Floors and Corridors

David Beckworth: Hey Macro Musings listeners, today’s episode with George Selgin was taped in front of a live audience on January 15, 2019. Since this time there has been an important development on the topic of our conversation—the Fed’s operating system. So stick around at the end for an update.

David Beckworth: Today is a special Macro Musings episode. We are in front of a live audience at the Cato Institute to chat with George Selgin about his new book titled, Floored! How a Misguided Fed Experiment Deepened and Prolonged the Great Recession. George is the director of the Cato Institute’s and monetary and financial alternatives and as a previous guest on the podcast. In fact, George, I believe you hold the record for the number of appearances on the show, number four now. Is that right? 

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

George Selgin: That's right, David. I'm determined to be a number one in at least this. 

David Beckworth: Okay. Alright, we're glad to have you on again and today we're here to talk about your book. This is Floored and of course George wants you to stop by and get one on the way out if you don't have one already. And this book deals with the Fed’s operating system, it switched in 2008 from a corridor to a floor system. And George, I want to begin by asking you, when did you first start thinking about this transition and implications for the economy? 

George Selgin: That's a good question David. I know I was thinking about it before I came here to Cato, which was in the Fall of 2014, I had been struck by the Fed's decision in 2008 to start paying interest on bank reserves. I thought that was a very strange thing for them to be doing at such a time but it wasn't until I was asked to testify on this subject before Congress in 2017 that I really began on the path that took me to writing this book because I ended up writing a 22,000 word testimony which I then had to summarize in five minutes before Congress. But ultimately I chose to do the opposite by expanding that testimony into a book. So it was that call to Congress, I think that really got me on the road to writing an entire book on the subject. 

David Beckworth: That was a great catalyst because it provided for a nice timing. Your book just came out and as we know, the FOMC, the Federal Reserve is now talking about this very issue. So the last two FOMC meetings, the minutes have come out. They've shown us that this has been a big topic they're considering, we know in June they're going to have a conference. Now, this conference may or may not touch on this issue, but it seems like there's a lot of conversation going on right now about a number of questions facing the Fed. So this is a well time book. 

George Selgin: I hope so. It isn't clear to what extent the Fed is really seriously considering alternatives to the current system, including switching to a corridor system. We'll talk about what the differences are later, I'm sure. It's not clear just how open they are, but the point of the book is to try and make them open to serious discussion of these alternatives instead of allowing  sheer momentum to turn this post crisis arrangement into the Fed’s arrangement for all time. And I think that that discussion is extremely necessary. 

David Beckworth: Well, I'd like to think that the discussions we were reading about somewhere in there they have seen or heard some of your arguments on this issue, but why don't we go ahead and define what a corridor system is and what a floor system is and maybe even the distinction between the type of corridor system that existed before 2008 and what you would actually envision the Fed going to if they did go to a corridor system. 

George Selgin: Sure. Both a corridor system and a floor system, they're both ways of, of setting interest rates of having the Fed regulate interest rates, and by doing so set the stance of monetary policy. Under a corridor system, a version of which is what the Fed used before 2008, the Fed regulates the federal funds rate, which is a market rate that is the rate at which banks lend reserves to one another overnight. It's an unsecured overnight market interest rate. And the way the Fed tried to regulate it was first by setting a desired target level for that federal funds rate. A level that had thought would be consistent with having a policy stance that wouldn't be too inflationary or it wouldn't cause unemployment. Then if the actual rate at which banks were lending to each other differed from that target, they would create reserves or reduce the supply of reserves to move the federal funds rate toward the target. 

George Selgin: So if they wanted to raise the actual federal funds rate, because the target was above its current level, they would sell bonds in the open market off of their portfolio, which amounted to buying reserves back from the banking system, that would make reserves more scarce and the federal funds rate would rise towards its target. In the opposite situation where they felt that the Fed funds rate was higher than where they wanted it to be, monetary policy was too tight. They would do the opposite. They would buy bonds by creating fresh reserves and that would make reserves more plentiful. So that was the corridor system. And we had a version of that before 2008. In the floor system, the Fed controls interest rate by interest rates by paying banks a certain rate of interest on reserves and the interest rate it pays banks on reserves indirectly through arbitrage type mechanisms also will govern interest rates on various short term markets. 

George Selgin: And in that kind of so-called floor system, the Fed first floods the system with reserves and then pays banks to hold on to reserves. That's what interests on reserves does. And as a result of that, there is no scarcity of reserves. So it's not just the case that the fed is regulating interest rates without changing the supply of reserves because it can do so by changing the rate of interest it pays banks to hold reserves. It's also true that changing the supply of reserves doesn't actually make much of a difference. That's how you could have quantitative easing, for example, where they created trillions of dollars of reserves, but because they were operating a floor system, that didn't really obviously affect the stance of monetary policy, or at least not by changing interest rates. The interest rates stayed at 25 basis points where the interest rate on reserves had been set and it stayed there for many years. 

George Selgin: So those are the basically the two systems, one has scarce reserves and no interest on reserves, or at least interest rate on reserves isn't regulating this stance of policy. The other has plentiful reserves and the stance of policy is set through changing the interest rate on reserves. 

David Beckworth: Okay. Now there've been many arguments made for the floor system. There are maybe the original arguments, some academic arguments made for them in the past. There's been some newer arguments made by Fed officials. So why don't you run down the arguments for a floor system, kind of the old and the new. 

George Selgin: One of the arguments for floor system… Well, let me step back a bit and say, if you have a corridor system like we used to have, sometimes there'll be situations where that system has to turn into a floor system. And I actually believe that a floor system is a good idea when you have no choice but to have such a system. When does that happen? It happens when the market interest rates fall so low that they fall below to the level where the Fed can't make its own policy rate any lower. Traditionally that was assumed to be zero. So if you get a situation where interest rates fall to zero, the federal funds rate falls to zero in corridor system, it turns into a floor system. Because at that point, of course there's nothing to play with. 

