Powered by 
The Federal Reserve’s balance sheet may sound technical, but it lies at the center of major questions about monetary policy, market function, and central bank independence. How large should the Fed’s balance sheet be, and what happens when policymakers try to bring it down? In this conversation, John Cochrane and Darrell Duffie examine what the Fed owns, why it matters, and why reducing it is more complicated than it looks.
Cochrane and Duffie also explore why the Fed has found it so difficult to shrink that balance sheet after years of quantitative easing. They discuss how post-2008 liquidity rules, supervisory pressures, and changes in market structure have increased banks’ demand for reserves, making balance-sheet reduction more complex than a simple asset runoff. The conversation also considers why the composition of the balance sheet matters, especially when the Fed holds mortgage-backed securities, and how a narrower role may help preserve central bank independence.
Read more from Darrell Duffie:
Darrell Duffie, “Is the Fed Bigger Than it Needs to Be?” Freedom Frequency. April 14, 2026.
Darrell Duffie, “The Payment System Puts a a Floor on the Fed’s Balance Sheet.” Brookings Papers on Economic Activity. March 26-27, 2026.
Transcript
John Cochrane: Hi, I’m John Cochrane, a senior fellow here at the Hoover Institution, also sometimes known as the Grumpy Economist. I’m joined today by Darrell Duffie, also a senior fellow at Hoover and sometimes professor of finance at the GSB.
We know him as the polite economist because I’ve never seen Darrell be grumpy about anything, even when he had plenty of reason to do so. Welcome, Darrell.
Darrell Duffie: Thank you, John.
John Cochrane: We’re here because Darrell has written a very interesting new paper for the Brookings Papers on Economic Activity about the Federal Reserve balance sheet. Darrell and I started a long discussion about this paper. At one point, Darrell said, “Well, John, we should videotape this and send it out through Hoover’s channels. This is getting fun.”
And I said, “That’s a great idea.” And the Hoover marketing team said, “That’s a great idea.” So here we are. And we all think it’s such a great idea, this conversation will likely inaugurate a sequence of conversations of the sort. So we hope you like it, and let’s get going. Darrell, welcome.
Darrell Duffie: Thank you. Thank you, John. There’s almost no better interlocutor on any subject in economics than you, John, so I’m looking forward to this.
John Cochrane: Well, we’ll find out.
So our topic is the Federal Reserve balance sheet. It’s in the news. You may have seen our other colleague, Kevin Warsh, who is nominated to be Fed chair, wants to reduce the balance sheet. Lots of discussion about the balance sheet, but it may seem like a little bit of an abstruse topic.
So I can just start, Darrell, with the simplest question. Can you help our audience understand just what the Federal Reserve balance sheet is? And then we’ll get to the question of: Is it too big? Is it too small? And what to do about it?
Darrell Duffie: Sure, John. Happy to do that.
So let’s start with the fact that a central bank like the Fed is a bank. So it has assets and it has liabilities. I think when Kevin Warsh was speaking about this topic and complaining about the size of the Fed’s balance sheet and how it’s become a kind of lightning rod for the Fed, he was more focused on the asset side of the balance sheet. Those are securities that the Fed owns.
And today there’s about six and a half trillion, a little bit more, about two-thirds of which are US Treasury securities and about one-third of which are agency mortgage-backed securities. So they’re sort of quasi-governmental because the government owns those agencies, but they’re riskier than Treasury securities.
Now we have the liabilities. That’s the other side of the balance sheet, for those of you that do accounting, and those are the things the Fed owes to others, primarily deposits. And those deposits break down into two large categories.
The first category is the main subject of the paper you mentioned that I wrote recently, and those are deposits by commercial banks at the Fed. Those deposits can be withdrawn on a moment’s notice, just like any commercial bank deposits.
The other large deposit category is a special deposit account that’s held by the US government at the Fed, so that it can pay its bills very quickly and receive its tax revenues very quickly. That deposit account is held by the Treasury Department. It’s called the Treasury General Account.
Right now, there’s about three trillion of commercial bank deposits at the Fed. Those are called reserves or reserve balances. And there’s approaching a trillion of US Treasury, a little bit less right now, around 900 billion, but it’s probably going to hit a trillion when we get into tax season in a few weeks.
