Hanno Lustig on Dollar Dominance, Dollar Safety, and the Global Financial Cycle

Hanno Lustig is a professor of finance at Stanford University, and a senior fellow at the Stanford Institute for Economic Policy Research. Hanno joins David on Macro Musings to discuss his work on dollar safety, safe assets, convenience yields, and more. More specifically, Hanno and David discuss the dollar dominance in global financial markets, how the US’s status as the world’s safe asset provider reinforces its exorbitant privilege in money markets, whether the countercyclical demand for safe assets can help explain why US inflation has been so low this past decade, how years of low interest rate policy might have contributed to the growing wealth gap, and much more.

Read the full episode transcript:

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: Hanno, welcome to the show.

Hanno Lustig: Well, thanks, David. Thanks for inviting me and I'm really excited to be here.

Beckworth: Yeah, it's great to have you on. We were kind of joking before the show got started how I've had a number of people from Stanford on the show over the years. Just a few names, John Cochrane, Darryl Duffy, Nick Bloom, Matteo Maggiori. So we're going to have to build a wing out there at Stan for the Macro Musing West Coast Division Wing, and just have a studio set up. You guys can just step in and record shows. So speaking of Matteo Maggiori, you recently did a piece on him. He won some award, and you wrote of the Journal of Finance. So tell us about that.

Lustig: That's right. So Matteo won the prestigious Fischer Black award, which is awarded to sort of the best young financial economists. I think the cutoff is 40. And it's awarded biannually, so it's kind of a big deal. And Matteo won the 2021 award. And what I got to do, and this was really a fun task, is to sort of write a summary of his work to date. It's also quite humbling because he's a very young guy, and when you look at everything he's accomplished in a very short period of time. I think he graduated from graduate school at Berkeley probably 10 years ago. It's really quite impressive. So it was also one of these things where you do the write-up and then you say, "Okay, now I better go back to work."

Beckworth: He's inspiring for sure. He's done so much. Yeah, so it's great to have you on the show and to continue this conversation. We'll provide a link to your article in the Journal of Finance that you wrote up on Matteo. We'll provide a link to the show we had with him as well in the show notes. But you're doing interesting work yourself. A lot of interesting work. In fact, our paths almost crossed at Jackson Hole 2019. Apparently we were in the same room, listening to the same papers. I didn't meet you in person. I did meet your co-author, Arvind Krishnamurthy, during that time. In fact, we were on a walk, that little walk. Did you go on that walk through the woods in the afternoon?

Lustig: Oh yes, I did.

Beckworth: Okay, so I must have walked right by you. But in any event, I was actually talking to your co-author, Arvind, and interestingly, he made this comment that really resonated with me. And it was about Mark Carney's speech. He had just given this speech, if you remember, about the synthetic hegemonic currency idea. And I've had Mark Carney on the show, and we talked about this. But Arvind made a great observation that really just stuck with me, and that is he could not see this idea working given the scale of dollar assets out there in the world. There's just so much, in order to scale the synthetic hegemonic currency, it would just take a lot of work, a lot of commitment, a lot of leverage on balance sheets at central banks. And he didn't see it as plausible, so it really struck me. And I started to go look at the numbers, and he was right. There's just so many dollar denominated assets around the world, in the US. There's kind of a first-mover advantage of sorts. But any thoughts on that?

Synthetic Hegemonic Currency

Lustig: Yes. That's a very good point. If I remember correctly, Carney was sort of effectively suggesting that the dominance of the dollar, especially in debt markets, is not healthy. And he was saying maybe we should have like potentially a digital alternative. And I think that's certainly an interesting idea, but I would tend to agree with Arvind, that it's hard to see how this could work in practice and there would be huge issues associated with scaling it up.

Lustig: Arvind has an interesting paper that you recently cited on Twitter with Zhiguo He and Konstantin Milbradt where he sort of really tries to think carefully about what makes a reserve currency, what generates safe asset demand. And there is sort of a self-fulfilling aspect to it. As more and more investors believe that the dollar is the right reserve currency, and that the US should be the provider of safe assets, that in turn makes US safe assets safer. And that, of course, explains why it is very hard to dislodge the US as the safe asset provider.

Lustig: Now, of course, if there would be a big shock, that could happen. And it has happened in the past, right? If you take a broader historical perspective and you look at the UK, arguably before the first World War, I think most economic historians would say they were in the same position the US is now. And of course, that came to an end after the first World War, for a variety of reasons. So I guess the short answer is I agree with you. I think it's very hard to see how a digital alternative, a synthetic alternative, would pose an immediate threat to the dominance of the dollar. On the other hand, if you take a broader historical perspective, we have seen that eventually, these things do come to an end. And it's important to think about that, especially for the US, I think.

Beckworth: Yeah. That's fair, completely fair. And to be fair to Mark Carney as well, because when I asked him about it on the show, he said his motivation wasn't to come up with this radical idea. That was kind of like the telling of his talk. His real motivation was what you mentioned, was the fact that all this dollar debt is disruptive around the world, the global financial cycle we'll talk about. And then also, about the time, I guess Facebook had announced the Libra. And so he was kind of like, "Let's get ahead of Libra. Let's also address this cost of having this dominant currency."

Beckworth: Well let's use that to segue into your work. So you have a lot of articles I want to cover. I want to start with one that speaks to exactly these issues we've just been discussing. And this is a paper you co-authored. And it's called “Dollar Safety and the Global Financial Cycle.” And you have six facts that you and your co-authors mention up front, and that motivates you to come up with a model to explain it. So maybe we could kind of just skim over the six facts, just quickly mention them. And if you want to add something to explain why they're important, please do. So the first one is dollar funding advantage. So any thoughts there?

