Standing Repo Facility, Safe Asset Shortage, and the Fed's Low Inflation Problem

David Andolfatto is a vice president of the St. Louis Federal Reserve Bank and is published widely in the field of monetary economics. He also blogs at MacroMania, and has recently published on issues such as the zero lower bound, the symmetry of the Fed’s inflation target, a new standing repo facility, and MMT. David is a returning guest to Macro Musings, and he joins the show today to talk about these issues.  David and David also discuss the safe asset shortage, average and flexible inflation targeting, and the legal, political, and economic restraint surrounding negative interest rates.

David Beckworth:  Our guest today is David Andolfatto. David is a vice president of St. Louis Federal Reserve Bank and has published widely in the field of monetary economics. David also blogs at MacroMania and can be found on Twitter providing interesting commentary and discussion. David is also a returning guest to the show and has recently published on some interesting issues like the zero lower bound, the meaning of asymmetry of the Fed's inflation target, a new standing repo facility, and MMT. 

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

Beckworth:  David joins us today to discuss these issues. David, welcome back to the show. 

David Andolfatto:  Hi Dave. Thanks, it's a pleasure to be back. 

Beckworth:  Glad to have you on, and for all our listeners, I do have to put out a disclaimer that anything David says reflects his views, not the Federal Reserve's views, much to my dismay. I wish that these views would reflect the Fed. But David is here on his own volition, speaking his own mind. Is that fair? 

Andolfatto:  Absolutely fair. Thank you Dave. 

Beckworth:  Okay. So with that out of the way, I want to begin by talking about an issue that you and I both like to talk about a lot. And that's the safe asset shortage issue. And it has implications for the inflation rate in the US. And one thing I think we haven't done on this show in the past is to cleanly lay out the linkage or the mechanisms through which this elevated demand for Treasury securities does affect the inflation rate. And I thought you'd be the perfect person to help us understand it. 

Andolfatto:  Alright. So maybe we should step back and lay out the land firstly. So to begin, we should be clear about what economists generally mean by safe assets, right? I mean a safe asset is not necessarily a risk-free asset. So it's a bit unfortunate that the labeling suggests that it is. But instead it's an asset that can be basically readily accepted as an exchange medium or basically used as a form of money. 

Andolfatto:  So the cash in our pockets is a perfect example of a safe asset. The cash in our pocket is not risk free. I mean, the price of your favorite good or service might suddenly jump tomorrow and that would whittle away at the purchasing power of the cash in your pocket today. So that's the sense in which a safe asset is not necessarily risk free. But we like to hold these monetary instruments, these safe assets. They help us facilitate exchanges, and we can basically buy and sell things very easily with them. 

Andolfatto:  The same thing is true for a much broader class of securities. So a firm needing to borrow cash, for example, can readily do so on the repo market by extending certain types of safe securities as collateral for cash loans. These securities often take the form of safe government debt, like US Treasury securities. But they might also consist of privately issued debt instruments like asset-backed securities. 

Andolfatto:  But whatever the case may be, to the extent that these securities are easily convertible into money, either through their use as collateral or because they can be disposed of easily on secondary markets, the fact that they are easily convertible into money kind of renders them attractive as exchange media. 

Andolfatto:  So to put it another way these safe assets are valued not just for their income stream, but also for their liquidity properties. And this enhanced demand has the effect of raising their price. You know, think how much more valuable your iPhone would be if you knew you could use it as cash whenever you wanted. So the effect of this liquidity premium for securities is to basically lower their yield. 

Andolfatto:  So, to kind of get back at how this ties in to low inflation, let me point out to something I think most people would agree is an empirical fact, is that for a long time now, probably starting in the 1970s, and certainly since the 1980s, there's basically been a secular boom in the demand for safe assets. US Treasury debt is used extensively as collateral in credit derivative markets. It's used extensively as collateral in shadow banking, both of which have grown very rapidly, especially since the 1980s. 

Andolfatto:  US Treasury debt and US dollars are used extensively as safe stores of value for emerging economies, especially after the emerging market crises, financial crises of the 1990s. The financial crisis in 2007 and '08 further enhanced the demand for US Treasurys as safe security. And at the same time that the crisis destroyed a large supply of previously safe assets in the form of private-label asset-backed securities and even agency debt. 

Andolfatto:  The European sovereign debt crisis kind of had the same effect. And since then we've had what you might think of as an added regulatory demand for safe assets in the form of Basel III and Dodd-Frank regulations. So there's been this long-running secular boom in the demand for safe securities. And arguably the supply of these objects hasn't been keeping pace. 

Andolfatto:  So one obvious manifestation of these secular forces is a long-run decline in the yield on these securities. And this is what we've been witnessing in the data. But now imagine that there are some forces in place that kind of keep these yields from falling, say because the central bank is targeting an interest rate, or perhaps because some zero-lower-bound considerations are coming into play. 

Andolfatto:  As the demand for these safe assets continues to expand more rapidly than the supply, you know the pressure can't manifest itself as further declining yields. It has to manifest itself in other ways. And what are these other ways? I mean in reality there could be many ways. But one way in particular is that the price level that will fall. I mean, it should put, theoretically this should put downward pressure on the price level. 

Andolfatto:  And a way to think about this I think is if we just think of these safe assets as just basically monetary instruments, in the limit ... When we teach undergrads we say, "Consider money demand shock. A positive money demand shock." And you know, and money is a fixed interest rate security. The demand for money is growing and the supply is fixed. You'd expect one way that the market pressure would manifest itself is as a decline in the price of goods and services, as people would scramble to sell their goods and services more cheaply to acquire the safe assets that are presently in short supply. 

Andolfatto:  So at the end of the day I think you can really just think about it as basically a money demand shock. And to the extent that money demand has been growing very rapidly, where money here is broadly defined to include a large class of Treasury securities and other safe assets, that one manifestation of this has been for a secular disinflationary force. 

