Welcome to this podcast. We are joined today by Professor Cameron Harwick, an economist in New York with interests in monetary theory, institutional evolution, and folk music. He is an associate professor of economics at the State University of New York at Brockport. He is interested in how humans coordinate with each other using tools like money, language, and norms, and his research uses tools from macroeconomics, game theory, and computational modeling. The subject of discussion will be two of his papers. The first is titled “Bubbles and Broad Monetary Aggregates: Towards a Consensus Approach to Business Cycles,” published in the Eastern Economic Journal. The second is a chapter titled “Finance in a Theory of Money.” Welcome to the podcast, Cameron.
Thank you for having me.
Why don’t you begin by talking about yourself, your upbringing, where you were raised in the US, and what motivations or influences led you to pursue a career as an economist?
I grew up in North Carolina. I started out in political science, because one unfortunate thing about American high school education is that you get the impression that politics is the organizing principle of how things work. Going into college, I quickly realized that political science is not a real field; it’s three fields in a trench coat. Economics, though, with the first class I took, I realized this is what I was looking for. Since then, understanding things in terms of incentives has been a very fruitful way of looking at the world. Later on I got frustrated with the formalism of the field, and after graduating I found myself reading a lot of economics. I realized that many of the papers I got the most out of were by authors at George Mason, so it was a no-brainer to go there for grad school. What I love about that program is that there is a real excitement about economics. A lot of places squeeze that out of you in graduate school, but Mason cultivates the excitement of learning how to understand the world. That’s what I really appreciate about that approach.
George Mason gives you a rather different view of how economics is done, because it is considered a non-orthodox research program. Is that correct?
What I appreciate about the Mason approach—and it is difficult for me to imagine being excited about economics any other way—is that it’s theory-first. That has been unfashionable in the past. There’s a Heinrich and Muthukrishna retrospective on the replication crisis in empirical psychology where they argue the reason these empirical fields imploded so much is that they didn’t worry enough about theory. They didn’t nail down the concepts before going to the data. That’s the approach Mason takes: nail down the concepts, then go to the data and understand the data.
So it is not a methodology-driven research program. PhD programs are more about learning the methodology and then tweaking the empirical work. Of course there is some of that, and that can be valuable, but if you focus on the empirical methodology without the concepts…
Yes, and we’ll talk later about a lot of the ways that has failed in economics.
Absolutely. And George Mason in particular is an Austrian School of thought. It is one of the two universities in the US.
I would not necessarily cop to that label, just because there is so much good and bad under it. What I do appreciate about the Austrian approach is the theory-first way of doing economics. That doesn’t always go in directions I feel good about, but as a broad approach to economics I find it very useful.
Because it actually allows you to trace the evolution of the economic thought process. I think that is very critical.
Correct. You’re very focused on real-world referents. If you read a lot of theory papers in mainstream macro, you are often wondering where the real-world referent is. Sometimes you can get too hung up on that, but with the Austrian approach you’re never left wondering, what’s really going on here, or where are the real people in this model?
And the other thing, which we will get to in some of the questions that follow, is that economists of the previous generation wrote very objectively and to the point. Their research is easy to follow as an economist.
There’s definitely a way that graduate training can make you a worse writer.
Cameron, we come now to the next question. This is an objective analysis of Austrian business cycle theory. They focus on artificially cheap credit initiating malinvestment over the course of the boom, which must be liquidated during the bust as the natural rate of interest reasserts itself. You have a paper on this where you talk about the structure-of-production framework, and that investment in industries farther down the supply chain is more sensitive to changes in the interest rate. Can you throw some light here?
That’s the orthodox Austrian business cycle story, which my paper actually criticizes. The structure-of-production framework can be imagined as all capital being goods in process: a brick, or a piece of steel that goes through successive stages – mined or refined, cast, installed, and finally sold as a final consumer good. This was intended by Hayek as a pedagogical example of how interest-rate malinvestments could in principle work. The problem is that most capital is not like that. A lot of the important capital in an economy is durable capital – things where you can’t just attribute certain units of output to certain units of input, because they give you a whole stream of output over time. Things like tools and factories.