Dollar Safety and the Global Financial Cycle

Lustig: Right. So if you look at the returns on dollar bonds, what you notice is when you do sort of a careful apples to apples comparison is that they're low. So what we've done in this paper and in other related work is we, for example, take a US treasury and then compare the yields on US treasuries to the yield on a German bond, where we hedge to currency risk. So we manufacture out of that German bund a synthetic treasury. And what you notice almost invariably is that the yield on the synthetic treasury that we construct is higher than the yield on the real thing. People want the real thing. And what's interesting about this is ... Perhaps it's not too surprising, but what's interesting about this is that the gap occasionally is quite large. So, there is this sense that there is a dollar funding advantage. This is not just true for treasuries, by the way, but it's also true if you look at sort of close substitutes for treasuries. For example, if you were to look at triple-A rated US corporate bonds, and you did the same exercise, there's an economist by the name of Gordon Liao at the Fed, which is who's done exactly this in a very nice paper in the JFE. He finds that you see similar gaps for highly rated corporate bonds as well. So there's this broad sense that there's a dollar funding advantage.

What we've done in this paper and in other related work is we, for example, take a US treasury and then compare the yields on US treasuries to the yield on a German bond, where we hedge to currency risk. So we manufacture out of that German bund a synthetic treasury. And what you notice almost invariably is that the yield on the synthetic treasury that we construct is higher than the yield on the real thing...what's interesting about this is that the gap occasionally is quite large. So, there is this sense that there is a dollar funding advantage.

Beckworth: All right. Second fact, dollar debt dominance. We spoke about it, but if you want to add anything to it, please do.

Lustig: Yeah, so we've spoken about this. So the dollar is really dominant in debt markets. And when the Euro was created, there was a sense initially that perhaps the Euro would be able to contest the dominance of the dollar. But that hasn't really happened for a variety of reasons. And so one quick footnote that I'd like to add here, David, is sort of the following.

Lustig: So my colleague Matteo Maggiori, who you've had on this show, has a very nice paper with Schreger and Neiman where they actually show that bond investors around the world have a home currency bias. So it seems like they're extremely reluctant to buy corporate bonds that are not denominated in their own currency. So it's not a home bias per se. It's not about who issues the bonds and whether they live in your country. But it's about the currency they were issued in. And the only exception to this is the dollar. So it's very stark. Foreigners aren't willing to buy bond denominated in currencies other than their home currency, except for dollar denominated bonds. So I think that's kind of a neat fact to keep in mind with respect to dollar debt dominance.

Beckworth: Yeah. Very interesting. And we will come back to this as we work through all your research. All right, your next one is flight to dollar safety. And I know you and Arvind have had papers on this. And I believe your 2019 paper touched on this as well at Jackson Hole.

Lustig: That's right. We have a 2019 paper, and then we've also done some joint work with a bright young economist who's at Northwestern, Zhengyang Jiang, who was also our student at Stanford. And what we've noticed is that whenever there is sort of volatility in financial markets, there is a flight to the safety of the dollar, which actually causes the dollar to appreciate. That's kind of an important stylized fact. And it is one that in a second I'll explain when we talk about the model. The model will kind of replicate that. And other models kind of struggle to replicate that fact. So it's kind of a key fact, the third one, I'd say.

Beckworth: Yeah. And I'm going to just tie that into your next observation, which is the global financial cycle. So I think a lot of people know about Helene Rey's work on this. When the Fed tightens, when the dollar, independently, the Fed changes, it affects the global economy because of all these dollar denominated assets out there. But I guess the question I had is this previous point about this flight to dollar safety, if I recall correctly, you guys, you and Arvind and your co-author call it the global dollar cycle, which kind of builds upon this notion of a global financial cycle. Is that fair?

Lustig: That's right. We were obviously building on Helene's work here. And what we're saying is part of the reason you see this co-movement across a whole bunch of financial variables, across borders, across countries, is the role of the dollar, and the fact that because there is this dollar funding advantage, what's going to happen is that corporations, governments around the world are going to seek to take advantage of the dollar funding advantage by issuing that in dollars. And that's going to be sort of the propagation mechanism that generates our version of the global financial cycle, which we think is partly also a dollar cycle.

Beckworth: Yeah, it's such a powerful thought, or amazing thought that when times get bad, the dollar gets stronger.

Lustig: Right.

Beckworth: Like 2008, you would think the dollar might have gotten weaker because US was the epicenter of the housing boom and bust cycle, but instead, the dollar got stronger, and just reinforces this tendency that you outlined. Isn't that striking to you?

Lustig: It's striking. And it actually ... I think you're absolutely right. It's sort of striking and surprising. And it took us awhile to figure out what the logic is, but once you approach this sort of from an asset pricing perspective, it makes a ton of sense because the dollar exchange rate is going to price in all the future convenience yields that foreign investors perceive to get when they hold dollar assets. And so when bad things happen in the world, and these convenience yields go up, and the forecasts of future convenience yields go up, that's immediately priced into the dollar. And that's what's driving it. And as a result then the expected returns on holding dollars going forward are lower. The dollar instantaneously appreciates. So it's sort of the beauty of standard asset pricing logic. The dollar is sort of ... You can think of it as a security, loosely speaking, that gives you a cashflow that consists of convenience yields on holding safe dollar assets.

Beckworth: Yeah, I know, very convincing story and great angle. Let me throw one other observation out there and see how you feel about this. So I wrote some popular pieces over the past year where I made the case that the Fed's interventions in dollar funding markets has only reinforced the global dollar cycle. In fact, I linked to one of your papers in at least one of my pieces I wrote. And I argue look, the Fed effectively opened up branch offices across all the major central banks of the world. It backstopped dollar funding markets. Therefore, going forward, on the margin, investors are going to hold more dollars in their portfolios because they know it's safe. They know the Fed will backstop this global dollar funding market. So this global dollar cycle arguably was reinforced by what the Fed did last year. Is that a fair interpretation?

