Well, the Fed said that it was fine with the overnight repo rate, the overnight reverse repurchase facility getting drained. And it also wanted to normalize in its own mind the bottom end of the effective Fed funds rate target with the overnight repo facility. They don't want to have the overnight repo facility supporting rates from below anymore. They think they can steer the effective Fed funds rate better when there's no upward pressure coming from the overnight repo facility.

That will be a relatively innocuous effect because, as you mentioned, there were only 100 billion or so in the overnight RRP in any case. Various speakers at the Fed, including governors and presidents, had said that it would be fine to let the overnight repo facility go down. And I'm not arguing that there should be a lot of balances in the overnight RRP. I actually think it's a reasonably good idea not to rely on it.

That said, we now know when it's at zero, there's no more buffer available for reserve balances. So the Fed will have to provide reserve balances

without relying on the banking system to pull those balances out of the overnight RRP. The Fed will have to do it instead when balances are needed. What did you think of the paper by Nareel Rabini and Stephen Moran? Yeah. What were your thoughts on that? I thought it was they were reaching for an hypothesis for which there was almost no evidence. I thought it was, I mean, it's a very creative hypothesis.

Interesting idea. Technically, the administration could do a substitute for monetary policy stimulus by increasing the supply of bills or reducing the supply of bills if they're going the other way. But as a matter of practice, the government doesn't do that. It knows that its job is to manage the debt at the lowest cost to the taxpayer. It had been increasing the supply of bills, but for technical reasons.

Actually, it increased quite a lot, roughly tripled over the last few years. But the technical reasons had to do mainly with getting better cash buffers, especially for the possibility of the government hitting the ceiling on government debt and the Treasury Department needing a way to manage its cash in that event. So it needed an extra cash buffer.

That makes sense. Two things that I'll probably agree with you on is calling something stealth quantitative easing. I mean, I recently interviewed someone who I've got a ton of respect for. He has used a effective say of stealth quantitative easing. To my mind, quantitative easing has a very specific meaning.

per definition, and things can be stimulative or have a QE-like effect, but that doesn't mean that they're QE. So I don't agree with calling something quantitative easing. And the other point is I defer to you that it is not part of the Treasury's intent to be stimulative or do shadow easing at all. They are focused on cost to the taxpayer. And intent in a legal sense

matters a great deal. If person A runs over person B with their car, was person A trying to do it? Was it an accident? Intent matters a lot. I grant that. But in terms of the actual effect, Daryl, is treasury bill issuance more stimulative than treasury coupon issuance because it is a less risky instrument to banks and therefore requires less balance sheet capacity and I guess risk capacity than a 30-year bond, which could move 3% in price a day, whereas a one-month treasury bill, it's 100 bucks. It's always going to be 100 bucks.

Sure. Well, in order to answer that, let's go back to what quantitative easing does and then see the analogy. So when the Fed wants to stimulate the economy, one of the ways it can do it is to buy up long term Treasury securities, removing them from the market. Then investors say, hey, these long term securities are more scarce. I'm going to have to pay up for them. That lowers long term yields.

And when long term borrowing rates go down, the economy gets stimulated. Corporations, municipalities, states, people can borrow for long term at lower rates because the Fed did quantitative easing. Now let's go to an alleged stealth quantitative easing at the Treasury Department. If the Treasury were to reduce the quantity of long term bonds available to the market,

by providing instead short-term securities, treasury bills, then the market would have the same reaction. Gosh, there's not as many 10 years, five years, 30 years available as there used to be. I'm going to have to pay more to get the amount that I need. Investors will drop yields. And again, corporations and local governments will say, good, we can now borrow money at lower long-term interest rates.

to open new factories, hire more employees, invest in more capital stock. So it has a similar macroeconomic effect, even though, as you noted, it's a completely different operation. In both cases, you're reducing the quantity of long-term treasury securities available to the market and increasing the supply of short-term safe assets available, whether it's reserves in the case of monetary quantitative easing or it's treasury bills in the case of

so-called stealth QE. All of that said, it sounds like a recipe for the government to do its own version of monetary policy, but the government doesn't do that as a matter of practice. And I talk to people in the treasury and banks and in the Fed all the time. I'm very confident of that. And secondly, it's not a good idea because we don't want to have two central banks. One is plenty. I agree with that. One point, Daryl, I never understood on lowest cost to the taxpayer is the

Why is borrowing at the short end more favorable than borrowing at the long end, especially if you have the inverted yield curve? Now, I asked someone who happened to be on the Treasury Borrowing Committee, the T-BAC Committee, and he said, well, we're taking into account the fact of the curve. But if the curve is always right and you always believe the curve, wouldn't wherever you borrow at the curve, whether it's all overnight money or it's all 30-year money, always be the same cost to the taxpayer? I never understood that.