Even things we use in our daily lives are like that.
Absolutely. One would hope you don’t have to buy all the productive things at the point of use all the time. The issue is that when we think about durable capital, the stages-of-production construct doesn’t make a lot of sense. Hayek knew this and really struggled with it in The Pure Theory of Capital. He did not adequately solve the problem. Lewin and Cachanosky have a lot of papers where they introduce the idea of financial duration. You can think of that as the interest elasticity of the present value of an investment project: how much does the value of a project change when the interest rate rises or falls? What that captures is the potential for these kinds of interest-rate-driven malinvestments. The problem is that once you start thinking in terms of financial duration rather than the roundaboutness construct, the actual Austrian story looks a lot different. The Austrian story in its canonical form really depends on particulars that do not generalize. What I show in the paper is that interest-rate-induced capital malinvestment just is not enough to explain actual depressions. You need some sort of aggregate-demand explanation. You can bring those together to some extent, but you are not going to be able to do without an aggregate-demand explanation.
Is this connected to the main assumption of most business-cycle approaches, where they start with aggregate demand and supply in equilibrium and then things diverging from there?
Yes. The big difference among a lot of the business-cycle approaches is how aggregate demand gets out of whack in the first place. The canonical Austrian story doesn’t really think about that at all. Aggregate demand is supposed to not be that important. This gets us into the credit-cycle and quantity-theory approaches, which we can talk about later. A big part of the question is how aggregate demand moves. Is it just a question of monetary policy, or can it move endogenously through money supply or money demand?
Yes, we will come to that in greater detail. I will ask you about the economist Leland Yeager. He was one of the foremost exponents of the Austrian School, is that correct?
I don’t think Yeager would call himself an Austrian, although he certainly engaged with Austrian thinkers. He himself was not.
He has a book titled The Fluttering Veil: Essays on Monetary Disequilibrium. There are four parts: Part One is Monetary Disequilibrium and Its Consequences; Part Two is Monetary Misconceptions; Part Three is Cantillon Effects and Other Diversions; Part Four is Avoiding Monetary Disequilibrium. You refer to his books in quite a few places in your paper, and I have read the book. It goes straight to the point. These writers are very concise in what they are writing and how they write it, and the point they want the reader to grasp.
You’ve picked up on the same thing I appreciate about Yeager, which is that he is a very clear writer. He’s somebody who wants to cut through a controversy by breaking it down into the simplest possible pieces and then building back up to an answer until he’s satisfied. A lot of the time that gives you the best statement of a position that exists. For example, the cash-balance interpretation of depression, which is in that compilation, says a lot of the same things as a lot of classical monetarists, but in my opinion it does so much more simply and straightforwardly, in a way that, coming from zero context, you could get a good sense of the actual mechanics of the business cycle. He has another essay, “Individual and Overall Viewpoints in Monetary Economics,” which is a very basic point but such a good statement of it that you keep coming back to it. The point is basically, thinking of things from my perspective as an individual in the economy looks very different from looking at things from an overall perspective. Assets and liabilities are going to net out in an overall perspective, where they do not from my individual perspective. What that means is that a lot of the representative-agent models of macroeconomics are not going to generalize, because you’re mixing up those individual and overall viewpoints. That is something I keep coming back to. For example, what does savings mean? What is the role of liquidity in an economy?
What exactly is the meaning when we say money supply and money demand are not equal? You have a paragraph on this idea, right?
That is Yeager’s whole idea of monetary disequilibrium. As an individual I can build up my money balances by not spending, but then other people have to spend in a compensating direction so that I have something to accumulate. The problem is that if everybody’s trying to build up their money balances, where do those come from if they’re not accommodated by monetary policy? That’s what forces down the price level and leads to short-run real effects.
You don’t want there to be too much of a disequilibrium between money supply and money demand.
Exactly. Because that short-run disequilibrium – it is short-run, but it can have major self-feeding effects in that short run.