The dollar exchange rate is going to price in all the future convenience yields that foreign investors perceive to get when they hold dollar assets. And so when bad things happen in the world, and these convenience yields go up, and the forecasts of future convenience yields go up, that's immediately priced into the dollar. And that's what's driving it. And as a result then the expected returns on holding dollars going forward are lower. The dollar instantaneously appreciates. So it's sort of the beauty of standard asset pricing logic.

Lustig: Yeah, I think that's an interesting point, David. I mean, this is a bit outside of the model and the analysis we've done. But it certainly sounds right to me that what the Fed effectively has done through these swap lines that they've made available is to sort of cement their role as the world's safe asset supplier, I think, and sort of signaled to the market, "Well, if there are dollar shortages going forward, in a crisis, you can count on us to alleviate those." So I think that's right. I think that is probably the right way to think about this.

Beckworth: Okay, last two facts. US exorbitant privilege, dollar risk factor.

Lustig: Right, so there's a huge literature on the exorbitant privilege of the US and the key players here are Pierre-Olivier Gourinchas and Helene Rey, who's written numerous papers documenting the facts, and also offering what I think is a compelling explanation. They sort of made the point, look, if you look at the US balance sheet relative to the rest of the world, it's really remarkable, because the US almost looks like a highly levered hedge fund. They have issued a whole bunch of risk-free assets, and then they invest all of this in risky assets abroad, in equities. That's been, broadly speaking, the pattern from the 1950s, 60s, all the way to the 2000s. And then the question is well why is that? Why is that the equilibrium outcome? That's been, I think, one of the key questions that people in international finance have been wrestling with.

Lustig: And Pierre-Olivier and Helene have sort of offered one explanation, which is sort of a ... Maybe we can go into detail later, but they've offered sort of a risk sharing explanation. And Matteo Maggiori actually also has a paper in that vein that he may have talked about with you. And we're going to offer a slightly different, I think partly complementary explanation, which is more about just being the world's safe asset supplier, broader than thinking about the mechanics of risk sharing. So we can go into the details a bit more.

Beckworth: Yeah, let's get into your model and your explanation for the facts we just listed.

If you look at the US balance sheet relative to the rest of the world, it's really remarkable, because the US almost looks like a highly levered hedge fund. They have issued a whole bunch of risk-free assets, and then they invest all of this in risky assets abroad, in equities. That's been, broadly speaking, the pattern from the 1950s, 60s, all the way to the 2000s. And then the question is...Why is that the equilibrium outcome? That's been, I think, one of the key questions that people in international finance have been wrestling with.

Lustig: Okay, great. So I think it's helpful to sort of contrast our story with what I think up to now has been sort of the standard story, which builds on the work of Helene and Pierre-Olivier, and also Matteo and others. Their view was, "Look, this is basically risk sharing." A typical US investor is either less risk averse than a foreign investor, or maybe the financial sector is more sophisticated. That's Matteo's story. And so it's efficient to just allocate more aggregate risk to US investors, or to US financial intermediaries. So it's a risk sharing arrangement. And then what you should see is that when there's a bad shock, there's sort of a transfer of wealth to the rest of the world. So exorbitant privilege in that view means the US is the world's insurance provider. We're willing to take on a little bit more risk than the rest of the world, and they pay us a premium for that. And that's why we get to run current account deficits. We're collecting an insurance premium.

Lustig: I think that's part of the story. Our view is a little bit different. We're saying well it's not just a risk sharing arrangement, it's also just a fact that the US has this monopoly on providing safe assets to the rest of the world because of the unique position of the dollar, and that generates seigniorage revenue. And when bad things happen in the world, actually that exorbitant privilege only increases because there's more demand for safe assets. And as we just talked about, that pushes up these convenience yields. So it pushes up the gap between the yield on a treasury, and a synthetic treasury that you construct from a German or an Italian bond.

Lustig: And that causes the dollar to appreciate in turn. So that's kind of the key fact. Our mechanism offers sort of a simple explanation for why when bad things happen in the world, why the dollar instantaneously appreciates. This insurance story, that's the one part the insurance story struggles with. What has to happen for goods markets to clear is you have to have the dollar actually depreciate. Why is that? Well, it's because there's a home bias in consumption. You just made foreigners richer, but they prefer to consume their own stuff.

Lustig: We've got all this US stuff lying around, so the relative price of US stuff has to go down. You're going to have to have the dollar depreciate. And that seems slightly counterfactual. So that's the tension that we, I think, address by having the safe asset mechanism kick in. And then to wrap things up, what is exorbitant privilege here? Well, it's not so much the insurance angle. But it's more just the ability to sort of manufacture these IOUs that we give to the rest of the world, and they're willing to pay a premium. And we collect seigniorage revenue, essentially, as Americans, on all of this.

Beckworth: Hanno, that's a great explanation of those facts that we listed. Very fascinating. We'll provide a link to the paper, “Dollar Safety and the Global Financial Cycle.” But I want to ask you a question that I've asked many other guests on the show, since you really are in this literature. You've done a lot of work, you're doing original thinking in it. But if we could run earth's history again, let's say a million times in simulations, with different shocks applied each time. Of course, we don't do that because we are ethical and believe in humanity. But if we could do this, would we end up with such a dominant currency, a dominant reserve currency like the dollar? Is it inevitable that we tend to end up with a money of monies, where convenience yields emerge as you say? Or is it just luck? Is it just path dependency?

