Raghuram Rajan on the Impact of the Ratcheting Effect of The Fed’s QE Program

Raghuram Rajan is a finance professor at the University of Chicago and leads the Group of 30. Previously he was the chief economist at the IMF and the governor of the Reserve Bank of India. In Raghuram’s first appearance on the show, he discusses his famous 2005 Jackson Hole speech, how he righted the ship on India’s emerging economy, the consequences of zero-sum thinking, the differences between being a policymaker and an academic, the ratcheting effect of QE on the Fed’s balance sheet, and much more.  

Subscribe to David's new Substack: Macroeconomic Policy Nexus

Read the full episode transcript:

This episode was recorded on January 20th, 2025

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected]. 

David Beckworth: Hi Macro Musings listeners, this is your host David Beckworth with some really exciting news. This episode is the very first Macro Musings’ episode with full-length video. Yes, you heard that right, we are now doing the podcast in video as well as audio formats  The full-length videos will be posted on our YouTube Channel @MacroMusingsDavidBeckworth. There is a link to that channel in the show notes, and it would mean the world to me if you subscribed to it and shared it with others. Not only will that channel have full length videos, but it will also have some fun behind the scenes content. Additionally, the full-length video from each episode will also be posted on our X account @Macro_Musings, so make sure you are following that account as well. We are so happy to be giving you even more macro bang for your buck. Now onto the show.

Welcome to Macro Musings, where each week we pull back the curtain and take a closer look at the most important macroeconomic issues of the past, present, and future. I am your host, David Beckworth, a senior research fellow with the Mercatus Center at George Mason University, and I’m glad you decided to join us.

Our guest today is Raghu Rajan. Raghu is a finance professor at the University of Chicago. He is also the former governor of the Reserve Bank of India and formerly was a chief economist at the IMF. He now leads the Group of 30. Raghu is also the co-author of several important papers that have looked at the causes and consequences of the ratcheting effect of the Fed’s QE programs on the Fed’s balance sheet. He joins us today to discuss these papers as well as some of the highlights from his career. Raghu, welcome to the program.

Raghuram Rajan: Thanks for having me.

Raghu’s Career

Beckworth: Let’s talk about your career before we get into your papers. You’ve had an amazing run. You’ve worked at the IMF, the Reserve Bank of India, also at the BIS, and now the Group of 30, and you’re an accomplished academic. In fact, you won the Fischer Black Prize in 2003, which goes to the best finance researcher under the age of 40. A lot of accomplishments, a lot of places we could go, but let’s go to the IMF when you were there as a chief economist from 2003 to 2006. What were the big challenges and issues you faced during that time?

Rajan: Those times seem so gentle relative to today. Big issues. Initially, we were at the tail end of what has come to be known as the Great Moderation. The emerging markets were recovering. This was a period when China was going gangbusters, huge demand for commodities, and emerging markets sell commodities. Many of them were actually recovering from the various crises of the late 1990s. The IMF had a problem, which I didn’t think was a problem.

The problem was this: Nobody wanted to borrow from the IMF. What that meant was the IMF didn’t know where its income was going to come from because it made its money by charging people. I remember some surreal conversations where some of the directors wanted to keep borrowers on, even though they had graduated, so to speak, because it needed the money. Of course, better sense prevailed. Eventually, of course, the world goes back to a situation where countries get into crises, and there are big borrowers, and the IMF has enough money to pay its staff. That was one interesting aspect.

The other interesting aspect was really the difference in attitudes about the financial sector. One part of the fund was celebrating all the developments that were happening. Another, I think, of which I was a member, was getting more worried about the increasing financial risk that seemed to be being absorbed by the financial sector. That was an interesting tension because the Panglossian view, so to speak, from one side, was being attacked by a more alarmist view from the other.

Beckworth: Well, since you’re speaking of that, let’s go to your 2005 Jackson Hole speech, famous now. You gave a warning based on these concerns. I was not there, but I heard the stories that it was not well received. There were people in the audience who really gave you hard-time pushback, but you now stand vindicated about it. Tell us about that experience.

Rajan: I was asked to write a paper for Greenspan’s last Jackson Hole conference. I remember he made some of his more memorable speeches at Jackson Hole. This was the send-off to a person who most people at that time thought was, if not the best central bank leader in history, maybe the second best. That was the nature of the debate. I was asked to speak about the financial sector. Of course, my expectation when I agreed was I’d talk about what wonderful things had happened during his term.

As I started thinking about it, my natural academic instinct started taking over, which is you can’t say the obvious. Now think of what is nonobvious. Then I started looking at the issue of risk. Where had all the risk gone because everything seemed benign? A little bit like today. Markets were very, very buoyant and credit spreads were down. Where is all the risk? Then slowly I understood the risk was tail risk and was being absorbed on bank balance sheets.

Essentially, what banks and institutions like AIG were doing was writing insurance to the rest of the system and collecting massive premia because the tail risks hadn’t hit, but eventually they would. Remember, for example, AIG was writing credit default swaps and claiming nothing had really happened. Nobody had defaulted in a big way. This was just money for free. Of course, there comes a time when you have to pay the piper. I started worrying that did these entities understand how much risk they were taking? At some point, the risk would come to roost and they would have to pay out enormous amounts. That looked very difficult even from the vantage point of 2005.