When we talk of monetary disequilibrium, the word “monetary” means money, so it is very important that we have a basic premise of how we define money. People should be cautious or circumspect when they do that. In your paper you say: “To call something money is not to make a statement about the good itself but about the expectations in a community that might lead individuals in it to employ the good at a certain juncture in their plans.” So we are thinking of money not as a good, but – this aggregate, if we measure money like this, then this would carry the meaning of the total nominal value of liquidity services being provided in the economy. This is the way you define money, and it is also how you define liquidity.
Liquidity is the service provided by money
But that service also depends on expectations. Why don’t you throw some light on this?
Connecting this to the last question, this is me grappling with a claim by Yeager. What is so great about Yeager is that even when you disagree with him it is very fruitful to think through why. Yeager has this paper, “Essential Properties of the Medium of Exchange,” which is also in this compendium, where he wants to draw a bright line between “money” and “not money”. And what puts something in one category versus the other. Throughout the paper he’s thinking of money as some particular good that we use as a medium of exchange. What I’m doing in the paragraph you mentioned is saying: what if we don’t start with the assumption that money is a particular good? This is part of that theory-first approach to monetary economics that I mentioned earlier.
It goes all the way back to Menger’s conjectural account of the origin of money.
Right. So I’m trying to define money as rigorously as possible in terms of categories of choice. There’s an early Hayek paper where he talks about what makes something an economic good. What’s the difference between a chair and a non-chair? It’s nothing physical about the good itself or the object itself, but how we use it. As economists we don’t care about physical things at all. We care about the wants and needs that they satisfy, and about the relation that they hold to people’s intentions and plans. So if we apply the same logic to money, everybody knows the definition of money as a medium of exchange. But let us go a step further, because “medium of exchange” has a lot packed into it. It means not only that I can spend it, but also that other people will accept it. So we have to ask why people would accept it. What we come to is that in order for anything to be money there has to be a structure of expectations throughout the whole economy that we have coordinated on some good, or set of goods, that we will accept in exchange for things. Because now indirect exchange is a routine part of people’s plans, and the economy can coordinate resource use as opposed to – and I have another couple of papers on this – contrasting monetary exchange as an institution as opposed to, for example, kin networks as an institution for allocating resources.
When you say institutions, you mean institutions like banking networks?
What counts as an institution is one of the big fundamental questions in institutional economics. I would define institutions as a routine, recurring mode of interaction. We have expectations that when I interact with you this way, you’re going to respond this way. When I give you money, you’re going to accept that and give me a good or service. When we schedule a Zoom call, a podcast would be a kind of institution, because we’ve structured the way we are interacting with each other. We have mutual expectations
So we are coordinating with each other.
Basically yeah. Institutions are basically a way for people to coordinate with each other. What we usually colloquially think of as institutions, like banking institutions—Doug North has a paper where he talks about institutions versus organizations. We’re not usually very rigorous about that distinction, but organizations are a formal structure of control, and that might correspond better to what you were talking about.
I was actually referring maybe to an organization of the banking system, the evolution of the banking institutions where all these investment banks have come into the picture. We will get to that maybe. I just want to quote Leland Yeager again here, because anyone listening to this podcast should actually know who a monetarist really is. He has this quote from Part Two, Monetary Misconceptions: “A monetarist is an economist convinced by the evidence that the quantity of money and changes in it dominate the total flow of spending in an economy. The government budget and so-called real factors in the economy, including investment incentives, are distinctly subordinate influences when not paralleled by the behavior of the money supply. The view attributed to monetarists that only money matters is a straw man invented by critics. The monetarist view of what determines total spending is rather that money matters most.” Do you have anything to add to this? I think this sums it up very well.
I think this is a fine definition. People think about monetarism as associated with the equation of exchange, MV=PY, and that’s tautologically true. That just tells us that income equals expenditures. But tautologies are useful when they help us slice up observations of the world in useful ways. And frankly, if velocity or the demand for money were erratic and all over the place, the equation of exchange would still be true, but it wouldn’t be very useful for us because it is not slicing up the world in useful ways. I think that’s kind of where the profession has settled right now. They see that the velocity of M2 since the late 70s and early 80s has been all over the place. And the now-standard New Keynesian framework arose out of that.