Andolfatto:  So this may be one of the forces that are contributing to the kind of low inflation experience in the United States. 

Beckworth:  Yeah, I would argue it's probably one of the more important ones. And that it's a sustained, ongoing, kind of persistent force. And one of the charts that I think is very convincing and it underscores this point is to look at the yield on a 10-year Treasury and map that against, or plot that against the velocity of the MZM money supply. 

Beckworth:  So the MZM stands for money of zero maturity. It's a bunch of liquid assets from cash all the way up to I believe institutional money market funds, retail money market funds. Anything that can be quickly turned into purchasing power without any kind of penalty. And those two things look almost identical. And so you see this decline in this important benchmark safe asset yield, and a decline in the velocity of that particular form of money. 

Beckworth:  To me it's a very convincing story to be told. And I guess more importantly, back to the original question it's a key force that the Fed has had to wrestle with, right? The Fed is trying to hit its target but it can't, or it's struggling to at least, and this is one of the reasons why. 

Andolfatto:  I think that's right. I mean, the Fed is limited in its ability to create, to supply these safe assets. And to a first approximation, what the Fed is able to do is basically alter the maturity structure of the outstanding supply of these US Treasury securities. This is basically what we're doing with open market Operations or QE. And the responsibility ... If this theory is correct kind of the responsibility for managing the supply of this broadly defined money resides with the fiscal authority. 

Beckworth:  Okay. So the relative dearth or lack of supply of Treasurys, which to some listeners may be shocking. 

Andolfatto:  Yes. 

Beckworth:  They're thinking like, "Are you kidding me?" I think the headline number is 22 trillion. It's really around closer to 16 trillion. But how can this possibly be? But again, this goes back to what you said earlier. There's just this global supply for some safe store of value and the US Treasurys happen to be the best form of it right now. 

Beckworth:  But with that said, right, so there's apparently this relative shortfall of Treasurys, compared to the demand for them, and it's putting downward pressure on inflation. So I had a post up, I think you saw it earlier this week, where I acknowledged that. But then I go on to argue that the Fed could in theory offset that. And I gave this counterfactual, what if the FOMC was made up of clones of Neel Kashkari? I put a picture there, Neel Kashkari looking at himself. 12 Neel Kashkaris at [the] FOMC. 

Beckworth:  So my question to you, am I being too harsh on the Fed? I mean, if you had Neel Kashkaris running the Fed, could they have made a difference? Or am I expecting too much? 

Andolfatto:  That's a very good question. By the way, I love that graphic. I mean, 12 Neel Kashkaris. I mean, it was a ... 

Beckworth:  Yeah. He actually responded to it. He had a fun tweet. I think he said, "While I don't endorse a committee of clones I do like the idea," and he goes on to explain it. But yeah. So tell me, am I being too harsh on the poor Fed? 

Andolfatto:  So, I think that you may be. But I mean a lot of this kind of depends on what sort of theory one has in mind about what the ultimate source of the shortage of safe assets is. You know, what sort of tools does the Fed have at its disposal. I mean, it has interest rate policy, right? I mean, so to what extent can its interest rate policy influence the safe asset shortage? 

Andolfatto:  I mean, arguably you know in some models actually raising the policy rate actually has the effect of reducing these shortage, because by increasing the rate of return on these securities you come closer to basically the Friedman rule, if you know what I mean. And if you want to think about the Friedman rule as a policy prescription that essentially eliminates the asset shortage, right, it makes liquidity plentiful. 

Andolfatto:  The Friedman rule, sorry for the listeners who don't know what the Friedman rule is, but you're going to have to go and google it. But to the extent that one believes in that mechanism, arguably one thing the Fed should be doing is raising the interest rate. Now, I think there's a whole number of reasons why perhaps we might not want to take that too seriously. 

Andolfatto:  But at the end of the day, you know even in terms of balance sheet policy, what can the Fed do? I mean, it literally has to take as given kind of the deficit and choose how much of it essentially to monetize. And to the extent that the private sector views interest-bearing reserves as very close substitutes to interest-bearing Treasurys. You know, the Fed messing around with the composition of that debt between reserves and Treasurys is probably inconsequential to a first order. So at the end of the day I do think it's up to the fiscal authority to influence or manage the supply of safe assets. 

Beckworth:  So if President Trump is listening and he wants to add some stimulus to the economy it's in his hands, is what you're saying. 

Andolfatto:  I am afraid so. 

Beckworth:  Okay. 

Andolfatto:  I shouldn’t have said I'm afraid so. But I mean, I think that's right. 

Beckworth:  Yeah. Yeah. Well, let's move on then and talk about something else you've written about recently, and that is the nature of the Fed's inflation targets. So the Fed says it has a symmetric inflation target. Many folks including myself have questioned what that actually means. And you have your own take on that. So how do you view symmetry for the Fed's target? 

Andolfatto:  Yeah, so you know, for your listeners, if you don't know the Fed actually only formally implemented an inflation target in 2012 of two percent. I mean, one could argue that we had one before implicitly. But the formal establishment was in 2012. And ironically, kind of just after it was formally announced inflation has basically been undershooting that target ever since. And this has led people like yourself to question whether the inflation target is symmetric. Indeed, I mean Jay Powell in a press conference actually questioned it as well. Which was what motivated my post. 

Andolfatto:  So people are suggesting that the Fed seems to be treating two percent more like a ceiling rather than a symmetric target, because the Fed seems unwilling to let inflation overshoot the target. And so I reflected on that thought and I said, "Well, but you know inflation overshooting to me is not necessarily inflation targeting. It's what you'd expect from a price-level targeting regime." 