So those are the two primary liabilities, and there’s a smattering of other things we probably won’t get into today.
John Cochrane: So I think of the Fed as a giant money market fund.
Darrell Duffie: Oh, please. I forgot to mention one very important liability that we often don’t talk about because we take it for granted, and this is paper currency. And there’s a large amount of that. The Fed has about 2.4 trillion of paper currency outstanding.
John Cochrane: Yes, you and I don’t have any of it, but somewhere out there, there’s a lot of cash. I gather a lot of it is $100 bills in mattresses around the world that we never see.
So I think of the Fed as a giant money market fund. It issues largely interest-paying shares. It backs those shares by mostly very safe Treasury securities, as opposed to a bank, which backs the shares by risky loans and so forth.
And great, we run a large money market fund, transforming in some sense Treasury debt into something that banks like to hold and that is more liquid, if you will, easier to buy and sell than Treasury debt. What’s the problem?
Darrell Duffie: Okay. Well, I mean, Kevin Warsh identified the problem as the Fed having gone through a number of rounds of quantitative easing. Let me unpack that.
Quantitative easing is when the Fed is trying to rescue the economy and support the economy or the financial system by buying more of these Treasuries or agency mortgage-backed securities. And in the mind of Kevin Warsh, it’s been too much of that. In fact, that’s the reason that he left the Fed, resigned as a governor more than 10 years ago, is that he felt that was excessive. And since then, it’s gotten much, much bigger.
And in his view, it’s too big. It suggests in his mind that the Fed could be viewed by others as being adventurous and going out and buying stuff at its whim and maybe having too big a footprint on the US economy, and maybe even to the point at which others would point to the Fed and say, “Well, we need to clip their wings. They’re being too adventurous.”
Now independence of the Fed starts to come into the picture, and it has been, as you know, quite recently.
So that’s the complaint by those that are complaining. There are some more technical issues related to a very large balance sheet. For example, when the Fed pays for these securities that it buys by providing banks with more of these deposits, the banks are now holding so much of these safe deposits that are earning a nice interest rate that they’re less anxious to provide loans.
So this is a so-called crowding-out argument, and the banks are subject to capital requirements even for those Federal Reserve deposits. And so there is a sort of crowding-out story that’s a bit of economics.
And then there’s a very vague piece of economics that I don’t really understand the cost-benefit analysis very well, which is called, quote-unquote, footprint, meaning the Fed is so attractive as a provider of liquidity in money markets that the interbank money markets are atrophying. The banks themselves are not that active anymore borrowing and lending from each other, and that leads to a kind of so-called malaise in the interbank market, which is very inactive right now in terms of Fed funds, the federal funds interbank borrowing market.
So those are the complaints on too large a balance sheet, and then there are some complaints going the other way.
John Cochrane: Yeah, yeah. So the heart, I should say, this is still background. The heart of Darrell’s paper is not about the merits of a large balance sheet, but about the technical problems that we’ll run into if we try to reduce that balance sheet.
Darrell Duffie: Exactly.
John Cochrane: And I am making a note to myself to get there sooner rather than later, but I do want to spend a little time on the merits.
In part, as Darrell knows, I love the big balance sheet, parts of it. So a big balance sheet backed by short-term Treasuries. Let’s get to the composition of the balance sheet later.
But in the old days, the balance sheet, the reserves, were sometimes as low as $10 billion. Now they are in the trillions, and they pay interest. So banks put no effort whatsoever into managing all their cash reserves.
We have what Milton Friedman called the optimal quantity of money: assets that banks hold for other reasons happen to provide liquidity. Liquidity is costless to the economy, said Milton. Give us as much as we want.
I think Milton never went to advocating for that regime, paying interest on lots of reserves, because he worried that then the money supply couldn’t control inflation and it would all spiral out of control. And it didn’t.
So this strikes me — I point this out especially because here at Hoover, we’re sometimes a little critical of the Fed — but they made this huge innovation to large amounts of reserves that pay interest, and it turned out that had really no effect on inflation. Our inflation came much later from totally other causes and removed all of these games that the banks were playing.
Why should banks borrow from each other, really? Why is a footprint bad than that? So that’s always struck me as a good thing. And I’ve always been skeptical of the notion that QE did anything other than to give the Fed something to talk about when it felt it needed to stimulate the economy, because the Fed buys Treasuries and issues reserves.