There was a pair of op-eds in 1983. Milton Friedman wrote one and Bill Barnett wrote one on the same day. Friedman’s looking at M2 and says, “Oh my gosh, M2 is exploding; we’re gonna see a return of inflation,” and Barnett says, “Well, no, probably not. M2 might not be the right thing to be looking at.” Unfortunately Friedman was wrong. Well, fortunately we didn’t get inflation, but unfortunately because it kind of discredited this view that money really matters, even though there were people at the time who were saying, “Hey, maybe this is not the right money to be looking at.” This is one place that the rubber really hits the road on this theory-first approach, because the monetary aggregates we are usually looking at, like M2, do not make a whole lot of conceptual sense. Friedman kind of knew this; there are brief places in his 1960 Monetary History of the United States where he hints at this, but mostly it is a punt. He likes M2 because up to that point M2 had a good statistical fit. So if you build your monetarist theory with a statistics-first approach like this, it’s gonna break down. It did break down in the 1980s, and that’s why there are not a lot of monetarists anymore. The proper way of thinking about the money supply has not gained as much currency, no pun intended.
Maybe we can say that the orthodox way in which liquidity was defined, has not served the profession well.
Exactly. I think a lot of times people are not even thinking about liquidity services; people are just thinking about the money supply as a stock of dollars, or whatever other currency, and liquidity services are just completely implicit at best.
So M2 in that case can lead to an overestimation of the liquidity services.
It could go either way. On the one hand, M2 has some somewhat illiquid things like mutual funds in it that shouldn’t count as much as they do, but on the other hand M2 is also missing a lot of things that provide some liquidity services. I could imagine at various points it could be an underestimate or an overestimate.
It can be both. So, there are sort of two approaches to how you approach the business cycle. One is the quantity-theoretic approach, which I think makes sense now that we have talked about it, and there are the credit-cycle approaches. Can you talk a little bit about the credit-cycle approaches?
That distinction comes from Bob Hetzel. These are basically two ways of explaining the business cycle. What he calls the quantity-theory approach, is something like your standard aggregate-demand story, where sometimes the central bank is going to overshoot relative to money demand, sometimes it is going to undershoot, and you get a recession. Credit-cycle theories, on the other hand, are gonna tell a much more micro story. Usually these are going to center in financial or credit markets. And there are a lot of different varieties. In my paper I classify both the traditional Austrian story and the traditional Keynesian story as credit-cycle stories. There are a lot of different varieties, but they’re all going to tend to see interest rates as the transmission mechanism, and the money supply or money demand is going to be less important. It is another way of slicing up these observations. What I do in the “Bubbles and Broad Monetary Aggregates” paper is use this constructive definition of money, where we define the money supply as the total value of liquidity services in an economy, regardless of what asset provides those services, and we can tally those up. By defining money that way, what we can do is connect these micro movements in financial markets to aggregate demand in general. These things are going to affect not just money demand but even the money supply as well, because a lot of that is not under the direct control of monetary policy; it is under the indirect control of monetary policy.
I come back to the next question, which we were talking about very briefly. This is connected with Menger’s 1892 conjectural account of the origin of money out of barter exchange, where he focused on the process by which a single good becomes far and away more salable or liquid than other goods. What exactly is he talking about here?
Menger tells a story where we start in a barter economy, trading goods for goods. The problem we run into is that in order to make a trade, not only do I have to have something that you want, but you have to have something that I want. That’s inconvenient, because how often is that going to happen? So Menger tells a story where people in this economy are gonna look for the most salable good. Even if I don’t want it – maybe we will pick shells or something – even if I don’t want shells, I know other people want shells, and so I want to acquire them so I can get the things that I need. Then all of a sudden I am willing to accept shells, and now I am one more person who accepts shells. By the end of the day everybody has coordinated on this idea of shells, or some good, as money, as the most salable good.
