Feb 21, 2022

Will Diamond on Safe Assets, Risk-Free Rates, and Convenience Yields and their Implications for Policy

Risk-free discount rates in the US are especially low among G10 currencies, and the US’s central position in the global financial system can help explain this.
David Beckworth Senior Research Fellow , Will Diamond

Hosted by David Beckworth of the Mercatus Center, Macro Musings is a new podcast which pulls back the curtain on the important macroeconomic issues of the past, present, and future.

William Diamond is an Assistant Professor of Finance at the Wharton School of Business at the University of Pennsylvania. Will joins David on Macro Musings to discuss safe assets, convenience yields, bubbles and public debt and the implications for policy. Specifically, David and Will get into competing theories of interest rates and the rise of New Keynesian thinking, the role of the dollar in the global financial system, the drivers behind the growth in US debt, how the construction of risk-free interest rates unaffected by convenience yields on safe assets can improve our understanding of the financial system in times of stress, 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 macromusings@mercatus.gmu.edu.

David Beckworth: Will, welcome to the show.

Will Diamond: Thank you so much, David, for having me on. This is a great podcast, I've listened to it quite a bit and excited to be a part of it.

Beckworth: Well, I'm glad to have you on because as listeners know, as you know, Will, from Twitter and from listening, I'm a big fan of the safe asset literature, and you are very active in that area. You have some real interesting papers that we're going to talk about later. And I just am looking forward to this conversation. In fact, you've been kind enough to give me some of your time in the past, I was working on a paper and I had some questions about modeling and the right approach. And you actually took out time to do a Zoom meeting just solely based on help that I needed. And I knew that you knew this because of your work and our interactions on Twitter. So, a real delight for me too. Now, Will, before we get into safe assets and all that other good stuff, tell us a little bit about yourself. How did you end up where you are and in this field where you're looking at safe assets a lot?

Diamond: Well, I guess there's two different answers to this question. One of which makes me sound very principled. And one of which makes me sound very lazy. The lazy answer is basically my whole family is econ and finance academics, so this was the path of least resistance in life. My sister is an economics professor. My dad was a finance professor. My mom was one at one point too. So, there was never any pressure for me to study these topics, but in terms of it being a very obvious career path, which to most people, it's a relatively obscure one, academic econ or finance was always something on my mind is possible. And if I hadn't done this, I guess the rebellious thing would've been becoming a statistics professor or something extremely different like that. But that aside, I did actually consciously choose this field on some level too.

Diamond: And really, I would say watching the 2008 crisis happen was the key moment for me, more than anything else I've ever seen watching, particularly Bernanke. But the whole crowd of people involved in the financial crisis rescue then was people who had studied topics which were very intellectually interesting and important for their whole careers. Learned important things about them, and then got to apply them in public policy in a way that was directly about large real world issues, but still tied to decades and decades of serious intellectual work. And it's extremely lucky for somebody like Ben Bernanke to have that perfectly choreographed career. I don't think anyone can expect to replicate that exactly. But in terms of this being an area that's both intellectually deep and directly applicable to decisions where economists are directly involved, made it seem the right mix of application and interest.

Beckworth: Well, that's interesting. I've had a number of guests on the show who came of age of source during 2008, the crisis, and it really shaped their interest, what they wanted to do. You're another example of that. I've had other guests, some international guests who grew up in countries where they had financial crisis or hyperinflation, and that really propelled them into this field as well.

Beckworth: So, it's interesting to see this generation of young scholars like yourself entering this field and making contributions. And you've had several interesting articles on the Journal of Finance, the Journal of Financial Economics, some other journals, and we're going to get to them. But one thing I like about you Will, is that you're engaged on Twitter. Twitter is an interesting place to go. And I've said on here before, it's a place where I have found my people, so to speak, my tribe, unlike you, I don't come from a family full of econ and finance professors. So I imagine when you go home for Thanksgiving, you guys can talk shop, you can talk QE, large scale asset purchases and such. I am-

Diamond: A little bit too much.

Beckworth: Okay, there you go. Whereas when I go home, we have great conversations, but usually if my work comes up in my area, it's more of me trying to explain it to family members. We talk about other things, sports, politics, religion, but you actually are engaged in conversation on Twitter as well. And I admire it because someone like yourself, who's at a top school you're at the University of Pennsylvania. You're an assistant professor. You're going after tenure. Lots of pressure on you.

Beckworth: And just looking at your record, you already have some great publications, but there might be a temptation not to do Twitter. It might be seen as a distraction or to be more precise some might see it as a substitute rather than a compliment to your work. But I'd be interested to hear your side of the story. When you engage on Twitter, do you find it gives you more ideas, connects you to people, makes you a better scholar, what are your thoughts?

Diamond: That's a good question. And I could probably spend an hour answering it, so I'll try to keep it to the requisite podcast length. I would definitely say for me, I get a lot of interesting ideas talking to people on Twitter. They're sometimes not directly connected to what the field is doing right at this moment. So, I originally signed up for Twitter, the pandemic started, everybody stuck at home and it wasn't a substitute for work it was a substitute for the social life that we all lost. And I've realized over time that there's a lot of interesting economists on there, but also interesting non economists there too. And I would say if you read my main feed of tweets, I'm mostly talking to economics. If you click over to the and replies tab, I'm talking to all sorts of people, usually academics or academic adjacent people of some sort or another.

Diamond: But I would say, yeah, I've gotten a lot of intellectual breadth and opportunity from just meeting all sorts of random people on Twitter. I think there were some papers written, they did a randomized trial of when you advertise your work on a Twitter account with sufficiently many followers, it does course it to get more sites. So there is some professional benefit I would say, but I would guess most of the academic economists who are highly active on Twitter, they can justify it because of the professional benefit, but really they have this inner need to argue with people about ideas.

Diamond: I think Twitter is a replacement for the lunchroom, which got taken away during COVID lockdowns. If you can have your colleagues arguing the Fed is being too dovish or too hawkish, people get a little bit too into it and a bit heated. Maybe the downside of this is it's recorded on the internet and everybody can see what you're saying. So somebody can go through it a year later and find your five most offensive quotes or something where you called your colleague stupid. But I think it's really great for just, I think ideas travel much faster in the profession now, which maybe has some downsides, but overall I would say it's great that we just quickly learn about new things.

Beckworth: Yeah, I agree. I found it very useful for me when I was still in academia and working on getting tenured and I found it useful just the interaction like you said with people, but I also found it useful not just with people you might know, but it opens up the world. So, even if there hadn't been a pandemic Will, you'd be able to talk to people in Europe or in Asia, scholars who you wouldn't otherwise have met.

Diamond: Totally.

Beckworth: And that's definitely true for me. A lot of the guests on this podcast have arisen because we have had conversations on Twitter. So, it's been a real joy for me to be on there. And I know Twitter can be something else if you don't approach it in the right manner. But I have found it very productive. And preparing for this show I came across an interesting article that speaks to some of the uses of Twitter for scholars.