George Selgin: Increasing the supply of reserves doesn't make any difference because the interest rate isn't going to go any lower. So in a sense you get a floor system in extremis because there's nothing else you can do. So in that circumstance, it's crazy to argue against a floor system because you're stuck at the floor no matter what. 

David Beckworth: You’re almost forced into one, right? 

George Selgin: Yeah. And it's a material floor, because you can't go any lower. So you have a floor system because you can't go any lower. But we have had a floor system where you could go lower, but the interest rate on reserves is set above zero and the system is flooded with reserves at that above zero rate. Okay? The argument that’s made for that kind of floor system is that having plentiful reserves is desirable in itself. That it's great to have a banking system that's sitting not on a few $100 billion of reserves, which is what was the case before 2008 but on trillions of dollars of reserves because there's all that liquidity and banks don't have to worry about being short of liquidity and therefore they're safer and so on. That's one of the key arguments they make is plentiful liquidity and therefore safer banks. 

George Selgin: The other argument is there's no muss or fuss in adjusting interest rates because it's just a matter of changing the interest rate paid on reserves, that's a very easy thing for the Fed to do. Under the old system In contrast, the Fed had to figure out how to change the supply of reserves in order to move interest rates on the market by doing that. And that involved a lot of estimates about the demand for reserves and the elasticity of demand and all that. So the folks at the New York Fed had to do a lot of homework in order to regulate interest rates and set the stance of monetary policy. That homework is no longer necessary. That effort's no longer necessary. Those I think are the two main arguments, so not the only ones but they’re too. 

David Beckworth: So just to maybe phrase that last one a little bit differently. The Federal Reserve with the floor system can adjust its target rate, set it there and not worry about what it’s doing with its balance sheet. Is that fair? 

George Selgin: That's right. 

David Beckworth: It can expand the balance sheet, put lots of liquidity into the banking system. So if there's a banking panic, it doesn't have to worry. It has extra levers, extra hands free to move around and that's one of the big appeals. the original appeal, my understanding, was that this extra lever [is] for monetary policy. 

George Selgin: Yes, they'd like to say that with interest on reserves, the floor system, they had two things they could play with separately and they weren't tied together. One was they could set the interest rate through the interest on reserves. And the other was they could just change the balance sheet but they weren't linked together. I make an analogy in the book where let's say we're talking about driving a car. In an ordinary car the steering and the accelerator or the gas, there's a sense in which they're not completely independent. That is, the faster you go, the more careful you have to be about where you steer. And you certainly don't want to try to take a really, really sharp turn while going 80 miles an hour say, so there was a connection between these two settings and they weren't quite independent and what the Fed people are saying essentially is, well we've under this floor system, we've got a situation where we can steer the interest rate as much as we want, but we can stomp on the reserve creation gas as much as we want. 

George Selgin: But those things, how much we do of one doesn't affect what we can do with the other one. And they see this as great because they can control liquidity in the banking system any way they want to, but it doesn't affect controlling the stance of monetary policy, which is set independently and they see this as a big virtue. I see it as a big problem and I also extend the analogy to explain exactly what I mean by that. 

David Beckworth: Okay. What about the argument that a floor system improves interest rate control? I mean, as you mentioned, the interest on excess reserves as an administrative rates so they set it, it stays there. Arbitrage tends to push the other short term rates toward it as opposed to a corridor system. Is there any merit in that argument? 

George Selgin: Well, there is in the sense that you can influence interest rates that way and so certainly if we speak about the fed funds rate is the target, you could say that setting the interest rate on reserves is one way to influence the fed funds rate. But actually it can be very deceitful. This is one of those places where the Fed story is a little bit deceitful and that's because this gets a little subtle, but in the old system, right, the fed funds market is very active because the reserves are scarce. Some banks have to borrow them and others don't. So we have a lot of trading in the federal funds market, you typically had like $200 billion a day traded in federal funds market. And the interest rate in that market was definitely a meaningful interest rate. 

George Selgin: So changing the supply of reserves could move that interest rate around in a meaningful way, and that's how the old system worked. In the floor system, there's a sense in which the interest rate on reserves perfectly controls the fed funds rate. But it's in a trivial sense because in a perfect floor system where all banks have plenty of reserves, there's no need for any activity in the federal funds market. There is actually no effective federal funds rate, interbank rate. What you see today is not an interbank grade on the fed funds market. it's how much banks are lending to non-banks on that market for. And that's a side story we don't need to go into. So in a pure floor system, you have no activity in the fed funds market. So there is no fed funds rate. So the central bank can say they're kind of perfectly controlling a rate that doesn't actually do anything where there's a market that's not active. 

George Selgin: And so what really matters is whether setting the interest rate on reserves influences active market rates, and it does, it does influence other short-term market rates, not particularly well. See, it's when we gauge it that way we can ask whether it works better than the old system, and I think the answer is no, it doesn't work any better than the old system. It might work as well for regulating interest rates. It's problems have to do with its consequences for other things, not for whether it can set interest rates that way or not. 

David Beckworth: Well, I raise this question because at the last FOMC meeting where the minutes have come out, they bring up this issue of interest rate control. Some of them apparently are worried that if we go to a corridor system we'll have far less control over short-term interest rates. And I think this speaks to maybe a misunderstanding and I hope you can speak to on the difference between the corridor system we had pre-2008, what a symmetric corridor system would be. 

George Selgin: So first of all, the old system, they control the interest rates quite successfully for decades with that old system, we know that it worked. It wasn't a so called symmetrical corridor. I should explain that the reason a corridor system is called a corridor is because you have a target interest rate, which is the Fed funds rate. Then you have below it some lower bound which could be zero and was before 2008, and above you have the Federal Reserve's emergency lending rate or discount rate and the idea was that in principle the actual effective federal funds rate could bounce around between those upper and lower levels but could never exceed them. But if the Fed open market people were doing their job, it didn't deviate very much from the target and that's what all those open market operations are doing is trying to keep those deviations small. 