I think there is a good and a bad aspect to this story. The good aspect is that what Menger does here is focus us on what money does. The service that money provides is its salability, and what I call in this paper liquidity. Liquidity is salability; it is the ability to accomplish an exchange with some good.
So we can say that is the primary function of what money is supposed to do?
I think it’s a tautology to say that any medium of exchange is going to be money at least to some degree or another, and we’ll talk about how exactly we disentangle that later. Menger also shows that this is an incentive-compatible process. Nobody has to direct it; it’s something that everybody has an incentive to coordinate on. Once you’re part of a monetary network, it doesn’t do you any good to break from that. I do think the Menger story has two limitations. First of all, this story is ahistorical. It’s a thought experiment; it is not an actual history, just because the barter economy that he describes could not possibly ever exist. It would spontaneously sprout money, as he talked about, very quickly before it was even established as a barter economy. So for the actual history we will have to take another approach, and I have other papers about that. More importantly for us, a lot of people have taken Menger’s story to imply that there will be one good that we coalesce around, and that one good will be money. You mentioned this Kroszner quote. Even though whatever money good this Mengerian process centers around becomes more and more salable as people join the monetary network, Kroszner’s point is that that also makes other things more salable too. It is not like other salability decreases. Think about it this way: if I want to sell a bicycle, because we live in a money economy we can calculate what is the value of this bicycle. Then I can ask, if you want to bake me some bread, how many loaves of bread is going to be about equivalent value to this bicycle. Now we’re not using money; we’re bartering; we’re trading goods for goods. But the only way that it’s feasible to barter is because we live in a monetary economy where we have that unit of account. So now all of a sudden money has become more salable itself – money is universally salable – but it has also made bread and bicycles more salable even in barter transactions. I think this is really important for understanding financial markets, because it’s kind of trivial or unimportant if we’re trading goods for goods, but it’s not trivial if we’re talking about financial assets.
Absolutely. Kroszner was skeptical about the way we define or use money; it might actually, bring about the end of money itself as a medium of exchange, by increasing the marketability of other goods, like financial assets.
What’s funny is that Kroszner mentions this as a hypothetical possibility, and Yeager even has a paper with Greenfield where he talks about some hypothetical futuristic economy where we can separate the medium of exchange from the unit of account. But we kind of already live in that economy. There is not one thing that we use as a medium of exchange. Part of what I am trying to do in this broader project is that there is no sharp distinction between money, money substitutes, or money and non-money. There is a whole gradient of things that substitute better, or things that provide more liquidity services, things that provide fewer but still some liquidity services. If they help you exchange, that is liquidity services. Even things like Treasury bonds are used as collateral in a lot of financial transactions. That is not as liquid as cash; you can’t spend that at Walmart, but you are doing monetary-like transactions with it, and so it seems like that ought to count to some extent in our measure of the money supply.
I think in his book, Barnett, page 118, conjectures this. High-end financial assets as a Ferrari. He says: “When clearly the price of a Ferrari is too high if valued solely for its transportation services, and a Ferrari cannot be solely for recreation since Ferraris are seen on public highways, Ferraris must be a joint product providing both kinds of services. The same has been true of money since many monetary assets began bearing interest long ago.” We are talking about the liquidity which you talked about, and the other is investment return, which is the financial assets. Investment return is actually not monetary service; we can say it’s not liquidity.
The Ferrari example is such a good example because it shows that we have the tools in economics to think about these kinds of joint goods that provide multiple kinds of services, and the price reflects multiple services that they provide. Suddenly we just forget that when we think about money as just its own isolated kind of good.
In this case we are overestimating the liquidity services if we are talking about broader monetary aggregates.
M2 is going to overestimate the liquidity of things like mutual funds, and then underestimate things like Treasury bonds.
Maybe it’s underestimating liquidity of the broader monetary aggregates.
Another economist which you connect to this is Perry Mehrling. He has a paper on the asset-liability structure of money. He has this hierarchical structure, which is of critical importance over the business cycle, as the quantity of each layer constrains the issue of assets in the layer above. Why don’t you talk about this?