Andolfatto:  So targeting the price level and targeting the rate of change of the price level are two subtly different things. Inflation is targeting the rate of change of the price level. So inflation targeting in my mind means getting people to expect that inflation will eventually return to target from below if inflation is presently undershooting, and from above if inflation is presently overshooting. 

Andolfatto:  And a symmetric inflation target in this context means that the rate at which inflation is expected to return to target is the same, whether inflation is presently above or below target. 

Beckworth:  Okay. 

Andolfatto:  So I think it's basically too early to judge whether we have a symmetric inflation targeting regime, because since 2012 we've only been undershooting. The true test will be imagine sometime in the future when inflation is overshooting. Will the Fed at that point be as patient in letting inflation crawl back down to two percent? If the answer to that question is yes then you can say it looks like the inflation target is symmetric. 

Andolfatto:  If on the other hand the Fed responds very aggressively to bring inflation down then you could argue, "No, it looks like the inflation target is asymmetric." That seems like they want to bring it down a lot faster than they're willing to let it go up. But we don't have enough data to tell I would argue at this point. 

Beckworth:  Okay. Well, that's a fair take on what symmetric could mean. I wonder what would happen if the Fed moves towards average inflation targeting. So there's a lot of talk about average inflation targeting and if I had to look into my crystal ball I'm guessing that's what the Fed is ultimately aiming for in this year of review of its strategies and policies. 

Beckworth:  When I think of average inflation targeting, though, I think of a different kind of symmetry. How would you explain average inflation targeting, and would it be different than your definition of [a] symmetric inflation target? 

Andolfatto:  Yeah. I think that, well notice that in my symmetric, in my description I mean the Fed would be on average hitting [the] inflation target over long enough periods of time. So the question is really what sort of time period are we talking about. And I think that's the issue. How quickly do you want that inflation to return to target. That's a conceptually different question than being symmetric, right? Because symmetry means if you want it to move to target more quickly, I mean presumably that's true on the downside and the upside if you have a symmetric target. 

Beckworth:  Yeah. So with the average inflation target, now correct me if I'm wrong, in a certain case it could be very similar to a price-level target. 

Andolfatto:  I think that's right. Exactly. Yes. 

Beckworth:  Okay. 

Andolfatto:  So why not advocate then for a price-level target? 

Beckworth:  Right. That's what I'm wondering. Is this maybe, again you're speaking on your own volition here. You don't represent the St. Louis Fed or the Federal Reserve organization. But I'm wondering if this is an easier way to get something like a price-level target without saying so? 

Andolfatto:  Yeah. Possibly. And indeed, the notion of flexible inflation targeting is kind of getting at that idea too. And I can see the merit in a price-level target, and indeed even a nominal GDP targeting regime. But I think that as a practical matter I kind of question its practical implementation. I mean, it could work. But it's one thing to say that we can make inflation overshoot in our models, right? And our models are kind of well behaved. Our models, we have a very clear understanding of the economic forces that are at work. And so we're quite confident in the ability of our policymakers and our model to fine tune policy. 

Andolfatto:  But you know, reality is considerably messier than our models, and hence our ability to fine tune policy in this manner I think is much more uncertain. And so to my mind the main thing is to have long-run inflation expectations anchored. And inflation targeting regimes have been relatively successful at achieving that goal. And I think many of the things people wish monetary policy can do beyond that, I think in my view might be better relegated to the fiscal authority. 

Andolfatto:  But I guess as an academic I would be very interested to see how a price-level targeting regime might work out. It would be an interesting experiment for sure. 

Beckworth:  Yeah. Absolutely. Now that's where I thought you might be going. It's up to the fiscal authorities to provide the oomph to maybe get us to this average inflation targeting, price-level target. So I was actually talking to someone just yesterday about this, how it would be almost strange but fitting in the Trump administration if ... You know President Trump does not have a lot of love for the Federal Reserve System right now. 

Beckworth:  And it would be interesting if he proposed or did something along these lines where he made a rule, his own rule, we'll call it the Trump Rule, where maybe it's two percent inflation, maybe it's a price level target, maybe it's a nominal GDP level target. But you pick your target and he creates this system where Treasury automatically, if the target's being missed, Treasury automatically creates new bonds, deposits them at the Fed. The Fed gives the Treasury the money and the Treasury sends out these checks until its target is hit. 

Beckworth:  So kind of an automatic helicopter drop rule. We'll call it the Trump Rule. That would be very credible. It would be a runaround. I think two things, it would be a runaround to the Federal Reserve's doing what it's supposed to do. But I think it would also highly incentivize the Fed to do something knowing that Trump is in the background lurking, breathing heavily over them. 

Beckworth:  But this goes back to your point about fiscal policy carrying a big stick, carrying a lot of weight. 

Andolfatto:  Yeah. I think that's right. I think these are the type of alternative rules that people are thinking about. You'd have to worry too, it's easy to kind of in the present environment, undershooting inflation, it's easy to kind of promote policies that kind of expand the supply of money or safe assets. 

Beckworth:  Right. 

Andolfatto:  But the challenge of course is always when- 

Beckworth:  You want to have guardrails on both sides- 

Andolfatto:  ... inflation is heading above target you're asking for some sort of form of fiscal austerity. 

Beckworth:  Yeah, I want to be very clear. Yeah, to all our listeners I want to be very clear. I do not want to return to 1970s inflation. I believe there needs to be some important guardrails. That's why even with the Trump rule presumably hopefully there would be kind of constraint there as well that would keep it on target. A symmetric Trump rule, how about that. 

Andolfatto:  That's right. 

Beckworth:  So President Trump, if you're listening, make it symmetric buddy. Okay. Well let's move on from there to a really fascinating piece that you've written for the St. Louis Federal Reserve's blog. And you actually have a new piece out today. You have two articles on this. I think you've co-written them. And this has to do with the Federal Reserve creating a standing repo facility. So tell us what would the standing repo facility be and how would it contribute to monetary policy? 