Lustig: That's a great question. So I think my answer is going to give a typical two-handed economist answer. And say that it's a bit of both. I think there are fundamental forces that will drive one to having a sort of a dominant reserve currency and a safe asset provider because of what we talked about early on in the show. If you go back to that paper that Arvind wrote with his two co-authors, they clearly show that once you have safe asset status, every additional investor causes your assets to become safer. So there is this sort of force that makes your assets safer just by virtue of being a safe asset supplier. And so that naturally pushes you to have a monopolist, like the US now, or the UK before that, and maybe the Dutch if you go back far enough.

Lustig: So I think that's certainly one force. The other thing you mentioned is path dependence. And certainly I think there's a lot of that too, actually. And there's sort of ... Clearly, it's hard to dislodge the US as the world's safe asset supplier. If you think about where we are today, there are a lot of reasons why you might doubt that the US will continue to be the world's safe asset supply, particularly, we've talked about this on Twitter as well, the US's fiscal situation seems, at least to me, perhaps somewhat precarious.

Lustig: The real question though is who's going to be stepping in, right? And that's where I think there is no clear contender right now. The interesting thing about safe asset supplier status is that what sort of matters is your relative macro fundamentals. That's a nice thing that comes out of Arvind's work there. It's not so much what the fundamentals are, but as long as you sort of look okay relative to the other contenders, it's almost like a beauty contest, then you're probably safe. So the question is does the US still look okay relative to the contenders? And my tentative answer would be, maybe. I mean, the euro certainly has its own issues. The Euro zone is a monetary union, but as we all know, it's not a fiscal union. And that creates its own set of issues. So maybe investors will be reluctant to try the euro as safe asset supplier. Then there's China obviously, but China brings a whole host of other issues related to governance. They obviously have the right size. So that I think is a tough one.

The interesting thing about safe asset supplier status is that what sort of matters is your relative macro fundamentals...So the question is does the US still look okay relative to the contenders? And my tentative answer would be, maybe. I mean, the euro certainly has its own issues. The Euro zone is a monetary union, but as we all know, it's not a fiscal union. And that creates its own set of issues. So maybe investors will be reluctant to try the euro as safe asset supplier. Then there's China obviously, but China brings a whole host of other issues related to governance.

Beckworth: Well, let's segue into another paper, Hanno, that you have written that's along these lines. And I found it very fascinating. We had an exchange on Twitter about it, so I want to bring it up here. But your paper, it's co-authored, is titled, “Manufacturing Risk-Free Government Debt.” And in this paper, you highlight there's a trade off between insuring bond holders with the issuance of these safe assets versus insuring tax payers who might go through an adverse macroeconomic shock or recession. So you say there's a trade off between insuring those two parties. So walk us through that story.

Manufacturing Risk-Free Government Debt

Lustig: Right, right. So let me describe the trade off. As economists, we're all interested in trade offs. And what we want to point out is that there's an important trade off here that has been often overlooked in this literature. In this literature on sovereign debt issued by countries like the US, people sort of automatically assume that it's risk-free. But actually, if you think about it, what you're really buying is you're sort of buying a claim to future tax revenue, minus, you're sort of taking a short position in government spending. And then you have to think about the risk properties of these two cashflow streams.

Lustig: And so our basic point is, well look, you can't really have it both ways, except under exceptional circumstances, which we can talk about later. But in the baseline case, there's a trade off, because if you insure tax payers by allowing tax revenue to decline in a recession relative to GDP, which is what the US and other countries do, other experienced and developed countries, that's going to tend to make your sovereign debt riskier. It's not going to be risk-free. That's just basic finance. There's just no way around that. And so as you provide more insurance to taxpayers, you're shifting aggregate risk to the bond holder.

Lustig: And the bond holder, well, she's going to say, "I want to be compensated for that. I want to earn more than the risk-free rate." And so we describe this trade off. And it's kind of a very stark trade off. Basically what we say is look, your ability to insure taxpayers is quite limited as a government if you want to keep the debt close to risk-free. There's sort of only one rabbit that you can pull out of a hat, and that's the convenience yield that we talked about earlier. And I think that's what, again, makes the US potentially somewhat different because if you look the US, it seems to be able to do both. It seems to be able to insure taxpayers and keep the return on its entire debt portfolio very close to the risk free rate. We've done the numbers and we've also compared the numbers to numbers computed by George Hall and Tom Sargent, who've actually done a lot of work on this, by the way, on sort of getting good data on the entire government debt portfolio, all treasuries, going back to the very beginning of the United States.

Our basic point is...there's a trade off, because if you insure tax payers by allowing tax revenue to decline in a recession relative to GDP, which is what the US and other countries do, other experienced and developed countries, that's going to tend to make your sovereign debt riskier. It's not going to be risk-free. That's just basic finance. There's just no way around that. And so as you provide more insurance to taxpayers, you're shifting aggregate risk to the bond holder.

Lustig: And if you look at the post-war period, to give you the number, the actual return is less than 100 basis points above the risk free. So it's really a small premium that you get. Nevertheless, the US actually provides a lot of insurance to taxpayers, more than most countries actually. In recessions, tax revenue really plummets, which is our version of saying you're insuring the taxpayer, because you're saying, "Okay, well you don't have to pay me now. You can pay me later." That's what the treasury is telling the taxpayer. So then the question is why. And actually, we don't have a complete answer.