Now, those Treasuries might have been held by a private money market fund. So just to make it extreme, you hold a money market fund that holds Treasuries. The Fed buys the Treasuries in return for reserves, and now you end up holding an interest-paying bank account that holds the Treasuries through the public money market fund.
Why is the Fed money market fund so much different than the private money market fund that this has any stimulative effect whatsoever?
So all of those stories you told — now, I have to admit, if I didn’t think it did anything on the way up, then reducing it, likewise, I have to be held. Well, it’s not going to be terribly harmful to reduce it until we get to some of the technical problems you’re into.
But you expressed some skepticism as to the importance of reducing reserves because it’s overstimulating the economy. And I wonder if you’re on board with me that maybe that isn’t quite as strong stimulus as often mentioned, especially in Wall Street.
Darrell Duffie: I think even the original proponents of quantitative easing, meaning buying a lot of assets to support the economy or the financial system, I think those proponents have somewhat a buyer’s remorse now because it’s turning out to be hard to reduce the size of the Fed’s balance sheet.
And if you take seriously any of those complaints, then, as many of the Federal Open Market Committee members have, they are now in a situation where they want to reduce the balance sheet, but they’re having some difficulty getting it below a certain level. And that’s the genesis of my paper.
John Cochrane: Yeah. And we’re going to get there with one more question.
Darrell Duffie: Sure.
John Cochrane: Because the other part that you mentioned is the nature of the assets that the Fed is buying. And here I would join in some of the criticism.
The Fed bought long-term Treasuries. Your average money market fund buys short-term Treasuries and really just makes them a little more liquid. But the Fed bought long-term Treasuries, thereby effectively shortening the maturity structure of government debt so that when interest rates went up, our government had to pay more interest costs on the debt swiftly overall. And that was something many of us were warned about ahead of time and happened in the past.
And it also bought mortgage-backed securities, thereby deliberately funneling money to the mortgage market. Well, that sounds nice if there’s no budget constraint, but that means that you’re not funneling money to other markets.
And our dearly departed friend Charlie Plosser always made this argument. It wasn’t so much the economics of it, but when the Fed gets into the business of buying securities in order to prop up specific markets, that’s an invitation to political interference that the Fed, in order to remain independent, really should stay away from.
Darrell Duffie: Yeah, I totally agree, John. And I think that’s where the greatest buyer’s remorse lies, is in the mortgage-backed securities.
And you can tell that that’s the case because as the Fed was recently reducing its balance sheet, it was getting rid of mortgage-backed securities faster than it was getting rid of Treasuries. And it has professed, or members of the FOMC have professed, a desire to entirely rid themselves of mortgage-backed securities and to go to an asset side of that balance sheet, which is composed of longer-term Treasuries only to the extent of paper currency, because it doesn’t run off very quickly, and then Treasury bills to back the Federal Reserve deposits because they can fluctuate in size quite a lot.
And by going in this direction, the Fed would be largely immune from fluctuation and interest expense. So maybe a mild bit of seigniorage on the paper currency and then a very sleepy story about the Fed spending so much money on interest for the banks.
John Cochrane: That does seem appropriate that we need an accord between the Fed and Treasury, who is in charge of interest exposure of the US debt, and that should be the Treasury.
And actually, I was cheered to see the Treasury buying mortgage-backed securities, not because I think it’s a good policy. I think it’s a terrible idea. But if it’s going to be done, it should be done by the Treasury, the politically accountable branch of government who hands out subsidies right, left, and center, not by the Federal Reserve that wishes to stay independent.
Darrell Duffie: Yeah.
John Cochrane: That’s the alternative. If you think that we need to buy mortgage-backed securities, well, then let the Treasury do that.
Darrell Duffie: Yeah, I agree. And it’s not only a fiscal response in its own right, the Fed buying mortgage-backed securities, but it invites a loss of independence because Congress may get the impression that it can lean on the Fed to buy mortgage-backed securities to support the housing market.
And supporting the housing market, that’s maybe a very good fiscal objective, but the Fed should not be involved in that. And I think the Fed, everybody in the leadership of the Fed now recognizes that that’s unfortunate.