Andolfatto:  Right. So I've written a couple of blogs. I've coauthored with Jane Ihrig, who's visiting the St. Louis Fed from the board for the year, so we've had a very fruitful year together discussing a number of issues related to monetary policy. And Jane and I got to talking about this issue of why are banks holding so many reserves. And we noted that in the FOMC's normalization principles and plans in 2014 we know that the Fed wants to operate a floor regime, which basically means we want to keep the supply of reserves abundant or ample in the banking system so that banks don't have to worry about operating on a daily basis in the federal funds market. 

Andolfatto:  So in the statement of principles and plans though is this kind of interesting point, is that while the Fed is wanting to operate a floor system, it wants to do so with a minimum amount of reserves necessary to operate that floor system. And so what exactly that minimum amount of reserves is nobody knows for sure. I mean, people at the Fed, at the board I should say and other places, I think maybe the New York Fed, who interview banks kind of estimate this minimally abundant level of reserves to be north of one trillion dollars, which is pretty high by historical standards. 

Andolfatto:  Prior to the crisis reserves were only 10 or 20 billion dollars. So while reserves are considerably less today than they were at their peak in 2014 they're still at 1.5, 1.6 trillion dollars. So why are banks holding so many reserves? You know, one reason is well, of course they have to, because the feds injected it in there. But the question is what determines the demand for these reserves? And in part the demand is determined by the fact that the Fed pays interest on reserves. So we all know that. 

Andolfatto:  The question is for a given interest on reserves what further considerations determine the demand for reserves? And what would be the minimum amount of reserves that would be necessary to operate a floor system efficiently and effectively? That's the question that Jane and I are asking. It's not sufficient in our view to just go ask banks what they feel like the amount of reserves they need, because the answers that the banks are providing in these survey questions are conditional on the existing operating regime. And the existing operating regime consists of a regime that does not consist of a standing repo facility, you know apart from the discount window which we don't want to consider. 

Andolfatto:  So why are these banks preferring to hold interest-bearing reserves rather than interest-bearing Treasurys? I mean, one might argue that liquidity regulations are at play here, but the liquidity coverage ratio regulations are probably not relevant because the LCR, as they call it, liquidity coverage ratio, runs off a formula that treats Treasurys and reserves equally. They are given equal weights. 

Andolfatto:  So what we discovered is that what we think is happening is that regulators are pressuring banks to hold reserves over Treasurys for resolution purposes. So regulators are not treating Treasurys as cash equivalent for resolution purposes. And we think that ... We're speculating a little bit here, but I mean, we're thinking that in part this could be because, you know if you take a look at the discount window and the parameters that govern the discount window, you know there's a 50 basis point penalty rate. And all forms of collateral are given haircuts. 

Andolfatto:  And in particular, perhaps surprisingly or not, US Treasurys are given significant haircuts at the discount window. So I think like there's a five percent haircut on a 10-year Treasury if you bring it to the discount window. 

Beckworth:  That's interesting. 

Andolfatto:  I mean, this is huge. And so of course for resolution purposes we think regulators are suggesting that the Treasurys that these banks are holding are not very good for resolution purposes. You have to be holding more liquid reserves. And what Jane and I are arguing is, listen, if we could just set up a standing repo facility, kind of apart from the discount window, it could serve this dual purpose. First it could serve as a ceiling tool, which is kind of a separate issue, but you know, there's some reasons to kind of want a ceiling tool in there in any case, to kind of enhance interest rate control. 

Andolfatto:  But apart from that we think that the existence of this standing repo facility, that would accept just US Treasurys as collateral, and offer a relatively low lending rate, that just the mere existence of this facility should lead regulators to kind of relax on what they consider to be suitable securities to hold for resolution purposes. 

Andolfatto:  I mean, if the Feds stood ready to monetize US Treasurys on demand, at no discount, why shouldn't Treasurys be acceptable for resolution purposes? They should be. And so our idea is that apart from interest rate control, a ceiling tool, the mere existence of this facility could go a long way to reducing the minimum amount of reserves that are necessary to operate a floor system efficiently and effectively. 

Andolfatto:  And just how far the Fed could permit reserves to drain is unknown, but the good thing about having this facility in place is the Fed wouldn't have to estimate that number. With the facility in place you could just let reserves drain to however low they have to go. And then if reserves no longer became abundant you'd see that by banks coming to the facility to basically monetize their ... to ask for loans using their Treasury security. 

Andolfatto:  So at that point the Fed would know that it's at that minimally optimal point and it could inject an additional supply of reserves at that point to kind of send interest rates back to the floor. So that's the basic idea. 

Beckworth:  Yeah. So just to recap, the way it would work is the banks could come to the Fed, take their Treasurys, and quickly and relatively cheaply convert them into reserves when they needed them, is that right? 

Andolfatto:  Correct. Exactly. 

Beckworth:  Okay. Now, what's interesting about this, and there's several things actually interesting about this. One is it's kind of like the flip side, or you call it in your paper a complement, but it's kind of like the completing of the overnight reverse repo facility they now have, right? So that now does the opposite, right? 

Andolfatto:  Correct. 

Beckworth:  So tell us about how those two things would complement each other. In what way does it make it a whole? 

Andolfatto:  Well, I mean what we're basically talking about here is running a form of a corridor system. You know, much like other central banks do. So the overnight reverse repo facility could be thought of as providing a floor. And the complement to that is the lending. So that's a deposit facility where banks and a wider range of counterparties, in fact today, can kind of deposit funds at the Fed. And the repo facility would just be the counterpart, the lending facility. 