George Selgin: And they did a good job of that in the old days.  Now in a symmetrical corridor, you can pay interest on reserves and if it's symmetrical, the interest rate on reserves is the same distance from the target rate as the emergency lending rate, the discount rate, whatever you want to call it. And so that's the meaning of a symmetrical range. Now you can set that range as narrowly or as broadly as you like, that is the authorities could have a narrow corridor or broader corridor. A floor system is kind of an extreme case, but if you have a corridor where it's narrow enough, there's no way the interest rate can move outside of the confines of that corridor because it's either going to turn into a floor system and it hits the interest rate on reserves or the banks are going to go to the discount window in theory before they borrow for anything above that rate. 

George Selgin: So that's one point. The second point is, I don't think it's any harder for the open market operations to manage keeping rates near the target. It's not going to be any harder than was before, the Fed officials claim otherwise because they say, well if you look at the volatility of factors that might cause the fed funds rate to move, they're very high now. But that's because things have changed because we don't have a corridor system. If you look at the history, as soon as they put this new system in place, you had a lot more volatility and reserve demand, but that's a consequence of having a different system. You go back to the old system, you won't have that volatility again. There's no reason we can't reverse all the changes that we've seen though we can also still choose to have a different kind of corridor system instead of the old system. 

David Beckworth: Okay, so those are the arguments for a floor system. What are the arguments against it? 

George Selgin: Okay. This is where we can start talking about the important thing. 

David Beckworth: In you book, yes. 

George Selgin: Yeah. Because these are all, so far we're just talking about technical issues about interest rate control and they're not that interesting because yeah, technically these things can all work more or less equally well, but there's big more political economy issues that go with the floor system and some of them are more political. Some of them are there more economy. The first thing about a floor system is it necessitates a bigger balance sheet, all that liquidity. Now the Fed people say that's wonderful. They treat it like a free lunch and in the book I say, well there's no such thing as a free lunch, there's is no such thing as a free liquidity lunch and that's certainly true today. We have the  balance sheet of upwards of 4 trillion and reserves proportionately higher and what that means is that's all part of the floor system apparatus. 

George Selgin: That means that we have the public diverting, putting money in banks, but the banks are putting a lot of that money in the Fed. Before this new system went into place, if you looked at bank deposits and bank loans of all kinds, they were approximately 100 percent of bank deposits, approximately 100 percent. They were a little less most of the time because banks held some reserves including excess reserves with very small amounts relatively speaking, and that was true for long, long period of time those two values moved almost in lockstep. Since 2008, since this new system, bank loans or 80 percent of bank deposits and reserves are about 20 percent. Now it's true, that means banks are a lot more liquid, but it also means there's fewer savings going into loans of all kinds, bank loans of all kinds. 

George Selgin: Instead, those savings are being shunted into mortgage backed securities and long-term Treasury bills, which was the stuff the Fed has been buying. So that's an important change in the allocation of resources and the Fed likes to not talk about what it used to say about this. If you go back as recently as 2002 you find Fed statements about how proud they are that among the central banks in the world, they have one of the leanest balance sheets. That is our monetary system operated with a relatively small central bank balance sheet, precisely designed to make sure the Fed wasn't soaking up immense amounts of private savings and the Fed people said this is a good thing because we want to leave resource allocation, the allocation of savings to the private sector. Why is that good? Well, the presumption is that bank lending uses savings more productively. Banks could also buy government bonds, but they presumably are looking for the most productive uses, the greatest returns, and that ultimately contributes to economic growth. 

George Selgin: Now the Fed doesn't talk about that anymore. They don't talk about lean balance sheets. They talk about how liquid the banking system is. 

David Beckworth: Let me ask a question- 

George Selgin: That's one of the several problems. 

David Beckworth: Let's come back to this, but I'm going to ask a question on that first critique. Some advocates would say you're right, there's a bigger financial footprint. They'll look at the Federal Reserve's earnings, it's huge, think of all the money it's sending to the federal government. It's helping bring down the deficit. Why not? Why not have the Fed operate as a huge, large fixed-income hedge fund? 

George Selgin: Well, first of all the Fed’s remittances are not increased much by the policy of paying interest on excess reserves. In fact, on the contrary, as a result of that new policy, two things are true. First, interest is being paid on required reserves. And since banks would have held those anyway, that difference between paying interest on required reserves and not paying interest on required reserves is a net interest loss to the Treasury that's a net interest loss. As for excess reserves, well, the banks are getting the interest for the most part that's being generated by the extra assets that the Fed has. Well then the question arises, how come we had some very substantial remittances to the Treasury? Which we did. We had exceptionally large, of course other things equal, if the Fed is earning any net interest at all, it's going to earn a lot more with a $4.5 trillion balance sheet than it would be earning with an approximately $1 trillion balance sheet, which is what it had before. 

George Selgin: So yes, you're bound to have some more remittances, but most of the increases in remittances that people are talking about is just associated with a very temporary boost in remittances because the Fed bought all these long-term, relatively high yielding Treasury and mortgage back securities, whereas before it bought lower yielding Treasury bills and it didn't increase its interest payments on reserves to the bank's proportionately. The interest rate on reserves was not as high as the interest yield on these long-term securities. However, there's another side to that. As interest rates have been rising, of course the Fed has had to take capital losses and now all those interest payments are going way down and they're going to keep going down, the interest remittances to the Treasury. So that was just a rollercoaster ride. That's not a long-term increase. And to the extent that it is a long-term increase, it's just because of a shift in the duration of the Fed's balance sheet. That's what's doing it. And so we should not think this is a great windfall for the Treasurys. 

David Beckworth: It was a combination of the floor system with the large scale asset purchases going towards long-term Treasurys- 

George Selgin: Longer-term Treasurys with high relatively high yields and the Fed not having to risk adjust its remittances to the Treasury, the Treasury gets it all. And then when the risk hits, when the interest rates rise, the Treasury is going to get less. 

David Beckworth: Let's go back to the other critiques you have of the floor system. 