Beckworth: And this is an NBER Working Paper by Francesco Bianchi, Thilo Kind and Howard Kung. And the name of the paper is “Threats to Central Bank Independence: High-Frequency Identification with Twitter.” And what they do is they take a high frequency approach to analyze the effect of president Trump's tweets surrounding moments when he did not pal things that he said. And they go, they find these significant effects using event studies a little window around whenever the tweet took place and find that not only do asset markets respond, but they actually find that monetary policy responds. They do event studies, they do VAR analysis. And I think it's useful to note that, President Trump gets on Twitter because we're on Twitter, right?

Diamond: Absolutely.

Beckworth: So, we're part of this, I almost want to say a public good we're creating information, creating feedback, this idea of generation place. So I think it's a great place to be. And again, I'm happy that you're on there and that we got to interact. Let's move on to your research. And I want to talk about safe assets, safe asset yields, and you've got several papers I'm going to mention in a minute, but let me just begin by motivating with some facts that I find very striking.

Beckworth: So currently, and we're recording this in early 2022, the debt to GDP ratio for the US is close to 100%. Our last reading for the CPI was 7%. The CBO projects primary deficits, as far as the eye can see. Today, January 27th, real GDP came out and it was above expectations at 6.9% growth. One of the highest real growth rates in a long time. If you look at nominal GDP, it's at a blistering 14.33%. These are annualized rates. So we have a smoking hot economy, it's booming. And yet I checked the 10 year treasury of the day and it was at 1.79. It's been down at 1.5, it's come up, but still remarkably low given inflationary pressures, expectations of deficits, robust economic growth, and it's a puzzle. And that's part of what you try to explain in your research. Is that a fair way of characterizing your work?

Diamond: Yeah, I'm definitely interested in that collection of facts. I would say, just as the nature of the peer review process, I slice off little pieces of this in each paper. What you've posed is maybe the holy grail of macroeconomics. How do we understand how all of this can be occurring jointly, collectively, maybe 100 papers, slice off little pieces of that and add up to a reasonable understanding. But yeah, I think there's some deep mysteries in all of that. How can we have such low real rates of return when the economy seems to be growing pretty rapidly, which would, one, push rates of return up; and two, if we think that there's some fiscal crisis in the US or some inflation crisis in the US coming up, that would also lead to high nominal rates of return either because there's an expectation of credit risk or because there's an expectation of future inflation. So, we're stuck in really a strange situation.

Diamond: People compare our current situation to Japan sometimes, but they had low growth, low rates of return for long periods of time. And maybe if you average over decades and get rid of these cyclical fluctuations, we might look similar. But I agree. I can't think of a historical period that looked quite like this before. And I don't even know how to ask the question is the US government solvent to the next 50 years? I don't even have a meaningful framework I can write down to ask that question of could the US default on its debt, could the dollar lose its position at the center of the global financial system? And I guess one thing I learned over time is in macro questions make sure you know the things you actually know, condition policy on that, but try to be very, very modest about everything else.

How can we have such low real rates of return when the economy seems to be growing pretty rapidly, which would, one, push rates of return up; and two, if we think that there's some fiscal crisis in the US or some inflation crisis in the US coming up, that would also lead to high nominal rates of return either because there's an expectation of credit risk or because there's an expectation of future inflation. So, we're stuck in really a strange situation...One thing I learned over time is in macro questions make sure you know the things you actually know, condition policy on that, but try to be very, very modest about everything else.

Share this

Beckworth: Great points. And yeah, the idea that we could face a fiscal crisis in the future, there's been a lot of conversation about that fiscal dominance. It hit me today actually that there's been a lot of fiscal dominance talk and many of the same people telling us to worry about fiscal dominance have also said the Fed needs to get its act together, needs to tighten. It should have tightened last year and now the Fed is doing it. It struck me I was like, "Well, if the Fed's able to tighten, then there isn't fiscal dominance yet." If fiscal dominance was really here, we wouldn't have the Fed being able to do what it does. So there's a lot of interesting questions going on. And again, I think it's fascinating to try to wrestle with this. Now, one of them I think starting points and again, we'll come to your research in a minute is what is your theory of interest rates?

Theories of Interest Rates

Diamond: Yes, that's a big question.

Beckworth: It's a big question. And I would like to hear yours, let me just throw out a couple that I think are very prominent and popular. And so, I'll tell about what I think is the dominant macro standard economic one, and it has many different names, but basic idea is that fundamentals ultimately drive, at least over the long run, real interest rates. So, everything from desired saving relative to desired investment and underneath all that would be expected productivity growth, population growth, some measure of risk aversion. Some fundamentals are driving the preferences that cause investors want to hold more treasuries for example. So, those forces, which could be demographics regulations scarring from recessions, those things is not something that can be set by policy. So they're independent to what the Fed does.

Beckworth: And if you take that approach, then you're saying the Fed really can't control interest rates over the long run. It's a subject or it's a slave to the fundamentals. Now, there's a different camp on the opposite side that says, well, actually Central Bank does determine the path of interest rates. If you take the expected path of the short rate that helps set the long term rate, traders are looking at what they think the Fed's going to do. And you get a lot of people on Twitter, I think that embrace this. The BIS embraces a version of this and maybe I could summarize it by saying that view is often called the money view, the liquidity preference view, it's set money markets and then expectations based on the money markets. And then the first view I outlined would be the savings estimate view, the fundamental view. What are your thoughts on that? And where would you come down on it?

Diamond: This might be a bit of a cop out answer, but I have to say both on some level. And on some level, I would also say that was one of the first key insights in Keynesian and macro. The both answer is the IS-LM model, the IS is fundamentals, LM is liquidity preference. And people realize that these two markets have to jointly be clearing an equilibrium. That you don't have one equation and one unknown and macro, you have multiple forces, which simultaneously have to be an equilibrium together. And I think despite a textbook IS-LM model, just being two equations and two unknowns, and you can memorize your way through it as an undergrad in macro. I totally didn't get the IS-LM model, conceptually when I learned it, but I could replicate the formulas on the final exam.

Diamond: I think that's actually one of the deepest points in all of macro, is that you have to be thinking about these things jointly. And then, I guess about the ways in which interest rates can be pushed away from what you might call a frictionless benchmark of pure fundamentals. I guess I'd say there's two categories here, which get the most attention in academia. One is sticky prices and sticky wages that's been, I guess since maybe the 1980s or 1990s, a big force in what they call New Keynesian macro. That there is a pure fundamentals movement, what they called freshwater macro, which won a sequence of Nobel prizes. People realize if there's stickiness in wage and price setting, the Federal Reserve can have some short term control over the economy. Everybody always agree the Fed controls is the no nominal interest rate. They can just say, "We're paying 4%. That's what we're doing." But as a 4% nominal rate, if you increase that to five or reduce it to three, are you doing anything to the real rate is really the question.

Diamond: And the sticky prices and wages got to the idea that maybe over a couple quarters a year or two years, the Fed can temporarily manipulate real interest rates to control the economy. What some of my work and not just mine, this has been a pretty broad topic in post 2008 macrofinance is separate work, which is really bringing the LM model back into our IS-LM way of thinking. So sticky prices is not money, liquidity preferences, banking that was still removed from macro. And one of the ways to summarize this best is if you go back to the '50s, there was a very famous macro textbook by Don Patinkin, which I think the title was Money, Interest, and Prices. Then in the New Keynesian revolution, there was a textbook by Woodford, which was just called Interest and Prices, which is a very conscious reference to we got rid of all the money. And what I would say if I would summarize my work, plus plenty of other people in the post 2008 era, is we realized that LM curve actually is a very real object. And it's something you can see in the data.