John Cochrane: Well, whether in the presence of supply constraints, it is a good idea to funnel money into demand for housing is a topic that we’ll take up for another time, maybe with a Zillow tour of real estate in Palo Alto to make the point.
So now let’s get on to: For some reason or other, we want to reduce the balance sheet. You’ve done remarkable research showing hiccups, let’s call them, in the Treasury markets and in reserve markets the last couple of times they tried to get down to the balance sheets.
Darrell, we used to have $10 billion in reserves. Now we get anywhere near a trillion, and all of a sudden, things happen. What went wrong? Why is it hard to reduce the size of the balance sheet?
Darrell Duffie: Well, in the first instance, it was already difficult for the Fed to manage with a small number of billions of reserve balances to run the entire financial system. And they had already asked Congress for permission to pay interest on reserves so that they could have somewhat more reserves and not have such a scarce quantity.
They were driving monetary policy by scarcity of reserves, which many have questioned. I mean, that’s consistent with the Friedman idea that why would you starve banks of a source of useful liquidity?
John Cochrane: If I may just stop. So what you’re saying is, it is remarkable. We ran this entire economy on basically $10 billion, which is a tiny amount relative to a $36 trillion economy.
Darrell Duffie: Yeah.
John Cochrane: Most of the cash isn’t cash. $10 billion is just the stuff that goes back and forth at astounding speed. What you’re saying is the banks did this because they had to forego interest, and so they had very smart people trying to use as little reserves as possible.
And the Fed said, “Well, if we could give them some money, maybe they’d hold a little more and not basically run to the ATM machine every 20 milliseconds in order to make the transactions.”
Darrell Duffie: Yeah. Recently, Federal Reserve Governor Chris Waller, when asked why doesn’t the Fed reduce its balance sheet back to the old days, he said — this is almost verbatim, to the extent that I can remember his exact words — “You don’t want banks scrounging under the couch cushions looking for money.”
He said, “That’s massively inefficient and stupid.” And I agree with him.
John Cochrane: That’s the grumpiest thing I’ve heard you say yet.
Darrell Duffie: Well, I’m agreeing.
So this idea that you should starve banks for reserves to reduce the balance sheet is a silly idea. The objective should be, if they want to reduce the balance sheet, to make it possible to have abundant reserves, no scarcity of liquidity, but with a much smaller demand for reserves.
Meaning, make the institutional features of the monetary policy implementation and the plumbing of the financial system such that even a trillion less or more, or even much less reserve balances than the banks have today, would be plenty more than they would ever need on a given day.
And in order to do that, my paper dives into what would those changes need to be and how difficult would they be.
John Cochrane:
Okay. So why is it now, when we try to reduce reserves under a trillion, we start to get into hiccups in money markets? There’s kind of a vision, well, maybe 100 billion that pay a decent amount of interest would be enough to do the mechanics of paying, but we seem unable to get anywhere close to what we were able to do back in the old days.
Darrell Duffie: Yeah, there are two basic mechanisms at work here.
One is that along with all the other changes that came with the great financial crisis, the global financial crisis of 2008, Congress got religion on forcing the largest banks to be self-sufficient for liquidity. They didn’t like the idea that JP Morgan or Citibank or Morgan Stanley or Goldman Sachs, those large, large banks, would come as supplicants to the Fed and get additional liquidity.
So they introduced requirements that the banks be self-sufficient. And the Fed piled on additional regulations for safety reasons, regulating banks so that they must be self-sufficient.
Now, when they did that, they didn’t really have in mind days on which the banks really needed more reserve balances, but would be afraid to go to the Fed to get more.
Even today, there are supervisors in the Federal Reserve system that would send a bank executive a nasty note about, “Well, why did you have to go to the Fed yesterday to overdraft your account or to go to the discount window to borrow additional funds? You’re supposed to be self-sufficient.”
So banks are reluctant to do that. And because they now are not going to go to the Fed to get extra liquidity, they stack it up on their own balance sheet. They stack up the reserve balances, and they’re very reluctant to give them up when the Fed tries to reduce its balance sheet.
John Cochrane: So let me sort of try to make this concrete. The minute you say liquidity, I know we’re going to put people to sleep here.
You’re a bank, and you’ve got money coming in and money going out all day long, but you think today would be a day, there’s so much going on, you need a little more in your bank account at the Fed to match these transactions.