Lustig: What we do note is that if you earn these convenience yields as a safe asset supplier, that provides a countervailing force because as we've just talked about, convenience yields go up in bad times. In bad times, in recessions, investors around the world want more treasuries, T-bills, what have you. And so the treasury earns more revenue, seigniorage revenue. So it is true that tax revenue plummets in a recession, but that's potentially offset a little bit by the increase in the seigniorage revenue. So that's the exorbitant privilege at work here. What we struggle with though is the numbers. Quantitatively, this channel doesn't quite seem strong enough to completely account for what we see in the data. So that's kind of the puzzle aspect of this whole thing.

Beckworth: Yeah, very interesting. And I think this paper ties together a lot of the other papers and ideas that we've been talking about. And ultimately, government is a form of insurance. It provides insurance to the population. And there's two groups you identify here. So I think it's a fascinating way to think about this. In fact, my whiteboard here in my home office, I did kind of a two axis, and I had insure taxpayers on one axis, insure bondholders, think of this function. And what you're describing is maybe a shift out on that when a country has high demand for safe assets. So another way of saying this is if you have a strong counter-cyclical demand for your safe assets, so when recession hits, and the demand for treasuries goes up, it provides this additional source of funding to offset the loss of funding from taxes. It's a very powerful idea, kind of puts everything together in a nice package.

Lustig: It alleviates the trade off.

Beckworth: Alleviates, okay.

Lustig: For the US. But of course the thing to keep in mind is that this wouldn't necessarily be the case for other countries. And in fact, if you look at ... There's some interesting work on that that we cited in the paper. If you look at developing countries, emerging market countries, there you see this trade off much more starkly. It is very hard for emerging market countries to provide insurance to their taxpayers without seeing a big penalty applied by bond market investors.

Beckworth: Yeah, so the US has this kind of unique advantage. And can we say, to some lesser extent, other advanced economies do as well? So like Germany, for example. Germany might be able to tap into this to some degree?

Lustig: Yeah. Great question. Yes, to some degree. Although, when we looked, for example ... We looked at the UK because the UK has amazing historical data made available by the Bank of England, going back centuries, actually. You can go back 300 years, or 400. It's quite something. It's called the Millennium Project. The Bank of England makes the data available. And so when we look at the UK, what we see is that in the post-war era, after the second World War, they do some provision of insurance to UK taxpayers, but it's not nearly as pronounced as what the US does. The US Treasury does seem to be quite a bit more aggressive in insuring their taxpayers than the UK has been. Now, of course, that could be because after the first World War, we talked about this, the UK lost its sort of status as the world's provider of safe assets, and they face a steeper trade off, I imagine, than the US does.

Beckworth: Okay. Yeah, it's very interesting. And again, we'll provide the link for the listeners. I encourage them to check it out. One more question on this, before we move on to another paper of yours. If the counter-cyclical demand for the safe assets, and let's focus on the US, is strong enough, and maybe even persists beyond the recession, then it would mean increased seigniorage flows into the US, correct? And then, could that, in turn, be part of the explanation for the low inflation, say the lower than target inflation over the past decade prior to the pandemic?

Lustig: That's a very good point, one that we have not explored. But I think you're on to something there. It's entirely feasible that if there's a shortage, or possible that if there's a shortage of safe assets, that the way you kind of restore equilibrium, and sort of increase the supply of safe assets in real terms, is by having lower than expected inflation in response to a shock. I think that sounds plausible to me. Again, it's not something we've explored, but actually, after we discussed this a couple of months ago, I've been thinking about this. And I think it'd be interesting to pursue this in a fully specified model where you have a role for monetary policy, which is, to be clear, something we have not done yet in our paper. I should say that this is joint work with my long-time collaborator, Stijn Van Nieuwerburgh, who is at Columbia. And I always joke that he still works with me because I'm the only one who can pronounce his name. And also with Zhengyang Jiang and with Mindy Xiaolan. So wanted to make sure I mention my co-authors there.

Beckworth: Yeah, very nice. So this is a nice segue into the next paper we want to talk about. And this is “US Government Debt Valuation Puzzle.” So you mentioned in the trade off in the previous paper that the traditional measure of seigniorage flows isn't big enough to offset the decline in tax revenue. And this is kind of a puzzle, or part of the puzzle you addressed in this paper. So walk us through. Why is there a debt valuation puzzle for US government bonds?

The Puzzle of US Government Debt Valuations

Lustig: Right. So this is actually, as you said, closely connected to the previous papers. I'm going to be able to build off of what we just talked about. So, I think what economists sort of done when they talk about treasuries, they sort of think, "Well, these things are roughly risk-free." And what we say is well, if you actually sort of do the asset pricing carefully, what you realize is you're buying a claim to US tax revenue, minus government spending. US tax revenue, we've talked about this, is highly pro-cyclical, plummets in recessions. So that's risky. Government spending, on the other hand, especially the transfer part, goes in the opposite direction. It's counter-cyclical. What we do is we spend more as a fraction of GDP in recessions, giving more transfers to help people who end up in dire economic straits. And other developed countries do the same thing. But as a result of that, what you end up buying is actually a pretty darn risky claim, at least from the business cycle perspective.

Lustig: The long run economic risk, you can sort of roughly think of taxes and spending as sort of being in the long run a constant fraction of GDP. So now in the long run, you have output risk. If GDP grows much faster over the next 50 years than we anticipate today, that's great. That means there's going to be more net cashflows to back up our treasuries. But on the other hand, if output comes out well below expectations over the next 50 years, then that's going to be bad news obviously. So as a treasury investor, you're bearing all that risk. And so what we do, to put it simply, is we sort of price these claims the way you would price any other claim. We price a stock. We make sure that our model can price a whole bunch of assets, not just bonds, and we price the whole bond portfolio, which is kind of the novel aspect

Lustig: And then what we quickly realized, as we said before, it's really hard if you come up with sort of realistic forecasts of future tax revenue and future spending. It's just awfully hard to get anywhere near the market valuation of treasuries. And one way to sort of rephrase that is to say well, if you were to look at the probability of a huge tax increase to close the entire gap, we look at a two standard deviation tax increase. So what would that probability have to be in order sort of for everything to kind of make sense? And the answer is over the past 20 to 30 years, it would have to have been between 20% and 30% per year. So it's not quite consistent with investors being fully rational is our conclusion. It seems like they're constantly expecting a huge fiscal correction that just doesn't happen.