What I would say if I would summarize my work, plus plenty of other people in the post 2008 era, is we realized that LM curve actually is a very real object. And it's something you can see in the data.

Share this

Diamond: And some of my results in my papers are directly about this other people too. But the nominal interest rate itself is quite important, because what is on some level, what you forego when you hold cash, let's say you think cash is a special liquid asset, treasuries are a little bit less liquid. You can't use treasuries to pay for something your local store. If treasuries are paying 3% a year, you are effectively losing 3% a year as the price you pay for the liquidity value of cash.

Diamond: So, these liquidity-based transmission mechanisms are purely about the level of the nominal interest rate, which is distinct from the sticky price story, which is about cyclical fluctuations in the real interest rate. And I don't think I've ever seen a model that puts all of this together and ask the question, suppose we put liquidity frictions back in, in a meaningful way on top of our old macro framework. What does that do to what we should think about for optimal policy? I'd like to write a paper on that at some point, I think that's one of these things that would take three years and be a little bit too slow and difficult for a, well you're on the tenure track.

Beckworth: Fair enough. But we look forward to that paper Will, in the near future after you do get tenure, and you'll come back on the show to discuss it as well. We'll make a point to do that.

Diamond: I'd love to be back whenever you'd have me, but don't mean to overuse your time.

Beckworth: No, no, absolutely. A great observation about the return of the LM curve or return of, I would say more generally money and liquidity into these models. So, it probably was easy enough to get by without it. I have a friend and a colleague named Josh Hendrickson, and he has really helped me appreciate the fact that the standard model, new Keynesian workhorse model, there's LM curve, but they drop it.

Beckworth: And that's because there really isn't exchange in the model, there's a Walrasian auctioneer, who magically gets everything to work. And so, one of the big missing ingredients is exchange. And I saw a prominent economist, speaking of Twitter, he recently posted one of his papers that had 16 different equilibrium values in a pure exchange economy. And I went to look at it and it was of all raisin based model. There was no actual exchange taking place in the model. So, I'd love what you're doing, but others like you're doing, you're getting back to this transaction exchange part of the story. And I guess what you're saying is it took the great financial crisis to remind us or slap us in the face to get back to that understanding.

Diamond: Yeah. And I think this is, I potentially am overstating here, because I'm just looking at the recessions, which I've seen while I've been paying attention and whatever goes wrong with the economy is the mechanism that policy makers and academics have temporarily forgotten about. In 2006, we didn't believe in financial crises. In 2020, we didn't believe in inflation. And of course that's the thing that comes back. So, I don't think macro policy is often a super intellectually deep issue. You can write very deep papers about it. You can always understand the economy better, better data and better models. But I would say the first order to being a responsible policy maker is know your history relatively well and just have a pretty long mental list of potential ways the economy can screw up. And what have simple remedies been in the past? If you get the very basics right, then I think you're already in the 90th percentile.

The first order to being a responsible policy maker is know your history relatively well and just have a pretty long mental list of potential ways the economy can screw up. And what have simple remedies been in the past? If you get the very basics right, then I think you're already in the 90th percentile.

Share this

Diamond: And yeah, I think we've seen this twice. And I guess potentially before this, the tech recession in the 2000, 2001 that you could put in this as well that people are writing, I don't need to name names but there are some rather amusing papers written right before that recession happened saying is the stock market overvalued let's compute a present value of cash flows and not think about risk previa and time varying expectations and interesting things like that. And no it's not overvalued. And that was actually the best possible forecaster I would say, when people are writing papers saying, "No, there's no stock bubbles."

Diamond: That's when you're having a stock bubble. And right before 2008, I don't want to name the authors of this paper, it's a famous, highly guarded paper and within its literature, it is very good work, but there is a famous paper written, I think it was either 2006 or 2007 that claimed with just sticky wages and sticky prices, but many different rigidities in different sectors they finally had a very good quantitative fit to the whole economy. And I would say when somebody's saying that that's the best possible predictor that a financial crisis is going to happen.

Beckworth: Yeah. Very interesting. And I like your point about generational influences and understanding what the current or next crisis is going to be. So you mentioned academics and policy makers in particular, they should have a good long history. I remember after the great financial crisis, many commentators saying that the people who were trading mortgage back security, CDOs that all this stuff backed by real estate, that they had very recent experience, they hadn't known any national downturn in housing. All they thought was housing prices had to go up. And I recently interviewed Paul Crewman and he made a similar point about bond traders today.

Beckworth: He goes, "David, you're putting too much confidence in what the bond market is forecasting for inflation." He goes, "There's a bunch of 20 year olds trading government bonds right now and they didn't live to the 1970s." So he said, "You want to take it with a grain of salt," which I think is a good point. So all of us, policy makers, academics, traders, we all need to know of our history, I think is an excellent point. Well, let's go to your papers and I want to start with one from 2020 and it's “Safety Transformation and the Structure of the Financial System.” And I like it because it helps motivate safe assets with a general equilibrium model. So, walk us through your paper and what you find.

Safety Transformation and the Financial System

Diamond: Sure. That paper had two motivations that happened to tie together quite tightly. One was looking at the pre 2008 financial system, how it differed from things before in particular, the fact that the banking system was so heavily dominated by securitization. If we think about the theories we have of why banks exist, previous work emphasized bankers who are making illiquid loans to borrowers with whom they had a relationship because bankers had this store of private information that they'd acquired, they wanted to fund themselves in a relatively information and sensitive way. And that was the explanation for why banks would issue liquid deposits while holding illiquid loans. It always had something to do with the banker knowing more than the market about the value of some of their underlying assets. And relationship lending is not over, but I think it's decreased dramatically over time. And I wanted to write down a model that still is able to replicate the basic facts of banking.

Diamond: Why are banks invested almost entirely in debt? They don't hold equity securities, for example, except a tiny bit for market making. Why are they so highly levered? Why are they funded primarily with deposits with just a little bit of bank stock issued to bear the risk in their asset portfolio. And what I realized is if you think banks are designed to create safe assets, this is the most efficient way to do it. So, the model has two basic ingredients. One, I assume that there's a demand for safe assets. I don't have a deep answer for why money-like assets are demanded in this paper. And I actually think much of the literature has totally gone around the question of where does money demand come from? We just can observe it in the data, fit a curve to it and say, "Here's how big it is."

Diamond: But taking that as given suppose riskless assets, which are roughly what a money-like asset would be. If there's a special demand for that, then you want to structure your banking system to create as many riskless assets as possible. How should a bank do that? Well, they should invest in a portfolio that's relatively low risk. And that's why I say banks invest in debt securities rather than equity securities. And then, because the debt securities they invest in have some risk, they need to issue some stock as well. But if banks find it either costly to expand their size or costly to issue stock because of maybe transactions costs of bargaining with the investment banker who does the road show of getting your stocks placed, you want to minimize the amount of stock you issue while maximizing the amount of safe assets you create.