Andolfatto:  And these would be administered rates that would presumably be adjusted as macroeconomic conditions justified. So the standard kind of Taylor Rule  sort of properties could be applied to them. And the purpose of putting the ceiling there, and the floor, is basically to gain interest rate control over money market rates. Basically money market rates are going to be bounded between that corridor. 

Andolfatto:  And by the same token, it basically releases the Fed from worrying about the size of its balance sheet. I mean, the balance sheet is going to vary to whatever size that it needs to vary to accommodate interest rates in that corridor. And indeed, that corridor is consistent with operating a floor system because the floor system is just flooding the system with sufficient reserves to make sure that the money rates fall to the floor. So I think that's how the two rates kind of tie together and help the FOMC implement monetary policy. 

Beckworth:  Now, you've thrown a lot at us there. I'm trying to unpack that a little bit. But so right now the overnight reverse repo facility, which currently exists, I know it's not being used as much as it once was. But right now it exists and banks are able to take reserves and swap them for Treasury securities. Is that right, kind of on a temporary basis if they need them. Is that correct? 

Andolfatto:  That's right. So the Fed kind of releases Treasurys. 

Beckworth:  Yeah. So if a bank needs some collateral for whatever reason they can do that. And all you're saying is, "Hey, let's go full circle here. Let's allow the opposite transaction to take place." 

Andolfatto:  Correct. It's just a symmetric… 

Beckworth:  Yeah. So it's not that much of a stretch. And you mentioned in your paper, the Fed has done this before too. The Fed has done this type of transaction before, what would be done in the standing repo facility. 

Andolfatto:  Well. Back up a second. I mean, the thing is the Fed does it all the time. I mean, you could imagine instead of a standing repo facility, you could imagine a discretionary action by the New York Fed to operate, to inject reserves for Treasurys in open-market operation essentially, on a discretionary basis. 

Andolfatto:  So these actions and facilities are kind of in place. The distinction here is that it's a standing facility. It's a standing facility. So it's not like, you know, especially in a resolution you don't have to worry about, I don't know, Treasurys being accessibly discounted if a large amount of Treasurys were brought to the market. So I think the fact that it's a standing facility is very important for us. 

Beckworth:  Yeah. So it's predictable, there's procedures in place, you know what to expect, there's no stigma. It's just kind of a standard tool the Fed would put in place. And the more you think about, at least the more I thought about I, I was like, "Why doesn't this already exist. You have one half of it there, guys. Let's go one step further." 

Beckworth:  And again, I mean the points you make, and of course I'm very sympathetic, I'll tell why in a minute, but the points you make, I mean it would dramatically ease the life of the New York Fed. Its market desk wouldn't have to try to estimate reserves or deal with trying to run a floor system. It would just make life a lot easier. 

Beckworth:  You also mentioned, we talked about earlier, no stigma, better optics. But here is the kind of the big thing I really want to get to, is this would effectively, if I understand correctly, turn the Fed's floor system into a symmetric corridor system like we have in Canada. Is that a fair take? 

Andolfatto:  Right. I mean, I think to be fair every system is a corridor system in a sense. I mean, there's already the discount window for example, that could be interpreted as a ceiling rate as well. So in that sense the Fed is operating a type of corridor system. So this is just another type of ceiling tool that like you said I think this would be preferable because presumably the stigma would be less of an issue. 

Andolfatto:  And even though it is a corridor, I mean it's still not inconsistent with operating at the floor of that corridor, if that's what the FOMC desires. 

Beckworth:  Yeah. Well, I guess we've had this debate on Twitter, you, George Selgin, and me, and others, about floor versus corridor systems. But I look to Canada to me as kind of like the ideal example, the best real-world example. I'm sure there's other ones out there. The one I know the best that I think works the best is the one in Canada. And they had a symmetric corridor system before 2008. They temporarily went to a floor system. And then they came back. 

Beckworth:  And if you look at their target rate and their overnight repo rate, they track each other really closely. So they've done a great job with interest rate control. They have a smaller balance sheet. I mean, you mentioned there's the whole political optics reasons to like this as well. But I like the way you frame it, and this is why I think, as I said initially, this is a great way to bring together advocates of the current system, as well as advocates of a symmetric corridor system. There's a kind of a bridge I think that would link the two and lead to a nice transition forward. 

Beckworth:  So my next question is how has this proposal been received? Have you got any feedback, any interest from Fed officials or other interested parties? 

Andolfatto:  Right. So there was a lot of commentary in the press, a lot of industry experts kind of weighed in with their views. I think that it's fair to say that there was a lot of ... People were largely sympathetic with the idea, although a lot of people expressed kind of concerns about whether or not the ceiling tool would put a hard cap on secured overnight funding rates, things like that. 

Andolfatto:  Most of the commentary was centered around the use of the facility as a ceiling tool. What Jane and I have done in our follow-up piece, which was just published this morning on the St. Louis Fed blog, was to kind of highlight the other dimension of the tool, kind of as a ... It could be conceptually used quite apart as a ceiling tool. It's just a mechanism literally to drive down the demand for reserves for resolution purposes. 

Andolfatto:  And the reason for that would be simply to be in accord with the FOMC stated desire to operate a floor system with the minimum amount of reserves possible. 

Beckworth:  Yeah. And let me ask this question. If the Fed did go to this facility, it did have a standing repo facility, and we did ultimately converge to something like what they have in Canada, would it not make sense to abandon the whole charade of a federal funds rate target and just go straight to a repo target interest rate? 

Andolfatto:  Yes. I think so. I mean, logically that's I think where to go. But there's subtleties there and delicacies, mainly political in nature. You know, how close are we coming to actually directly targeting the yield on US Treasurys for example. At least these are the issues that come up when I discuss the issues with my colleagues. I say exactly that. Why don't we just go and target the repo rate. The arguments invariably are not economic in nature. The arguments are largely political. 