Of course, some people agreed that that was indeed happening. Some of them asked, so what do we do? The housing is taking off. There are all these other risks in the system. If we pull the plug now, we risk bringing it down like a pack of cards or like a ton of bricks, pick your analogy. There were others who were, as I said, the Panglossian crowd, which there was also in the IMF, who said, everything’s wonderful in this best of all worlds, and this is all financial development taking us to a higher plane.

It turns out that we had to worry about the risk. I still don’t know what we could have done about it at that point, which has led to my greater sense that the best time to nip risk-taking is at the beginning. Yes, we don’t fully know whether it’ll turn out or it won’t, but given the extent to which industrial countries are now extended in terms of debt, in terms of the oncoming liabilities that they’re exposed to, there’s not a huge amount that you can play with at this point. Better take less risk than more risk as a nation. That’s why I have, over time, become more conservative in my advice to central banks. Don’t push things that you can’t control or that you don’t understand. It comes back to bite you.

Beckworth: That ties into our discussion we’ll get to later on the Fed’s balance sheet. One of the questions I always ask is, is there a limit? At some point, can they continue to expand, expand, expand? They may want to save some of that dry powder for when they really need it, when markets truly are dysfunctional.

Going back to this Jackson Hole speech, I just have to ask because you’re being very charitable calling them the Panglossian crowd. They were upset. That’s my reading of it. This was Alan Greenspan’s farewell. You come and crash the party with this somber speech, hey, things may not be so wonderful. I understand Larry Summers was not pleased. Alan Greenspan was not pleased. You made that stand, so that’s very admirable. You were vindicated, ultimately. 

Rajan: Absolutely. Look, yes, some of the people there were mad. Fortunately, I had talked about what I was going to speak about with my boss at the IMF, Anne Krueger. She was very supportive. She thought that was exactly what I was brought into the fund for, to talk about financial risks. She thought I was doing my job. I had at least one backer in the place that mattered, which was the fund. I think the problem, to some extent, in these international settings is, first, you don’t want to upset the apple cart. You don’t want to also get the reputation of being a Cassandra, always screaming things are going to go down.

There’s a third reason, which is, yes, everybody knows the sky is going to fall down at some point, but what can we do about it? What are you suggesting? That’s where, because I was not a central banker then, I was more on the advisory side with the IMF, I could say there’s a risk the sky will fall down in a few years, I don’t know exactly when, but it looks ominous, without suggesting what had to be done. I think 2005 may have been the last possible time we could have put in measures.

See, we had already bought into the Greenspan doctrine, which is you can’t prevent these kinds of episodes when they’re building up because we don’t quite know when it is this and not the rising productivity and all the good things coming together, and so let’s wait till it collapses, then pick up the pieces, just as the Fed did in 2001. The big difference was in 2001, there wasn’t a credit bubble along with the stock market boom and bust.

2007/2008 was much more painful. If you think about the consequences of 2007/2008, it was not just the tremendous amounts of money that had to go into the financial sector to bail it out, as well as the tremendous amount of fiscal spending, but also the tremendous change in the political landscape as a result of the sense that they really did not know how to manage the system. That may be the longer-term damage done from that episode.

Beckworth: That comes to a question I was going to ask you later, but let’s jump in now. Even today, we have this world of zero-sum thinking. It’s evident in our trade policies and our approach to immigration. A lot of issues are zero-sum thinking, us against them, as opposed to positive-sum thinking, the pie is getting larger. I wondered to what extent had we been able to avoid the Great Recession, the global Great Recession? Could some of that populism been minimized or at least dialed back some?

Rajan: You’re exactly right, because in the Great Moderation, the lead-up to the Global Financial Crisis, it certainly seemed a world where there was enough for everybody and more, and that everybody benefited from everybody else doing better. It wasn’t zero-sum game. Remember, this was the time that China was accommodated into the international markets. China was growing gangbusters. Double-digit growth was a given for China at this time. At the same time, other countries were doing well.

As I said, the emerging markets were benefiting. The US was growing strongly. The world, those three or four years before the Global Financial Crisis, was growing at unprecedented rates as a whole. Even the conflicts in Africa were coming down and Africa was able to grow. This was a good time, but it’s precisely at those times that we have to look to the risks and say, what could upset this? We have to be Chicken Little, if you’re anywhere in a position of authority because it’s the good times that breed the bad times. That’s why being paranoid at that point is not a bad idea.

Beckworth: There was a paper that came out a few years after the crisis. The author’s name’s escaped me, but we’ll link to it in the transcript. They looked at financial crisis going back 100 years, cross-country, nice panel data set. They looked at the consequences for populism. They showed, whenever there’s a financial crisis, not just a garden variety recession, but a financial crisis, has this long-lasting effect. How you respond to that can maybe make it even last longer. I do think it’s pivotal, not just to preserve the financial system, but for democracy, to avoid populism. You’re right. Sometimes it pays to be Chicken Little. Now, in your case, you were a serious academic. People took you seriously. You weren’t always saying, hey, the world’s going to end. You called it at the right time. That’s pivotal, so people do follow it. 

Let’s transition from there into your career at the Reserve Bank of India from 2013, 2016. That’s a few years later. You went at a very interesting time because 2013, we got QE3 going on in the US. We got a taper tantrum. We got people talking about currency wars. You’re on the other side of this. You’re actually at a central bank in a large emerging economy. Tell us your perspective on these developments and what else you faced at the RBI during that time.