Diamond: And the optimal way to do that is for the bank to hold a diversified portfolio of debt securities, which looks like holding mortgage back securities, the more modern securities based banking system. The second motivation was for thinking about policy in particular, thinking about quantitative easing. There had been worries at the time I was working on this paper that quantitative easing might have either been good or bad for financial stability. I presume your listeners are relatively macro types, but just to make sure, quantitative easing is when the Central Bank would purchase assets such as treasury bonds or mortgage backed securities. And then they would pay for it by creating what are called bank reserves, which is effectively a type of money that lives inside the banking system. This is sort of the Central Bank doing banking on its own. It's buying bank assets and issuing money, just like a bank would.

Diamond: And what I wanted to do with this paper was have a complete model of the financial system where I could then ask questions about what does this do to the riskiness of banks? What does this do to how much leverage the non-financial sector is taking on? And the simple answer that you basically get out of it is if the Central Bank is doing banking itself, that crowds out the need for private sector banking and you substitute to a smaller and less risky financial sector in response to this. So, I was finding in the model actually that quantitative easing is good for financial stability as long as the government itself is stable enough to hold risky assets and be able to back everything with credible taxation when needed and so forth.

Beckworth: Very interesting. And I was going to come to this later when we talk about QE, but you just brought it up. Your model says that QE crowds out the services that would've been done by the private sector. Is that right?

Diamond: Yes, correct. That's right.

Beckworth: That to me speaks to the critique of QE that's often brought out, the Modigliani–Miller or QE irrelevance or Wallace neutrality. I've had a number of guests in the show recently we've talked about this. So, I hope listeners aren't tired of hearing about Wallace neutrality, but that's fascinating because what you're are saying is that QE is just really the Fed doing what would've been done otherwise by financial services. But the difference is it moves the risk and puts it on the government's balance sheet.

Diamond: That's right. Then the question is fiscal risk for the government versus financial sector risk in the private sector. And if you wanted a complete model of how much QE to do, you would have to think about this trade off and that would get into how credible, the ability of the government to raise taxes and so on would enter into that picture.

Beckworth: Yeah. That's fascinating. I didn't realize an implication of your paper. So it underscores this critique of QE, what it actually accomplishes. But there's another implication of your paper I believe, and correct me if I'm wrong, but it really does explain why the US is such a financialized economy, it speaks to why we provide this service to the world and there's related literature, Helene Rey and P.O. Gourinchas has this paper about how the US economy as a whole is a banker to the world. They go in the States's as a venture capitalist to the world. But basically if you look at the consolidated balance sheet of the US economy relative to the rest of the world both public and private, it looks a lot like a bank's balance sheet.

Diamond: Absolutely.

Beckworth: And so your paper, your model can be generalized to the US economy as a whole, and it helps us understand why finance is so important in the US. Is that right?

Diamond: Yeah, totally. I don't mean to claim credit for that point personally, because I can point to many others who've written about that. One paper I had single out is by Caballero, Farhi and Gourinchas in the AER 2008 where they took a model where there's multiple different countries who have, they assume to have different degrees of financial development and they get the result that the country with the most developed financial sector naturally takes on this role and is able to match a range of style aspects, both about the financial system, but also say the size of trade deficits as well in exchange for providing monetary services, the rest of the world sends goods to the US in exchange for that.

Diamond: And yeah, I think that's a general point that, I haven't seen very good data about this, but my sense is other currencies that became the global reserve back when the British pound was a global reserve currency. I think this is a natural combination, that you have to be running a disproportionately large share of the global financial system as well. Whether that's good or bad from a welfare perspective, I think is super difficult question. But just in terms of here are facts about the world and here's a framework that can explain why that's true. I think academia has done pretty well in understanding that point.

That's a general point that...other currencies that became the global reserve back when the British pound was a global reserve currency. I think this is a natural combination, that you have to be running a disproportionately large share of the global financial system as well. Whether that's good or bad from a welfare perspective, I think is super difficult question.

Share this

The Global Role of the Dollar

Beckworth: Yeah. We've tried to wrestle with some of the welfare implications in the past. I'll just throw a few out there. You don't have to comment on it, but I'll throw a few out there. We've had Matt Klein on the show and he has a book out on all the trade imbalances and the argument he makes along with Michael Pettus, who's a professor I believe in China. He's American who works in China, is that this special role that the US economy plays to the world has led to an erosion of traditional jobs, manufacturing jobs, this shift into finance, the dollar tends to be a little bit higher in value than otherwise would be the case because of the demand for safe assets. And they look at it and say, "Look, this has a real cost on society. Not everyone goes to work on Wall Street.

Beckworth: And so there's real tradeoffs there. And I think that's a fair critique. There's a flip side to that as well though, from a US perspective, it definitely provides real resources that the country wouldn't have otherwise received. And I would take a step back and I'm just conjecturing here Will, so you can stop me if you think I'm way off. But I think all the good that globalization has done, so the billion people without poverty in Asia, globalization could not have happened without a global medium of exchange. Maybe it's not a sufficient condition, but it's definitely an important part of the story. If you didn't have the dollar facilitating trade, there'd be a lot more poverty in the world because there wouldn't have been as much globalization. So, I think that you're right, it's difficult to maybe do the bottom line and calculate the net outcome. But I think there's some good things you can point to as well as some of the concerns. Any thoughts?

Diamond: Yeah, I would say overall, I totally agree with that. Both the pluses and the minuses, I guess if we had to make lists of people who gain and lose because of the global role of the dollar, the losers would potentially be people who hypothetically would be making cars in Detroit that would be exported to the rest of the world, which the strength of the dollar can work against. Quantitatively, how big that effect is, I have no idea. I'm sure somebody has tried to answer it, but that seems fundamentally a difficult thing to get at. But yes, broadly speaking, having a reserve currency makes your currency strong, crowds out the need for people who export goods from your economy, because you're essentially exporting finance instead. And then on the other side, that's a very deep question, to what extent was the dollar itself crucial for globalization versus say international trade agreements versus say a low risk of wars and things like that. I think these things are probably very hard to disentangle from each other.

Diamond: And I guess as we start to converge from macro, which is highly uncertain to international relations, which is extremely difficult and uncertain, my confidence levels gets even lower. But yeah, I would definitely say that the Deng Xiaoping era in China, which is probably the fastest decrease in poverty anywhere in world history, if there isn't a working global trade and financial system that becomes much harder to do because exporting was such a big deal for that. And relatedly, I should say, I  don't know if you've ever read Dani Rodric, he's a professor at the Kennedy School he's argued for a while that having a weak currency is useful for economic growth in developing countries because it can help you build up your manufacturing base. And this is getting far from my research. So, just to make sure the audience take this with an extra grain of salt for me.