Lustig: Now, coming back to our discussion earlier, this does have that seigniorage term in it. So we try to carefully measure the seigniorage that the US earns, which is non-trivial, because that convenience yield can be as high as 200 basis points per annum. So it's potentially a big number. But we find that it's not big enough to completely close the gap from a quantitative perspective. Certainly qualitatively, it goes the right way. It matters, but it's not big enough, doesn't give you quite a big enough cake.

Beckworth: So just to summarize, you're looking at the value of US government bonds. And you're looking at future payments, future flows to the holders of those bonds, and future government spending activity. And when you add it all up, you can find kind of a net present value of all those future claims in today's terms. It just doesn't add up to the price in the market, doesn't make any sense.

Lustig: That's exactly right, David. This is sort of a standard net present value calculation that a first-year MBA student would be able to do. Now, an important question people always have is, "Well, wait a minute. If I buy a 10-year bond, I don't particularly care about all these other cashflows coming 30, 40 years from now." And I think that's a compelling argument. People always have that question, even when we present at seminars. And so the answer is the reason you care about all these future cashflows is because to ensure that the Treasury can roll over the debt in all future periods, you have to take all future cashflows into account. At the end of the day, unless there is some bubble in debt markets, it has to be the case that the value of debt today is backed by future cashflows. And so that's the exercise we're doing. It's a standard MPV exercise.

Beckworth: So Hanno, this is very fascinating, and it makes a lot of sense. And it's a question I've wrestled with too. I say look, I know the bond market knows more than I do. They've got skin in the game. They're investing portfolios. They have jobs they want to keep. They have returns they want to make. They know more than I do. And look at the wisdom of the markets. But man, what are they expecting? Some massive fiscal reform in the future? And so it just doesn't add up. And the implication, and correct me if I'm wrong, the implication is there must be a bubble term, some unexplained part of that equation that we can't capture with traditional asset pricing metrics. Is that right?

Lustig: That's right. So the conclusion that we reached tentatively, based on the work we've done, is that there's some evidence that potentially, treasury investors are overconfident, that they keep predicting large surpluses that actually don't materialize. And one tentative piece of evidence that supports this notion is if you look at the Congressional Budget Office's projections. Now, to be fair, these are projections based on current law. They're not actually forecasts that take into account changes in legislation that could occur in the future. But if you use these projections, you do see that what actually transpired, the actual realized deficits have consistently been much higher than what the CBO has been projecting for the past two decades or so.

The conclusion that we reached tentatively, based on the work we've done, is that there's some evidence that potentially, treasury investors are overconfident, that they keep predicting large surpluses that actually don't materialize. And one tentative piece of evidence that supports this notion is if you look at the Congressional Budget Office's projections...you do see that what actually transpired, the actual realized deficits have consistently been much higher than what the CBO has been projecting for the past two decades or so.

Lustig: So if you sort of connect that back to our bond market investor, they could be forgiven potentially for making similar mistakes and having been too optimistic. Truth be told, of course, we have seen two huge shocks in the past 20 years that nobody could have anticipated, the great financial crisis as well as the pandemic. And obviously, both of these events have contributed to this divergence between the CBO projections and what actually happened. And by the way, I just mention this as one piece of evidence. And I want to emphasize that I think the CBO does provide an invaluable public service in making these projections, which is incredibly hard, and in making all the inputs that go into these projections available.

Beckworth: Yeah, but it's a fair point that for two decades, effectively bond investors have been wrong or they've miscalculated. So there's a systematic error here, right?

Lustig: Yes.

Beckworth: If you go, again, using asset pricing framework, there's a systematic error. And maybe instead of calling it an error, we need to call it a bubble. So I'm going to tie this into a friend of yours, Markus Brunnermeier's work. He has a paper called “The Fiscal Theory of the Price Level with a Bubble.” Or he has another one out, “Mining the Bubble.” So as I read his paper, and you can correct me if I'm wrong, he looks at this bubble term. And maybe you can help me define it, but what I read from his paper, I got from his paper is he looks at convenience yield, this money-ness of assets. And he looks beyond kind of traditional money. He looks beyond like the monetary base or reserves. He looks at the convenience yield on everything and treats that as the bubble. So how would we define this bubble and try to reconcile it with what we see in markets?

Lustig: That's an excellent summary, David. I think the way Markus and Yuliy and his co-author, the way they try to attack this puzzle is they use a broader notion of convenience yields, a bit broader than what we've been using, to argue that that seigniorage revenue that the Treasury receives every year is potentially quite a bit larger than what we estimate it to be. That's the way I would reconcile Markus's findings with what we find. So they have sort of a broader notion of what is comprised in that term. And I should mention that there's sort of a contemporaneous paper that pursues a similar approach by Ricardo Reis, who's at the London School of Economics. And I suspect you've also had Ricardo on the show.

Beckworth: Yeah. Yeah. We talked about that, “R less than G less than N,” I believe.

Lustig: Exactly.

Beckworth: Yeah, that's a great, great paper. It was a great conversation. So let me ask this question here, and this may be a naïve question. But so what they're saying is the future flow of all these convenience yields is much bigger than we originally thought. And if we kind of take the net present value of all those future convenience yields, given that treasuries are used as money beyond just simple investment vehicles, we get the net present value of all those future seigniorage flows, broadly defined. You get this term, and maybe it's viewed as an asset, that's sitting there on the right hand side of that equation, which kind of fills the gap.