George Selgin: Well, one of the other critiques has to do with the point I made earlier that it, whereas in a corridor system, banks are actively lending to each other overnight in this unsecured federal funds market. In a floor system, you've shut that down. Now that actually matters a lot according to quite a few economists, including at least one Nobel Prize winner who believe that the presence of this unsecured market, the banks participate in all the time is extremely important for eliminating contagion risk. That is the risk that that bank failures will spread throughout the system. And the way it does that is, banks that know they're going to participate in that market routinely have a strong incentive to do homework, to decide which other banks they want to deal with. So they have an incentive to learn which other banks are safe and which ones aren't, and to keep their eye on the situation so that if anything develops anywhere in the banking system, they're not gonna make any more loans overnight to any troubled bank. 

George Selgin: So the overnight unsecured market is like the canary in the coal mine of bank insolvency problems. Anything that happens there, any troubled bank will be shut out. And that's from that fact, information can flow to everybody about which banks are in trouble. And what that does in practice is to really limit the spread of panic because if a bank is ... If you know that banks are lending to some other bank in the interbank market, you know that bank is safe. That's where the first sign of trouble will happen. So you don't have to run on that bank or pull your money out of that bank. Other banks would, if they do it, then you'll want to do it. But otherwise they're doing the homework to know who's naughty, who's nice and all that. When you shut that down, there's no information now, nobody's monitoring other banks who has a strong incentive to do so and that can create problems. 

George Selgin: It's interesting that in all the other countries that have flirted with floor systems, they have all, with a prominent exception of the US and the UK, they have all worried about this issue and most have abandoned floor systems and gone back to a corridor type systems of some sort because they did not want to have killed their interbank markets. The Fed is the only central bank that hasn't even discussed this. They act like it's not an issue. 

David Beckworth: So let me ask a follow up question related to this lack of price discovery or this information that's missing now that was formerly brought to us by the interbank market. What is a substitute for, what is the Fed doing in place of that so we don't have that signal? Is there any kind of recourse in the floor system to kind of get around that problem? 

George Selgin: I think the Fed would say that part of the answer is well, banks are sitting on so much liquidity that, now we don't have to worry, but of course that's not right for two reasons. First, the liquidity isn't evenly distributed. Most of the excess reserves that have flooded the system are being held in a handful of New York banks and branches of foreign banks. Other banks are sitting on extra reserves too because they can't borrow in the interbank market, but they're not sitting on huge amounts. Their liquidity ratios haven't gone up sky high. But even putting that aside, liquidity isn't the same as solvency. Banks can still fail. We've heard that Citibank is having problems again, for example, who knows with their track record. But I wouldn't count on the presence of a bunch of excess reserves to solve the problem of bank failures and therefore to solve the risk of contagion effects. And I fear that those effects are going to be multiplied now that we don't have banks at least keeping an eye on each other and they don't have the incentive to do it, that they used to have. 

David Beckworth: So what you're saying is we're substituting some notion of comfort in having large amounts of reserves, but also regulators filling the role of the market here. They are going in and inspecting bank balance sheets and we're trusting them. 

George Selgin: Oh, well, yes. If you believe that that's a good substitute, I have a bridge to sell you, but no, it's really not a good idea to remove incentives for bankers to keep well-informed. We've already, remember, we've removed the incentives for ordinary depositors to worry about which banks are safe and which ones aren't. That we did with deposit insurance which now covers the vast majority of depositors. So if banks aren't monitoring each other's riskiness, nobody is, and in order for them to have an incentive to do that, they have to be dealing with other banks in markets where their dealings are not perfectly secured also and the fed funds market was the prime example of an unsecured interbank market and it is essentially moribund right now and has been for some years. 

David Beckworth: Now, another critique you've brought up in the past I've heard you've talked about, is another political economy consideration. And that is it creates a temptation for Congress to rob the piggy bank. 

George Selgin: Oh yeah. 

David Beckworth: Tell us about that. 

George Selgin: This one actually scares me more than the others do and they scare me plenty. So in the old days, the good old day, tempted to say, because there was a connection between the size of the Fed's balance sheet on one hand, and the stance of monetary policy on the other, the Fed couldn't monetize government deficits or buy securities from government and agencies willy nilly without promoting, ultimately triggering inflation. Therefore, the Fed being an independent agency charged with dual mandate that includes limiting inflation, had a powerful answer to special interests, including the Treasury, to presidents who would say, "Why don't you buy more of our debt." And that answer was, "We can't, we have a responsibility to avoid inflation and acquiring a large amount of your debt will ultimately trigger more inflation than we're supposed to allow, and that was the answer. Very, very good answer. 

George Selgin: Well, they can't say it anymore because, and we've of course had a very good historical illustration of this the last few years, but this continues to be good even at full employment. Now reserve creation doesn't change the stance of monetary policy. Therefore, buying assets doesn't change the stance of policy that depends on the interest rate the Fed pays on reserves, the balance sheet can grow and grow and grow and there's no necessary inflationary consequence. And that means that it's a free parameter, it means the Fed cannot say the size of the balance sheet is going to affect our monetary policy stance. And then the question arises, well then, okay, what determines how big your balance sheet should be? Who should decide? And I believe the answer is all kinds of people are going to want to decide and that's not good. 

George Selgin: And that means all kinds of special interest in government and out are going to want to try to influence the Fed's decisions about the balance sheet. And of course, many of them seeking to sell securities are going to want the influence to be positive. 

David Beckworth: But this is not just a theoretical concern, you're alluding to some actual developments within the past few years where Congress has tapped into the Fed's balance sheet. 

George Selgin: Well, yes and no. Congress tapped into the dividends that the Fed pays Its member banks. 

David Beckworth: Okay. 

George Selgin: And that was a portent, I think, of the kind of thing that Congress- 

David Beckworth: Could do. 

George Selgin: ...Is willing to try to do. And now I see the balance sheet as a bit of a loose cannon up there where it is something that as soon as people understand that the Fed no longer has any inflationary motive for fighting attempts to get it to buy this and that, it's going to have a hard time resisting and knowing that people are going to put more pressure on it. 