Diamond: But manufacturing exports never have quite functioned like a perfect market. Every country that succeeds at having export led growth often seems to be involved in export oriented industrial policy to various degrees. China's one example, I guess a lot of the so-called East Asian tigers, Korea, Singapore, and so on have thought about these questions too. That doesn't mean that they didn't have free trade. But the idea of say infant industry getting subsidies or strategic sectors being treated differently maybe if you make inner immediate inputs, which are useful for things which you export on a final level loosely, I would say the rapid expansion and growth stories tend to have some of that involved. On the flip side, I would say countries that have not done well in globalization also claim to be doing industrial policy too.

Diamond: I guess Latin America has talked quite a bit about import substitution as a growth model, which is consciously designed. I would say most Latin American economists, which I interact with on Twitter, that's something which they joke about as not being very smart repeatedly from I guess, personal experience. So yeah, government management of the economy is difficult. I'm not a believer in the purely frictionless efficient market where a perfect government can't improve on things in many ways. But if we're comparing real world governments versus real world markets, only the very best governments I think are ever able to consciously intervene in ways to make international trade work in their favor.

Beckworth: Yeah. Very complicated question instead of issues to go through, but it's always fun to do it on the show. We pretend we have it all figured out and we don't. But one other point before we move on to your other papers on convenience yields and safe assets, you alluded to a paper, I believe you said by Caballero and maybe Farhi and Gourinchas that speaks to how a reserve currency emerges, how you get this dominant currency in the world.

Broadly speaking, having a reserve currency makes your currency strong, crowds out the need for people who export goods from your economy, because you're essentially exporting finance instead. And then on the other side, that's a very deep question, to what extent was the dollar itself crucial for globalization versus say international trade agreements versus say a low risk of wars and things like that. I think these things are probably very hard to disentangle from each other.

Share this

Beckworth: And I find that fascinating. There's been several papers along those lines. There's another one by Arvind Krishnamurthy and another co-author. And they show that once you get something like that going, you can end up in a world where you have nowhere else to go. That's the term they used, nowhere else to go equilibrium, that the country providing the most safe assets or debt fixed income securities are relatively safe. It's like a lock in effect, path dependence lock in effect. And there's been several papers like this. And I have conjecture that I think it's means it's going to be really hard to displace the dollars role in the global economy more so than maybe even the pound. We often look at the pound as a comparison, but the network effects continue to grow.

Beckworth: In fact, I know you're familiar with the BIS's work and the data they collect and they show overseas dollar creation continues to grow. In fact, during a crisis, it grows even more. Some of your papers we'll talk about speak to this. Your work shows that the convenience yields actually increase during the crisis and you see dollar creation outside the US during crisis. So, it'd be interesting to see where this all ends up. And I've always asked people this question and so Will, I'm going to ask you since you're a modeler and you think about theory and you think about safe assets in the dollar. If we could run the world a million times through simulations and we tweak things just a little bit here and there a little shock here, a little shock there, the massive Monte Carlo simulation. Do we always end up at an equilibrium where there's one dominant currency? Do these path effects, these lock in effects, would this be what you think we would normally see?

Diamond: I would say normally, but not always. And if you look at the transition from the pound to the dollar, they both were global reserve currencies to some degree in the intermediate stage. I guess there's this, I could be getting the words of this wrong, but a quote from Rudi Dornbusch who's an old MIT macroeconomist that I think crises happen way too slowly and then all at once. And that's my sense of how reserve currency transitions could happen. That gradually some other currency starts to be used in international trade and in international finance. And only once they've gotten a minor market share is there any sense of them rapidly growing to a major market share. But yeah, I think it took huge shocks to have the dollar involved. And this is a one liner, which I sometimes facetiously like to use is what American exceptionalism really is, is we had an ocean between us and the world wars.

Diamond: So Britain was just absolutely demolished as was everybody else in Europe, I guess, maybe Britain less so because they had a channel at least, but still that huge amount of damage is what it took for them to lose their reserve currency status and for the US to enter instead. And in 1950s US just completely obvious, we have a huge economy relative to the rest of the world because everything else seemed to get burned down. If something like that happened to the US, absolutely, we could lose the reserve currency too, but it's not something that's going to happen overnight. And in this lock in effect you talk about, one of the really crucial things that that paper emphasizes, which I think is quite true is you don't just have to be very financially stable in order to be a good reserve currency.

Diamond: You actually, in some levels have to be a little bit financially propagate too. There has to be a lot of debt denominated in your currency out there in the world. Otherwise there's just an enough to go around for people to circulate and use it as money. And I guess the ideal thing would be a country that is both financially sound and has a gigantic amount of debt. Maybe the way to do that is you always run government surpluses, but you have a gigantic sovereign wealth fund invest in equities to make sure that your debt is backed by a tangible asset, I think Singapore does things like that, but most countries have not tried anything of that sort. But until that happens, I think you need a country that is really big and has lots of debt out there in the world in order to start competing with the US.

Diamond: The other aspect of this is the linkages between trade and finance. If you are, let's say a Japanese firm trading with a Brazilian firm right now, most likely even though neither of your countries uses the dollar, that's going to be a transaction which is denominated in dollars. You'll send an invoicing in dollars, not in either of your local currencies. And as a result, if you're doing all of your trade in dollars, you probably want to have a bank account in dollars as well. So these lock in effects, it's within finance, but it's also the finance trade connections. And even as we were getting it before, I would say the security international relations questions start to enter here as well. That's far beyond what serious quantitative scientific or quasi scientific economics can be about. But it always seems to be the case that the political hegemon of the world is connected with who has the reserve currency too.

Growth of the US Debt

Beckworth: Yeah. And what's interesting is the implications for that hegemon. So the US currently is that, questions about public debt sustainability really are much more complicated and simply looking at the stock of debt. I mentioned at the beginning of the show we have a 100% of debt to GDP ratio, but that number is reduced form. It really doesn't tell us what's driving the process. Is it the demand for debt? Is it reckless supply of it? Combination of the two? And I would go even further and this is the point I've made in some of my work is that it may even change the signs on the relationship. If the US wasn't supplying all this debt to both domestic and foreign markets, there might be additional financial crises that emerge, which would lower real growth, lower job creation, all those things. But it's a real tough story to tell. And it's counterintuitive, politicians don't get that, even some fellow scholars, that's a big pill to swallow that the signs actually change on that. Has that been your experience too then?

Diamond: Yeah, I would say I'm broadly in your camp on this and I think opinions on this range quite a bit. And I think it comes down to, on some level, what is the value of having the printing press for the reserve currency. In a purely real model where the value of an asset is a present value to cash flows, that doesn't matter at all. In an extreme model where you can never substitute away from the dollar printing dollars is a dramatically powerful instrument for controlling the global economy. And I guess you've had both Hanno Lustig and Nathan Tankus on your show previously, I think.

I think it comes down to, on some level, what is the value of having the printing press for the reserve currency. In a purely real model where the value of an asset is a present value to cash flows, that doesn't matter at all. In an extreme model where you can never substitute away from the dollar printing dollars is a dramatically powerful instrument for controlling the global economy.

Share this

Beckworth: I have, yes.