Rajan: What we had at that time in India was a couple of things. One, the macro wasn’t looking that great. Our fiscal deficit was large. Reforms had come to a halt, partly because the government couldn’t get along with the opposition. The opposition was preventing parliament from functioning. There were corruption scandals also that they were lambasting the government with.

In general, even though it was growing at a reasonable rate—remember, this is post-Global Financial Crisis, all countries have slowed down somewhat—relative to its past, India’s not looking that great. Plus, because there was a lot of exuberance in India also pre-financial crisis, lots of projects had been undertaken which now were stuck. Bank balance sheets were not looking that great. All in all, India was vulnerable.

Then, Ben Bernanke hints at the possibility of ending the period of accommodation that had emerged post-financial crisis, starting with the tapering of QE, and all financial markets go berserk. What we had in India was a period when the exchange rate depreciated about 25%. Now, you may think, what’s the big deal with the exchange rate moving? It turns out that everything is connected in India with the exchange rate, at least at that period, because we were importing a lot of oil. If the rupee depreciated, that meant we spent more on oil. Because oil was heavily subsidized, that meant a bigger outflow form the government budget.

The government budget also looked really shaky, which created yet more depreciation in the currency because there was a sense that, oh, do we really want to hold government bonds in India? Let’s sell them and so on. Inflation was picking up. That also contributes to currency depreciation. It was an unholy mess. Things were happening all at the same time. Investors were getting very pessimistic about India. It was not just India. We were part of the fragile five. That was a term coined by some smart analyst in one of the investment banks. That was being used against us. All sorts of entities were betting against India.

At around that time, I was appointed governor of the Reserve Bank. I had a couple of months before I actually took over to think about what I was actually going to do to remedy the situation. The old-style intervention in the currency, et cetera, hadn’t quite worked because we had run our reserves down and the markets were getting to perceive that. Of course, at some point, they bet against you and then no amount of reserves will work.

What I decided was that it was very important to give an agenda for reform, which would stabilize perceptions about India. This had to be done from the reserve bank, the central bank, because the policy establishment was relatively paralyzed. I had the full support of the prime minister on this. I went to the reserve bank and said, “tell me all the things you’ve been thinking about that we could do tomorrow instead of in due time.”

Also, my friends from the Latin American economy, central bankers from there told me, look, the way to stabilize the exchange rate is to say you will handle inflation. Once you get inflation under control, the exchange rate will stabilize. Don’t talk about the exchange rate, talk about inflation. Great advice from Arminio Fraga, from Agustín Carstens, whom I knew. That’s exactly what I did. We set out an agenda, both of reform, but also on a process of inflation targeting, that we would bring inflation down and then once it got to the target zone, we’d implement inflation targeting. It was a two- or three-year program.

Miraculously, everything worked out. We brought inflation down. The exchange rate started stabilizing when people saw we were serious about inflation. I raised the interest rate a few times to send the message, but not so high that it was unsustainable. Not to the 500% that other countries did. Maybe 75 basis points we raised it. That was enough to send the signal that we were serious.

Of course, we had a little bit of luck. Oil prices fell. Put all that together, inflation came down from double digits to the middle of our zone, which was 2% to 6% by the end of my term. It was around 4%. We moved into an inflation targeting regime, which has served us very well since. We also implemented a huge number of reforms. I like to say, for an academic who’s been thinking about banking and so on, to be in charge of a central bank is like being a kid in a candy shop. There’s so many things you want to do, you have to figure out, I have only so much time. What can I do?

We cleaned up the banks or started the process of cleaning up the banks. We licensed some new banks. We created payment banks, which were getting the telephone companies into banking, which was an attempt to get more people access to the system. We started what is called UPI, the Universal Payment Interface. Last I checked, there was something like 14 billion transactions a month being done on that, which is really a bank-to-bank payment system. Long before FedNow came in, we started it in 2016. FedNow started in 2023.

We did a fair amount in those three years. I thoroughly enjoyed that process because in a central bank, there’s really nobody looking over your shoulder and saying you can’t do this and you can’t do that, which is what happens in government. Even if you’re the prime minister, there’s somebody below you telling you can’t do this and you can’t do that. In the central bank, you’re the boss. So long as you listen to your people and hear what they have to say, there’s a lot you can do. I think we managed quite a bit. By the time we finished, the rupee was quite stable. Inflation was down to 4% from near double digits. There was a different sense. Of course, a new administration also helped, but this was a good experience.

Beckworth: You did quite a bit. You brought inflation down, as you said, from double digits down to 3.5% or so. That was remarkable. You successfully implemented inflation targeting in an emerging market. I remember when I was at the US Treasury Department, I worked in international affairs. I actually worked with the IMF at the time. It was 2003, 2004. When you were at the IMF, I was this lowly civil servant, a grad student just out of school working there. To get the job, they asked a bunch of policy questions, which most of us PhDs had a hard time answering.

One of the questions they asked me was, “Imagine you’re in an emerging market and you want to adopt inflation targeting. Under what circumstances could you successfully do it?” Well, you showed us. You’re the textbook case now, how to do it successfully. That’s the answer to the question that was posed to me many years ago.

Policymaker Versus Academic

You talked about being a policymaker versus being an academic. You had a lot of freedom at a central bank, but you also faced constraints, I’m sure, budget tradeoffs, stuff like that. Looking back, both at the IMF—and again, there’s limits to what you can do at the IMF, but at the Reserve Bank of India, you’ve also advised people. You now lead the Group of 30. What is the difference between being an academic and a policymaker? What are the realities one runs into when you make that transition from being an academic to a policymaker?