You’ve got some Treasuries, but for institutional reasons that I know you don’t like and I don’t like either, you can’t turn the Treasuries into cash. It takes a whole day.
Darrell Duffie: Unless you go to the Fed.
John Cochrane: Unless you go to the Fed. Well, you could borrow money from another bank against your Treasuries in a repo transaction, or you can give the Treasuries as collateral to the Fed, get some money, and that money will last just through the day, and then you’re good for it.
It’s the Treasuries. They’re there, you know? It’s not like you’re going to welch on the loan. That’s, I think, your vision: You can get liquidity from the Fed when you need it, but the Fed doesn’t like banks doing this at all, even though it’s a completely collateralized loan with Treasury securities as the collateral.
Darrell Duffie: Well, the Fed is of two minds on this. In fact, depending on whom you speak to in the leadership of the Fed, they would much prefer that when banks run short of reserve balances, they would go to the Fed and borrow, whether from the discount window or overdrafting their account or from other liquidity sources that the Fed has set up.
So there’s no question about that. They’re actually quite disappointed that the banks are stigmatized from using these facilities that the Fed offers.
But as I said, there are others in the Federal Reserve system that look askance at the idea that a bank is not meeting its own liquidity requirements, its need for reserve balances or other sources of liquidity, from whatever it already has on its balance sheet.
And as I said, that just gives the banks a huge incentive to keep a lot of reserve balances and makes it difficult for the Fed to —
John Cochrane: So in some sense, this is the tension between monetary policy and financial regulation. I now see it.
The monetary policy people, or the financial markets sort of people say, “Yeah, if you need money for a day, just give us the Treasuries and you can have money for the day and we’ll settle up tomorrow.”
The regulatory people say, “Oh, bank, you need money. Is there something wrong that we should come take a look at?” And the banks don’t want any part of that.
Darrell Duffie: Yeah. Even if the conversation inside the Fed is no longer that way, and I think they’re getting coached to cool it on that, the banks are nevertheless stigmatized. They don’t like the look, or someone, a manager inside the bank, doesn’t like the look of going against the spirit of those liquidity regs.
And curing them of that stigma has proven to be extremely difficult.
John Cochrane: Now, my understanding is that the European Central Bank operates much more freely this way, that banks routinely post collateral with the European Central Bank, get loans, both overnight and term loans, and with the full cheering and no stigma and no problems whatsoever.
So that certainly seems to indicate that this would be possible if we could get over this sort of psychological barrier to borrowing money from the Fed when needed.
Darrell Duffie: Indeed. Some of the research in that Brookings paper, that’s based on other work that I’ve been doing with University of Toronto expert Shristi Singh and the University of Pennsylvania expert Chaojun Wang, shows that if the banks were willing to overdraft a little bit, that would save a lot on the quantity of reserve balances that they need to do their daily business.
John Cochrane: So let’s talk about overdrafts, because this sounds to the average listener, “What? You’re allowed to overdraft your account? If I overdraft my checking account, I pay a $50 fee.”
But in fact, intraday overdrafts was one mechanism that banks used to settle their vast amounts of payments without having to hold a lot of money.
Can you explain an intraday overdraft in a way that it sounds like a good idea, not some terrible skullduggery by the banking system?
Darrell Duffie: Well, John, you’re absolutely right.
Back in those days you described, when there was only $10 billion available of reserve balances and banks had to make literally trillions of dollars a day in payments, they would often run bare. They would not have enough reserves already on their balance sheet. And in the middle of the day, they would just run a negative balance with the Fed.
And as long as it was fully collateralized, the Fed was absolutely fine with that. And the largest 10 banks were doing $130 billion a day, no trouble. And now, ever since these more onerous regulations have come in, the largest banks are simply frightened of drawing on those.
And with one exception, COVID, a day on which the Fed said, “We’re not looking at overdraft today,” the largest banks have hardly used those overdraft facilities at all. In fact, the smaller banks continue to use them to some extent.
John Cochrane: Yeah. And to some extent, it’s how — when you and I were young, if you wanted to go out to dinner on Saturday night, you had to go to a bank and cash a check and have cash.