David Beckworth: But just to make this concrete, I'm thinking of examples, for example, maybe the state of Illinois says, "Hey, we need a bail out." I mean, that would be a great example, wouldn't it? 

George Selgin: Oh, well, I can give you a better example in a sense so we have Donald Trump complaining that the Fed should not be raising interest rates. Partly perhaps mainly because that affects the terms in which the Treasury is able to finance debt. But he might just as well be moaning, not about prices, but about quantities. He might be saying, why don't you buy a lot more debt at the existing rate of interest and that would be a better tactic because the first complaint the Fed’s answer is no, we're responsible for the stance of monetary policy. So we have to set interest rates in order to meet our targets. 

George Selgin: The second complaint, it's have the well-balanced sheet. There is no answer. There's no clear simple answer to the effect that buying more of the Fed’s, the Treasury securities will violate any of the Fed’s responsibilities. The Fed might not like it. The Fed may not want to do it, though I'm not even sure about that. Bureaucrats like big balance sheets, but it would not have a clear ... It wouldn't contradict its mandate in a clear fashion for it to load up on more Treasury securities and go from four trillion to six trillion or eight trillion. Of course, it would have macroeconomic consequences. At least if we believe the theories behind quantitative easing, but we know that the inflation consequences could be very, very muted. 

David Beckworth: Well, let me bring up one last objection. We'll move on to another part of this discussion, but that is the bad optics it creates. So the new chairwoman of the House Financial Services Committee has said that she is really bothered by the interest on excess reserve payments to banks. And in theory if those were equal to Treasury rates, it wouldn't really be a subsidy, but at least it creates the image or perception of one. Do you see this being just another way of stirring up the hornet's nest? I mean, is it something that Fed should be mindful of moving forward? 

George Selgin: I think it is though the optics are bad. Not just because we now have banks earning interest for just sitting on reserves, but because most of the banks that are earning that interest are six or seven New York banks and a handful of foreign bank branches. And that doesn't look good to a lot of people. But it must be said that this isn't really a subsidy. Those banks are financing the Fed’s big balance sheet that is the basis for the interest it's earning and then paying to them. So they're like other savers in that sense. The Fed is saying if you put a ... We'll pay you interest, you keep, you hold on to our reserves that will allow us to finance a bigger balance sheet create and finance a bigger balance sheet. So it's a sort of Indian giving situation. 

George Selgin: The Fed creates reserves and pays the banks to hold on to them, which is to say, to lend that much back to the Fed. So the banks are not really in a sense, they're not getting something for nothing. Nevertheless, it does look bad and the optics are not very good. I don't want to hang my criticism of the floor system on merely optical phenomenon that is illusions. I want to hone in on what I think are the genuine and very serious flaws of the system. But I think the optics are certainly something the Fed officials would be unwise not to consider. 

David Beckworth: All right, so let's look to the future now. There was a survey that was conducted of the senior financial officers in banking systems and they suggest that they would be willing to depart with enough reserves till they got to an aggregate number near 900 billion. And I'm extrapolating, it's actually 600 million. But if you extrapolate to all the banks you 900. So close to a- 

George Selgin: Or 1.2 actually. 

David Beckworth: Well close, let's say it's a rounded trillion around number. Right now the reserves are around 1.6 1.7. So at the rate that the Fed is shrinking its balance sheet, we could be there pretty soon. If we assume that number is the amount that they're going to hold them at or what we could find overnight market rates jump above interest rates its reserves. It could be a few years if the Fed decides to go even lower than that, it'd be longer. But do you see the Fed just continuing on autopilot with this balance sheet shrinking down, down, down until by default they trigger a corridor system? 

George Selgin: Well that's a little tricky because, no, they won't trigger a corridor system. They won't go to that extent. I don't believe that'll happen. My personal guess for some time now is that the balance sheet, we'll never get below 3 trillion and may not get much below 3.5 trillion, which is not still- 

David Beckworth: How about the overall balance sheet or the reserve? 

George Selgin: The overall balance sheet. 

David Beckworth: Okay. 

George Selgin: Now I'm not sure. I have to think about how many, I think that that translates into about a 1.5 trillion of reserves or 1.2 something like that. 

David Beckworth: Not much more reduction than in the balance sheet. 

George Selgin: No, but it won't be because they've triggered a floor system. It will be because there's certain pressures in the money markets and some banks may be a little shorter reserves and they'll use that as a reason to stop the balance sheet reduction well before anything cataclysmic or gigantic happens. I don't think they want it to be that small to begin with and they're just anxious to have an excuse to stop the unwind prematurely. But I don't think that what the bankers say is the limited amount, how low they'd be happy to see the system go. You can't go by those survey numbers because of problems of a fallacy of composition, right? Each individual bank says, "Well, I would be unhappy to hold less than this, et cetera." But it all depends on what other banks are doing. And if all banks are aiming for the same thing, first of all, there are many cases where they can’t all have it. And then there are situations where what the banks want will depend on the Fed's policy reactions. 

George Selgin: Clearly if we have a corridor system, if you said to banks, how many reserves would you want if you could readily borrow in the overnight market without any difficulty? They would answer differently than they do when they know that that market is constrained. So the survey doesn't really tell us anything. It tells us about some of the bankers' beliefs and those beliefs are all perfectly reasonable by the way. They agree with my priors, but it doesn't give the Fed its marching orders on what its ultimate goal for reducing system reserve should be. The Fed cannot draw any meaningful conclusion about that optimal value job by polling bankers. 

David Beckworth: Well, I would mention this survey does also say near the end, a different scenario. How low would overnight market rates have to go relative, excuse me, how high would they have to go relative to the interest and excess reserves? So if the Fed allowed market rates to go above IOER and banks in that situation begin to change their story a little bit. They're willing to part with more. 

George Selgin: That's right. All the answers in the survey makes sense. They're all reasonable answers. If I were a banker, I think I would have answered similarly but the Fed has to, it has to be careful what policy conclusions it draws on the basis of those answers. And the public has to be careful not to necessarily swallow the Fed’s arguments when it points to the surveys or reason for doing this, that, or the other thing. 

David Beckworth: What about international experience? You've alluded to them already, but we've got a neighbor to the north who tried the floor system for a brief period and they went back to a corridor. What can we learn from these other countries using these systems? 

George Selgin: A general rule regarding Canada is that we should try to do more things the way they do them. We should have done it in the 19th century and now we should try to do it today. Canada floored with a floor system during the crisis, but it was for a very short period of time. They went to a floor and then they went back to a corridor, which as I said earlier, something that in a corridor system, you occasionally might have crises where you do that, that's fine, that's part of our corridor system works. But they never pumped up preserves on such fast to scale as we did in the United States even proportionately speaking. They usually operated with very few reserves. The Canadian Central Bank was actually a leaner, meaner central bank than the Fed in that regard. And now they've gone back to being lean. 

George Selgin: And it was it was a good example of using a floor temporarily in the middle of a crisis but not sticking to it. Whereas in the US, our story of how we got to the floor system is absolutely incredible. If you don't want to read about the floor system for any other reason, you should read about how and why the Fed turned to it in October 2008. But it was all about tightening monetary policy or not wanting to loosen policy. And this was at a time when the US economy was circling the bowl, so to speak. And they started paying banks to hoard reserves because they didn't want interest rates to fall any lower. Of course that was a futile effort, but they were trying to prop them up and this only made the ... This is the point about saying that it deepened the recession. Implementing a floor system deepened the recession was a wrong policy at the wrong time. 

David Beckworth: What about other central banks like the European Central Bank? What are they doing and any lessons from them? 

George Selgin: We can't learn much from the European Central Bank. They have resorted to negative interest rates. So I have several other central banks and there's some something to be said for negative rates if you can do it. But we can learn, I think from New Zealand and the Bank of Japan, what a lot of countries have done actually that's interesting. I'm not saying this is a good policy, but it's a good answer to that liquidity requirement. They have so called tier systems where they said, okay, we'll pay banks to hold reserves up to a certain fixed amount, but any reserves that the excess reserves they have beyond that, we're not going to make it attractive for them to hold it. In that way they can have a certain extra cushion of liquidity and reserves that they introduced banks to hold. But at the margin, if the central bank creates more reserves, the banks don't have any incentive to hoard them. And that means that interest rates can still be controlled the old fashion way in those systems, even though you have this extra liquidity. 

George Selgin: I don't think we need to do that. I think we've already got reserved requirements. Now there's a new Basel liquidity coverage ratio that we enforce. There's no reason to have any further devices for encouraging banks to sit on reserves, we already have a suspender and a belt. We certainly don't need anything more. 

David Beckworth: Can you speak briefly to the liquidity coverage ratio you just mentioned, how does it complicate any transition away from the floor system? 

George Selgin: So this is a new Basel innovation that resembles other Basel regulatory innovations in at least one respect, which is that it's stupid. But what it is a new so-called liquidity coverage ratio that's based on a complicated formula, but it basically says that banks have to hold a certain percentage of so-called high quality liquid assets. And there are two different tiers or types of liquid assets and they have to hold certain amounts of each based on their liquidity needs which are measured by outflows and how much, what kinds of assets they have and the volatility and it's very, very complicated. But for our purposes, there's one essential thing that has to be made clear about this. People have said, "Well, we might as well stick to a floor system because under the new liquidity coverage ratio requirements, banks have to hold a lot of reserves, which are high quality liquid assets. And they couldn't dispense with them even if we didn't pay interest on reserves because of this new requirement." 

George Selgin: So they're essentially arguing that there's no going back to a corridor type system with or without interest on reserves given these new requirements. Well, that's just wrong because Treasury securities, among other things, also are high quality liquid assets. Now what these commentators have said is, "Yes, but look, the banks seem to be lately preferring reserves." They'd rather have reserves. Reserves are better. But that's not true. In fact, it's becoming obviously untrue lately, but banks had been preferring to hold reserves to meet their liquidity coverage ratios. But that's because the interest rate on reserves was being kept above the rate of return on short-term Treasury securities, in a corridor system that wouldn't be so, so they would prefer a Treasury security. 

George Selgin: Well, there's a twist to this. I wrote about this the other day because apparently Federal Reserve officials have told their supervisors to tell the banks that they want them to hold some definite amount of reserves, use a definite amount of reserves to meet liquidity coverage requirements, even though the requirements themselves don't call for that. So it seems that the Fed is using their supervisory, its supervisory powers to create an artificial demand for reserves that would make going back to a corridor hard. But that's purely an arbitrary regulatory, discretionary, action that has nothing to do with any official existing regulation. And this is sort of thing Fed people do all the time. They tell their regulators, "You tell this bank to do this or to not do that." And it doesn't matter if there's a law or not. The bankers are going to do what their supervisors tell them to do unless they're big New York banks. Then they tell the supervisors what to do. 

David Beckworth: Well, in that very cynical note, I want to step back because this whole discussion is very much in the weeds. It's very technical. I want to step back, pull back and ask maybe more general question. So this system you've argued in the book and elsewhere, it leads to kind of a low inflationary environment. It kind of throttles back with the creation of money assets, it has an effect on the level of inflation, nominal spending in the economy. And so my question is, is this turned to a floor system, a symptom more than a cause? Is this more of an epiphenomenon? Is it reflecting the Fed’s desire for low inflation environment? Notice the Fed is so committed to low inflation that it by default had it turned to something like a floor system in 2008 and if the answer's yes, isn't the real issue the monetary regime versus getting caught up in the operating system itself? 

George Selgin: Well, no. There are two different issues. One issue was the stance of monetary policy. That always matters and you can get that wrong in a floor system or a corridor system, you have that money too tight or too loose in either system. So we all agree and we should have monetary policy just right. Is anybody against that? No. Okay. See how easy monetary policy…so that's a question of the stance of monetary policy of where you set as it where the interest rate that you're controlling that's standing for the monetary policy stands. Now the problem in 2008 for example is it was really that the monetary policy stance was too tight, but the introduction of a floor system and interest on reserves was the device by which they implemented this overly tight policy which weren't able to do otherwise and been for interest on reserves. 

George Selgin: The Fed was creating a lot of reserves through emergency lending, et cetera, before quantitative easing and that would have naturally loosened monetary policy. Turns out that would've been a good thing in retrospect. In retrospect, we know what spending was collapsing then now and the Fed use this turned to a floor system to put a stop to it. So, it didn't have to. In principle in a floor system, there's always an interest rate that's the right interest rate. It just so happens that that Fed policy back then was dedicated to keeping interest rates higher than they should have been and used the floor system to do it. Now the problem to me that's distinct, the problems that are distinct to a floor system are separate from the problems of getting monetary policy right. 

George Selgin: And they have to do with killing the interbank market bank monitoring. They have to do with a gigantic instead of lean balance sheet and the inefficient use of savings that results from that, they have to do with the political economy of a balance sheet that's not obviously the business of the size of which is not obviously a matter of monetary policy. And so must be a matter of something. And with optics of the sort you mentioned those, those I think are the real issues with the floor system. It may be that there are some technical respects in which a floor system makes it harder to pursue monetary goals. And I think there's something to that. And I'll give you an example of it. Throw out an example of it but I don't think those are the main issues. 

George Selgin: The sense in which I think a floor system does pose challenges for monetary policy because it relies only on interest rates on prices rather than quantities. In the old system, right? If you wanted to ease monetary policy, just talk about that. You throw more reserves in the system. The banks have reserves that are excess, they don't want to hold on to them. And interest rates are going to fall because the banks are going to then turn around and place those reserves in short-term markets, including the Fed funds market. But notice here that the easing of monetary policy involves not just changes in interest rates but essentially involves spending, lending, and ultimately growth of broader monetary aggregates. In the new regime, the interest rates are changed, but quantities aren't changed. And some people believe that it makes a difference for what ultimately happens for whether inflation goes up, for whether spending goes up, for what happens to other kinds of economic activity. 

George Selgin: And I think that, I think they're on to something. So I do think there's a problem with this mechanism for monetary control quite apart from the other issues I've raised. Nevertheless, I think we should be clear about what are problems unique to a floor system and what are problems that are related to having a bad monetary policy stance. 

David Beckworth: All right, well thank you George. We will now turn to the audience and I'll take questions for George and just a few guidelines in terms of asking questions. Please wait to be called on. Wait for the microphone. Identify yourself and please stick to a question, not a speech or a statement. I'll start up here first. I think I've got a question in the front. 

Norbert Michel: Norbert Michel, Heritage Foundation. George, do you see any sort of catalyst, any sort of event that would be served as a catalyst to changing this system at all? And if so, what would be needed to change that system? Do we need this? In other words, is legislation going to be needed to change to revert back to a corridor system or do you see the Fed just doing this on their own after being pressured? Or am I just way off here? Because I'm thinking of this as a Fed with an administration that is interested in an infrastructure sort of a program for lack of a better term. And I'm just wondering where this might go. 

George Selgin: Well, I think there are two ways in which we might end up going back to a corridor system. One is by persuading important people at the Fed that that's something they should be considering. And I do think there are people at the Fed who are open minded though I think they're not enough of them and there's a lot of persuading yet to be done. So the optimistic view is we just keep talking about this and get the Fed to talk about it. And maybe they'll see things they'll see things the way I see them that would be nice. On the other hand, I do think there's a role for legislation, but it's really a rather minor role because if you read the current statute on interest on reserves, it really doesn't allow a floor system or it barely allows a floor system. 

George Selgin: If you go by the spirit of the thing yet the Fed has used, it has interpreted in a way that that allows it to use it to support a floor system. What I mean is, the original 2006 law that grants the Fed the ability to pay interest on reserves and was put into early effect in 2008 by a separate piece of legislation specifically says that the interest rate on reserves is not to exceed the general level of short-term interest rates. Now, in a floor system, the interest rate on reserves is generally and in our system has almost always been above many short-term rates, including the federal funds rate, which is an overnight rate and but it's been over for most of the history, not very recently, but for most of the history of the floor system, it's even been higher than one year Treasury bill rates. So, and which are not a really short-term. 

George Selgin: So, what the Fed did in 2015 taking advantage of the wonderful notion of Chevron deference, the details of which are not worth going into, especially before dinner but basically they got to write up the rules about how the statute would be implemented and for that purpose they said the short-term interest rates will be defined as, and then they included in the rates as consisting of rates like, and then they included a bunch of longer-term rates, term rates instead of overnight rates. But the kicker was they said, and the Fed’s primary credit rate. Now the primary credit rate is the official rate name for what we call the discount rate. And the Federal Reserve as a matter of policy sets the primary credit rate way above, and you can read this in the Federal Reserve Board’s publications, the primary rate is set above the general level of short-term interest rates. 

George Selgin: Now if you're following all this, what this means is that the Fed can abide according to its own interpretation of the statute. It can set the interest rate on reserves at a level that doesn't exceed the general level of short term interest rates. By setting it below an interest rate, it also administers that's always way above the general level of short-term interest rates. Got it? So if I were to recommend Congress provide a nudge to get the Fed to consider going back to the floor system, I would have it do so by simply saying, "No, you can't interpret the statute that way. Let us tell you exactly what we meant." And that would force the Fed to either abandon the floor system or do some more illegal fancy footwork. 

David Beckworth: All right. I believe this gentleman here was next. 

Doug Carr: Hi, I'm Doug Carr with R Street. Isn't one of the big issues with the balance sheet that goes with the floor system, the changes in it have such a destabilizing impact on the financial markets during the crisis? They actually sold 600 billion from their portfolio in a short period of time coincided exactly with the stock market fall and with the drying up of repo money, then they go into QE and we all know what happened with stocks and QE. Now we've got QT.. Net, the central banks are taking more money out than they're putting in. And obviously it's been tough sledding since that happened. 