Lustig: That's right, David. I think the one thing that I would say here, it's sort of a caveat, is that one thing we know about convenience yields is that if you increase supply, they're going to go down, right? This is true with money. If you supply more real money balances, that's going to happen. But it's true with these money-like assets more generally speaking. And this is, I think, a key point. If you're just sort of loosely comparing R to G, and that's not what Markus and Ricardo are doing, but other economists have sort of gone down that path. If you just sort of compare R to G and you say, "Well, R's consistently been quite low relative to G," that's quite tricky. Because if you keep supplying treasuries, R is going to go up because convenience yields go down. That's inevitable.

Lustig: So that's, I think, a key caveat. We have to be mindful of the fact that in the US, we can exhaust the exorbitant privilege. So I think the key point here is we need to think and do more work about what the relation is between these convenience yields that we earn, which are real, and Q, the quantity of treasuries that we supply to the market and that have to be absorbed. That's sort of what I think requires a lot more work from us.

Beckworth: I guess another area of work is how to actually measure this bubble, right? That sounds like a challenge. I mean, you could just treat it as the gap, the missing pieces to kind of solve for the missing part of the puzzle. It seems like an interesting research agenda for some enterprising grad student or researcher to kind of come up with a way to quantify that bubble term.

Lustig: I think that's a great topic to be working on. And I agree with you. I really encourage young researchers, especially those who are sort of interested in macro and finance to think more about these issues. One thing that's happened in finance is that sort of financial economists have tended to not think too hard about these sorts of issues. And we've left all the hard work to macro economists. But I think we, as financial economists, I'm sort of putting my financial economist hat on now. I think we can add a lot. Because we've learned a lot about how assets are priced, how risk is compensated in financial markets. It is shocking how asset pricing, there's thousands and thousands of papers on the valuation of small cap momentum stocks, for example. But there's very little on this topic, valuation of treasuries, the whole portfolio, which just seems like a first order question for everyone.

Beckworth: Right, right. Seems like a very fertile area for sure. Let me tie together some of these strands here into a question that I've been working on. And what is the link, and we touched on it earlier, but what is the link between safe asset demand and the low inflation that we saw pre-pandemic? And I mentioned this already with your manufacturing safe asset paper. But there's several channels we could take with this understanding. So one story, one transmission mechanism, and it's told by like Ricardo Caballero, Emmanuel Farhi, and others in their safe asset work is that as yields go down on safe assets, on treasuries, eventually they go down below the equilibrium rate, R star. And then via like a Phillips Curve, that interest rate gap creates an output gap, and then via the output gap you get low inflation. And that might explain some of the low inflation they had in the decade prior to the pandemic, when everyone was wondering why can't the Fed, why can't the ECB hit its inflation target? So that's one story you could tell.

Beckworth: Another story would be the one we've just been talking about, that this bubble term is increasing the seigniorage flows, the net present value of seigniorage flows, broadly defined into the US and other advanced economies, and that's creating some kind of dis-inflationary pressure. A third story or channel you could tell would be kind of just the standard monetary demand supply story. So to the extent treasuries are acting as money, they have the convenience yield, they provide transaction services. And increased demand for them is going to lower the velocity, their use, and that's going to lead to lower spending, and thus lower inflation. So there's three stories there. Maybe they're all complementary. How do we see safe asset demand translate into low inflation?

Lustig: Yeah, this is an intriguing hypothesis, and one that deserves to be looked at more closely. But certainly to me, it sounds right that if investors perceive larger seigniorage revenue increases the real value of treasuries that are outstanding. And one way to sort of enforce the intertemporal budget constraint, and it's for the price level to not go up as much as people originally anticipated, and to have low inflation. I think that's a really interesting hypothesis.

Lustig: One thing that I think what we need to think more about in this context is well, shouldn't we then sort of expect to see low inflation predominantly in countries like the US that have sort of safe asset supplier status. And there's a case to be made perhaps that low inflation has been sort of a more widespread phenomenon in the broader set of countries. So I think that would be one sort of tension that one would have to confront empirically if you were to push this idea. But it certainly sounds right to me that that might have been what's going on. So I definitely look forward to seeing your work on this.

Beckworth: Well, I'm looking forward to seeing other's work on this as well. It just seems to me there's been an absence of, or a lack of connecting the dots here. There's been this huge work on safe assets, yours, Arvind's, Caballero, Farhi, all these great economists working on safe assets and the implications. And there's also been this discussion, at least, again, pre-pandemic, of low inflation mystery. Why has inflation been so low? And I've been to several conferences where they discuss everything but safe assets. I even raised the question at Q&A. And no one really kind of took the bait, I guess. And so it just seems to me there's a clear connection.

Beckworth: I know Emmanuel Farhi, the late Emmanuel Farhi and Ricardo Caballero have a paper. I believe it's called “The Safety Trap,” something along the lines of a deflationary trap that emerges, from several years ago. But they actually explicitly modeled this. They put it in a Phillips Curve, and they actually show low inflation. I just haven't seen anyone make the case that's part of the story. Maybe not the only story, but part of the story for the decade prior to the pandemic. Of course, going forward, maybe a totally different story with the run up in debt and changing some of those dynamics.

Beckworth: Right, in the time we have left, I want to go to one more paper of yours. And this is related to kind of the secular decline and trend in interest rates. So we know rates on safe assets in advanced economies have been going down for many decades. And why it's been going down's been a big point of discussion lately. But you also have a paper that discusses an implication of them going down. And the paper's title is “Financial and Total Wealth Inequality with Declining Interest Rates.” So walk us through that paper.