This is an idea with some pedigree, but Mehrling, like Yeager, has such a good straightforward statement of this. That is what I jump off from in this paper. The idea here is that you have what is called the pyramid of credit. A lot of older economists had tried to distinguish what is money and what is credit, and Mehrling is trying to cut through that. His idea is that you have an asset-liability structure in financial markets. At the base you have base money, which is why it is called base money. That’s an asset to everybody and a liability to nobody. On top of base money you have bank-issued money. This is going to be your deposit balances and things like that. This is an asset to you but a liability to your bank. What the bank does, because of fractional-reserve banking, is the bank is actually going to be able to issue a larger quantity of deposit balances than the base money that supports it. So you get a small base, and then you have this inverted pyramid with bank deposits on top of that. Mehrling generalizes this story, because that is a very old idea, but in addition to those bank deposits you also have further layers of this pyramid that expand and expand based on leverage ratios. For example, investment banks will redeem to you in terms of bank deposits; they will wire money to your bank; they will not give you cash. They’ll redeem in terms of bank deposits, and then they will issue their own liabilities on the basis of that, and it will be a higher quantity but less liquid. The usefulness of this is that it gives us a way of talking about the endogeneity of the credit structure. What happens if the supply of base money contracts? The layer above it, the bank deposits, has to contract even more than in proportion if banks try to maintain a certain reserve ratio. This is what in the past has been called the perverse elasticity of credit: when the demand for it goes up, the supply actually goes down because of this kind of pyramiding. The same thing is going to be happening in the upper layers too. We can kind of see this in 2008: the upper layers contracted a lot because of a small contraction in the lower layers that reverberates and expands, and so you get a larger contraction than you otherwise would. For me this is a way of connecting the credit-cycle and the quantity-theory explanations for business cycles that we talked about earlier. Mehrling doesn’t do this, but what I want to do is think about the liquidity services provided by all of the layers of the pyramid. If we do that, that contraction we talked about is going to be a contraction in the money supply and not just money demand.
I think the idea actually goes back historically to the work of Ralph Hawtrey in the 1920s. The origin of it lies in his book Currency and Credit. Because you are mentioning that Divisia aggregates are a way of reconciling this, and your paper was dedicated to this issue. Why don’t you talk about the conceptual underpinnings of Divisia aggregates in the way that the asset-liability structure is designed?
This is where we are actually able to take our conceptual idea of what the money supply ought to measure in terms of people’s actual plans, choices, and expectations, and now we have a guide to how we take that to the data. Divisia aggregation is basically a way of doing justice to that definition without assuming a singular money asset. Your example of a Ferrari was actually a really good example. Divisia aggregation is a way of pulling apart these kinds of joint assets. If we think of a Ferrari as providing joint status services, transportation services, it’s kind of stylish – what if we were able to pull out the transportation services and aggregate that with the transportation services of a bus or of anything else, a sedan –
Or even walking.
Yeah, all of these are ways of providing transportation services. We could come up with a price index for transportation, but we don’t want to put the whole value of a Ferrari in there, because a large part of that is not about transportation. Basically a Divisia aggregate is going to do the same thing for the money supply. There are some things that are mostly money – this is going to be cash, no investment return – and then there are going to be other things like Treasury bonds or money-market mutual funds that do give me a return; there’s some investment there, but it’s also liquidity. So the idea here is: what would make me willing to accept a lower-than-market-rate return on some asset? Why would I hold something if it’s making me less than the risk-free rate? The answer here is because it’s providing implicit liquidity services. That kind of arbitrage logic allows us to pull apart investment on the one hand and liquidity on the other hand, and we can take liquidity from all these other assets and aggregate that into at least an index – not a dollar value, but an index over time – of the liquidity services in an economy, regardless of what asset is providing those liquidity services. I think that’s going to be a lot more economically meaningful as a money aggregate than something like M2 that just takes this and this, and your Ferraris and your Nissans or something, and adds them together.
The link was lost during the 1980s. M2 was the standard go-to measure of money supply during Friedman’s days like the 1960s and 1970s.
Which is exactly what we should expect. Prior to the 1960s and 1970s there was something called Regulation Q in the United States that limited the investment services that a lot of these assets could provide. It was basically a cap on the interest rate for a lot of these assets. The effect of that was to prohibit a lot of these joint investment and liquidity assets, so you had investment assets on the one hand and liquid assets on the other, and could not really combine them legally. When that was repealed, of course M2 is not going to be as sensible an aggregate anymore, just because now you get a lot more financial innovation; you get a lot more of these joint assets. If you are just adding them together in a simple sum kind of way, that’s going to track liquidity services less and less well as we innovate financially more and more.
You use the word “economic meaningfulness” in your paper, the economic meaningfulness of monetary aggregates.
Exactly, because we want our empirical aggregates to correspond to something meaningful. What does meaningful mean? Meaningful means in terms of the choices, plans, and expectations of the people actually in the economy. M2 just has not been that since the end of a lot of these financial regulations in the US.
This is connected with the network structure in the construction of monetary aggregates across countries. You draw some literature on this also in your paper. What are these network structures?
In the chapter that you mentioned earlier, the network idea generalizes this inverted pyramid idea of the asset-liability structure of the economy. In the same way that the structure of production was never supposed to be a general model of the entire economy, because you just can’t operationalize that. Similarly, the pyramid of credit that we talked about earlier is a great way of explaining things, but you can’t really operationalize that in empirical work because it is too simple. What we don’t see is financial institutions issuing one kind of asset redeemable for an asset below it in the pyramid. We see some of that; that is kind of one way of understanding what fractional-reserve banking does. But especially once we get up into the upper layers, we see very complex securitization; we see lots of diversification and bundling; and we even see less liquid assets like mortgages being securitized and packaged and reissued as something that would be in the pyramid above it, but it is actually more liquid because you’ve diversified away a lot of the risk. It is not so simple as a pyramid structure, but if we think of it as a network structure, that lets us do justice to the complexity of this structure across institutions, across country borders even. That lets us use certain tools in network analysis and also operationalize this for empirical research. Now that we have nailed down the concepts, now we have the right tools to apply, now we can do that.
One way this can be used is to measure the inflation of the OECD countries. In an empirical paper, Ciccarelli and Mojon observed that the inflation of 22 OECD countries have a common factor that accounts for nearly 70 percent of their variance fluctuations at business-cycle frequencies. You say this is linked with the network structures of the measuring of liquidity. Is there some empirical evidence on this in the way that how they are connected to each other?
What you’re referring to is this weird result that domestic inflation in a lot of countries is actually not very well predicted by the money supply in that country. What it is predicted better by is global liquidity. If you take some measure of M2 across all different countries, that actually predicts domestic inflation better than domestic M2 in a lot of places. That is a little puzzling. There are a lot of models and empirical papers trying to figure out what exactly is going on here. What I suggest at the end of my chapter is that thinking of the network structure of financial markets across countries, and aggregating correctly the liquidity services of a lot of the issues of these institutions or organizations, can suggest an explanation for this. I’ve floated this explanation to some of these economists, and they all tell me the data’s not there. There is a lot better data now for Divisia aggregates in a lot of countries that there was not five or ten years ago.
Slowly it is becoming more accessible.
So I am hopeful that this will be testable in the near future. The idea here would be that if we think of a network of financial institutions, there’s a connection between some institution in terms of the assets that it issues and then some other organization that holds its issues. There are going to be assets that are assets to one and a liability to the other. This is basically the pyramid idea, just with more complex linkages. The insight here is that if you think of a lot of these linkages going cross-country, you could have financial institutions in one country holding assets denominated in another currency on its balance sheet. So if there is monetary expansion, for example, in the US, now US institutions become more liquid; now US institutions issue more assets themselves, more liabilities to them; and these assets get held, for example, by European banks. Now some of this liquidity spills over. This is on the balance sheets of a lot of European banks; they become more liquid; and even if the ECB does nothing, maybe they are able to issue now more liabilities themselves that provide liquidity services. So US inflation is going to spill over into a lot of other places. This is exactly what we see. This literature has been observed for a long time, but we saw this very dramatically after COVID. A lot of places – the whole world, with some high-inflation exceptions – the whole world had been struggling to get inflation up, and then after COVID all of them overshoot to a large extent. Is that because all of the central banks are doing the same kind of monetary policy at the same time? To some extent it was; there was big fiscal accommodation. But also because a lot of these linkages go across countries and liquidity will spill over. That is something that central banks are going to have to be aware of in terms of international monetary policy coordination.
You are linking this with some potential lines of research that one can pursue, given the fact that the development of liquid financial instruments has taken place over the course of the business cycle. You call for identifying the particular shocks that affect the switch in particular bubbles from stabilizing to destabilizing speculation.
Because what’s going on in this network structure – one unproductive response that some, for example, narrow-banking advocates call for is they say, well, we have got all these liquid assets being issued by private organizations; this makes monetary policy harder, so let’s reinstitute something like Regulation Q; let’s split apart these hybrid assets between investment and liquidity, and that’s supposed to make monetary policy easier. First of all, I don’t even think that is true. There is a nice Belongia and Ireland paper where they show that that’s not true. In fact, what that does is distort the structure. You’re not going to get some smooth sides of the pyramid where you can choose how much investment, how much liquidity, and vary that smoothly. If you make that a big discrete step, you’ve made people’s lives worse. You’ve taken away options from people, and if somebody wants to move from here to here, now they have got to make a big jump rather than a small jump. The problem with financial repression as a solution to these endogenous dynamics is that it forces people to make these kind of big destabilizing jumps.
That being said – I’m not going to say it makes monetary policy harder – but it does suggest that monetary policy is going to have to change priorities, which to some extent it already has. Part of the issue with some of these endogenous dynamics is that expectations can become self-feeding. You get positive feedback loops. When expectations become unanchored, the economy looks very unstable; it looks like you have animal spirits and things are going crazy, and you need the central bank to manage things very actively. I think that’s generally a mistake, because then we have to ask, why would expectations be self-feeding? What we see is that when expectations become unanchored, especially when a central bank loses credibility as far as its future monetary policy, that’s the situation where expectations become self-feeding. So to maintain stability, honesty and credibility become much more important. The gold standard was one way of ensuring the credibility of the future path of monetary policy, anchoring expectations. The 70s and 80s were our experiment with unanchored expectations; it didn’t go so well. And what we’ve discovered since the 80s, and especially since the 2000s, is that the credibility of the central bank in terms of setting expectations is critical. And it’s more critical when we have this huge variety of assets satisfying liquidity needs, because those are endogenously provided. And I think central banks kind of realize this, even if they are not thinking in terms of monetary aggregates, they do realize that credibility is important, that setting expectations is important. This is where we get forward guidance, for example. This is a big part of a lot of the New Keynesian literature. If you nail down the expectations, then you can solve the New Keynesian model. Obviously I think this would be better if they were looking at monetary aggregates as well, but the focus on the expectations I think is a beneficial development in monetary policy over the past few decades.
Hayek has this book, Monetary Theory and the Trade Cycle. In one of the paragraphs he actually says that the monetarist theorist should actually pursue a more profound and fundamental effect of the process by which money is introduced into the economic system. I think this is what we are failing to take account of, the effect of the process by which money is introduced into the banking system.
Absolutely. Monetary Theory and the Trade Cycle – what was that, 1933? – Just imagine how much more well-developed and more profound that process is today than it was in 1933.
I think this is where we would like to end, but then, having discussed all these, any point, Cameron, that you feel you can talk about as you wind up?
I do want to thank you for the thoughtful questions. I think we have had a lot of good things to talk about. Thank you for looking at the papers and really thinking about the arguments. I hope this has been interesting and useful.
It certainly has been very interesting, because I think it is very important to have a proper measure by which we define money, liquidity, and investment.
You’ve talked about doing empirical work on the basis of these kinds of theoretical work, and I will look forward to seeing your future work too.
Subscribe
— If you enjoyed this post, get new posts and papers delivered to your inbox.
Leave a Reply