Diamond: I seem to have the same argument with both of them about this, but from the opposite side. So, I would say what extreme is the modern monetary theory camp, which thinks there's a very big power from having the printing press Hanno was trained as a more classical asset pricer. And he's gotten interested in these money related questions, but his toolkit often is based on taking real present values of cashflow is generated by the government and comparing that to values of debt as they're price in the market. And I think the reason why there can be so much disagreement here is we really don't have a good model of what it is for an asset to behave like is money. So in my work, that's the part I skip over entirely. Why do people demand money in my job market paper? The safety transformation paper you mentioned, money in the utility function. I assumed it, that's it. And the reason why I did that is I realized nobody else is much better on this either.

Diamond: There's some work by Randy Wright and some of his co-authors that gets very into the micro of transactions, but that literature hasn't gotten into these macrofinance policy issues so much. And I think when you have these arguments between, I guess, Hanno and Nathan would be very good representatives of each point of view, one will say the fact that we have debt in our own currency is extremely important here. The other might say, well at the end of the day, an asset value is the present value of the cash flows it pays.

Diamond: And on some level, these statements both have to be true, but I don't think the boundary between the implications you could get from thinking both of these ways is well mapped out at all. We don't have a good story for when are people going to stop viewing the dollar as a money-like asset, how irresponsible would fiscal and monetary policy have to be for people to say, "Nope, I'm going for either the Euro," or I guess if I'm going to take a Twitter favorite, "No, we're all switching to cryptocurrency," or whatever that might be that I'd say just a very big known unknown, which I think is very... Even turning that into a well defined research question is quite difficult to get started.

Beckworth: Yeah. Two points to that. First, you mentioned Randy Wright. And I mentioned earlier, my friend and colleague Josh Hendrickson, and he follows in the vein of Randy Wright. He does monetary search model so he motivates from a micro foundation and he was gracious enough to allow me join him on some papers where we did monetary search models. And we sent him off to some good journals and I can tell you the reception was, what is this? They know what it is, but that's the sentiment, like this is not the main approach, this is not standard, we don't like these monetary search models. They want the more familiar. So, I totally have experienced firsthand the fact that that approach really isn't gaining ground. Maybe it is, maybe on the margins, I'm not in steeping that literature.

Beckworth: Second thing though, is this point you're raising about how do we value treasuries and public securities from the main provider to the world. And I was going to get to this question later, but since you brought it up, let's tackle it now, we will get back to your papers. But this is really a question about how do you explain the gap between the market value of US treasuries, and in a fundamental valuation. So if you go out, you can literally get estimates of primary deficits or surpluses, bring them to the present, discount them to the present, which is what Hanno has done. And there's a huge gap. And so, I've also had Markus Brunnermeier on the show, which you know, and he has a paper where he tries to reconcile this by looking at bubbles. So, his paper is about take the fiscal theory of the price level, add in a bubble term, in his case, he derives a bubble term and that resolves it for him.

Beckworth: And since then I've come across several other papers, do similar things. David Domeij, and Tore Ellingsen 2018 had a paper in the journal Monetary Economics titled “Rational bubbles and public debt policy.” And then Chris Waller, one of our governors has a paper in the journal of Economic Dynamics and Control titled “Liquidity Premiums on Government Debt and the Fiscal Theory of the Price Level.” And they all reached the same point. The way you can explain it is this bubble term that emerges because of the moneyness, liquidity of these securities. And so, I've brought this up in other contexts, and the feedback I get is well, is that term quantitatively important, maybe it's theoretically important, but can you really explain the entire gap between market value and this present value of cash flows from the government, can you explain it with that term? And I don't know. To me, it's the only explanation, but have you thought about that? How do we reconcile that?

Diamond: Yeah, this is a very difficult question you're raising. In general I would say, explaining the level of any asset price is quite difficult. And a lot of the tools that were applied to thinking about comparing, say primary surplus is paid out by the government versus the market value of debt it's related to another literature on why is the stock market value to how it is. And I hope I'm not getting myself in too much trouble here, but my one liner to summarize that whole range of literature is all of the false viable theories are false. And what we've ended up with is ways of explaining why say the stock market outperforms risk free short term debt. What's explaining that risk premium. Some people say it's about consumption going very far into the past. That's called the habit formation class of models. Other people say it's about the relationship between consumption today and consumption, going very far into the future. That's called the long run risks class of models.

Diamond: And other people say it's about extremely rare events that we don't actually observe in the data. And I would say these are the three different ways where you can hide your explanatory variable in stuff, which is hard to pin down with actual data. And I think that the difficulty we've had with pointing to a direct tool for saying, this is why the stock market has a 6% instead of a 7% rate of return on average or whatever would happen to be, I think applies as well to this literature if we're thinking about government debt valuation.

Beckworth: All right. Well, let's move on to your two papers that we're going to cover before the show ends. And these two papers speak to convenience yields and risk free interest rates. I'm going to go ahead and mention them Will, and I think we can cover both of them together, but you have a 2022 journal financial economics article with some colleagues and the title is “Risk Free Interest Rates.” And the 2021 working paper, “Risk free rates and convenience yields around the world.”

Diamond: That's correct.

Beckworth: Okay. So maybe to start us off with some definitions, what is a risk free rate and what is a convenience yield?

Risk-Free Rates and Convenience Yields

Diamond: A risk free rate is just the interest rate you would earn on a purely safe asset. It's the exchange rate in some sense between money today and money in the future. If I know next year you're going to pay me a dollar, how much that dollar is worth to me today is directly connected to the level of the risk free rate. If it's a 3% risk free rate, a dollar today is worth the same as a dollar plus 3 cents of a year. A convenience yield is one of the determinants of risk free rate levels. In a pure model with no money demand, no financial frictions the value of money is just based on what you can purchase with it. So, I would buy a treasury bond just because next year it's going to pay me a coupon payment and I'll use that to buy my coffee. But if you actually value the treasury bond as a safe money-like asset directly in and of itself, that's what we call the convenience yield of it.

Diamond: So these two terms, the terminology I like to use is the time value of money is the first conventional piece. What is the implicit exchange rate between consumption now and consumption in the future? Or I guess to be more exact, a dollar now and maybe a dollar plus some interest paid next year. But then in addition to that, the convenience yield is the amount of value you get actually from holding the treasury bond. For example, because you can post it as collateral in money markets or because it allows you to satisfy your regulatory capital requirements with the bank regulator, things like that. And these papers are about one way of measuring that convenience yield by backing out a risk free rate from an asset that we argue does not provide any money-like services, the way the treasury bonds do.

Beckworth: Just to be clear on the convenience yield. So, it's the return one gets from holding a liquid asset or alternatively what you're willing to forgo in terms of interest because you value liquidity service, it provides you.

Diamond: That's right.

Beckworth: So, it's safe to say that even the physical cash in my wallet right now has a convenience field on it.

A convenience yield is one of the determinants of risk free rate levels. In a pure model with no money demand, no financial frictions the value of money is just based on what you can purchase with it. So, I would buy a treasury bond just because next year it's going to pay me a coupon payment and I'll use that to buy my coffee. But if you actually value the treasury bond as a safe money-like asset directly in and of itself, that's what we call the convenience yield of it.