And now you go out to dinner on Saturday night, you just use a credit card, which is, in essence, a promise. We’ll settle up at the end of the month, and that seems to work well, which is how your and my financial system works. I think I have 20 bucks that’s been sitting in my wallet now for three months, with using essentially no cash whatsoever.
Darrell Duffie: Yeah. The difference in this case is that these facilities that the Fed provides — overdrafts, the discount window, and another one called standing repo operations — those are fully collateralized, so the Fed is not exposed to the credit risk of the banks.
John Cochrane: Yeah, you’re just getting over the fact that the collateral, the Treasuries, take a whole day to settle, whereas you might need cash right now. And for the moment, the Fed is the place to get cash right now if you can’t borrow from another bank.
Interbank lending would help a lot, but then that adds to the complexity of the whole business.
Darrell Duffie: Yeah. Interbank lending has a certain amount of friction involved in it now, and banks have largely avoided it.
It’s quite remarkable that these days, there’s around three billion or so of daily interbank borrowing in the United States federal funds market. That used to be a vibrant market, but now that banks have stacked all these reserve balances on their own balance sheets, they don’t need to go to the Fed funds market except for a tiny, tiny —
John Cochrane: Yeah, just to explain, back in the old days, you used to have to have reserves against your checking accounts and deposits. Every other Wednesday, I think it was, they added up the reserves. And if you came up a little short, you would borrow the reserves from another bank.
Every now and then, the interest rate to do that would spike because there weren’t enough reserves. But that interbank borrowing was mostly to meet your reserve requirements. I’m not quite sure why we’re sad for it going away.
The point of a central bank is to have liquid markets. It’s free to do it. There’s no risk to it. But anyway, that is —
Darrell Duffie: Yeah, absolutely.
John Cochrane: One of the big changes.
Darrell Duffie: That is a big change.
John Cochrane: So the third one — so why do we need so many — our conversation, in case you forgot, why do we need so many more reserves these days?
Part of it is we don’t have intraday overdrafts, so you need to bring more cash to the market every day. And the other was it’s harder to borrow from the Fed when you show up at the market, at the butcher’s, and you don’t have enough. You can’t just borrow from the Fed and do it that way.
And the third you mentioned is liquidity regulations. So banks are required to hold a lot more reserves just sitting there, which I think of sort of as the joke we told after the financial crisis. The Titanic hits the iceberg, and somebody says, “Put the lifeboats in the water.” And the captain says, “No, no, no. Regulations say we must have 10 lifeboats on the boat at all times.”
So the liquidity regulations and capital regulations force the banks to hold a lot more reserves than they would have back in the old days.
Darrell Duffie: Yeah. Technically, they’re not forced to hold reserves. They could hold almost any kinds of assets in sufficient quantities, but the cheapest way for them to meet these requirements and the easiest way, because reserves are so liquid, they’re instantaneously available to make payments, is to just hold a lot of reserves.
And on top of all of that, they get a nice big fat interest payment. And some of the banks have stacked even more reserves onto their balance sheets in order to collect those interest payments.
John Cochrane: Yes, it is a little strange that the Fed pays more interest sometimes on reserves than you can get on Treasuries or overnight repo and so forth. For a market rate, not necessarily a subsidy to being a bank, but perhaps that’s a minor issue.
It certainly isn’t coming out of loans. It’s coming out of holding Treasuries instead of holding reserves.
Darrell Duffie: There is an option there. In fact, I discussed it in my paper, and I’m modeling it with some others as well.
This option was presented to the Federal Open Market Committee back in 2008 when they originally decided to start paying interest on reserves. A special interest rate working group provided a secret memo to the FOMC saying, “One of the ways to do this is to pay the interest rate you need to control inflation and conduct monetary policy for the first amount of balances that a bank holds.”
But then, once they get beyond that necessary amount, drop the interest rate that you pay them. This is called tiering, so that they’re not anxious to hold more than what they need.
And that option would — I’m not suggesting it would be easily reintroduced — but it could be introduced in the United States, and it has been introduced to create a much lower demand for reserves, as it has been in New Zealand, Norway, South Africa, and many other countries.
John Cochrane: Now that strikes me, I’m the grumpy economist, as a terrible idea, because that means that the central bank has to decide how much reserves each bank actually needs.