George Selgin: Yeah. You're, referring to the taper tantrum, right? And that first example- 

Doug Carr: Recently, the Powell… 

George Selgin: Yes. Yeah. And that one too, it's true that that any big change in the Fed's balance sheet is going to be destabilizing. That of course would also have been true under the old system. The difference is the scale of potential changes of potential asset sales now or purchases is much vaster. And so the potential for the Fed to do mischief by messing around with its balance sheet is very great. So that is a problem with the huge balance sheet and indirectly you could say a problem with the floor system, though, it must be said that if anything, right now, the Fed officials are too timid about selling assets or unwinding their balance sheet. I wrote an article called “Operation SNAIL,” criticizing there balance sheet reduction plan. I think at the time they hadn't started yet, but snail stood for “Stall Now and Inch-along Later.” I think they could have been more aggressive than that but you're right that it can be very bad if they are too aggressive in reducing their assets. 

David Beckworth: All right. We had a question in the front row. 

Gary Towns: George, great book, Gary Towns and just a person. 

David Beckworth: We like people, it's good to have one or two. 

Gary Towns: Now, Bernanke was famous for dismissing any policy, practical problem that came out of the policy is idiosyncratic. And I'm just curious whether you've looked at what happened to interbank lending abroad, particularly in the Libor markets with the crisis these fell apart. They then morphed into banks really, we're required to continue to be able to provide an interbank rate even if the marks weren't good. This then translated into allegations, at least they had mispriced and that then became suits and it became large monetary results. And I'm just curious whether you've looked at that as well. 

George Selgin: Well, there's issues wrapped up in that question, Gary. One is what the Libor scandal itself really was what was involved in it. And the other is what exactly was going on in these interbank markets during the crisis. As for the first, I think you alluded to the reality that the bankers involved in this scandal well, the real scandal isn't that they manipulated Libor, the real scandal is that they were told to do what they did by government officials who then went got to all ethical on them. But it really was the bankers who were told during the crisis, "We don't want markets to panic. We need to be quote rates, blah, blah, blah, blah, blah." And that's really the story behind the story of the Libor thing. 

George Selgin: For us, the more fundamental question is, was what was going on in 2008 and what was happening to activity on these markets. I can talk with more authority about the Fed funds market than I can about Libor, but in doing so will allow me to address a criticism that's been made of my points of my arguments. It has been said that the reason interbank lending crashed in 2008 was not because of interest on reserves, but because of panic that suddenly banks didn't trust each other. And that's why they pulled out. And there's some truth to that, but it's a small part of the true story. Of course there was a panic and for a while, and you could detect this with something called the TED spread. And but that subsided very quickly for a couple of reasons, but one of which the lesser well known is that the government stepped in and said, "We're going to guarantee all these interbank loans." 

George Selgin: So whether you like that policy or not, counterparty risk was no longer after that step which lasted for a year or so. It was no longer driving activity or determining activity in the Fed funds market, which nevertheless never recovered. So if you want to know why that activity never recovered, you can't appeal to the risk. It had to be something else that something else was the Fed's interest payments on reserves. And so it isn't the case that we can't restore an active interbank market just by not paying so much interest on bank reserves as there's no evidence for that. 

Gary Towns: We referred what you do here as the great recession of 2008 and nine. You also have in your book, good discussion of 1893 and 1907. Why wasn't this described and why isn't it typically as it appears to me to have been a panic? 

George Selgin: Oh, I don't know. You're referring to my other book Money: Free and Unfree. And by the way, this one's different and I don't think I mentioned those other ... I think you could call a 2008 a panic. It was certainly a panic on certain markets. Who knows how these names, some people do call- 

David Beckworth: I think some people do call it the great financial crisis and others call the great recession. 

George Selgin: ... Great financial crisis. Panic maybe seems to have a little bit of an old fashioned sound to it somehow. I don't know. But strictly speaking, I don't know why you couldn't call it a panic. It was certainly a panic. The central bankers were panicking among other people. 

David Beckworth: Well, I'll be old fashioned here myself and call the time. I think we're out of time. However, we have a reception afterwards for all the unanswered questions. George is dying to hear them and eager to engage you. So let's give a hand for George. 

George Selgin: Thank you all. And if it got too weedy, the book is perhaps clearer than I am. I hope so. 

David Beckworth: Okay, listeners, here is a quick update. On January 19, just a few days after the recording, the Fed’s Federal Open Market Committee (or FOMC) decided they would stick with their floor system. The FOMC had been discussing the future of its operating system during the past few meetings and the tone of these discussions was decidedly pro-floor system. So it was not a surprise to see them make a decision in favor of the floor system.

David Beckworth: This decision, though not surprising, is debatable. First, as George Selgin mentioned in the show, this decision makes the Fed’s balance sheet an attractive temptation for Congress and the President. The Fed’s balance sheet now amounts to a piggy bank that can be tapped without worries that it will create inflation. After all, the whole point of a floor system is to separate the size of the central bank’s balance sheet from the stance of monetary policy.

David Beckworth: Second, one of the big motivations behind the FOMC’s decision was its belief that the floor system creates better interest rate control. However, the Bank of Canada shows that a symmetric interest rate corridor operating system provides great interest rate control. In fact, in terms of overnight repo rates, the Bank of Canada has actually done a better job keeping rates on target with its corridor system than the Fed with its floor system.

David Beckworth: Third and finally, as Peter Stella explained on last week’s episode, the large stock of excess reserves amounts to a poor form of debt management by the U.S. government. Excess reserves are very similar to treasury bills, except that only banks and few other financial firms can use them. Everyone can use treasury bills. The Fed has effectively taken on part of the Treasury’s role as U.S. public debt manager and done so in a relatively inefficient manner. I suspect this decision will not be the end of this debate.

David Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. If you haven't already, please subscribe via iTunes or your favorite podcast app. And while you're there, please consider rating us and leaving a review. This helps other thoughtful people like you find the podcast. Thanks for listening.

Photo credit: Istituto Bruno Leoni