Share this

Diamond: Yes. And the very old IS-LM model was the first actually to think about this, the convenience yield on your cash, if you think about treasuries as just the interest rate, things have gotten more nuanced since then, if treasuries pay 4% and your cash pays zero, the convenience deal was 4%. That's the first benchmark to start with.

Beckworth: Very nice. Okay. Let's go into your findings. What did you uncover? And I like the fact that you went around the world in one of your papers, but in the other paper where you stay home, you develop a term structure of convenience, which is also cool. So, you look at convenience yields across different maturities. So, maybe start with the term structure first. What did you find in terms of terms structure in the US of convenience yields?

Diamond: The first thing we found is that across the whole term structure, convenience yields are relatively similar in size. It's almost exactly 40 basis points whether we're looking at three months, six months, one year or two year maturity. What that would mean is for every dollar you invest in treasury bonds, you are giving up, I guess that would be 40 basis points. That would be 0.4% of interest in exchange for the money-like services of that bond. And the way we back this out is we want to come up with what we say is a risk free asset, which is not money-like which is a little bit self contradictory because any asset which is safe enough to ignore credit risk or anything like that, you can easily start posting that as collateral in money markets as well too. So, our approach was to find risky assets, which jointly could be used to create a risk free payoff. And we were very fortunate that there's something called the put-call parity relationship in option pricing.

Diamond: Where if you look at options written on a stock index, for example, there's a trade where you can buy some options and sell other options called the box trade, which replicates a riskless payoff. So, combining four options buying two of them and selling two others that's where the box comes from, because there's four to make up the box. You can create an option trading strategy that has a riskless payoff. And then we compare the yield on that versus the yield on treasuries. And the difference is our convenience yield estimate, which averages about 40 basis points difference.

Beckworth: Will, does the convenience seal disappear if you were to go beyond that, say five, six, seven, eight, nine, ten years out?

Diamond: Unfortunately what disappears is our data set beyond that point at least in this first paper. So, as far as we can look, it's accepted extremely short maturities where they're a little bit higher like treasury bills, for example, have a higher convenience yield than treasury bonds, bills are special, short maturity debt that's treated differently in some regulatory ways, money market funds can hold it for example. But once you throw out that little piece, it's flat across the whole sample. And the newer international paper, we have some of maturity data in Europe. I haven't looked at it yet, but hopefully at some point I can tell you what the five, six or seven year convenience yield looks like in Europe. I'm hoping we still get our flat result too, because it was nice and elegant.

Beckworth: So Will, what you're saying is very, very short maturities, there's a greater convenience yield, but after that it's relatively stable.

Diamond: Yes, that's correct. It's very flat from three months on, extremely short maturity safe assets are almost exactly money, on some level you can say a dollar is an overnight riskless asset because you always have that dollar with you whenever you want it. And things which are a very, very close substitute for actual cash are a little bit extra special. But beyond that, that's exactly right.

Beckworth: Well, I would argue that maybe even the 10 year has some convenience yields on it, 10 year treasury, because it's lower than equivalent 10 year AAA corporate bond. So there's still some moneyness on a 10 treasury and it's also benchmarked. So it's clear.

Diamond: And then there's the difficult question of the AAA treasury spread. A little bit of that is probably credit risk for the AAA bond. A large portion could very well be this convenience yield and why we liked our approach using options prices we felt this was an example where we could say there's no credit risk here. And at least for shorter maturities, we could disentangle those two. But I agree with you, if you could say remove the credit risk from a private AAA bond, there's going to be some convenience yield left over too. It's just hard to measure.

Beckworth: Okay. Let's talk about your other paper where you go and look at, I believe nine of the G-10 countries and come away with convenience yields for them. What were some of the big findings you found there?

Diamond: One of the things we were really interested in in this paper was trying to quantify in what ways is the dollar special in the global financial system. So we know the global banking system uses a lot of dollars. We know international trade uses a lot of dollars, but is there anything directly observable where dollar denominated asset prices are behaving differently from say Euro or yen denominated asset prices? In terms of convenience yields, actually we find that the dollar is not special at all.

Diamond: If you look at the cross section of convenience yields in different countries, it's very well explained by the level of interest rates in countries. So, Australia has high interest rates and a large convenience yield, there we find about 60 basis points. Japan has very low interest rates negative even sometimes. And we actually quite often find that Japan is a negative convenience yield too. So this idea getting back to the old LM curve from IS-LM that the nominal interest rate is what you forego when you hold cash. That explains the cross sections of convenience yields very well. So the spread between cash and treasuries, that's one notion of the convenience yield. Treasuries are still somewhat money-like, but maybe cash has one unit of money niche treasuries have 0.25 units of moneyness, now that that number is precise, just some fraction of that.

Diamond: And if there's more or less money-like assets like that, then just the relative rates of returns between these have to satisfy this pattern. This gets back to a paper by Stefan Nagel, who showed that this explains data quite well in the US, that the level of the nominal interest rate explains the time series of US convenience yields quite well. And what we showed is the cross section of different countries. The level of the average nominal interest rate in each country lines up almost perfectly with our convenience yield estimates, and the US has a middling convenience yield and a middling level of interest rates on both four bigger and four smaller of the countries we have. So in this sense, the US is not seeming to be special in the global financial system.

Beckworth: Okay. When I read that, I was a little surprised. Here it'd be my way to recover to make the US special. So, the US, as I read, had a convenience yield of about 34 basis points. Am I reading that correctly?

One of the things we were really interested in in this paper was trying to quantify in what ways is the dollar special in the global financial system. So we know the global banking system uses a lot of dollars. We know international trade uses a lot of dollars, but is there anything directly observable where dollar denominated asset prices are behaving differently from say Euro or yen denominated asset prices? In terms of convenience yields, actually we find that the dollar is not special at all.

Share this

Diamond: Yep. Depending upon the sample, sometimes it's 34.

Beckworth: And versus Australia that has 60 basis points. And then in some cases, Japan and Switzerland have negative basis points. I want to come back to them too.

Diamond: And I'll note, we recently added Denmark. Denmark is big and negative too, and they also have more interest rate, so that seems to be robust, that negative convenience yields are a thing when you have low interest rates.

Beckworth: I want to come back to that. I think that's very fascinating, but here's the way I try to reconcile this with this notion the US is banker of the world and we get some convenience yield and I do it as follows. Well, it's middle of the road or middle of the estimates, 34 basis points, it's more than made up for in terms of volume, right?

Diamond: Yes.

Beckworth: So, if you think of total profit or if I might use the term total seigniorage, the US is still just raking it in with even that modest amount.

Diamond: Absolutely.

Beckworth: Okay.

Diamond: Yes. The other thing I will say here is this convenience yield that we're estimating, that's a measure of in your country, a safe interest rate that doesn't provide monetary services versus a safe interest rate that does, there's a separate question of comparing interest rates across countries. So, a convenience yield the way we estimate it is two interest rates denominated in the same currency to spread between them, that tells you what's given up the lower interest rate here has to be more money-like, what are you foregoing instead of holding the less money-like asset? Another thing you can do is take foreign interest rates and use currency derivative to swap that into what they call a synthetic dollar interest rate. So you can take, say a Euro bond, combine that with currency forward transactions and turn that into a dollar interest rate exposure, and then compare that to dollar interest rates in the US. That's called a measure of covered interest parity or when it doesn't line up, that's called a covered interest parity deviation.