Wealth Inequality with Declining Interest Rates

Lustig: Yeah. Thanks, David. So the point we make in that paper is a really simple one. We say well, what we've seen is sort of a unprecedented decline between say 1980 and 2020 in the long term real interest rate in the US, and in other countries. And it's about anywhere between 350 to 400 basis points. So that is a huge decline. I think it's almost unprecedented. And what we say is well, let's explore the implications of this decline in real rates for wealth, for how wealth is valued in the marketplace, and how it's distributed for wealth inequality.

Lustig: And the point we make is all the assets that Americans have in their portfolio have some duration, some sensitivity to these interest rates. If you have an asset that is a claim to cashflows that accrue very far in the distant future, that asset has a high duration. And so when interest rates go down, the value of that asset's going to go up by a lot. You're going to see big capital gains. What are examples of assets like that? Well, obviously if you buy a 30-year bond, that would do it. But a stock, for example, has a pretty high duration. If you have a private business that is still growing quite quickly, that also has high duration.

If you have an asset that is a claim to cashflows that accrue very far in the distant future, that asset has a high duration. And so when interest rates go down, the value of that asset's going to go up by a lot. You're going to see big capital gains.

Lustig: Americans who own these sorts of long duration assets will see large capital gains in response to declines in rates. Whereas Americans who don't own assets that have high duration, for example if you only have a bank account, then you won't actually see large capital gains. Now, what that implies if you put it all together is that if I sort of mechanically feed these big declines in interest rates into a simple model, and then I can mark to market the wealth of all the households in my economy. And what I'm going to see is that wealthier households who tend to own more high duration assets in the US, this is a pretty strong pattern in the data, they'll tend to see much larger capital gains than the poorer Americans.

Lustig: Now, important point to add there is that doesn't necessarily mean that they'll be consuming a whole lot more in the future because the thing is even if their consumption in the future, if that consumption path was unchanged, their wealth would still have to go up a lot just to be able to afford the same consumption. When rates go down, you need more wealth. But in any case, what we show is that we can sort of account for a pretty sizable chunk of the increase in measured wealth inequality. So this is only about wealth. We're not talking about income inequality here.

Americans who own these sorts of long duration assets will see large capital gains in response to declines in rates. Whereas Americans who don't own assets that have high duration, for example if you only have a bank account, then you won't actually see large capital gains.

Lustig: Just through this channel, so we think this is sort of an important thing that economists ought to be thinking about more. At the end of the day, financial wealth metrics are valuation metrics. And what we've seen is that valuations across asset classes have gone up dramatically, and that's exactly what you'd expect when long rates go down. And the implications of that fact, they differ in the cross-section. And that has important implications for inequality.

Beckworth: Yeah, so your paper is really fascinating. And for me, it sheds light, helps kind of round out some of the other discussions going on right now about low interest rates. A recent paper that was discussed at the Kansas City Economic Symposium this year was one called “The Saving Glut of the Rich” by Atif Mian, Ludwig Straub, and Amir Sufi. And they kind of approach ... Now, they're looking at income inequality, not wealth inequality. But they kind of look at causality almost from the opposite direction. They say. "Hey, the inequality's given rise to all this excess savings by the rich. And that's driving down rates." But I think what you show is you have to be very careful, because causality could go the other direction.

Lustig: That's right. So theirs is sort of a different direction of causality. As you said, they said, "Look, if you increase income inequality, you could expect to see declines in rates." And the mechanism they have in mind is one where rich households have sort of different preferences and just save more. They're sort of what economists call a non-homotheticity. Excuse me. So rich households are not just scaled up versions of poor households. They're fundamentally different. They're consumption technology's different and they save more. And if you sort of increase income inequality, and you give a bigger fraction of the pie to the rich, that's going to mean that there's more savings in equilibrium that have to be absorbed, and that drives down real rates. That's the channel.

Lustig: And there's sort of a lively discussion going on in macro right now about whether this is right. For example, Ben Moll has a paper looking at Norwegian data, where he sort of finds that this is not really the case. That at least in Norway, richer households don't seem to display this behavior. So I'd say this is sort of an interesting ongoing debate and a very important question. And I should also mention, and you've kind of pointed this out as well, that the channel that Atif, Amir, and Ludwig focus on is potentially complementary to what we're discussing. So it's worth trying to figure out what's driving this decline in the real rate, where taking the decline in real rate is given, and sort of saying, "Here are the implications for wealth inequality." As we're kind of going in the other direction.

Lustig: Our mechanism doesn't really hinge on where exactly this interest rate decline comes from. So you could have it be generated by a whole bunch of different things. And we've talked about other potential drivers. You've mentioned the work by Laura Veldkamp and her co-authors on scarring potentially being a driver. So that's one potential answer. Then there's fascinating work by Adrien Auclert at Stanford and his co-authors that emphasizes demographics. The truth is I think it's very hard to precisely identify which of these components accounts for what fraction. That's sort of a hard question to answer. But I suspect that all of these play a role at the end of the day.

Beckworth: Yes. So a lot of channels that may be contributing to the decline in interest rates and whether they will persist is a interesting question. Will we continue to see low rates going forward after the pandemic, after the run up in public debt? So a lot of interesting questions we could consider for many years. But I'm grateful to have you on today, Hanno, to discuss your work. And we have come to the end of the show. I just want to thank you for coming on. Our guess today has been Hanno Lustig. Hanno, thanks so much.

Lustig: Well thanks, David. This has been a lot of fun. Thanks for reading our work and for publicizing it. I really appreciate it.

Photo by Peter Dazeley via Getty Images