The idea of telling Jamie Dimon, “No, sorry, you only need 932 billion, not 937, so we’re going to stop paying you interest on this stuff,” and he has to hire all those cash management guys to go find the pennies in the couch — that seems like a suspicious idea to me.
In part also because I’m still in the abundant reserves fan club. What harm does it do that you’re holding interest-paying stuff that is backed by short-term Treasuries, rather than holding a money market fund that is interest-paying stuff that’s backed by short-term Treasuries? They look like almost the same security to me.
I know the money market fund isn’t tradable instantly, but from the grand scheme of things —
Darrell Duffie: Well, John, I don’t think there’s any threat that tiering interest rates on reserve balances in the US is going to happen —
John Cochrane: All right.
Darrell Duffie: Anytime soon. It’s a very unpopular idea, especially among the banks, because it would reduce their profitability quite a lot.
But it is one of the ways that if the Fed did decide that it was sensitive to a very large balance sheet, it could significantly reduce the size of its balance sheet and still leave reserves abundant. In fact, so abundant that banks would be saying, “Don’t give me any more of those reserves because now I’m beyond the point at which I’m getting a full market interest rate, and I’m getting a measly low interest rate, and I’m going to eject those extra reserves as fast as I can from my balance sheet if I don’t need them to make payments.”
John Cochrane: Yeah. That’s, I think, sort of where we want to go.
So what you’re saying is, we had an economy that used to get along on $10 billion of reserves. Somehow we’re stuck at about the trillion level, but a lot of that is because we have passed regulations that require banks to do things in a fairly inefficient way if you regard reserves as being something costly and you want to get rid of them.
So come on, let’s just burn up the rule book and get to lower reserves if that’s what we want, eh?
Darrell Duffie: Not as simple as it sounds because, first of all, you have to convince banks to draw from liquidity. You have to talk them into tiering their reserves or retool the payment system so that it’s not such a reserves hog, another subject of my paper.
And these are all conversation stoppers for many people. It all comes back to how much effort and cost the Fed wants to put into this.
John Cochrane: So the bottom line is, in order to reduce reserves below something like a trillion, trillion and a half, you need a major rewrite of financial regulation and operating practice.
Darrell Duffie: We’re at three trillion now.
John Cochrane: Oh, sorry.
Darrell Duffie: Yeah. And already, we’ve hit rock bottom, and the Fed is adding reserves because it now understands that the financial system needs more.
This brings us to a very influential Jackson Hole paper by Acharya, Rajan, and Steffen that described what they called a ratchet effect, by which every time the Fed does quantitative easing and expands the quantity of reserves, it’s very difficult to get it back down to where it was.
So it used to be, a trillion was abundant and no one could ever imagine that the banks would need more than a trillion. Then there was a blowup in September 2019 when 1.4 trillion was not enough. And now we’re at three trillion is not enough.
John Cochrane: So let me ask this slightly larger question.
First of all, reserves are wonderful from a financial stability point of view. We actually almost are at narrow banking right now. Narrow banking is where you put a deposit in a bank. The bank turns around and invests that in interest-paying reserves. Such a bank can never fail because it’s just holding reserves. There’s no way it can ever lose money.
I’ve always been a great fan of narrow banking for just that reason. Strangely, the Fed has basically put the kibosh on narrow banks, although our current banks are sort of becoming narrow banks. It put the kibosh on segregated accounts, which is another way to have large, uninsured, interest-paying accounts that are 100% backed.
But now we have crypto coming along that may make fast payments between people not going through the banking system, not making reserves happen.
Similarly, money market funds — I tried to cash in a large amount recently to make a purchase, and it took two days for Vanguard’s money market fund to deliver to bank. Why can’t a money market fund give me almost instant payments? Because they’re not allowed to have accounts at the Fed and make interest payments?
And as you know, I’m a great fan of why doesn’t the Treasury offer directly overnight fixed-value floating-rate debt that is instantly transferable? It seems, in the larger scale, the mechanisms of making very fast transactions, even outside of the banking system, are unnecessarily hobbled and could take up a lot of this slack.
Darrell Duffie: Well, that’s a controversial topic among central bankers. It’s their purview to provide what would be called a settlement version of money, something that’s always worth one and would be readily accepted.
John Cochrane: No, I think they should provide such a thing, but they don’t allow other people to provide stuff to them.
Darrell Duffie: Yeah. Well, many don’t want others to provide it because they feel like they’re running monetary policy implementation. And if others start providing similar services at different kinds of interest rates, then the central bank becomes less and less relevant in setting interest rates.
John Cochrane: So you think that there’s still a feeling that their control of the payment system is vital for them to fix the interest rate?
Darrell Duffie: Well, that’s a good question. If stablecoins ever actually do come into the economy in a major way, about which I’m quite skeptical, but if they do, then central banks will be put to the test.
And as you know, under current US legislation, stablecoins are not allowed to pay interest. And that was at the strong suggestion of the banking lobby.
John Cochrane: As I think the ban on narrow banks was at the strong suggestion of the banking lobby as well. In fact, the Fed pretty much said so. This would undermine the profitability of the bank. Really? That’s your job? But that’s for another day.
So bottom line, I think, if the — let’s try to get to bottom line. If our friend Kevin Warsh goes in and really wants to reduce the balance sheet, I think we’re forecasting that that won’t really change much stimulus.
The idea that it’s creating financial market bubbles and so forth is overstated. So I wouldn’t look to huge effects on the economy, but it will require a lot of regulatory fixing to do, and there is some question overall to what benefit. What’s your bottom line?
Darrell Duffie: Well, I’m open-minded, and in my paper, I don’t take a stand on whether the Fed’s —
John Cochrane: You’re very good at that.
Darrell Duffie: Yeah. Well, you’ve said to me in the past, John, that economists shouldn’t weigh in on politics, but the Fed itself is very sensitive to the politics of a large balance sheet.
And I could see the Fed making a rational decision that it wants to reduce its balance sheet. But if Kevin Warsh comes in, or when he comes in, assuming the nomination is confirmed by the Senate, which I hope will happen soon, I don’t think you’re going to see him saying, “Start selling assets so we can reduce our balance sheet today.”
I would imagine that he would say, “We need to gather all of the available theory, evidence, and all the experts, and come up with a plan and see if this is feasible, to what extent we could reduce the balance sheet, how long it would take, how much it would cost,” and then put off the execution of that plan until it’s ready.
John Cochrane: I think you’re quite right. Knowing Kevin, he’s very familiar with how the financial and banking system works, so this is not something he’s going to tear through on day one.
I’m glad you brought up the politics of it because we have avoided the elephant in the room, which is the questions of Federal Reserve independence. I should advertise briefly: Our upcoming monetary policy conference will have a big focus on Federal Reserve independence.
But I think Kevin, as we do, has a view that the Fed, in order for the Fed to exercise its monetary policy function independently, it has to stay out of more politically fraught arenas.
So shrinking the range of what you do actually allows you to be more independent within what you do. And I think that’s a large part of the desire for a smaller balance sheet, whatever the economic and financial things we’ve been talking about today, and probably a worthy one.
Darrell Duffie: I agree.
John Cochrane: All right. Darrell, any last words? I think I’ve plumbed you enough for the moment.
Darrell Duffie: I just want to thank you, John, for the opportunity to talk about this today. It’s a really engaging subject, and obviously, you know a lot about it.
John Cochrane: Well, thank you, Darrell, and obviously, you know way more about it, which is why I always love talking to Darrell.
No, he’s not just the polite economist. He’s the person I know in the world who can explain how this stuff works better than anybody else.
I find it so hard to read about financial regulation and how banks do various things, then I just go ask Darrell, and this is all clear. So your talent for distilling this complicated stuff is one that I am always jealous of.
Darrell Duffie: Thank you, John. And you do not deserve the reputation of a grumpy economist.
John Cochrane: I’m not really. Thank you all, and I hope you enjoyed the show.
John H. Cochrane is the Rose-Marie and Jack Anderson Senior Fellow of the Hoover Institution at Stanford University. An economist specializing in financial economics and macroeconomics, he is the author of The Fiscal Theory of the Price Level. He also authors a popular Substack called The Grumpy Economist.
Darrell Duffie is a senior fellow (by courtesy) at the Hoover Institution, the Adams Distinguished Professor of Management and Professor of Finance at Stanford University’s Graduate School of Business, and professor (by courtesy) at Stanford’s Department of Economics.