Diamond: And in this sense we do find that the US is special, but what this is implying though, is it's not specifically US treasuries that are special. It's broadly US denominated dollar interest rates. So, triple a US debt, I think seems to earn something like this. There's another economist Gordon Liao who's looked at corporate covered interest parity deviations seems to show up there. We're showing in these option implied rates, the covered interest parity deviation seem to be showing up too. And so, if you look at every country in our data set, other than the US, if you swap their interest rates to dollars and compare that with the dollar interest rate, the dollar interest rate seems to be lower than all of the synthetic ones swapped in. So the US is special. The US financial system is special, but not uniquely in specifically US treasury bonds.

Beckworth: Okay, Will. I want to take that observation you just made and apply it to a issue we discussed earlier. And that is how do we reconcile the market value of US public debt with theoretical attempts to look at discounted present values of future cash flows for the US government. And there's been attempts to reconcile the gap between that looking at convenience yields, Hanno Lustig, we mentioned earlier, he has a paper Debt Valuation Puzzles. And at most, one of the versions of that paper looks at, it explains about half of the spread in some versions it's less than that, but what if we could reframe it this way and say, "What we want to discount back is not just the convenience yields on US treasuries, but all of the securities the US financial system issues."

Another thing you can do is take foreign interest rates and use currency derivative to swap that into what they call a synthetic dollar interest rate...And in this sense we do find that the US is special, but what this is implying though, is it's not specifically US treasuries that are special. It's broadly US denominated dollar interest rates.

Share this

Beckworth: We talked about the US being a special place. Now, then the question becomes, well, how does the US government tap into the seigniorage that's going into private firms? And one answer could be, is that the market understands or maybe investors understand if push comes to shove, the US government can impose financialization requirements that allows it to squeeze seigniorage from private firms. So in other words, if we're going to discount back both the primary surpluses as well as these convenience yields, we need to discount the potential that the US government could tap into, not just treasuries, but in a war time situation, for example, it could force firms to pay up more to share some of that seigniorage. So that would be one way for me to reconcile that, is that going too far?

Diamond: I think that makes quite a bit of sense. And I think there's two aspects to what you're saying. One, is that the convenience yields that we're measuring in the literature are too small in some sense, which I agree with. Really what you want to do is find completely generic assets, which are not risk free at all, and figure out hypothetically if you added a safe payoff to those assets, how much would their value go up by, just pure fundamentals of present discounted value of cash flows with no monetary frictions involved at all.

Diamond: And I think my co-authors and I, we made some progress by using options prices to do this, but I also think it's the case that dollar interest rates as a whole, if they're affected by the US being the reserve currency, even if you don't have to tap private sources of seigniorage too, you're already saying that we're underestimating the size of that convenience yield. So, if they got to 50% plugging in either my co-author's numbers or some of the other competing ones in the literature, to me that makes it sound like a real measure of the convenience seal.

Diamond: There's always this catch 22 if you write papers on measurable things and in macro, the questions always depend upon unmeasurable things, quite likely that convenience seal would be much larger. And I think one suggestive piece of evidence of this is if you look at the cross section of different stock returns, often risk factors there, at least in some papers people write are not very good at explaining the rate of return. So, one thing we can talk about is the average stock seems to outperform short term treasury bonds. That's a pretty robust finding that's called the equity premium.

Diamond: There's a separate question though, if we just look across different stocks and try to fit some risk model and extrapolate to what the zero risk interest rate would have to be, often that lies quite a bit above the interest rate for backing out from options prices here. Now, that's intrinsically a difficult exercise, because you have to know the true model of the world, which none of us ever would. But I'm inclined to say that this dark matter of what's unmeasurable of risk factors we can't quite quantify, there's quite a bit of hidden convenience yield lying in there too.

Beckworth: Very fascinating. I look forward to seeing more work being done on this, including yours. All right. One last question we're running out of time Will, but I want to go back to Switzerland and Japan, in your sample and you mentioned Denmark is added as well, but what's interesting about those two countries? Is there are also countries that have experienced really low inflation and in fact they've had deflation and if I focus on Japan in particular, they also have a huge, huge amount of public debt. Now, maybe I say huge too much. It's over 200%, I think debt to GDP. And one of the explanations one can give as well the real demand, the money demand for that debt is high. But what you're suggesting is that's not the case relative to say some other country, the convenience yield is actually negative, correct?

Diamond: Correct. Yes.

Beckworth: So, it raises some questions I guess, about what's going on in Japan and maybe to the lesser extent what's going on in Switzerland.

Diamond: That's right. But this again gets back to this old point that in IS-LM macro, you have to clear the fundamental side and the monetary friction side jointly.

Beckworth: Oh good point.

Diamond: What I think is true, definitely of Japan, but I also think of Europe is large savings, which pushes down not just safe money-like interest rates, but rates of return as well. So, if you compare say Japanese risk free rates to how their stock market has performed over the last few decades, just all Japanese rates of return have been quite low for 20, 30 years, maybe even longer. And if you naturally have low rates of return and low inflation, because people aren't consuming very much, the demand for goods is low, low consumption demand means there's not pressure to increase consumer prices, then you get stuck in this area of low nominal real rates, because you have both real rates and low inflation. And then I guess that's all the fundamentals, the IS side of the IS-LM model.

Diamond: And then, the LM curve is just telling you the nominal interest rate tells you in a liquidity scarcity sense what's going on with convenience yields and you naturally have this an unintended side effect. So, I don't think anybody is ever thinking about how many 80 year olds there are in Japan as a determinant of the severity of financial frictions but if you go through this relatively simple chain of reasoning, even two market general equilibrium thinking can lead you to these interesting indirect answers.

Beckworth: That's interesting. And this goes back to what you said earlier, how do you reconcile the US which seems to be in a boom with the low yields and that story there would be the convenience yield where in Japan, it's not having the boom, it's more stagnant, slower growth and therefore the convenience yields are getting compressed along with other return.

Diamond: Although in a longer time series US rate has not been zero the way it's been in Europe, but it's certainly gone down over the years. I guess I make a big deal out of around maybe the early 1970s there being a structural break where US productivity growth fell. And this is averaging out a lot of booms and bust overall. But I do think the US is in a low growth long term equilibrium. People complain about America is declining, we're no longer the global empire and so on, compared to the rest of the developed world, we're doing great, but definitely compared to who we were 50 or 100 years ago, the rate of growth is not remotely what it used to be.

Beckworth: Fair. Fair point. Okay. Well with that, our time is up. Our guest today has been Will Diamond. Will, thank you so much for coming on the show.

Diamond: Thank you so much for having me. This has been extremely fun and hope to talk to you both either here or on Twitter, whenever you're interested.

Beckworth: See you there.

Check out Ideas of India, and subscribe to Ideas of India on your favorite podcast app.

Photo by Stefani Reynolds via Getty Images

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate