Welcome and Policy Session 2: Evolving Monetary Policy Transmission Mechanisms
After welcoming remarks from Atlanta Fed president Raphael Bostic, the conference's second day began with a session on changes in monetary policy brought on by changes in government policy and technology. Moderator Frederic Mishkin, Alfred Lerner Professor of Banking and Financial Institutions at Columbia University's Graduate School of Business, led the discussion with panelists John Y. Campbell, Harvard University's Morton L. and Carole S. Olshan Professor of Economics, Lucrezia Reichlin, London Business School professor of economics, and Jefferies LLC chief market strategist David Zervos.
Transcript
Raphael Bostic: Good morning. One try; that was very good. Good to see you all. I hope you enjoyed yesterday as much as I did. I thought the sessions were quite good. I hope that you got some relaxation in as well. I want to again congratulate our program committee because the topics were very timely for the things that I've been wrestling with—thinking about bank stability and financial stability is really important. I really appreciated Vice Chair Barr's discussion of the overarching framework to really give people a sense of the totality of how he sees it all fitting together. That's an important perspective to have.
The experts, the panel on the discount window: there was so much detail and there was so much specificity that I actually got much more clear on all of the issues. I really want to express my appreciation for everyone who participated in that. The research panel on shifting supply chains really talked to me about the extent to which we are moving towards a new equilibrium about how global trade happens. It's something that I will take on board as I think about some of the implications for a broader macroeconomy.
Of course, we had last night's dinner with Professor Glaeser, talking about cities from a historical perspective. I will tell you I still have visions of him jumping off the stage, and I'm hopeful that we will not have any more leaping, or any of that sort of stuff—gymnastics—for the rest of today.
We have another great day today on tap. The schedule is packed with a presentation and discussions of really important and timely issues that will help shape the future of monetary policy, global financial markets, and the macroeconomy. I'm not going to go through the whole agenda for today; you have that. Hopefully you all have your apps activated and are ready to ask questions.
I will mention that tonight we have a special treat as the capstone for the program. I'll be on stage here with two of my colleagues, President Susan Collins from Boston and President Loretta Mester from Cleveland. We are going to try to keep up the history of the capstone event. I'm going to make sure it's interesting but also entertaining.
That's all I have today. Enjoy the day, continue to engage, participate, and ask questions. I'm going to turn this over to Frederic Mishkin, who will moderate the opening panel. Fred.
Frederic Mishkin: I strongly believe that introductions for the moderator should be very short. We have three distinguished economists. I'm not going to introduce them in terms of what they do because you can read their bios on your app. We're going to go in reverse order, so we're actually going down the line here with David first. Some are going to be sitting in chairs, some will come up to the podium. John is going to jump off the podium because he's also from Harvard, and that's what Harvard people do.
Each speaker has 15 minutes, then we'll have a go-around of them replying to each other, and so forth. I may have some questions to get things started, but you'll also have an opportunity to ask your questions. I've got my little magic iPad; everything is very tech savvy here so we don't have to look for you. Actually, the lights are bright I couldn't see you anyway, so this is going to work out very well. David, why don't you start us off?
David Zervos: Thank you, Fred. Thank you to everyone at the Atlanta Fed for having me back. It's always a lot of fun to come to this conference. I've been doing it for many years and have some wonderful memories and a lot of good friends in the audience and ex-colleagues as well, so looking forward to the panel.
I wanted to talk today about something that I don't believe gets discussed enough in central banking. I think it's at the heart of many of the miscalculations that have taken place over the last two years if not longer as we've embarked on this tightening cycle, and that specifically is the Fed's balance sheet. Somehow, some way, we tuck it away and we don't think about it, and it ends up on autopilot or some sort of roll off. People don't often think about what it's still doing in the background, what the vestiges of these large balance sheets are.
For reference, the Fed's balance sheet prior to the global financial crisis was typically about 5 percent of GDP and was made up of T-bills. In March of 2022, before this tightening cycle kicked off, it was 40 percent of GDP—eight times larger, at $9 trillion—and was made up of very long-dated, highly durated securities. I don't think we can sit back and imagine a world that's completely unaffected by balance sheets that are this large.
It's not just the Fed. The ECB's balance sheet was at close to $10 trillion, the Bank of Japan at $6 trillion. We had $500 billion here, $500 billion there, in the Bank of England and the Riksbank (the Danish Central Bank). You add them up, it's something like $30 trillion worth of assets that got taken out of the financial system, all long-dated.
One of the biggest stories of why this tightening cycle did not feel like a traditional tightening cycle was that the vestiges of that stock of QE in the system were not properly taken into account when thinking about the overall level of policy restrictiveness. We have now built a system where there are two dials for monetary policy: there is a rate dial, and there's a balance sheet dial. When we run out of room on the rate dial, we dial up the balance sheet. We've done it now twice. It's worked out well, twice. We've created recoveries in the economy. We've created a new move lower in the unemployment rate.
Those policies at the zero lower bound still have impact. That's what I want to talk to you about today. This is all postulates. I don't have a model for you. I don't think anybody has a model for the balance sheet. I don't think in FRB/US...$500 billion on the balance sheet. Is my mic on or off? Hello?
I want to get quickly into this idea, then talk about one other thing. My basic storyline for our clients at Jefferies has been, since the beginning of 2023, that if you adjust for the size of the balance sheet and you think about the stimulus that is still embedded in the economy from the balance sheet, that the level of interest rates that we went to and from is different than what you might have thought, a priori. We didn't go from 0 to 5.25. We went from -3, -4 to something like 3, 4.
Today, we don't sit at a 5.25–5.5 percent interest rate. When you adjust for the stock of QE, we sit at something that's much lower. That is a policy that's not nearly as restrictive as you might have thought. If that's the case, it helps explain why we've had seven or eight quarters now of growth right around 2 or higher, the unemployment rate has stayed below 4 percent, and this economy, from a financial side, has done pretty well. Actually, from all sides.
We didn't really destroy that much aggregate demand, not nearly as much as everybody thought, with 500-plus basis points of tightening. That was the original theory. I'll give you two channels where I think the balance sheet has made a difference. The first is on the asset side and the second is on the liability side.
On the asset side, when we look at what happened as rates rose, the Fed's balance sheet lost value. Bonds that it bought at par—Fannie 1.5s, Fannie 2s, all those mortgages when rates were low—dropped, and they dropped significantly. They dropped to $70 price bonds, $80 price bonds. The Fed's balance sheet lost over a trillion dollars in mark-to-market at the peak, probably closer to a trillion and a half.
In a world without big balance sheets...if we go back to a cycle like 1994, my first cycle after I left the Federal Reserve System, when I was sitting on a trading floor in London, we raised 300 basis points and lots of people lost a lot of money. They lost a lot of money in the mortgage market, in particular. That 300 basis points stung. It blew up Orange County, it blew up Mexico, it blew up Kidder-Peabody, it blew up large parts of the financial system.
It wasn't too long after that the Fed was cutting rates in '95 because there was a lot of pain. When the Fed raised rates this time, a significant amount of the losses that would have been taken by me and my friends on Wall Street were actually sitting inside the balance sheet of the Fed. Happily, not being marked to market, there were no shareholder revolts, nobody yelling and screaming about losses. There were winners when rates went up: those that locked in low rates, and there were losers in the non-big balance sheet world. The losers used to be insurance companies, bank pension funds, Orange County, Kidder-Peabody, David Askin, if anybody remembers that name.
Now, the biggest loser was actually the Fed's balance sheet. We had wealth effects; we had huge, positive wealth effects. Those who locked in low rates—in Europe, corporates, households in the US, governments—all won, and a big chunk of these losses were socialized inside of the central bank balance sheet. That wealth effect was not just the wealth effect of the trillion or trillion and a half of mark-to-market, but also what happens in markets is when you start to lose, your risk manager comes up and goes, "Hey, you've got to hedge." You get huge moves in financial markets, well beyond what normally would happen, because people panic. There wasn't that panic.
All of that asymmetric loss was in the system, meaning loss in the public sector but not in the private sector and gain in the private sector that was not offset. It was not a zero-sum game.
Then we have the second effect of the balance sheet, which I think is the more mercurial one and the harder one to analyze, which is: What does it mean to have reserves cash in circulation, reverse repo, that's eight times larger as a percentage of the size of the economy than it was in most normal periods prior to the global financial crisis? I don't know the answer to that.
I would imagine that if my old debating partner on this stage—God rest his soul, Allan Meltzer. We did a wonderful event with him here a few years before he passed—if Allan Meltzer was here, he would say, "That's really stimulative, having that many reserves in the system. You need to be careful with that." I imagine his coauthor, Milton Friedman, would say similar things, and I imagine Paul Volcker, who used to write about liquidity in the system incessantly, would be wondering what it means to run a central bank balance sheet eight times larger than you did at any time in the past at equilibrium.
What I want to postulate to you today is that we sit in a world where policy is not that restrictive, where the balance sheet is way further away from neutral than any definition of neutral you could think of based on the past. What that's allowed is a holistic structure to monetary policy that has a neutral real rate that is higher, an R-star that's now higher, and a balance sheet equilibrium level that is also higher, so that the overall stance of monetary policy is actually pretty close to neutral.
The only way to define neutral in a world with two dials for monetary policy like we've built in the post-GFC world, is to openly discuss what an equilibrium balance sheet really is, and not punt the question, not say, "Just because I can't put $500 billion of balance sheet into the FRB/US model, then I'm not going to talk about it."
It needs to be discussed, it needs to be thought about, it needs to be analyzed. Most of all, I think it needs fresh minds, PhD theses, research, to really think about what's happening on the asset side of the balance sheet as rates move, how stimulative that feedback mechanism might be, and what it means to run these large levels of reserves cash in circulation and generally high-powered money in the system. Is that something that we can now offset with a higher rate?
I'll give you one other outcome of that I think is important. It may be that the new equilibrium is big balance sheets and higher R-stars or higher equilibrium real rates, which really solves a lot of problems for us as central bankers because it gets us away from the unpleasantries of the zero lower bound and negative interest rates. That may be a very interesting positive outcome on the future, and it may be what I think we're headed to, which is a world where the Fed recognizes that QT needs to be tapered sooner rather than later, and that we may be having a higher equilibrium real rate structure.
That does seem to be the policy outcome of the last few months, after many months of the committee telling us that QT could continue to go on and on and on and on, even while rates were being cut. Somehow that changed in the last few meetings. We see in Japan rates being raised while QE is still taking place. There does seem to be an idea forming here that balance sheets will remain large. Rate structures can react to that in a different way than we may have thought.
I'll leave you with this: policy is not that restrictive. Demand did not get that crushed, the unemployment rate did not rise much, and the main reason that inflation did what it did over the course of this cycle had almost nothing to do with the demand side. It was a supply-driven shock. There were three consecutive supply shocks: the reopening and the supply chain disruptions, war in Ukraine, and zero COVID policies in China. We shot up to 9 percent in the summer of 2022—not at the beginning of Joe Biden's presidency, as he said. That happened a year and a half after he took office.
Since then, we've been unwinding those negative supply shocks. We've had strong growth and lower inflation as a series of unwound, negative supply shocks or positive supply shocks hit the system. The aggregate supply curve moved back out, prices went down, output went up—straight up—and the aggregate supply curve moves out. Prior to that, the aggregate supply curve was restricted.
In fact, if you go back and try to think about the inflation process that hit us, the big moves in inflation from the beginning of '22 to the middle of '22, as the Fed was moving its policy into the tightening realm of March, those are two consecutive quarters of negative growth. We had two consecutive quarters of negative growth in the first half of '22 with inflation spiking up to 9. It's hard to tell an aggregate demand story there. That sounds like aggregate supply to me.
Aggregate supply drove it, aggregate supply largely ended it. Do we have little vestiges of, "is policy a little too tight or a little too easy?"Do we have a little excess aggregate? Maybe. The big-picture story in this whole economic cycle was a massive supply shock, followed by the unwind of a massive supply shock, and a little bit of aggregate demand tinkering to take policy from uber-, uber-, uber-easy at QE stock-adjusted rates that were negative, to rates that were pretty close to neutral and today sit probably pretty close to a neutral level, given how large the balance sheet is.
Where the Fed decides to stop the balance sheet will matter in that discussion. To me, that sets up beautifully for the clients that I speak to at Jefferies about a world where they can take risk again, they can feel comfortable in the risk assets base. We have been pushing them that way over the last year and a half, largely because we never believed that policy was that restrictive and that demand would never get crushed because of the size and scope of the Fed's balance sheet and what that meant for the overall stance of monetary policy.
The lenses we need to look through have a duality to them, and it troubles me every day. I'm sitting here in a conference with a number of Fed presidents and folks on the FOMC. It troubles me every day how little we pay attention to the balance sheet or hear about the balance sheet. It is a tool that has become so important for us to solve the problems that we've faced in the global financial crisis and COVID and it doesn't just go away. It sits with us, and it's still in us; that painkiller that we put in, in March of 2020 and added all the way through to '22, that surrogate painkiller because we ran out of interest rate painkiller: it's still there. We still have a $7-plus trillion balance sheet and it still creates the stimulus in the system that we need to think about and manage, and understand when we're setting rates. I'll leave it at that and turn it over to the next speaker.
Mishkin: Lucrezia, are you up next?
Lucrezia Reichlin: Yes, but I'm going up to the podium. Good morning, everybody. Thank you to the organizers for inviting me to this fantastic place. It is quite hard to talk about such a flamboyant speech but I will also talk about the balance sheet and I will start by stating the obvious: that central bank's monetary policy, the operational framework, has really changed quite dramatically since 2008. The financial and fiscal footprint has increased greatly as can be seen by the percentage of the assets of the central banks over GDP.
Part of this is the result of the federal monetary policy at the zero lower bound, and part is the effect of emergency liquidity operations during periods of turmoil. Through this experience, it seems that central banks have also developed a new way to understand how to conduct operation in normal times, and my remarks will focus on this aspect: operations in normal times.
The first question I want to ask is: To what extent is the new operational framework a radical departure from the pre-crisis conduct on monetary policy? I will start by quoting a recent book by Ben Bernanke that reviews the Fed policy since the 1951 Accord, in which, basically, he asserts that no changes have occurred in the fundamental way monetary policy works, and in its objectives. In the book, he argues, yes, lessons were learned, but the Fed's monetary policy is still based on a modernized form of Keynesian economics, driven by the objective of leaning against the wind, and its financial stability policy is still rooted in the lender of last resort function.
These foundations have not changed since the 1950s. What has changed, he argues, is the context in which the Fed operates—large and stable financial system, zero lower bound, high public debt. Now, is this true? Is it the context, or is there something fundamental about how monetary policy is now working?
Since 2019, the Fed has adopted the ample reserve framework and, as it has been argued by many, this framework has a lot of advantages. First, it provides a good control of interest rate in the face of large liquidity shocks. It does not provide an incentive for banks to economize our reserves, so it makes the system more stable. Indeed, it corresponds to the famous Friedman rule applied to reserves. In fact, the Friedman rule is achieved when the market for reserves is saturated.
It also provides the central bank with two tools which allow it independently to control inflation: the interest rate and the quantity of reserves. Therefore, it allows the central bank to increase the supply of reserves for financial stability reasons with no effect on interest rate. This has been used quite successfully in a few episodes recently.
This makes so much sense that one may wonder: Why was this not done before? Actually, it was done before. There are a few examples, but it seems that this has been rediscovered in a big way. The main reason why this was discovered—this way of conducting monetary policy during the financial crisis—was because it became clear that financial frictions were pervasive and the old framework could not work because it required an efficient distribution of reserves. That with dysfunctional markets was impaired.
In that context, all central banks, not only the Federal Reserve, abandoned the scarce reserve system and injected abundant reserves to preserve financial intermediation and monetary policy, but for both objectives. After the crisis, there was a realization that actually these frictions, although they're not so huge as during the peak of the turmoil, are still with us. The consensus has changed in a quite fundamental way. The consensus on how well the efficient market hypothesis approximates reality. This consensus has changed, and this is really at the root of the change in the operational frameworks.
Crucially, changes in liquidity regulation, changes in the structure of financial markets such as mismatching the demand and supply for the services of market making and its consequence on the decline in liquidity in the money market, have generated high demand for excess reserves. In that sense, I think we can agree with Bernanke that the Fed adapted to changes but it did not depart from the standard way to understand how monetary policy works in these transmissions.
Now, who are the critics of this system? The critics of the new system say that this implies the central bank's balance sheet will be too large in normal times. The meaning of "too large" is not entirely clear. If one worries about being far from historical norm, it is sobering to look at the historical data produced by research by Ferguson et al. These are historians who have collected data from the last centuries.
You can see from the chart—this is from this historical research—that central banks' balance sheets today in advanced economies...this is for advanced economies, not just for the US. If you compare with the experience of the last few centuries, although large in relation to the GDP, it's not large in relation to the private sector. These are central bank assets, a share of total bank lending to the non-financial private sector. The ratio you see is below historical norm, although a little bit thickening up. This was before the beginning of QT.
The explanation is that as the size of the financial sector has increased, so has the safety net of the central bank. In a way, so has the reflection of structural changes in the economy. Given this observation, it is intriguing, however—and here, as a European, I have been invited probably to talk a little bit about Europe—that the ECB has made different choices and decided to opt for a return to a scarce reserve system.
Let me say a few words about the changes in operational system which have just been announced at the ECB. After many years operating essentially at the floor like the Federal Reserve, and much discussion between diverse views, the ECB has recently announced that it would partly go back to a reserve scarcity system where reserves are supplied at the margin through regular refinancing operations against collateral. This has been called a "demand-driven system," as opposed to the "supply-driven system" at the Fed. Realistically, the ECB has also introduced some hybrid features—a narrow corridor of 15 basis points, new longer-term lending operations, a small structural bond portfolio—and in the mind of the ECB these hybrid features are introduced to ease the difficulty of banks rolling over large structural liquidity needs on a weekly basis.
The new system is therefore a compromise between supporters of a larger role of central banks in markets, and those who want to return to a small balance sheet, what I call "narrow central banking." Now, the market has not reacted much about these changes, about this announcement, probably anticipating that if liquidity stress does appear the ECB will respond dramatically as it has always done. Still, time will test for the robustness of this framework.
In this chart—this is taken actually from the ECB itself—you can see how they project the unwinding of the legacy portfolio, which is in blue and is supposed to gradually be going to almost zero, and then credit operations start kicking in. At the end, you can see in red, that small structural portfolio of bonds that they will keep in the steady state.
It's interesting to ask in this audience: Why has the ECB gone in a different way? The ECB wariness of the ample reserve system is mostly fiscal, and how large is the fiscal footprint of the central bank today? This is a relevant question to ask. If we look still at Ferguson's data, centers of data for advanced economies, the same data set I showed you before, it appears that although the share of central banks' holding of government bonds relative to the total outstanding is not at the peak, the share of government bonds to total central bank assets is at the peak.
In a monetary union, we know of fiscal federation member states facing different degrees of sovereign risk. This is problematic, and this is what has motivated the ECB to go to a different system. What I want to stress here is that there are political economy issues related to these choices that cannot be overlooked, and the ECB debate, which was quite divisive and quite...whatever. Although, in a different context of an asymmetric federation, I think it is food for thought to the Fed because a large fiscal footprint comes with a complex relation with Treasury.
Although the ample reserve system makes a lot of sense, managing fiscal monetary interaction with a large balance sheet may turn out to be challenging, even for the Fed. It would require careful managing of balance sheet risk as well as interaction with governments, management functions. I suspect that a strong separation between central banks and finance minister is probably a thing of the past. We are returning to a closer interaction between these two functions.
That, of course, does not imply a threat to central bank independence, but it will have to be handled with care, and perhaps some changes in governance will be coming. In a way, the ECB experience in the future offers a natural experiment which will help us answer the question: Can central banks go back to a narrow balance sheet in the new economic context? We will see.
Now briefly, because I think I'm running out of time, the second question is: Does the change in operational framework have implications for the transmission of monetary policy? In a trivial sense, yes, because monetary policy cannot work without a functioning financial system. A more subtle question is whether it matters if reserves are borrowed, as they will be at ECB, or non-borrowed, as they are at the Fed.
In principle, the ample reserve system and the interest on reserves suggest that it should not matter, because we have two instruments. A recent paper by my former colleagues at the ECB (Altavilla, Rostagno, and Schumacher) suggests the contrary, and I want to show you something from the paper.
This is from their paper, and this is a work based on bank data. It shows the response of bank lending after a one percentage point increase in reserves in Euro-area data. You can see that what they find is a strong effect, a positive effect, when reserves are non-borrowed, and no effect whatsoever when reserves are borrowed. I find these results very intriguing because it goes against what I just said, the independence of the two: there is a quantity of reserves, and the interest rate.
It resonates a little bit with what David said. Perhaps more research has to be done on this issue. Now, it is interesting that this issue about borrowing on borrowed reserves is reminiscent of an old literature—which I remember, given my age—Christiano and Eichenbaum in 1992, Strongin in 1995, Bernanke and Mihov—who showed us significant effects or changes in the reserves and interest rate, and the real economy, could be found only if focusing on non-borrowed reserves.
Of course, in that context we were in a different world because there was no interest on reserves, but it seems that this paper is finding something similar in a situation in which reserves are remunerated. In the context of today, these results say that quantity matters, so even given an interest rate, quantity matters. Therefore, coming out now of the contractionary phase, the calibration between quantitative tightening and the interest rate will be actually important.
Now, let me conclude. As the structure of the economy and markets evolve, central banks must change the ways in which they operate, and they have. Operational frameworks have never received so much attention outside the circle of central bank experts, but now the situation is different. Everybody is talking about operational frameworks. Nobody had any idea a few years ago.
The reason is because it matters. It may matter for the transmission of money policy, although here I will put a question mark: more research? But it matters because it has implications for the institutional architecture, regulating the relationship between central banks, treasury, and financial regulators. This is made clear by the fact that the US and the Euro area are now embracing different practices, a divergence which is hard to explain from economic considerations alone, but that reflects different political constraints. Thank you.
Mishkin: John?
John Y. Campbell: Thank you very much. I will get as far as the podium, but no further. I'll stay right here. Last night you had fast Harvard, this morning you get slow Harvard. [laughter] The context of this panel is: How does the structure of the financial system that we live in affect the transmission mechanism of conventional interest rate monetary policy? Both David and Lucrezia have talked a lot about the size of the central bank balance sheet and how much that matters, and of course how much it's changed over time. I want to direct our attention in a slightly different direction, to think about the way in which mortgage markets work.
My main focus will be cross-sectional comparison rather than changes over time, since certainly the US and most other countries have had fairly stable mortgage market institutions over time. It goes without saying that mortgages are an important part of the financial system They are the largest household liability in the US and most other developed countries, and mortgage rates are the main direct channel through which interest rate policy affects household consumption.
Mortgage rates obviously also have a strong impact on house prices, and hence the construction industry. Then, of course, there's the financial stability side—problems with mortgage lending were at the heart of the global financial crisis and more recently affected the US regional banks in 2023. We were talking a lot about that yesterday. I want to make two rather specific points about our mortgage system. The first is I want to contrast a fixed-rate mortgage system (a FRM system) with an adjustable-rate mortgage system (an ARM system). Then, I want to broaden out from that and talk about how policy on amortization might be relevant for monetary policy.
Then I'll turn to something very relevant right now, which is what happens when you have a monetary tightening, and in our fixed-rate mortgage system the so-called "lock-in effect" where people stop selling houses and they stop moving.
First, the more general topic: What about FRMs versus ARMs? Let's think for a moment about the mortgage channel of monetary transmission. How does the level of the interest rate set by the central bank affect the economy? One of the channels is the so-called mortgage channel. This works very differently than the kind of "new Keynesian" model, Taylor rule—the stuff that we normally talk about in macro—because it's about redistribution across agents. There's a very nice paper by Adrien Auclert which really focuses on that.
The mortgage rate is going to change the monthly payments that borrowers make. Of course, it's also going to affect the payments that lenders receive. There's going to be an aggregate effect on spending, on consumption, if borrowers change their spending more than lenders do. Why might that be? One reason would be if the borrowers are domestic residents, and the lenders are foreigners. Another reason, more relevant for a large economy like the US, is that the borrowers have a high marginal propensity to consume because they're borrowing constrained, while the lenders are unconstrained, permanent income folks, rich folks. They have a low MPC.
Now, importantly, that second argument only works if the mortgage payment changes are temporary. If they're permanent, then everybody changes consumption one for one. Whether you're constrained or permanent income, if there's a permanent shock, you change one for one. Now, what this means is that the mortgage channel is stronger for adjustable-rate mortgages than for fixed-rate mortgages, for three reasons.
The first is that ARM payments are linked to the short rate. FRM payments are linked to the long-term mortgage rate, which typically moves less—less than one for one when the Fed changes the short rate). Secondly, ARM payments change for all borrowers automatically, but FRM payments change only for new borrowers, and on the downside, people who refinance.
Refinancing is often limited by credit problems that people have. This was a big effect in the global financial crisis, that in the hardest hit regions of the country people had negative home equity, so it was very hard for them to refinance. There's also a lot of evidence—and I've done a lot of work on this—that there's inertia among borrowers, and particularly poorer and less educated borrowers are often very slow to refinance. There's evidence of very troubling discrepancies by race also, in terms of the propensity to refinance. Stimulative monetary policy only reaches the fixed-rate mortgage borrowers who refinance.
Then the final point, getting back to what I said on the last slide, is that if the Fed cuts the short rate that's going to temporarily reduce ARM payments until the short rate goes back up. The change in the long-term mortgage rate is always more persistent because it's a forward-looking asset price. It's closer to a random walk. That means that fixed-rate mortgage lenders will adjust their consumption more and that will offset the effect on borrowers.
The point I'm trying to make is if we want a strong mortgage channel of monetary transmission, we're much better off if we have a higher ARM share in the economy. Now, why do we care? We might not care. You might say, "Look, if the transmission is weak in this country, the Fed can just step on the gas more aggressively, move the rate more. That's going to be fine."
Maybe, but the central bank may care if it wants to control the relative impact of monetary policy on households through consumption and businesses through investment. The mortgage channel is about households and consumption spending, so this is affecting where monetary policy hits the economy.
There can also be regional differences in the strength of the mortgage channel. In the US, we know the ARM share varies regionally, so that actually made a big difference in the recovery from the global financial crisis. Finally, of course, if we think about a European context or an international context, if the central bank is pegging its exchange rate to a foreign currency, like Scandinavians do to the Euro, but they have a different mortgage system, they're going to get a very different impact of monetary policy.
Finally, there may be circumstances where the central bank may want an even stronger mortgage channel than ARMs give you. For example, when the short rate's close to the zero lower bound, this channel gets turned off. So, one approach is to build forbearance provisions into mortgage contracts ex ante. We know we can do that ex post; we just did that in COVID. It's always troubling to change financial contracts ex post. That really kind of messes with the foundations of the financial system, and it might be better to have a mortgage structure that would have provisions ex ante that would define when this could occur.
I've got a recent paper in the Journal of Finance with Nuno Clara and João Cocco, where we use a structural equilibrium model to demonstrate that there can really be welfare gains from allowing zero amortization. In other words, people only pay interest in a recession. And very importantly in that model, we look at what happens to default, as well as consumption.
Now, I want to move on and quickly talk about a very currently relevant issue for right now in the US, which is monetary tightening and the lock-in effect. Everything I've said so far has been symmetrical between easing and tightening, but there's another mechanism by which a fixed-rate mortgage system can specifically weaken the effects of a monetary tightening, as we've been seeing recently. If it's disadvantageous to refinance your mortgage when rates have risen, then fixed-rate mortgage borrowers become reluctant to move, and that's called "lock-in."
A decline in homes for sale can prop up house prices which reduces the contractionary impact of higher interest rates. Lock-in can also reduce the efficiency of the housing market and the labor market by making people unwilling to move to better jobs. There's a very nice paper by Fonseca and Liu on this, and I've got a chart here from their paper. You might think it would be hard to show evidence of lock-in in data. After all, until very, very recently, we've lived through a 30-year decline in interest rates, so we haven't seen a big spike of the sort that we're living through now. So, how can we do research on this?
Fonseca and Liu however, point out that because there's a band, think about the rate falling between zero and, say, about 1.8 percent. Within that band, it's not worth refinancing if you're not moving. You haven't reached the threshold that justifies the fixed cost of refinancing. Within that band, if the rate has fallen since you took out your old mortgage, the more it's fallen the more you have an incentive to move. It's going to tip you towards moving.
What they show in the paper is that the horizontal axis on the figure there is the change in the decline in interest rates since you took out your old mortgage. It's almost always positive in their data because of the trend downward, a decline in interest rates. Between roughly zero and roughly 1.8 on that horizontal axis, the moving rate increases with the extent to which rates have fallen. It flattens out above 1.8 because if rates have fallen a whole, whole lot, then don't move, just stay where you are and refinance. There's no further effect on the moving propensity.
What we're living through now is we're way off to the left of that figure. We're in a region that we haven't actually seen before. Where are we now? Here's a FRED plot of recent mortgage rates. You can see the very sharp increase in the fixed mortgage rate in 2022 and then some further creeping up in 2023. In 2021 that 30-year fixed-rate mortgage average was below 3 percent, and now it's over 7 percent.
What's happened to house prices? Not much. There was a little bit of a dip in 2022, 2023. I'm using here the Case-Shiller National Index. That index is up over 50 percent since January 2019. There was really a very, very modest response in 2022-2023 and house prices have resumed rising.
One possible explanation is lock-in—that existing homeowners are reluctant to sell because they would give up their old, cheap mortgages. This FRED plot is the housing inventory, the new listing count in the US. You can see how that collapsed for pandemic-related reasons in 2020, but the important point is to look at the decline in new listings in 2022 and 2023.
It's a little tricky working out the equilibrium effect on prices because sellers are also buyers. If people are not selling, then they're not necessarily buying, but it is certainly possible to get an aggregate effect. I have a student working on this. There's a number of people who've been working this out and showing that you can, in fact, get effects on prices.
You can also get cross-effects, so that, for example, the parts of the country where people live who would normally be downsizing, for example, suburbs of Boston, with big houses suitable for bringing up kids, that's where prices rise a lot because the people who would sell and downsize are not doing so. Yet you still have the pressure of buyers who want to move into that phase of their lives.
There can be distributional effects on prices. Do we have to live with this problem? No. There are all kinds of ways to fix it and other countries use those mechanisms. One mechanism is assumability, where if you have a mortgage on your house and you sell it to somebody else, the new owner can take over the mortgage. That's standard, for example, in Denmark, which also has a fixed-rate mortgage system, but very uncommon in the US. In the US, certain government mortgages—Ginnie Mae, FHA, VA mortgages—can be assumable, but only if the new buyer also qualifies for the same government mortgage. So, it very rarely happens.
Another solution is portability. When you sell your house, you can take your mortgage with you to your new house. That's very common in Canada and the UK, but basically unknown in the US. A third mechanism is you allow borrowers to buy out their old mortgage at market value. That's what happens in Denmark, a very smoothly functioning system, unknown elsewhere. Of course, lock-in doesn't arise at all in an ARM system of the type that we were talking about earlier. I'll stop there; the back of the slide has some of these references I've been making but let me leave that with you. Thank you.
Mishkin: Thank you very much. Why don't I first of all open it to the panel to comment on anything that they've heard from anybody else, and then I'll use some of the questions from the floor. You guys want to comment on each other?
Zervos: Should we go in reverse order again? I guess I'd make one comment on John's analysis. It does strike me that in the fixed-rate mortgage system, one thing we've missed a little bit is the wealth effect, again from people feeling the liability side if their balance sheet has fallen in value with these higher rates, the value of their mortgage (as in Denmark, where you can buy it out at 80 cents on the dollar): they can't do that, but they still feel it. That's an important positive wealth effect that comes from higher rates, which is maybe behind some of the resilience in the consumer in terms of how they feel.
I remember my old friend Andreas Lehnert, who I presume is still in the system somewhere, and the Board of Governors wrote a piece on the virtues of adjustable-rate mortgages at the request of Alan Greenspan in the lead-up to the 2008 global financial crisis. If I remember correctly, Andreas probably wanted to tuck that as far away from his legacy and his CV as he possibly could. I would be very surprised if there was a big push, regulatory-wise, to go back to adjustable rates, because of what happened in the global financial crisis in '08. Those would be the two comments I make.
Reichlin: Maybe I have a question for you. Since I don't know if we said actually different things, I wonder: what exactly is the transmission mechanism that you have in mind? You have an interest rate instrument, you are on the flat part of the demand for reserves, so you can move interest rates without moving reserves. And you can also move reserves without moving interest rates. What is it? How does this pure quantity mechanism work, in your view? It seems like it's really a return to old-fashioned monetarism that you have in mind.
Zervos: On that front, it is. It's a return to some of the thinking that was around in the Friedman-Meltzer era that has largely been pushed aside. When I first got to the Division of Monetary Affairs in the early '90s, when I was finishing up my graduate work, the Monetary Studies section—which, I don't even know if it exists anymore, Monetary Studies, which did all the aggregates—that was the premier section in Monetary Affairs, estimating money demand equations, thinking about M1, M2, M3, M2 plus every kind of aggregate we could think of. I guess just a little more work on that, and thinking about how we transmit an ample or abundant reserve system into the real economy.
Reichlin: You have an interest rate on reserves now. You didn't at the time.
Zervos: I agree. They would probably be upset about that, certainly Friedman would be. He never liked interest on reserves. He thought you should never put a price on anything you can create for free, which was an interesting argument at all times, but...I'm not saying I have the answer, Lucrezia, I really don't. I'm looking at it...
Reichlin: Coming from the ECB, I have some experience with monetary alliances.
Zervos: Right, and you have many more that are going to veer toward the Austrian side there than probably at the Board of Governors, especially in the North. I guess where I feel the most confident in my view is on the redistributional side, the loss side, where the central banks are now taking large losses, and in particular large losses on these mortgage bonds, and private sector bonds in Europe, and corporate bonds. There's a gain in the private sector from the liability side of their balance sheet contracting, but there's no offsetting loss in the private sector. That loss has been socialized.
There are these positive wealth effects that were never there before in a big balance sheet world that I think we've misunderstood and misread, that can create offsets to the tightening that comes from higher rates. That I feel very confident about, and I think that could be quite large. On the transmission of reserves through the system and the old-fashioned, monetarist views: I want to believe it, I just don't have the mechanisms to put together. My mind's been numbed by being in the markets for 35 years so that my academic abilities are just not what they used to be. I'll leave that to the fresh PhDs, that are 25 to 35, to figure out.
Campbell: David, I wanted to ask you: you were using a couple of different metaphors, and metaphors—we shouldn't sniff at them. They're actually very, very powerful, and I will walk away with the image of the two dials. But also, the other image you used. The other metaphor was medications, painkiller. You said QE was a painkiller, and it's still in the system. I'm wondering, though, whether we should tweak that second metaphor. Is QE like a medicine that actually kills pain itself, or is it like one of those things where there's an interaction between different medications? You're often told, "Well, if you're taking this pill, you shouldn't take this other pill, because it'll prevent the first one from having its effect."
It seems to me you were sort of saying that a big Fed balance sheet actually turns off the effect of interest rate policy to some degree, right? It weakens it, because you were saying now, having taken the QE, now you have the big balance sheet. When the Fed tightens that you don't get this pain, so you don't slow down the economy as much as you thought. In other words, the presence of a balance sheet is like a medication that inadvertently turns off another medication. It would also follow then that if a recession comes and the Fed wants to cut the rate again, that that would also be less effective, right? Because some of the gains now are socialized.
Zervos: I think that's true, on the latter part, and I would agree with that. There is a dulling effect. I guess I thought about it differently, but I'll have to think about the way you put it because I like the way you're thinking about it as well. I always thought of the interest rate medicine as a kind of Tylenol. I've got a headache, I take my Tylenol, and it solves my problem and I know how to use it and I know how to prescribe it. Then, one day we ran out of Tylenol. It just didn't work anymore, so Ben got in the back room and pulled the curtains back and "Wizard of Oz" style created QE and the magic medicine appeared in 2009.
That's more like the advent of morphine, and all of a sudden now we're on this really powerful painkiller that we don't have a lot of experience with. It may have some very addictive side effects, or negative side effects. We've seen countries that get hooked on balance sheets go down in flames: Zimbabwe, Argentina, Turkey. People that abuse the balance sheet.
It's not that we don't know what this drug is; we know very well what it means to print. We know what it means to abuse the balance sheet. We don't know what it means to use it in small quantities when we run out of traditional painkiller. I think of it that way, and then I think I want doctors, like many of the FOMC members here, that know how to administer that painkiller, and know how to take it out of the system so that I don't get hooked on it, or get messed up, or get politically pushed into overusing it.
Your point is very interesting, and probably worth thinking about more. What's the interplay between the two? Is there a dulling of one, so that now I'm in the mode of, I've actually turned off the ability of the traditional mechanisms to work because I've introduced the unconventional? I've got to think about that more, but I think that's a very interesting alternative way to think about it.
I've looked at it as you run out of one and you bring in the other. But then when you go back to taking the Tylenol out of the system, you still have all this other painkiller in there, so it just doesn't work the same. That's not exactly the same as nullifying the effect. To me, we still have QE painkillers in our system. That's what I would argue, and that's why policy overall is not that restrictive.
Mishkin: I want to criticize something that David said, and it relates to another transmission mechanism which we haven't talked about, which is managing expectations. David's clearly a part of the "team transitory" view. He said it's all supply shocks, and that's what explains everything. I think that's a mischaracterization of what went on.
The supply shocks, clearly, are very important, and when you think about monetary policy it typically should look through supply shocks because they're transitory. There's a very strong case, in fact, this is the view at the Fed. It meant that they had made a huge mistake, and that there were very big supply shocks, but then massive demand shocks that occurred at the same time.
That leads to the persistence issue, because when you want to look through, if something's transitory, you just look through it. You don't really adjust monetary policy. It's typically why the Fed focuses much more on core inflation, in terms of thinking about monetary policy, than the transitory movements of headline inflation.
In this case, the Fed basically ignored that. We had huge demand shocks from two sources. One is massive fiscal policy, and that's some of the questions that have been asked on my iPad, and also that the unusual aspect of this COVID effect on the economy was the lockdowns, which is something that we had never, ever worried about before. Probably the last time it really happened in a major way was during the so-called Spanish flu—which had nothing to do with Spain but happened in 1918—where people couldn't spend. A remarkable case of where the economy ended up, with huge unemployment and yet people's balance sheets improved tremendously because they couldn't go out and spend.
I'm a foodie. My whole marriage is based on eating together, and we couldn't go out. Then finally we got shots, and I could spend again. It had huge effects, and then the absolute massive fiscal policy actions, both at the start of the pandemic, but then—I think, mistakenly—later on.
That then cements in more persistence, and also very importantly, affects expectations. If you think about success in monetary policy, a huge amount is not what you do but what you get people to expect that you'll do. Then you have to do it in order to actually confirm that. That a key part of the story of what occurred.
The Fed, thinking everything is transitory, looking through it, didn't react by raising rates. What happened? One of the things you start to see in the data is that inflation expectations start to move up. Also, there was a technical issue that average inflation targeting, which was a good theoretical idea, was extremely badly executed. The result of this was that we actually ended up having persistence from the supply shocks, which typically wouldn't have happened. This is my interpretation of what went on. Let me just finish.
On the other side, the Fed finally somehow woke up and did something extraordinary for the Federal Reserve. As somebody who's been inside the system, it's usually very conservative, doesn't move very fast when it makes a mistake, and that's very problematic. Instead of being the supertanker that takes 20 miles to turn around, they became a motor boat. You immediately see that effect on terms of managing expectations because the inflation expectations settled back down to the level that they were before.
When you think about the inflation process, inflation expectations are critical. This work by Nakamura and Steinsson, for example, is very, very cogent on this issue, that a lot of the movements of inflation are actually due to these movements in expected inflation. This issue of that transmission mechanism is actually also very important, and was very important in this cycle as well. I have another question about just a clarification, but you may want to respond to that.
Zervos: Well, you took objection to what I said, so I'll take a stab at objecting to the objection. I think the Fed did an incredible job, an absolutely incredible job, of managing long-term inflation expectations through one of the greatest inflation shocks of our lifetime. When inflation was peaking at 9 percent in the middle of 2022, 5-year/5-year breakeven inflation was still in a 2 to 2.5 percent range. It never broke. When the Michigan data started to peak at 3.3 for the long-term (5- to 10-year) inflation expectations, the number one survey that we have on long-term inflation expectations—Jay got into gear that June meeting, and he just went: 75, 75, 75, 75—four in a row. He saw the risk of a de-anchoring, and he went aggressively.
You're absolutely right that central banks are supposed to look through supply side shocks. That's right; that's the textbook answer. But you can't look through three persistent shocks that might dislodge long-term inflation expectations because your number one goal is to make sure that you anchor long-term inflation expectations. If those supply shocks last too long and they come too quickly, you have that threat. What I think the Fed did was actually almost textbook 100 percent correct.
Mishkin: So, we're in complete agreement on this.
Zervos: That you wait, you look, and if they last too long and they start to affect the anchor, you go and you go big. I say that's exactly what a textbook response should be, so I don't look back at any of this as a major mistake. I think that the mistake was defining "transitory" as some sort of monthly or quarterly concept. "Transitory" should have always been defined with the anchor in mind, that we didn't lose that 5-year/5-year breakeven at 2 to 2.5 percent, that we didn't lose the University of Michigan's 5- to 10-year survey, or the New York Fed's inflation expectations survey.
We can lose one-year inflation expectations, and we can lose two-year inflation expectations. Those are supply shocks, those are short-term movements and fluctuations. You're never going to want to or need to control them. But that long end of the yield curve, that inversion that we've had for the longest time since the 1920s all sits back with one incredible thing. The credibility of the central bank came through stronger than it's ever been, after one of the worst inflation shocks that we've ever seen.
I don't think my story of supply over demand is...I think it's a very cogent story. I don't think it sits at odds at all with how the Fed operated. I really think behind the scenes, as I listened to Jay speak through every one of his best press conferences, he seemed very comfortable with the supply side story all the way through and seems even more comfortable with it today than he's ever been.
This idea that demand was really a driver I think is starting to dissipate significantly, and if I hear one more person blame fiscal profligacy for the ills of the world I think I'm going to vomit. I mean, now they're going to probably tell me that fiscal profligacy starts volcanoes and earthquakes and climate change. It's just amazing how every CEO under the sun goes on and says, "Our number one problem is we have too much debt, and our deficits are too big and we need to stop that immediately or the world's going to end." It's nuts.
I come from a Rochester PhD and a Chicago school training where to me fiscal policy neutrality was always a lot more important. The idea that we have big fiscal multipliers just doesn't resonate with me. I don't see them in Japan after 250 percent debt-to-GDP ratios. They've spent more money than anybody could ever imagine via the government, and all they've done is crowd out the private sector, not grow, and create almost no multipliers that I can find.
I don't buy the fact that we missed this through fiscal. I don't think the fiscal was nearly as important as everybody says, and I don't think the "fiscal Armageddon" story that is being pushed by so many today is actually one that...I certainly don't push it, and I try to push my clients away from believing it. I just don't think it's the story.
The story to me fits beautifully with a big set of supply shocks that were negative and now we're enjoying the opposite side of that—how much more, I don't know—and that the demand side was really just a sideshow in the whole thing and really executed by the Fed to make sure that the anchor did not dislodge. They get kudos all around for keeping that anchor in place. That's how I would describe that. That's how I would respond.
Mishkin: Let me ask about the issue of the balance sheet and how it may affect the transmission mechanism. I see it as a positive rather than as a negative. One of the problems of using interest rates as a tool of monetary policy is that you can actually end up with nonlinear events and actually have financial stability issues. That's exactly when we've sometimes seen the Feds tighten. If the balance sheet effects are not socialized because they're not held by the Fed, all of a sudden you have sectors of the financial system which could become under a lot of stress.
One potential plus of the ample reserve reserves regime that we have now is that those nonlinear effects are no longer happening because, as you pointed out—and John has talked about this as well, and Lucrezia—the losses are now actually not taken in the private sector. They're in fact taken by the government, which basically has very little effect on the economy, and particularly in a nonlinear way.
When we think about this, it can tell us that we have this advantage of now that the interest rate tool becomes more usable because we don't have. This is a very interesting issue in thinking about how we would do this. One other thing...
Reichlin: Can I make a point?
Mishkin: Yes, go ahead.
Reichlin: We are having two discussions here. One is what happens when you go from non-neutral to neutral. It's perfectly plausible that there is interaction between the speed at which you decrease the size of balance sheet, and the interest rate; it's just the effect on different parts of the curve. So, it's a question of calibration.
But in normal times, when you are in a neutral state, the ample reserves are... as you say, the main motivation is financial stability. When you have interest rate and expectations, that channel, which has always been the textbook channel in the models that you write, it works as usual. So, nothing has changed, except that now we have this kind of big...
Mishkin: Right. There's an issue that we like to be in a linear world, so that what we've seen actually, which is a huge change, was that we basically would have pretty much a linear world for a very long period of time. That's very easy to deal with from the point of view of monetary policy. You have something go on. Think about the recession that occurred in 2000, 2001. Basically, it was very linear. There are some problems; what do you do? The Fed cuts rates, and the economy returns to recovery.
When you get a nonlinear world, you go off the cliff. So actually, I would teach this. If you like the Roadrunner cartoons, I said, "Here's a great economist explaining what happened during the great recession." It's of course the Roadrunner, and Wile E. Coyote sits there, and then goes off the cliff, and then he plummets down.
One of the things actually that we want to think about, and I think John's discussion of the mortgage markets is very relevant to this, is how we set things up so that we're in a more linear world, where the economy doesn't go off the cliff when monetary policy just can't solve the problem. That's a question of the design of our transmission mechanism, in effect, to make the world more stable.
That's a very key issue in thinking about these balance sheet things. It wasn't necessary, by the way, up until all of a sudden we went into a much more nonlinear world where there were these dangers.
Campbell: One of the issues there is the extent to which banks engage in maturity transformation. One of the things that gives you the nonlinearity is when there's a lot of maturity transformation in the banking system, and then—as we learned in the S&L crisis—it might raise rates, and you blow up part of the system. Now, the modern securitization system of mortgages in this country was intended to get rid of maturity transformation, so that we would have these MBS that would be held by long-term investors—bond funds, pension funds, insurance companies; get them out of the banks.
That was always the idea, and the problem is the banks keep finding ways to get the MBS back on their balance sheet; right? This was this was the problem with First Republic, and I'm not sure what we do about that. But certainly, one question is how banks are required to account for their holdings of MBS, and whether they're required to mark them to market or can treat them as hold to maturity assets. Very interesting issues there, which other people in the room know much more about than me. But I do think trying to get those MBS off bank balance sheets is a part of this building of a linear financial system, to use Rick's phrase.
Zervos: I do think there's a lot of deals happening today. I think TCW and Wells Fargo just engaged in a deal, but there does seem to be a move in the banking world to push banks to be more of a sourcer of loans, a sourcer of deals, a sourcer of product, for those with the long-term liabilities that better match that long-term asset class. If we can get our banking system to be more of a payments system, lending to the corporates that they know but then passing those loans back out and sort of finding good homes for them that aren't in places where they're short-funded, I think we'll have a much healthier, stable financial system.
I kind of feel like the regulators are pushing it that way, and maybe the Bill Nelsons of the world can keep pushing us that way. There's good people out there that know that story well, and we're going to have a safer and better banking system because of it. That's happening, and I see it now with the partnerships that are happening between long-term asset managers and banks, because the asset managers want the loans and then banks become the conduits for that.
I had a question for you on the ECB, because it strikes me that the ECB's view on the balance sheet is much different at the moment than the Fed's. There's a desire to bring it down, bring it down quickly as you said, and I see that as a real risk in the European system, as you might expect, and sort of a missing of an opportunity.
Reichlin: I thought you liked it.
Zervos: No. Just for the record, I really like big balance sheets. I think big balance sheets are great.
Reichlin: Oh, you do like them? Okay, great.
Zervos: I love them because they kind of get almost back to Rick's point, which is they bring us to a higher R-star. We get away from zero, we get away from nonlinearity. I would have thought the Europeans would go, "We don't want to go back to negative rates; we don't want to go to zero."
Every bank, every insurance company, every pension fund was banging on the ECB's door going, "This sucks." All of a sudden, you have this opportunity to dial up the balance sheet, and it allows you to keep the rate structure at a higher level because you're using more of the painkiller on the balance sheet side. You can pull back on all of that rate painkiller and it keeps you away from zero. It keeps you away from the nonlinear '80s. I think it solves an enormous number of problems.
The one thing, who I think it'll piss off, it'll be those that don't like the monetary financing of government debt, because what it means is the ECB will hold more BTPs, at equilibrium. The Germans will be upset about that.
Reichlin: Okay, but again, this is not about the zero lower bound. This is about what is the size in the steady state, the structural size. I think they don't want it, because it's not a federation. They don't like to hold bond, because that means bond of different states, which face a different sovereign debt, sovereign risk, and so on.
This is the reason, because as I said otherwise, this ample reserve system that I thought you didn't like makes a lot of sense. Now, whether they will be able to...they're not going to go back, exactly, to the pre-2008 crisis, because they also understand as far as I understand that, because I'm now out of the ECB, I think that they understand that this would be risky and that we live in a different world than the pre-2008 crisis.
They will go, ideally, to a smaller balance sheet, but they will go very slowly. Whether they will manage to go back to a relatively small balance sheet, we don't know. That's why I said this is a natural experiment. Can central banks today, with the financial market that we have today, with the type of liquidity regulations that we have in the market, with the mismatch between the demand for market making and the supply of market making...can we go back to a small balance sheet? We will find out, because the ECB is experimenting.
Zervos: Sounds dangerous to me.
Campbell: David, you mentioned your Rochester and Chicago training, and from that historical perspective the idea that the central bank should be owning a huge fraction of the private financial liabilities, and that that's a great thing, I'll just say it's ironic. It may be right, but it's ironic. [laughter]
Zervos: It is. I feel a little bit torn by that. I guess I go back to more of my fiscal policy roots with that, which was the lack of sort of more "Bob Barrow, our government bonds, net wealth" stuff, that keeps me away from thinking fiscal has an enormous impact. On the monetary side, I think I've become a heretic of sorts, and I'm okay with it. I've had my "come to Jesus" moment with the idea that monetary policy matters a lot. It matters more than we think, more than I was taught. I don't believe in monetary neutrality, and it's not really a Lucas Critique sort of problem for me. It's just, watching it every day for 35 years, being in the markets and seeing what impacts everybody...and maybe it comes back to what you were saying on expectations, and policy just has such enormous effects on expectations.
I do want to, one, clear the air a little bit on the balance sheet story. I think we've unlocked a bit of something really powerful, and something really beneficial to society and to policymakers, when we think about balance sheets and running them bigger on a more permanent basis like this, even though we haven't defined what ample reserves are, we haven't defined what any of these concepts are.
I think we're getting back to a world where we're more comfortable with rate structures that we're more comfortable with, because the balance sheets got us out of jail, they got us out of NERP, they got us out of ZERP. We were all going to be Japan, remember? That was the story in 2008, 2009. We were all going to go to liquidity traps and zero rates and never get out, and the managing of the balance sheet is what got us out, both in the GFC and in COVID.
Somehow, we got out with this tool, but nobody likes to talk about it. Once we're out, everybody's like, "Let's just leave that over there, and forget about it. I don't know how that fits in the model very well, and I don't know how that works exactly, and I don't know whether the asset side or the liability side is having an impact, but let's just let it dribble lower until Lorie Logan calls up and says there's a problem, and then we'll stop."
That's not the right answer. The right answer is to take the tool and use it for the bettering of monetary policy. I like to always tell clients, I say, "You guys, you tech funds and you guys in the pharma business, and you guys in every corporate suite we talk to at Jefferies, you think there's no technological advances on the policy side? Well, there are. There are massive technological advances." I think we'll look back 50 or 100 years in the history books at what was done in the 2010 to 2020 region as a huge, major technological advance in the administration of monetary policy, and how monetary policy evolved into something so much more than just tweaking 25 basis points or 50 basis points in an FRB/US model.
Understanding how we unleash that power and using that power constructively...I just hope it's the topic of many, many PhD theses at Harvard, John, I really do. It's just so underappreciated because we kind of just shoved money aside. We just said, "M1, M2, go away. Reserves don't matter. None of this matters."
Mishkin: I think that was actually right, to get rid of that. On the other hand, there is this issue of this tool, and it's now, by the way, in the toolbox of central banks because we didn't just use it once. We used it a second time. It's particularly effective in terms of when economies have a nonlinear hit. When you look at both the global financial crisis and COVID, these were huge nonlinear hits.
It also has a big effect in terms of the modeling that we do, so standard new Keynesian models are completely linearized models. That was the huge mistake, that and that they didn't deal with the fact that there are different agents and asymmetric information is a key part of the story.
That quantitative measures matter, I think, is something that's actually well accepted in the profession, well accepted in central banks. M1, M2...Milton Friedman, in one way, won the war about actually thinking long-term about monetary policy. But on this one, and Allan Meltzer, who is a very close friend...they were nuts on this. Only if you have horrendous policy—when you talked about the cases of like Venezuela and so forth, that's not what we're talking about for advanced countries. It's really not the key. I think that there is a different view here, but it's completely different than the old-style monitors do.
Zervos: What they were worried about was that once you try it, it's like morphine; you can't stop it. They were worried about that; rightfully so, but what we've also seen is that doctors with newly advanced painkillers use them and they use them well and they get you off them well. That's the technological innovation that I'm trying to impart when I make this discussion.
Mishkin: I want to make sure we at least get one question from the audience. [laughter] But it also relates to something we talked about, which is the issue of R-star. You talked about R-star rising. I'm not sure I understood the mechanism, but that's something that we can discuss.
Clearly, where R-star is, is very critical to thinking about where the stance of monetary policy is. Unfortunately, from a theoretical concept, we don't quite know what it is. From an estimation viewpoint, we don't quite know what it is. The question from the audience was: How much does fiscal policy affect R-star? It's a different kind of mechanism than the one that you and I had talked about.
On the other hand, where do we think R-star is? Why is it moving? Or maybe actually, to me, one of the important aspects of good policy and good monetary policy is admitting what you don't know. The big mistakes in monetary policy have been, we think we know what the natural rate of unemployment is—so, the U-star—and in fact, I think that was again part of the mistake that the Fed made in the early phases, and the Great Inflation is very much part of that story.
Also, R-star is very key. Now, one of the ways to do this, if you don't know, you can basically do it by the feeling approach. The art of monetary policy, the science, is actually important. As an academic who's been involved in research on the science of monetary policy, I'm very proud of what our profession has done to effectively professionalize monetary policy, but we also have to admit that we just don't know where things are.
The issue of R-star is a critical part of the story. The actual question that was asked is: Where does fiscal policy play a role in this? You also had a mechanism which maybe you'll elaborate on, because you mentioned it, and it came out from somewhere, and you obviously have some ideas where it comes from. Why don't we open the floor to this discussion of R-star so that the audience feels that I did my job well as a moderator.
Zervos: I hate to bring it back to the balance sheet, but I will because I think the balance sheet has fiscal implications. We're going to have a $7 trillion balance sheet. That's where the equilibrium looks like it's going to end up. We have $34 trillion worth of debt. Everybody gets on CNBC and goes, "Oh, my God, we've got $34 trillion worth of debt. We're never going to be able to pay it back, our kids have to pay that back." That $7.5 trillion that sits at the Fed will never be paid back. That will always be rolled over by the Fed. It's a perpetual liability, it doesn't have a maturity date, it doesn't have an end date.
In a sense, what running a bigger balance sheet has done is it's eased fiscal policy by allowing us to have $34 trillion of debt, but only go back to the market and roll $27 trillion worth of debt because the Fed will always soak up that extra $7 trillion. At every auction, the Fed will be there, and they will be replacing bonds that mature with new bonds, and they will be part of that mechanism.
What they have done is made the fiscal burden of the United States go down by running a bigger balance sheet. Again, I think a stimulative activity, which in a way is to me representative of a fiscal easing of sorts because that is the pure inflation tax win. That is government spending that has been funded by the Fed's balance sheet. It's not been funded by an excise tax, a sales tax, a VAT. It's been funded by the Fed, so that's a fiscal expansion.
Mishkin: It could actually work in the opposite direction. You basically now have given more fiscal capacity, in a sense...
Zervos: You do.
Mishkin: ...because the Federal Reserve has taken this on. In that sense, you worried less about either having to resort to an inflation tax, or fiscal consolidation, and so forth. And that actually could be something that could work in the opposite direction.
This is a top question: What the hell is R-star, and what drives it?
Reichlin: Can I say, on the fiscal: what matters is the consolidated balance sheet between government and central bank, so it's just a distribution of risk between one part of the government and the other. So, who cares?
Zervos: Who actually funds it, in the private sector?
Reichlin: No, I mean we care for governance, and this is the point: that's very important, because all our governance structure is based on the idea of the '90s: central banks' balance sheets were narrow, so there was a strong separation between the fiscal and the monetary. Once you open that door, then you can see that you may want to consider other types of governance design, and that's a very important issue. This is why I think that, although we may still do monetary policy in the same way, expectation and interest rate as main tools, we are not really in the same world that we were in the '90s.
Now, coming to R-star, I think it's interesting to compare Europe and the US, because our fiscal policy is very different and our growth rate is lower. I'm not saying causality here, but that there is an association. This maybe gives us some food for thought of how we should think about R-stars. Obviously, fiscal policy has an effect on R-stars for obvious reasons now. How much it is, I don't know.
But again, we will see the decline of inflation in these two parts of the world. The difference might be also the different fiscal positions. Again, there is an interaction between monetary and fiscal policy. We always knew it, but with big balance sheets these interactions are larger.
Mishkin: John?
Campbell: I think this is just an illustration of the fact that in economics the most important things are the things that are hardest to ever make progress on.
Mishkin: That's why it's fun.
Campbell: But it's fun.
Mishkin: This is the reason we became academics. This is a fun field.
Campbell: I do think that beyond the issue of the role of fiscal policy is very important, but we shouldn't forget the longer run, supply side determinants as well. Things like the rate of technological progress, the demographics, the declining birth rate, which is a very, very powerful trend that seems to have accelerated since COVID and has very, very fundamental implications. And then the question of the capital intensity of investments: are we going to get into an era where we have to make a lot of physical investments in things like climate change adaptation, green energy, and so forth, that actually require a lot of capital and drive up R-star? That's a possibility.
It's very easy to list the possible fundamental determinants. It's very hard to measure them, so I think empirically we end up falling back on looking at the data and trying to feel your way, as you said, Rick, and just figure it out as you go because it's so hard to use prior knowledge.
Mishkin: Yes, and admit that we don't know. What that means is that the process of monetary policy frequently has to acknowledge that we don't know, and that may mean that you have to turn things around. The key to success in policymaking is that you're a learning organization. I'm very biased. I'm a Fed-nick, I've been in the Federal Reserve system in two roles.
But the Federal Reserve is a learning organization. There are some aspects of the ECB on that, that I had some issues with [laughter] for a period of time, but I think they've been better lately. I think also that's part of the lesson, which is knowing what we don't know means that we're also going to operate policy in a different fashion. With that, it just clicked zero. Thank you very much for listening, and I hope we provided food for thought.