Beckworth:  Well, one last question along those lines, we got to move on, because this is getting in the weeds, but I don't want to bore our listeners too much. But if you have a corridor system, I mean you can target a repo rate. It gives you some flexibility around that. It doesn't have to be like one for one all the time, right? There is some market pressures in there as well. You just got to keep it within the corridor. 

Andolfatto:  Yup. Sure. 

Beckworth:  Okay. Okay. Well, I guess we're in agreement. We'll take that as given. So, David, one other feature about this that goes back to our earlier discussion about safe assets, and that is I think this would lead to much more effective, much better public debt management by the US government. And I'm thinking back to a previous guest I had on this show, Peter Stella. And what Peter Stella made very clear, really resonated with me, is that by the Fed having a large balance sheet ... 

Beckworth:  There's a number of issues someone could point to, but I think one of the key issues is that the Fed is effectively taking over some of the public debt management role that has normally been done by the Treasury. And it's doing it in a very inefficient way for this reason. If the Fed has a large balance sheet the way it did that is it issued a bunch of its own liabilities, excess reserves, or reserves. And it sucked up all these Treasurys. And of course also mortgage-backed securities. But let's ignore those for the time being. 

Beckworth:  And so what it's done, it's swapped reserves for Treasury securities. And the thing about reserves is that only banks and a few other financial firms can use them. Whereas Treasurys can be used by everyone. And so Peter Stella's argument for going back to a smaller balance sheet for the Federal Reserve is it's just better public debt management, it's better for the financial system, it's better for the safe asset shortage problem. 

Beckworth:  And I think your proposal, again, is just another step in that direction. It would help address this concern that Peter Stella has raised. Do you think that's right? 

Andolfatto:  I think that's right because the mirror image of the decline in the demand for reserves is a release of the Treasurys, I guess, is one way to think about it if I understand the argument correctly. 

Andolfatto:  But I think that that argument would have had more force prior say to the present administration's fiscal policy. I mean, the supply of these safe assets, it seems to be growing very rapidly right now. But I think that the general point is correct. I think also my former colleague Steve Williamson has harped on this, that a swap of Treasurys for interest-bearing reserves is kind of reducing liquidity in the market sector, because Treasurys can be used as collateral extensively. By non-banks they can be rehypothecated for example. 

Andolfatto:  So it's kind of ironic that by an open-market purchase of Treasurys can actually reduce the liquidity in the financial system. It's really quite an interesting idea. 

Beckworth:  Well, you're right. You know, over time the Fed's share of total outstanding Treasury securities is declining. As deficits grow, as the economy grows, this becomes less of an issue. However, as you said, the Fed's balance sheet has not gone down to the pre-crisis level. Or even after adjusting for like currency growth. But I would argue the real danger may not be like right now, or even the drag, the Peter Stella argument. Maybe there's not as much of a drag now, but given that the balance sheet is decently sized, it's more the future, right? 

Beckworth:  I'm almost certain the Fed's going to resort to QE in the future because interest rates are so low that it doesn't have much room to cut. And so I just see an ever-expanding balance sheet in the next recession and the recession after that, which is why it's good to get things in order now, before we get to that point and have to deal with these issues. 

Andolfatto:  Yes. You'd make a fine central banker, David. 

Beckworth:  Well. Right now I'm just a podcaster. Okay. Let's move on to something else you've written about. Very interesting. This is the year of review for the Federal Reserve on tools, tactics, communications, exciting year for people like you and me who ... Well, you work at the Fed but I follow the Fed fairly closely. And one of the tools that the Fed could consider, and maybe arguably is legal depending on how you interpret its current laws, is negative interest rates. 

Beckworth:  Now, you had an interesting post where you considered why hasn't the Fed used them. Is it a legal constraint, or is it an economic constraint? So what is your take on that? 

Andolfatto:  Yeah. Just, so I wrote this post. Yeah, that post was motivated by, you know I had recently attended a macro conference, and like many conferences that I sit in on, there's invariably these papers that are being presented, kind of in this genre, this New Keynesian model where there's a zero lower bound imposed. And the zero lower bound of course wreaks havoc on the ability of the central bank of stabilize the economy. And these models kind of live off of the existence of this zero lower bound. 

Andolfatto:  And I remember sitting in the audience and kind of thinking to myself, you know, these are typically younger academics. I was wondering, they've grown up in a world where all they talk about is the zero bound. I wonder if they treat it as kind of a force of nature, something like gravity. Is it kind of as natural to them as you and I treat gravity? 

Beckworth:  Right. 

Andolfatto:  Because every model they've ever seen has a zero lower bound. Of course there's a zero lower bound. Or do they understand it as more of an economic constraint or perhaps even a legal constraint. 

Andolfatto:  So it got me to thinking, you know the idea of a zero lower bound as an economic constraint is kind of semi-defensible. I mean, this is how I grew up thinking about it. I mean, it relies on the notion of the existence of zero interest-bearing cash. And that if a central bank tried to lower the interest rate on its electronic-deposit liabilities to negative territory, the demand for electronic forms of money would basically disappear as people and firms would just substitute into the zero-interest-bearing cash, which is a higher yield than zero. And so that threat of that arbitrage activity is what presumably makes the zero lower bound an economic constraint. 

Andolfatto:  But then I began to wonder about how practical it would be for banks and businesses to carry all of their money around in the form of 100-dollar notes. And that's the maximum denomination of the US dollar. And certainly this would be fine for people with small money accounts. I'm not talking about that. But I think there's presently about 1.5, 1.6 trillion dollars in electronic reserves in the banking system. I mean, if you want to try to visualize what that look likes in 100-dollar notes, I have a visual on my blog. It's about the size of a football field and maybe a couple of stories high. 