Rajan: I think some obvious differences. As an academic, if you’re writing something which has big import, it can have tremendous influence if people pick up on it. As a policymaker, you are in a somewhat narrower place, and there’s only so much which is within your reach, your realm. As I said, I found the RBI had a lot, the reserve bank had a lot under its realm, and so you could do a lot. There’s only so much you can do.

As an academic, you speak to the world. As a policymaker, you sit in a room which has limited spread. At the same time, it is easy for advice to just go over the heads of policymakers. You can yell as much as you want from the rooftops about this problem or that problem, but unless a policymaker listens to you, nothing’s going to be done. They have so much on their minds, it’s rare to find a policymaker who’s going to listen to you, unless it addresses a direct problem they face. The ideal is to have those ideas but also be the policymaker who can implement them. That was my situation in the RBI.

I had chaired a committee on financial sector reforms in India, which was most of the reforms were ignored when the report came out. Now I was in a position to implement those. It turns out that was very useful on that first day when I was looking for reform ideas, because a bunch of those were in the report. Being in the right place at the right time as a policymaker, you can have enormous influence.

As an academic, you hope that somebody in the right place at the right time listens to you, but that’s more a hope. Each place has its benefits. I love the fact that I’ve gone back and forth, that I’ve benefited. I’ve realized the benefit of academia when I go back and remember how constrained I was as a policymaker, including in what you can say. At the same time, sometimes it’s fun to actually do stuff.

Beckworth: One more question about India before we jump into the research papers about QE and its ratcheting effect. You had a book in 2024, I believe, Breaking the Mold: India’s Untraveled Path to Prosperity. India has this great demographic dividend, but a lot of challenges that lay ahead. What do you see as its prospects? What needs to be done, do you think, to truly tap in and capitalize on this?

Rajan: What we were worried about when writing the book was a few things. One, that the world was closing down on manufacturing imports. China was exporting too much. Even before the current Trump administration, protectionism was growing, certainly in the first Trump administration, but the Biden administration didn’t pull back on it. You could see the signs in Europe also. Given that, there was absolutely no room for a China-sized country, which is what India is in terms of population, adding to the manufacturing output the world had to absorb.

Manufacturing-led exports, the surest way for growth in the last 40 years, was simply not a path for India to develop. It could do some, but it couldn’t be a path it could rely on. Interestingly, Indian services have always been the job creator. Not necessarily the IT services, but even the services aimed at urban consumers, such as retail, plumbing, finance, carpentry, construction, all that stuff.

The book was saying India is a big country, we don’t necessarily need to rely on export-led growth. That would be nice. Let’s do it in high-end services—India exports services like consulting services, legal services, as well as IT—but let’s also focus on supplying services to the growing urbanized India: education, healthcare, retail, finance, things like that. If we can do that well, which requires an enormous amount of focus on skilling people, on education, on healthcare, on improving the human capital asset, then you can create plenty of jobs. It can be an alternative path for growth, not just for India, but for other countries in South Asia or in sub-Saharan Africa, or in many parts of Latin America. That was the thesis.

My sense is certainly in India, what we’ve seen is in the time the book was written, service exports have overtaken manufacturing exports. It used to be that services were the classic nontraded good. You could not export them, but today with communication links as they are, services like design, even in healthcare—yoga services. I know a lot of people across the world take Indian yoga lessons because somebody there is willing to provide them at some reasonable price.

The bottom line, I think, is we were offering an alternative path for growth, which would be less worrisome for the West, but which would also require India to upgrade some of its democratic institutions because when you’re buying healthcare from somebody, you want to make sure that the information is not going to get transferred to the government, which uses it to blackmail you. Privacy protections, data protection, all that would be necessary to advance this. That was what the book was about, how we need a democratic rethinking along with the service-led path. It’s being debated in India today, especially with the growth in protectionism that we’ve seen over the last few quarters.

Ratcheting Effect of Quantitative Easing

Beckworth: We look forward to seeing how things unfold in India and the great hope for that country. Let’s transition into a series of papers that you have written with your co-authors. I’ll come to them in a minute, but I want to motivate them by just reminding our listeners that we have just ended quantitative tightening, or QT, here in the US. Once again, the Fed’s balance sheet is at a permanently higher level in absolute terms. Relative to GDP, maybe you could argue it’s going to get back to some level that it was before the pandemic, but still it’s at an elevated level, at least prior to 2008.

My friend, Bill Nelson, who’s at the BPI, you may know him, he’s written a lot about this ratcheting effect as well as the work that you’ve done. He had a recent note, I just want to highlight some points from it. He goes back and he pulls out all the structural estimates of the size of the Fed’s balance sheet and how they have gone up over time. He starts with April 2008, and this is the Board of Governors staff estimates for the size of the balance sheet. Back then, $35 billion. Of course, this is before QE, this is before we went to a floor system, ample reserve system.

Then by November ’16, they came up with the new estimate, $300 billion. November ’18, they came up with another estimate, it’d be $1 trillion. October ’19, $1.4 trillion. Then just now, as we go into this new period, the implicit estimate of where reserves need to be is $3 trillion. It keeps ratcheting up. As a percentage of GDP, it’s now close to 10%. Bill notes that if you throw in currency and the TGA account, the Federal Government’s checking account at the Fed, we’re looking at a $6.5 trillion size of balance sheet. A clear ratcheting effect.