Okay, so this gets to the issue of whether you're a price-taking issuer like a corporation, or if you're a price-setting issuer like a monopolist or a monopsonist in the case of the Treasury Department can sell Treasury securities and keep selling them until it drives the price down, or it can stop selling long-term securities in such great quantities and raise the price. So it's not a price-taker, it's a price-setter to a significant extent. Of course, it's also...

relying on the demands by investors around the world to set prices, but it can have a big influence on prices on its own. This is an issue that PhD students and faculty here at Sanford are talking about all the time. What is the cheapest way for the government to fund itself

It's not a situation where it doesn't matter because the yield curve is always right. It matters because if you're the Treasury Department, you are having a big influence on the Treasury yield curve. Okay, so the reason that they issued more bills over the past few years, it's because they estimated that there would be a

high impact of, you know, if they issued 50 billion or say $100 billion of 10 year notes, that would widen the 10 year note spread, the 10 year would go up a lot more. I would say that they had a much more parochial cash management approach. They just wanted to have a whole lot of cash in their deposit account at the Fed in case they needed the cash. And so they issued a lot more Treasury bills to make sure they had enough cash.

There was also some evidence in the market a few years ago that treasury bills were way too scarce. And a number of economists, Jeremy Stein at Harvard, I believe he wrote with Sam Hansen, Robin Greenwood, and Adi Sundaram, a paper suggesting that the Treasury Department was issuing

too few treasury bills. And the reason for that is that there was a distortion in the treasury bill market. They were so scarce that treasury bill rates were way below the interest rate on reserve balances, indicating the market was hungry for a lot more. It would be natural for the treasury department to issue more bills until they met some of that high demand for bills. And then more recently, as I said, as a cash management

approach, just being risk averse. They want to make sure they have enough cash stored at the Fed. Were treasury bills so scarce? I know that the treasury issued a lot of bills in March and April 2020, but I guess were they scarce just relative to the huge amount of

deposits and reserves that was created in 2020 and 2021. Okay. Well, this is a really interesting interview. I'm glad we're talking about this. So going back many years, there had always been a view at the Treasury Department that they wanted an average maturity of outstanding treasury debt to be relatively long maturity so that there would be relatively low risk

that the government would have to roll over a lot of debt on short notice and suffer a major cost to taxpayers for issuing a whole lot of debt in a short amount of time. Imagine, for example, Jack, that almost all of the debt is in treasury bills. Well, in that case, you'd have to roll it over every few months. You'd have to roll over the treasury debt. And if markets didn't behave well,

you could imagine a lot of volatility in treasury rates and interest expense to the government. So the old view was, let's keep our average maturity very long and the risk of bad surprises, sudden high interest rates or a failed auction even worse, where you couldn't actually sell enough securities, that risk goes way down when most of your debt is rolling over very slowly at long maturities. That had been the prevailing attitude in the Treasury's Debt Management Office for some years.

But over the last, I would say, four or five years, I think the Treasury Department realized that risk was pretty low. And moreover, they were missing an opportunity to meet the demand of the market for more Treasury bills because Treasury bills had been very scarce. And then because of the repeated hit on the debt cap, the ceiling on U.S. government debt, they realized also that they needed a significant cash management account at the Fed. And the best way to manage that would be to have a lot of bills and a lot of cash deposits.

And so the debt ceiling in, yeah, I guess the big one I'm thinking about was 2023, where there was a time where treasury bills were trading way above 6%, 7% on treasury bills. And the highest the Fed funds got was 5.50 or close to there. And that was because literally there could have been a technical default. They would have been paid it back eventually, but they wouldn't have been paid

Thankfully, you know, Congress had it. Washington had his act together. So you're saying for cash management purposes, they wanted to have a lot of money. But why issue Treasury bills as opposed to those large

longer term notes. Why is that the best way for the Treasury to handle a debt ceiling thing where they're basically limiting borrow? If anything, I would think it would be the opposite. Well, cash management, you want flexibility, just like a corporation has a lot of short term assets and it has a lot of short term liabilities and it manages the two relative to each other because both are very flexible. You can roll over your short term liabilities and your short term assets very, very easily. They're quite fungible.

There's no big long-term planning exercises about how are we going to eventually get enough debt financed in the market to meet our short-term cash needs, or if we want to reduce our short-term cash, how can we manage that quickly? And one way to do that is to have

a substantial stock of short-term assets and a substantial stock of short-term cash. That makes sense. I said short-term assets. I meant short-term liabilities. So treasury bills. Earlier, Darrell, you said that there had been a view in the treasury that it's good to have long-term debt because if there's a financing problem, suddenly no one wants U.S. government debt, you're locked in. And

that over the past few years, that view has kind of no longer been there. Would the old view have believed in something called the loanable funds model? And I'm going to give an imprecise, but I hope accurate definition. And then you can share if I got that definition right, as well as your thoughts on it. And the loanable funds model is common sense sounding. It is that

governments need money to finance themselves. And if there's not enough money to finance the government, then that will risk, there's not enough money to fund other stuff. So if the US government borrows so much money, suddenly there's not enough money for Apple. There's not enough money for the bagel store across the streets.

And that is so accords with common sense. However, I would say just as you referenced at the very beginning of our interview, the U.S. Treasury stock has grown enormously. And probably the demand for U.S. Treasury debt is much higher now than it was in 1980 when the total amount of debt was a fraction of what it is now. Yeah. So it's absolutely right that as the government issues more and more Treasury securities, there is some degree of crowding out.

of corporate debt and household debt. If the government had borrowed much less money, then interest rates would be lower and corporations would have borrowed more and the economy would grow faster. This is what economists call crowding out. So the government is crowding out private market debt by issuing so much government debt because investors only want a certain amount of relatively safe debt. Investor-grade corporations

safe mortgages and so on, are going to pay higher interest rates than they would if the government didn't borrow so much. That's absolutely right. And it's getting to be an alarming situation where debt to GDP is now around 100 percent. And it's projected by the Congressional Budget Office to grow to about 150 percent debt to GDP ratio by mid-century. Right now, there's about 28 trillion outstanding of Treasury securities. The CBO projects

40.3 trillion in 10 years. 40.3 trillion in 10 years. There was 7 trillion outstanding just before the financial crisis. So that's five and a half times the amount of treasury debt. It's an unsustainable trend and therefore has to stop, as a famous budget director said in the 1960s. It's also being aided a little bit by, unfortunately, by inflation because

It's mostly not all nominal securities. So as inflation goes up, GDP grows with inflation in addition to normal growth. But the debt doesn't grow with inflation because it's denominated in nominal dollars. That makes sense. There's actually an increasing fraction of the debt that's treasury inflation protected. So that does grow with inflation.

Which could be a problem because if inflation is really high. There's more and more tips which are getting popular with inflation. We shouldn't get the idea that inflation is a great thing because it reduces debt to GDP. It's a terrible thing for other reasons. But nominal GDP growth higher than the growth of debt, that is a good thing to get debt to GDP down. No, that's absolutely right. So where do you think debt to GDP levels can...

So I agree it's unsustainable to continue growth, but what is the level at which we cannot return? I know that's an impossible thing to answer, but how are you thinking about that? Because 10, 15 years ago, there were people who said, oh, as soon as we get to a 90% debt to GDP ratio, boom, growth is going to suffer and the sun won't rise. And that did not happen. So how are you thinking about this?

Well, I'm older than you, Jack. I remember when the IMF said 60% that the GDP was the red line. You don't want to go above that, you're getting into extreme risk. But other countries like Italy and Japan have gotten into far worse situation, much, much higher than the US. And they're still financing themselves, although it's a tenuous situation. My view is that there are two breaking points. The first breaking point is when there's so much debt,

that the market structure can't handle it on a stressful day. And that happened when COVID broke in on March 12, 2020, the treasury market basically became dysfunctional. Even the Fed used that word in describing the market conditions. And it was because the quantity of treasury securities in the market was so large in mid-March 2020, the market just couldn't function properly. That's the first breaking point. And we're going to see that happen again unless

the moves that the government is making to improve the functioning of the treasury market are successful. That will happen before the really big one, which is when investors start to reject the idea of holding US treasury securities without a huge risk premium. For the time being, the US seems to have a very large fiscal capacity. Investors around the world seem really copacetic with buying as much treasury securities

as the government wants to issue. There was a signal last year, November 2023, when the 10-year yields backed up as the Treasury Department hinted that it might be issuing more of those. So it's not an unlimited demand. So far, we haven't really come close to cracking this point at which investors become reluctant to hold more Treasury securities and need to be compensated with extremely high yields.

And I think it will take quite a long time, but that's not a suggestion to the government to go ahead and borrow a whole lot more because they're digging a hole that eventually they're going to have to fill in and it's going to get harder and harder to pay off that debt. The obligations are being passed to future generations. It's not really fair to your children or my grandchildren to load them up with this much debt and they're going to have to pay all that interest expense

The interest expense, we talked earlier about crowding out other government expenditures, and it's also crowding out the household budget as taxes will eventually have to go up to pay off all this debt. So it's just really a bad situation. When it breaks, I don't know. It's very hard to predict. It could be a long time. The longer the government waits to correct the situation, the next administration has already indicated that it plans to make significant increases in its fiscal budget

the least net of taxes. The tax rates may go down for many Americans and borrowing shows no sign of going down, at least currently.

And whenever this red line is, we're not going to be like the IMF and make a prediction. But what will you be looking at as so as to measure that stress? You referenced, I believe it was late October or 2023 when the 10-year hit 5%. And then I think it was November 1st, 2023, the Treasury issued its plan and it involved a lot of Treasury bills. And then boom, the 10-year just went nowhere but down. And of course, that accorded with the

Federal Reserve moves and the like. But I'm just looking at term premium, which I, again, do not really understand, other than the very generic definition, is it was negative in 2020. And now it is, quote unquote, higher at 60 basis points. But if the term premium, Scott Besson issues a lot of 10, 20, 30 year treasury bonds, and the term premium widens to 1.7%, can we really say that the red line is here? Because that's where it was when I was born in the 1990s.

No, we can't say that. The term premium could get significantly higher without reaching what I would call a point where the bond vigilantes are setting the prices in the treasury market at very, very high rates because nobody else is willing to buy them. We're not even close to that point. And one of the reasons is that there's really only one significant safe haven in the world, and that's the US treasury security. Everyone says the dollar is the safe haven, but where do you put your dollars? You're not going to keep them in a deposit account earning nothing

There's an enormous demand for treasury securities because they're pretty much the only game in town when you're talking about a large stock of safe assets. And if you go to the next largest issuer, who is it? Any European countries, Japan or other Asian countries that are able to issue in size the quantity of safe assets that investors really need to manage their liability risk, pension funds, insurance companies, banks,

You and I, many people rely on treasury securities and we would have to become extremely frightened before we would stop buying safe assets and turn to something like, where would we go actually? I mean, treasury bills maybe, but those are also going to be quite volatile. So I think treasury securities are...

the safe haven of choice for years to come. But we shouldn't test the limit of that. And you said in the beginning, the Treasury market is much more

financed, I would say the word levered than in history. So a lot of these treasury securities are owned by people who are borrowing in the overnight market. And that is a risk free or semi, you know, very close to risk free market is not like, you know, you're buying, you know, crap mortgage backed securities and buying them overnight with leverage, you're buying us treasuries, but you know, it's still leverage.

Yeah, it's leverage. But as you know, the SEC has mandated central clearing and all of that financing of Treasury securities, not all of it, but almost all of it is going to be centrally cleared. So it's going to remain a safe asset. It'll be more expensive, but I think it's worth it. We want, you know, it should be okay to pay a small insurance fee to make sure that this market doesn't blow up. If you look at the non-centrally cleared market, there's very little in the way of margin requirements.

on that leverage, whereas centrally cleared treasuries have a margin requirement. And there's also a lot more transparency to the market when you know it's all being centrally cleared in one place. So I think there's very little risk that the treasury repo market is going to fail for some reason or become a source of counterparty risk to the economy. But the pressure on balance sheets to continue to finance it is going to continue to get bigger and bigger.

And that will eventually come back to the taxpayer. If you have to pay more to finance your purchase of treasuries, then you're only going to be willing to buy them if they yield a higher interest rate. And that comes right back to how much the taxpayer has to pay on U.S. debt. So we really do need to figure out a way not only around an increasing amount of government borrowing, but also the limited amount of space on dealer balance sheets for financing these treasuries.

Daryl, in the limited time we have left, I want to give you a chance to talk about your most recent paper, the title of which is Bank Funding Risk Reference Rates and Credit Supply. One central one, I'll summarize, just because I want to be respectful of your time, is that basically, since we moved from LIBOR to, in times of credit stress, there's no penalty for credit.

corporations who have tons of credit lines with banks where they can tap them at any time, there's no penalty to tap those SOFR rates because SOFR rates during an economic crisis, risk-free rates tend to go down. Whereas with LIBOR had a credits element to it, LIBOR rates could actually be quite high. So corporations wouldn't want to tap it. This actually has a risk to banks, maybe not a systemic risk, just to bank shareholders, because those credit lines are going to be tapped a ton.

The research you mentioned is with a different New York Fed economist, Stefan Luck, and two Stanford PhD students, Zachary Wang and Harry Cooperman, and a former Stanford PhD student, Professor Yilin Yang. And what we discovered is

First of all, it's kind of shocking. The total quantity of corporate borrowing from banks is more in credit lines than it is in term loans. I was very surprised to learn that. I was very interested. That's from the largest 20 banks. And sorry, that the untapped credit lines are larger than both the tapped credit lines and the term CNI lending. Yeah, the total amount of untapped credit lines available to corporate borrowers from banks is enormous.

And the risk you mentioned is that the next time the market goes into stress, the Fed will generally respond to a credit crisis by lowering risk-free rates. So all of a sudden, that huge amount of corporate credit lines are available as a source of funding to large U.S. corporations at a reduced interest rate at a time at which they will find it extremely valuable to drop.

because they want to store up cash. In March of 2020, when COVID hit, there was a huge precautionary draw on those lines that went into their deposit accounts and other safe assets. Some banks had to go out and get funding to meet those line draws. And when they did, they had to pay high interest rates.

The transition made a few years ago from LIBOR to SOFR as a bank lending reference rate was a good choice because LIBOR was really a corrupt reference rate. It wasn't reliable, it wasn't robust, it had been manipulated. So we needed something really robust and SOFR, I mean, there's a couple of trillion dollars a day in SOFR. It's virtually impossible to manipulate it. So that was a great move to get rid of LIBOR and go to SOFR. But now,

Banks have a big exposure to these massive line draws where corporations will draw at SOFR plus a pre-agreed spread that was established in normal times and draw a lot in a crisis. And banks will have to go out and get funding for that at potentially very expensive rates. It may reduce the supply of credit lines by banks. They might not be as likely to supply credit lines without charging more for them. At least that was our finding. The very largest banks tend to

to not suffer in these situations because they're viewed as safe by their corporate customers and the corporate customers will draw on those lines and turn around and put the money right back into their deposit accounts at the same bank. And so those banks, the largest, safest banks, don't have any difficulty in funding those line draws because they don't have to go out into external markets to get the funding. They get cheap deposit funding from the same corporate borrowers.

corporate borrowers who tap the credit lines, I think it was 94% of the money stays in those banks, which pays at best SOFR, probably maybe lower than SOFR, whereas the 6% or maybe higher for other banks and during the great financial crisis, they had to fund it in the wholesale borrowing market when it was

frozen. Well, that as well as all your other work is available on darylduffy.com. We will link to all the specific works that we spoke of as well as your general website. Daryl, thank you so much for sharing your insights with us. We so appreciate it. We wish you and everyone watching a happy holiday season and a happy new year as well as a great start to 2025. Thank you again. Thanks so much, Jack. Thanks. Thank you. Just close this door.