Andolfatto:  I mean, it's a lot of physical cash. And I invited the reader to kind of think about how this would work. I mean, how would these banks and businesses work with all this cash? I mean, this cash would have to be physically stored. It would have to be guarded. It would have to be transported. But not only that, it would have to be hidden from the authorities. Because if you don't hide it, it can be taxed. Not necessarily by the Fed, but there's a fiscal policy that can tax cash and effectively make the interest rate on this cash negative. 

Andolfatto:  And then I remember listening to an interview of Pablo Escobar's brother, and this is after watching Narcos on Netflix. 

Beckworth:  Remind our listeners who he is, just in case they don't know. 

Andolfatto:  Yeah. So Pablo Escobar was the leader of the Medellín drug cartel in the 1980s I guess. And of course he was killed in a raid. But his brother, in an interview afterwards, claimed that at the height of the cartel's operations they were spending close to something like $1,000 a week on elastic bands that they would use to tie up the bills, because of all of the cash that they were accumulating. 

Andolfatto:  And what they would do with these bundles of bills is they would bury them in large holes in the countryside. And he claimed that their accountants would regularly apply about a 10% appreciation rate on this cash, because much of it would be lost to rats and spoilage and stuff like that. 

Andolfatto:  So to me the empirical evidence suggests that the effective lower bound is probably much closer to negative 10 percent. And at a negative 10 percent the central bank probably doesn't have to worry about any lower bound on interest rates to conduct its stabilization policy. 

Andolfatto:  So then I got to thinking if this is the case then the zero lower bound is not an economic constraint. It's a legal constraint, if it's a constraint at all. And then I thought that if this is so, then why are all these young academics spending so much time, so much intellectual effort trying to come up with unconventional monetary policies designed to circumvent the zero lower bound. Why are we devoting so much effort to this endeavor when we should be rediverting our efforts to changes in legislation that permit us to eliminate the zero lower bound, which is clearly a legal constraint and not an economic constraint. 

Andolfatto:  So that's what got me to writing the blog. And I said, you know I concluded with a thought that either we take these models seriously, and the zero lower bound is really what's kind of really making it very difficult to conduct monetary policy, in which case we should just eliminate it. Or if we don't believe eliminating that zero lower bound would kind of lead to some sort of nirvana in terms of policy making, then I think we should dispense with this class of models. It's kind of like an either or. 

Beckworth:  Yeah. So we have a paper at Mercatus Center that is being worked on right now by Peter Conti-Brown and Miles Kimball that looks at the legal issue and whether it's possible. And addresses this issue as well. And I just, I want to maybe push this argument back another level, and that is, so it's not an economic constraint as you said. And maybe it's a legal constraint. And maybe legally it is possible. Let's say it's interpreted that it is legal. Let's say it goes to the Supreme Court or whoever, and they rule it legal. 

Beckworth:  So we've gotten past the economic constraint, the legal constraint. I would argue probably the real final constraint is a political constraint. And I'm thinking of people like Marvin Goodfriend, whose nomination was partly shut down because he was a negative interest rate proponent. I believe, if I recall the story correctly, Rand Paul decided he would not vote for Marvin because he was, you know for better or for worse, despite Marvin's massive track record of publications and scholarly work, he is known for this little area that he worked on, you know negative interest rates. 

Beckworth:  So there's a huge marketing ... It's called political hurdle to get over. You got to convince people this is an effective way to do monetary policy. And I think that's maybe the final barrier to getting it widely accepted. 

Andolfatto:  Well, that's right. I mean, like I said, there's one or two ways you can approach this. One, you actually believe that the zero lower bound is the problem. In which case that's right. I mean, you'd have a big PR kind of operation to kind of convince people, look, there's nothing magical about zero interest rates. I mean, interest rates are often negative in, like I said, I mean in the Columbian drug cartel. And also in many examples in history and in nature. 

Andolfatto:  But the other tact you could take is I don't believe that this is the fundamental problem of what inhibits stabilization policies. So, yeah, I do recall that story. I'm old enough to remember actually when Marvin first came out with that proposal. In fact, I recall talking to him and he told me that he was receiving death threats. 

Beckworth:  Oh really? 

Andolfatto:  Yes. So it's true. It's something, you know, if you can feel that strongly about it I guess. 

Beckworth:  Well, you know like you though, I think there's probably a more effective way to deal with depressed economies than negative interest rates. Although I think in theory you can make a compelling argument why you'd want to do this. And I want to do that just for the sake of this show here. And this is what, I've gone to Capitol Hill. I've talked to Republican staffers myself. And I've said, "Look, this is not my preferred approach, but here's the reason why you should at least give it a hearing, and that is if you believe in capitalism, if you believe in markets, and if you believe markets work through prices clearing the markets that's all this is. This is an argument for allowing a price, an interest rate, to clear a market." 

Beckworth:  It just so happens we have these institutional constraints that have prevented them so far, legal constraints so far. So if you believe in capitalism then you should at some point, in special cases, believe in negative interest rates. 

Andolfatto:  I think that's a very good way to frame it. You know, I could see people coming up against that of course saying, "Well, you know in a free capitalist society negative interest rates would never be an outcome," you know because ... And that the only reason it's going negative is because of some force manipulation by this large quasi-government agency… 

Beckworth:  Yeah. I know it's a uphill battle. I've argued with friends about this where like, "How can you do this David? How can you advocate this?" Like I'm not ... And I would always reply, "Look, what I advocate, again, is markets clearing." And really, the ultimate culprit, the ultimate perpetrator behind negative interest rates I would say is you my friend, because you're the one causing ... Your demand for safe assets, your demand for savings, that's what's driving rates down. 

Beckworth:  I know it's easy to point the finger at the Fed. We've talked about this too elsewhere. The Fed is really following where the market, where the fundamentals are going. It's not the Fed pushing rates down. 

Beckworth:  All right. In the time we have left I want to spend a little bit of time on an area you've been actively engaged in, and that's the modern monetary theory, or MMT discussion. And there's a lot we could talk about here. But I want to zero in on something specific, and that is yield-curve targeting. So MMT has a lot of policy proposals. Again, don't have time for it in the few minutes we have left. 

Beckworth:  But one of the interesting, maybe controversial proposals is that the Fed do something like the Bank of Japan. So the Bank of Japan is I believe targeting a 10-year yield, treasury yield or their government bond yield. And some of the MMT folks have advocated doing the entire yield curve. That means from short-term all the way to long-term interest rates on government debt. The Fed would explicitly target it. And you know, that struck me as bizarre. I think of Wicksell’s cumulative process. I think of things just blowing up. 

Beckworth:  But you directed me to a paper I hadn't seen before where these researchers attempt to model this and come out with a fairly benign and even favorable outcome that's stabilizing and feasible I guess. So make the case, or maybe explain it and then make the case, if you're comfortable, as to how this would work. 

Andolfatto:  Well, yeah. I don't know how much I want to ascribe this view to MMT necessarily. I mean, I think many people have monetary fears have kind of proposed kind of similar sorts of ideas. Even Neil Wallace, for example, I think. But I think, you know I don't have much time here to explain everything. But the basic idea, I mean one way to think about it is why should Treasury debt, at some trade at a discount. I mean, it's risk-free, right. 

Andolfatto:  This is a question actually that Neil Wallace asked 40, 50 years ago, in a paper that was called “The Inefficiency of Interest-bearing Government Debt.” In a sense the discount is applied because the Fed is purposely rendering those securities illiquid. And the question is, is why is that the case. You know, from an efficiency standpoint, in a first kind of best world it would make sense to eliminate the discount on Treasury debt and permit the Treasury to finance its operations kind of more effectively so that its debt isn't discounted. 

Andolfatto:  And so the Fed could actually undertake a policy like that, if it wanted to. And indeed, you pointed to the Bank of Japan as doing exactly that today. So that's kind of like briefly an economic justification for why one might want to do that. It would effectively render Treasury debt as money. And I think that just by saying that, I think you can see where a lot of the objections might come in. I mean, the objections are going to be largely kind of political in nature. 

Andolfatto:  I mean, are we going to trust the fiscal authority to manage this debt in a responsible manner, meaning essentially inflation, keeping inflation low and stable for example. So I think that's where the pushback comes. It ultimately comes, the reason for discounting or rendering illiquid some forms of government debt is that you want to have a monetary authority that's kind of held responsible for basically keeping inflation under control. But in our models a well-behaved fiscal authority can do the job just as well. 

Beckworth:  So just to make this concrete, the way this would work is the Fed would say, "One-year Treasury at a certain rate, five-year Treasury at a certain rate, and a ten-year Treasury at a certain rate. We're going to do whatever is needed to make sure the rates are there." And so they literally, they target these. And everything in between, they target them. The government supplies these. 

Beckworth:  And so it's kind of like the Fed's on autopilot, and then that kind of pushes the decision-making back to Treasury then, in terms of what actually becomes the inflation rate. 

Andolfatto:  Yeah. I think that's right. Yeah. Well, I mean but the Fed I guess would still retain control over the interest rate. 

Beckworth:  Yeah. 

Andolfatto:  So essentially what you're doing is flattening the yield curve. But you're still permitting the Fed to judge the level. And that level, you know now debt of all maturities become basically perfect substitutes. So they're indistinguishable. So, you know, what we're really talking about is the Fed is choosing an interest rate on the outstanding stock of government debt. And the Fed can presumably still be free, if this is so desired, to alter that, whatever that rate is, in its objectives to meet its dual mandate. 

Beckworth:  So I guess the concern would be would the parties involved stay responsible. And the first thing that comes to my mind would be like the 1951 Treasury Accord where the president, that would be President Truman, and Treasury wanted the Fed ... In fact the Fed was doing something like this back then, right? 

Andolfatto:  Yes. Yeah, exactly. 

Beckworth:  It was pegging Treasury bill rates. But, inflationary pressures were building, and the Fed said no. So I guess, maybe there's a political concern, but maybe a practical one that maybe it's, given who we are as humans, temptations, all those wonderful things, it's wise to have the monetary authorities separated in being able to say no. 

Andolfatto:  Maybe. I mean, it really depends on what one's views are about kind of what governs decision making, in legislative bodies. And to me there's really, it just depends on the actors involved. I mean, the Treasury accord episode is sadly an episode I haven't studied deeply. But I think that was around the post-World War II era, leading up to the Korean War. 

Andolfatto:  Again, there's a distinction between raising the policy rate and discounting Treasury notes of different maturities. And there's a distinction to be made there. And I think what you're talking about is, you know that the Treasury, much like today, you know the pressure coming from the administration, we want to keep the government's borrowing cost low. 

Andolfatto:  I mean, inflation is low today. Back then it was rising. But there was also a war to be fought. And there is a question of how to finance it, and whether or not a little bit of inflation might be part of the way that you finance these things. I mean, there's all sorts of ways countries finance war. I mean, from outright conscription of labor to printing more money. 

Andolfatto:  That's kind of, I think my hunch is that the Treasury accord era is a bit unfair. I wonder if something like that might not work more effectively in kind of outside of those turbulent times of the post war and Korean War era. 

Beckworth:  Okay. Well, with that our time is up. Our guest today has been David Andolfatto. David, thank you again for coming on the show. 

Andolfatto:  It's been my pleasure. A lot of fun as usual, Dave. Thank you. 

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: St. Louis Federal Reserve Bank