You have a series of papers with your co-authors where you jump into this. I want to mention the first two. I know there’s at least three of them. The first two complement each other. The first one is titled “Liquidity, Liquidity Everywhere, Not a Drop to Use: Why Flooding Banks with Central Bank Reserves May Not Expand Liquidity.” I believe this came out in 2022, at least the working paper one. Then you had a follow-up one which was presented at Jackson Hole, which really landed loud and made a big impact: “Liquidity Dependence, Why Shrinking Central Bank Balance Sheets Is an Uphill Task.” We’ll come to your more recent paper in a minute, but maybe walk us through these two papers, what they show, and why it’s important for Fed policymakers to pay attention.

Rajan: The first paper is with Viral Acharya, who has been a co-author on all these papers. He’s a terrific mind at NYU. He and I basically started with the question, why don’t we see bigger effects of QE? Because after all, the notion was that when the central bank expands its balance sheet by buying long-term securities from financial sector balance sheets, they will substitute by engaging in investment in more private sector long-term assets, maybe more long-term loans, maybe bonds issued by financial firms or industrial firms, and that should expedite growth.

Portfolio substitution is what this theory is called. The idea was this should be really beneficial, especially when the short-term interest rate is at zero and you can’t add more stimulus by cutting rates. Maybe you can add stimulus by expanding the central bank’s balance sheet. We know that for a long period, yes, the situation didn’t deteriorate, but it wasn’t as if the economy was growing gangbusters.

Our question was what really might be happening to offset some of these, and is it going to the right places, so to speak? What we concluded was that even as the central bank expanded its balance sheet and put perhaps the most liquid asset in the world, which is central bank reserves, on bank balance sheets, it actually might be increasing liquidity risk, certainly at times when the central bank tried to shrink its balance sheet. We came at this by walking through the various actions that banks would take to hold those reserves that the central bank was putting out.

On the one hand, as soon as a commercial bank has more central bank reserves on its asset side, it knows that these are paying interest. They were paying interest through much of QE, but they have to finance them. If you’re financing very liquid assets, perhaps the cheapest way of financing them is by issuing wholesale demand deposits because those, you’re giving them the promise of liquidity on demand.

Given that they have that moneyness, they’re willing to demand a relatively low interest rate. The spread between the reserves that you hold and the financing of those reserves through wholesale demand deposits is relatively small. You would not want to finance them with time deposits because time deposits are more costly. Those are buying you time, and that’s not particularly useful when you have such a liquid balance sheet.

The first prediction of that paper—we hadn’t looked at the data, it was just theorizing—was that you would see a significant expansion in demand deposits, wholesale demand deposits in financing bank balance sheets when they held more reserves. That was one. The second was they’re not satisfied with just that. They probably run down their time deposits because that’s longer duration. They’re also going to say, we have plenty of liquidity on our hands. What do we do with it? Why don’t we fund liquidity needs of firms and hedge funds and so on? Let’s offer credit lines. Let’s offer them liquidity on demand because we are so flushed with liquidity.

The point of all this was to say that when commercial banks get a lot of liquidity in the form of reserves on their assets, they want to sell it down very quickly to make money off it. That selling down of that liquidity by issuing demand deposits, which in the final analysis will require liquidity if they come for their money, or issuing credit lines to firms, which again will require liquidity if they come for their money, basically exposes the bank not to individual demands for liquidity that you can handle because you’ve got plenty of reserves, but to systemic demands for liquidity. When everybody’s running for their money, you are in deep trouble because you’ve written so many promises through your balance sheet.

What we were arguing is, in that kind of environment, you might actually see that the willingness to take long-term investments might be more limited because liquidity risk has gone up in the system, systematic liquidity risks. As somebody investing in five- or 10-year projects, are you certain that you will not experience a shortfall at some point before the project is finished and that creates additional financing needs, et cetera?

Bottom line was it wasn’t such a clear play when the Fed expanded its balance sheet for firms to go out and invest in real projects. Instead, what we thought was this might actually lead to more financial speculation as financial firms saw that they could get liquidity easily and entered various speculative arrangements. We can talk about that in a bit.

Beckworth: Just to summarize. Two big insights, at least, I got from those papers. One is that QE dramatically affects the liability structure of commercial banks, that they respond, not just mechanically because they’re getting reserves, but behaviorally. They’re adjusting, and they’re getting rid of, as you said, time deposits and credit lines, demand deposits, runnable funds, liquidity. What was really profound, the insight that you guys brought up, was this idea of net aggregate liquidity. Yes, QE expands liquidity on one hand, but then all these claims on it on the bank balance sheet.

You can’t just look at aggregate reserves. You’ve got to take that, subtract out the claims on it via the banks. The banks themselves are looking at those reserves as a backstop for all the additional liabilities they’ve issued. It’s a super fascinating point that you’re not just throwing liquidity in the system, you’re altering the liability structure of the banking system. The second point, you’re undermining the very purpose of QE. QE is to lower long-term yields, stimulate long-term investment, but banks, their response to it is to shrink the duration on the asset side. It’s a really fascinating point. Therefore, you get to this asymmetry that the Fed wants to shrink and get rid of the reserves. Banks can’t keep up because now we’re used to all this liquidity. It’s so fascinating.

In your Jackson Hole paper—again, it’s titled “Liquidity Dependence: Why Shrinking Central Bank Balance Sheets Is an Uphill Task”—you add some extra analysis to this, including the spread between the effective funds rate and IOR. You also tie it into some real-world experiences. The ‘29 repo spike and the actual dash for cash. Tell us about those applications.

Rajan: Viral was asked by the Kansas City Fed people to write a paper. He said, “Why don’t we test our theories? We have no idea what actually happened. Let’s go look at the data.” Surprise, surprise, the data were really very cooperative. They seemed to suggest precisely what we had talked about, that demand deposits, especially wholesale uninsured demand deposits, were going up, especially in the banks.

There’s a time series demand deposits going up through the system, but what we looked at also was a panel. We’re looking at the cross-section. We found that the banks that got the most reserves also were issuing the most demand deposits, but also running down the time deposits during the period of QE. Now, as you pointed out, there’s a mechanical reason for this. When an insurance company sells its bonds to the Fed, it acquires essentially a payment from the Fed, which it deposits in its bank. From the bank’s perspective, the bank’s reserves go up because that’s what the deposit into the bank means, but its deposits also go up.

Banks can readjust. They don’t readjust. It turns out this thing is going secularly up during the whole process of QE, but the adjustment is through the time deposits. They’re shrinking their time deposits and expanding their demand deposits, even as their balance sheets expand, but their credit lines are also increasing substantially. Then we looked at those periods of turmoil. The September 2019 was the first when repo rates go haywire, and the Fed comes in with QE again, but then we had the pandemic problem in March 2020, when again everything starts exploding, and the Fed comes in again with massive amounts of QE.

We looked at those episodes and asked whether the banks that were most exposed by writing these claims on liquidity or selling these claims on liquidity to all and sundry, whether they were the ones that were most hurt, and we did indeed find that. Liquidity risk was a real thing. What that also meant was that because the banks were stretched, when you withdrew liquidity, it didn’t come back as much. There were people who said that QT, which is withdrawing liquidity from the system, would be as boring as watching paint dry. It turns out it wasn’t because you did a little bit of QT, and the system had gotten used to this larger quantity of liquidity, and it backed up quickly.

That was what we saw in September 2019. The Fed resumed QE before the pandemic crisis. Then it went gangbusters on QE. More recently, what you have seen is that at a much larger balance sheet, the Fed has stopped reducing its balance sheet. QT is ended, but not just that. In the last FOMC, they actually resumed expanding their balance sheet, not through long-term bonds, so they were very careful to say this is not QE. They were supplying more reserves into the system, which effectively was expanding the Fed balance sheet again, this time by buying Treasuries.

My sense is we’ve not departed from this model of ever-expanding Fed balance sheets. The worrisome part is this kind of intervention. I think you asked the right question, which is, is there a limit to this? Some people would say, well, the Fed can keep coming in with expanding its balance sheet, so long as people are willing to hold Fed liabilities, it’s not a worry. Then ask the question, why is this not something every emerging market can do? Why does it think twice about expanding the central bank’s balance sheet?

Because there’s less trust that everything will come together, that the central bank will not be expanding its balance sheet at a time when there’s big concerns about inflation, there’s big concerns about the federal deficit. What is very important is you should have confidence in the finances of the government and the central bank combined. You should have confidence about inflation. If you don’t have confidence on either of these, I think central bank balance sheet expansion becomes much more fraught. That will be the time when you will say, “Maybe I don’t want to hold more US central bank paper or US government paper.”

Beckworth: There are a lot of questions that this brings up and challenges. You mentioned some of them already. I also worry about the Fed’s footprint in the financial system. We’ve completely hollowed out the federal funds market. It’s a shell of what it used to be. There’s no overnight, unsecured lending between banks. In my mind, that’s a missing market. There’s no price discovery, no counterparty risk. If you look at many other central banks, they’re actually trying to shrink their balance sheets and go back to something smaller to resurrect that interbank market. For them, they want to see some price discovery, something going on.

Also, the Fed effectively becomes the largest-ever fixed-income hedge fund when it has these big balance sheets like this. I’m not sure if that’s something we want the Fed to be doing.

Rajan: Absolutely. There are a bunch of worries, but there are worries because the ratcheting up of the Fed balance sheet—remember when the Fed started increasing its balance sheet, government finances were really in much, much better shape because they’ve just gotten worse over the last 15 years. The government balance sheet in 2008, when you said the Fed started expanding, which is just during the middle of the financial crisis, was still quite good. Importantly, inflation was not a problem.

Today, inflation is a problem. For five years, the Fed hasn’t brought inflation below its target. Government financing is a problem. When Secretary Bessent came in, he complained about the fact that the previous Treasury had been issuing only short-term bonds because they couldn’t really get people to buy the longer-term or were really worried. But what has the Treasury done? It has again emphasized issuing the short-term. What is the Fed buying? It is buying the short-term now.

When you think about it, when the Fed started expanding its balance sheet, government finances were not a problem, inflation was not a problem. Balance sheet expansion was a new instrument. Today, I would say it looks a lot more like the instrument the emerging markets got away from, which is financing the government. Emerging markets used to finance the government by printing money, essentially giving the government reserves to spend. With the Fed balance sheet so large, you mentioned the size of reserves. Look at the size of government Treasury securities. It was about $800 billion in the early days of QE. Now it’s $4.5 trillion. That’s 15% of GDP, maybe more, that is holding of government bonds. It’s just announced it’s going to buy more Treasuries. 

That’s serious government financing. Today—we’re talking on January 20th—Japan just told us eventually the bond markets wake up to unsustainable financing. Now I don’t know how this will play out for Japan, but you don’t want to be in that situation as the most powerful economy in the world, which is why I think what you started then—and again going back to the limits of balance sheet expansion—at some point it looks a lot more like government financing and you start worrying.

The other thing that you worry about is whether the markets have now internalized a Fed liquidity put. It’s that whenever liquidity runs short, the Fed will come by. Then all sorts of new speculative schemes become really au courant. They work because whenever things dry up, the Fed will come in. Take the bond basis trade, which is now trillions of dollars, it’s a form of financing for the US government, but it’s based on hedge funds who are dependent on the Fed continuing to provide liquidity.

We can go through the details, but it’s really a worrisome form of financing. It’s not just Fed financing. Is the Fed also supporting indirectly other forms of financing which are not the most stable? You have to worry that this cannot be a long-term solution.

Beckworth: I’ve advocated for shrinking the Fed’s balance sheet, maybe going to a demand-driven operating system, either a scarce reserve corridor or even a ceiling system. My colleague at the Mercatus Center, Tom Hoenig, who you know, he says, “David, David. Silly you, David. You know, at the end of the day, the Fed has to keep the US government solvent. You know primary deficits are forecasted as far as the eye can see. You may want to go back to smaller Fed balance sheet, but unless we get our fiscal house in order, the Fed’s going to be under a lot of pressure going forward,” just like you’ve said.

When you look at all of the expectations on the Fed, I feel sorry for them because they do have a lot on their plate. Even things like making the standing repo facility more robust, tweaking the supplemental leverage ratio, even stablecoins, they’re all motivated to some extent by this increased supply of debt. How can the Fed keep control of interest rates? How can the Fed keep control of Treasury market rates? It’s got to tweak these tools to accommodate increased growth and debt. I completely agree with you. There’s a bigger, more fundamental problem at work here. I often throw out these proposals, and the reality check is there’s a bigger problem. 

Rajan: Let me just respond to that, what you said. I think that’s exactly right. Great respect for Tom Hoenig. He was a wonderful doyen of the Kansas City Fed when he was there and, of course, his career since. One of the questions that arise in this environment is what if the rating from the government doesn’t seize? It needs that financing, but the financing gets riskier and riskier because it’s going to these entities. With the hedge funds’ financing, their holdings of Treasuries with overnight borrowing is not the most stable in the world and is dependent on constant liquidity promises from the Fed, but does the Fed have the duty to accommodate all liquidity needs?

Does it have to be in an environment of ample reserves? If it commits to that, does that encourage more of this speculative financing? Is it part of the problem as well as providing part of the solution? I think that’s a huge issue that will come to the fore going forward if the rating doesn’t stop. I’m completely with Tom when he says that, “Well, what choice do you have?” But you would want to be preaching from the rooftops, “Hey, this is not safe financing. We need to stop this because otherwise it’ll blow up. Even we will not be able to rescue the system.”

Beckworth: This goes back to your papers. You talk about liquidity dependence, which is in part based on the belief the Fed will stop QT, will step in and help. One other observation to tie this in, stable coins is a big thing right now, and there’s a lot of promise, a lot of challenges as well with stablecoins. One interesting development about it that I hadn’t thought about until I read Jim Klaus, used to be at the Board of Governors, he’s now at a think tank, he raised this interesting point that, let’s say, in the limit that dollar-based stable coins displaced most of US physical currency. Again, this may be a far-fetched scenario, but just for the sake of argument, that currency is an important source of funding for the Fed. It’s the one stable source of funding for the Fed. 

If the Fed were to lose its currency franchise, so to speak, then it’s only left with these interest-earning liabilities, which, depending if they do QE or not, there’s going to be more operating losses on the Fed’s balance sheet going forward or on its income going forward, and that’s going to politicize it. Even though it may not be a big deal to the implementation of monetary policy, just the appearance of losses is going to, I think, undermine its independence. I do think there’s a lot of challenges facing the Fed going forward.

Rajan: No, you’re absolutely right. Actually, this is a situation that a bunch of developing countries, central banks already face, that nobody wants to hold the domestic currency because they’ve been inflating it away, and so everybody’s effectively dollarizing. The way to dollarize today is not to hold those bills, but to move into cryptos. Earlier, it was volatile cryptos like Bitcoin. Now, it is the stablecoins, which at least will hold their value, and for a business person, that’s really important because it’s transactions media that they want, not some speculative medium.

In a number of countries, you’re seeing more dollarization happen. When you talk to central bank governors from these countries, that’s the great worry because, as you said, their revenue stream just goes out of the window because they really can’t print currency that people want to hold, and somebody who’s providing the stablecoin is making the money there.

Beckworth: Yes, seigniorage that would have gone to the Fed is going to the private sector. Let me summarize your last article here and tie it into a final question about policy implications. We’ve been touching on them, but you have another article. It’s really fantastic. This one’s titled “Liquidity Dependence and the Waxing and Waning of Central Bank Balance Sheets.” Here, you even go more granular into the data. You talk about how this ties into the March 2023 banking turmoil, how it’s small, medium-sized banks, it’s uninsured deposits.

It’s an interesting story we don’t have time for to get into, but I encourage listeners to check it out, we’ll provide a link to it. I’ll also note that I put this up on a Substack recently, Raghu, and interestingly enough, a bunch of people sent me some other articles that build off of this. There’s an interesting article from the Riksbank. They looked at what QE does to their banks. There’s some folks at the Fed have done some interesting work as well, but just bring this full circle here. What can we do going forward? What are the implications, and what can we do if you’re a Fed official?

You want to do QE, let’s say, if we were at the zero lower bound, there’s nothing else you can do. But you also want to be in a place where you can unwind it fully, and so that you’re not creating this perpetual growth in the Fed’s balance sheet. Do you have any insights and recommendations for the Fed?

Rajan: By the way, there’s one more paper. The paper you just cited is a development of the Jackson Hole, but there’s another theory paper that Viral and I wrote, which has the title, “When Is Less More? Bank Arrangements for Liquidity vs Central Bank Support.” This is with Zhi Quan Shu, a really smart PhD student at NYU.

The idea there—and that’s the reason I brought it up, because it’s an answer to your question—is central banks like to intervene. Liquidity support was always the reason they were set up. It wasn’t so much inflation targeting or currency management. It was to resolve the various crises that were permeating through the financial system in the late 19th century, early 20th century. The problem, however, is that pricing that intervention is really hard at a price that reflects the true value the central bank brings.

What we argue in the paper is that intervention is always underpriced. Because it is underpriced, the system takes too much liquidity risk. You can see it. Despite the advent of central banks, despite the advent of deposit insurance, crises have actually grown bigger and more damaging in the post-war world than in the pre-war world, where they intervened less often, partly because we were on the gold standard. Their balance sheets weren’t as elastic at that time.

The point we’re trying to make is we’re in a situation where the central banks can’t commit to be really tough and always come in. There’s political pressure also on them to come in. As you said, as Tom Hoenig said, it wouldn’t be possible politically to say, “Let the chips fall where they will, let the system implode.” You have to come in. That act of coming in creates the risk-taking, the so-called moral hazard that we don’t pay much attention to. This cycle can get bigger and bigger until you can’t come in anymore, at that point.

What do we do? There’s a very intrusive supervisory system that you could get into, which does the central bank have the knowledge, the information to do it effectively? I don’t think so. That’s one possibility. Thou shall not issue more than this amount of liquidity, liabilities. Thou shall not fund dozen such hedge funds more than you need to. All these things you could do, but it requires a level of intrusion.

The other way is to start thinking about how you charge more for the times that you have to come in. It’s not just the fees that you charge. There’s a paper in the Journal of Finance. I don’t want to say I advocate it, but it makes you think, which says, in any crisis, let the first guy fail. Bail out the second guy. Let the first guy fail, so there is some sense amongst the people taking risks that bad things can happen. Could you let a hedge fund go? After that, bail out everybody.

Now, given what happened with Lehman, I don’t think there’s any appetite to try an experiment like this. It may be that we need some way to convey the lesson that the bailout will not always be available. Otherwise, things can just get bigger and bigger. What that is, other than intrusive supervision and regulation, is not clear to me because just charging a fee is not going to be enough. Plus, it’s very hard to charge a fee when you’re intervening to rescue somebody in the midst of a crisis, but we never charge a fee later also.

Certainly, the experience during the Global Financial Crisis is no banker really paid a big price for getting their bank into trouble. Maybe bailing out the institution but not bailing out the people who led the institution may be part of the answer, but this is a tough problem. As I said, this is where central banks say, what’s your answer? It’s not clear there’s an easy answer.

Beckworth: Let me give my answer in closing, and you can respond to it. It’s not a panacea by any means, but just something for the Fed to consider. I got this from discussing on a previous podcast with an official from the Riksbank. What they do is they offer a term deposit, they call them CDs, at the targeted policy rate. Then the deposit rate is like 10 basis points below it. When you do QE, yes, initially you have reserves, but you have an incentive to turn those into longer-term deposits at the Fed, which then could minimize the bank’s incentives to create all this runnable liquidity on its balance sheet.

It also is a demand-driven system. The banks themselves determine how much reserves, but they’re incentivized to go into term deposits, and that might fix some of the problems. I’m just curious. Do you think that would be just a small fix?

Rajan: I think we’ve got to look for all the answers we can. I think quantitative constraints on how much liquidity risk they can take, which this seems like you’re moving them into more term deposits by limiting the demandable stuff they can issue, could be part of the answer. What we found on the pricing of liquidity in difficult times is actually the right price to charge is what the market’s willing to pay. The real deep discounted fire sale price for assets, implying a very high rate of return if things come back, that’s the charge the Fed should apply to get things right. Nobody’s going to be willing to accept that kind of charge, but that’s what the economic charge is to get people to do the right thing.

If we can’t do that, we’ve got to find every other way to constrain the kind of risks before they actually occur. I think this is one of the possible solutions that we need to put out there. 

Beckworth: With that, our time is up. Our guest today has been Raghu Rajan. Thank you so much for joining us today.

Rajan: Thank you, David.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. Dive deeper into our research at mercatus.org/monetarypolicy. You can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. If you like this podcast, please consider giving us a rating and leaving a review. This helps other thoughtful people like you find the show. Find me on Twitter @DavidBeckworth, and follow the show @Macro_Musings.