Bill Nelson on the Growth of the Federal Reserve

The Fed’s increased role in financial markets has created more problems than it has solved, and here’s what to do about it.

Bill Nelson is a chief economist and an executive vice president at the Bank Policy Institute. Bill previously was a deputy director of the Division of Monetary Affairs at the Federal Reserve Board where his responsibilities included monetary policy analysis, discount window policy analysis, and financial institution supervision. Bill also worked closely with the BIS working groups and the design of liquidity regulations. Bill rejoins David on Macro Musings to discuss his article titled, “I Don't Know Why She Swallowed a Fly,” which looks back at the significant growth of the Federal Reserve, both in its reach and in its size, since the Great Recession of 2007-09. Additionally, Bill and David discuss steps the Fed could take to return to a reasonably sized institution, conducting policy with a light imprint on financial markets.

Read the full episode transcript:

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

David Beckworth: Bill, welcome back to the show.

Bill Nelson: Thanks, David. It's nice to be back.

Beckworth: It's great to have you on, I know you need to provide a disclaimer before we get going. So why don't you do that now?

Nelson: Sure. Let me just take a minute to explain to people what the Bank Policy Institute is. So if the Bank Policy Institute is a trade organization owned by the nation's largest banks, including foreign banks that have significant operations in the United States. I'm their chief economist. And well, I seek to be impartial in my views. My reputation is my stock and trade, but I am working for this banking organization. If I'm ever caught being impartial, I'd encourage all your listeners to please let me know. With that said, these are my views and not necessarily the views of our member banks.

Beckworth: Okay. Well, with that out of the way, Bill I want to talk about your note and to help motivate the note. I want to go back to your experience at the Board of Governors when you were there and you oversaw bank supervision, you helped design liquidity provisions, as I mentioned earlier. And this issue is near and dear to your heart because of this past work that you did. In fact, you, and Ulrich Bindseil who's also a past guests of the show and he's from the ECB.

Beckworth: So you were representing the Fed, Ulrich Bindseil representing the ECB. You guys co-chaired a working group at the BIS that did a report on the interaction of monetary policy and regulations. And this is what your article really touches on that space. It came out in 2015. So maybe it'd be useful for us to hear you highlight some of your findings in that paper. And then we'll try to map those onto our discussion today from your new paper.

Nelson: Sure. Happy to, yes. So as you noted Ulrich and I co-lead this team of about 20 central bankers organized by two groups at the BIS, the Bank for International Settlements. One is the committee for the global financial system, which is a committee focused on the intersection between monetary policy and financial stability. And the other was the markets committee, which is a committee focused on monetary policy operations. The report did its work in 2014 and the study came out in 2015. Anyone interested can find it at CGFS paper 54. And as you know, the findings are very reminiscent or closely tied to what I write about in the note we're going to discuss today about the interaction between the bank regulations and monetary policy and how they can feed into one another.

Nelson: So we were particularly tasked with looking at the Basel III reforms and considering how they would impact monetary policy operations and transmission. I'll just highlight a few of the findings since they're so similar to not only what I wrote about, but what actually happened. So one of the findings was that there would be weakened forces of arbitrage across money markets, particularly because of leverage ratio requirements, which reduce the ability of financial institutions to take offsetting positions to arbitrage.

Nelson: And as a result, central banks would find themselves perhaps having to interact with a broader set of counterparties so that their monetary policy moves would be well transmitted. We noted that the regulations actually treated interactions with the central bank more favorably than interactions with other financial institutions and that would tend to make central banks larger other things equal.

Nelson: So in there, you can see it in the favorable treatment of deposits at the central bank reserve balances as a form of liquidity reserve as being one of the drivers of the explosive growth in the Fed to meet the demand for reserve balances. The report noted that just tighter regulations in general, which although, noting that those were certainly appropriate response to the regulatory problems highlighted by the global financial crisis would reduce aggregate demand.

Nelson: And in situations when interest rates were running low, that could mean that there would be more alternative forms of stimulus needed and in particular QE. And that more QE was going to result in more binding leverage ratios and potentially therefore, less bank credit. Precisely, the kind of cyclical cycle that we're going to discuss today.

Beckworth: Yeah. Well, your paper was prophetic, right? As you mentioned, a lot of this did come true and we'll talk about it some more. And we'll provide a link to it in the show notes. So you can check out the 2015 paper that Bill co-authored or co-chaired with Ulrich Bindseill. Bill, before we get going into your new piece, just general observation I've had over these past few months, really going back to the debate over the SLR exemption remember it came to an end and there was this question, what will the Fed do? Will it continue it? Will it end it? And what I saw just as I watched, I was engaged in conversations on the show about it, read all the commentary. I saw two camps emerge. And one camp comes from a perspective that they really are narrowly focused on financial stability almost to a fault.

Beckworth: And any tinkering of the ratio is a compromise, a step back. Even if you tweak the SLR, so you keep total capital at the same level, but you exempt reserves what the question was about, and they're willing to die on that hill, no matter what the consequences for monetary policy may be in terms of operational effectiveness. On the other hand, there's another camp that says, "Look, we got to make monetary policy effective. If we want to stabilize the economy, you want financial stability? You want the Fed to be able to offset business cycles, to respond, to swings in economic activity? It has to be able to do so. And if you have problems in the plumbing and financial regulation, you need to fix that."

Beckworth: And so you may do more harm to financial stability by keeping this SLR requirement for reserves in place than by exempting it. If your ultimate concern is about financial stability, there's more than one way to look at this. And the second group I put myself in the second group, and I think you too as well, but there's been a number of reports that have come out in support of this.

Beckworth: The Group of Thirty had a report out on the treasury market that said, "Look, the SLR exemption for reserves should be put in place." Brookings had a task force on financial stability, a number of papers have come out. And to me, it all makes sense, but it's just interesting to see these two camps emerge and have a very different view maybe of what ultimately drives financial stability. And how do you get there? Is it you do it just through financial regs or do you do it through monetary policy and financial regs? It's the latter group. And I don't know, is that a fair characterization of these two camps, do you think? Or my being maybe overly harsh?

Nelson: No, I think that's pretty fair. So here's how I look at it. If you go back to the open board meeting, when the supplementary leverage ratio was passed by the Fed and the video's available, I encourage everyone to watch it. Almost all of the governors, if not all of the governors sitting around the table indicated that they were supporting the higher than in the international norm setting for the supplementary leverage ratio, because they anticipated two things would come to pass. That would mean that it would not be binding. So generally, you want... And I think there's almost universal agreement among banking economists on this point, you want to leverage ratio which is not risk sensitive to be a backstop to your risk sensitive capital requirements, because otherwise you create some perverse incentives for institutions to make themselves a riskier.

Nelson: You also punish very low risk activities like market-making so ideally you've designed risk based requirements to be sensitive to risk. You want them to be the ones that are influencing banks decisions. So all of the governors sitting around the table said, "Well, we're supporting this because we realized that there are going to be changes to the risk-based requirements that will cause them to go up over the next year or so, in particular the G-SIB surcharge."

Nelson: And that came to pass. But in addition, the level of reserve balances that is currently elevated, because that was soon after the end of QE3 this is in 2014, I think, is expected to decline. So in particular, if you look at the most recent forecast of reserve balances, that the FOMC had been provided just a couple of weeks before you can see if you know where to look and read the footnotes of the blue book or Tealbook part B that the staff was projecting reserve balances to go from their level then of about one and a half trillion down to 25 or 35 billion.

Nelson: Which is the level that they'd been at before crisis. So they calibrated the LCR to a level that was appropriate for 25 billion in reserve balances. And so not surprisingly it's inappropriate when the banking system is forced to hold four trillion dollars of reserve balances. And it seems to me that one can be completely confident that the regulations were calibrated correctly. And yet nevertheless, see that under current circumstances which weren't like those that were anticipated it's important to redesign them to make sure that the leverage ratio isn't inappropriately binding.

Nelson: And doing so by changing the leverage ratio would be an action that would bring it into alignment with the calibration that was originally intended. So I don't see the same tension that others see between fixing the leverage ratio so that is acting as a backstop like it's supposed to. And maintaining stringent capital requirements in line with those that were designed by me and others coming out of the global financial crisis.

Beckworth: That's a great point. You've mentioned before, but just to repeat what you've said, the supplemental leverage ratio was not meant to be a binding constraint. It wasn't, but it is right now. And again, the danger I see is that if it becomes such a binding constraint, it impairs the effectiveness of monetary policy. And why do we care about that? Because if monetary policy is impaired, that can lead to more financial instability, which is the whole reason you start with this supplemental leverage ratio is to improve that. So you got to take a holistic view, I guess, and that's what you're presenting here. So let's jump into your article. And again, the title is, “I Don't Know Why She Swallowed A Fly.” So maybe first off, explain that title to us.

Growth of the Federal Reserve

Nelson: Yeah, I'd be glad to when I... By the way, I can't take credit for the title of friend of mine suggested it when I discussed the general topic. But nevertheless, I immediately fell in love with it. I was surprised that a lot of people don't get the reference. So the article is about how the Fed has perceived problems that then fix this by getting more involved, but then causes more problems. So it fixes those by getting more involved. And it's a reference to the nursery rhyme about the old lady who swallowed a fly, she swallowed a fly and then she had to swallow a spider to catch the fly, and she had to swallow... I forget, a cat to catch the spider and a dog to catch the cat and eventually a cow and then a horse. And when she swallows the horse she dies of course. And so-

Beckworth: Very fitting. Yeah. Very fitting, so wonderful title name, and a very interesting article. Again, we'll provide the link to it, but I thought what we would do Bill is I will read the introduction and then I'll go through the problems that you list and I'll read the problem to you that's developed, and then you can reply or answer with the solution the Fed has taken to the problem. So we have a number of problems we'll go through, I'll do the problem, you do the solution. And then at the end, you will present your preferred solution to all of these problems.

Beckworth: Some maybe simpler, cleaner fixes instead of going through the woman who swallowed the fly approach to solving problems. So we want to avoid that outcome that she had for sure. So let me start with your introductory paragraph here. And you write, “over the past 13 years, the Federal Reserve has consistently solved problems, whether they were partly entirely of its own creation by becoming larger and more involved in the financial system. That greater size and involvement has led in turn to still more problems, which the Fed has sought to fix by expanding its reach into the markets. This process has transformed the Fed from an efficiently scaled institution conducting policy with a small imprint on financial markets to a behemoth that is the largest borrower in both the unsecured and secured short-term funding markets. The Fed predicts that by 2023 its balance sheet will equal 39% of gross domestic product up from 6% in mid-2007. How did we get here?”

Beckworth: Question, consider the problems the Fed faced and the solution that adopted. So just to touch on that, the Fed is becoming a money market. The point you're making here, the Fed is the main counterparty in the money markets and that didn't used to be the case. So how did we get here, this huge balance sheet. So I'm going to read the first problem that emerged that pushed us in this direction. And then again, Bill, you respond with the Fed’s solution.

Variability of the Treasury Account

Beckworth: So problem number one, the treasury department's account at the Fed swelled from five billion before the global financial crisis to 1.8 trillion in 2020 with a corresponding rise in volatility, the treasury traditionally kept almost all its cash at commercial banks, but once the Fed started paying an above market rate on reserves, the treasury moved all its cash to the Fed, reducing the need for the Fed to borrow from banks and saving taxpayers' money. However, variability in the treasury account made it difficult for the Fed to conduct monetary policy with small balance sheet. All right, that's the problem. What was the Fed solution?

Nelson: So this one was conducted monetary policy using a giant balance sheet so vast that swings in the treasury account no longer matter. I do want to say before I forget that when readers go to the link and find this article, there'll be able to find that it's full of hot links to publications almost entirely by the Fed or ones by me that summarize publications by the Fed providing this information. So providing not just support, but literally the information that I'm citing here. So now I'm not-

Beckworth: What's wrong with that first solution? What's wrong with-

Nelson: So a little backstory here. So as you know I've always been a fan of the Fed operating with its old approach of a small balance sheet, keeping reserves scarce, having a light imprint on operating with very small repo operations each day that are unimportant to the counterparties. They're that small. Not operating in the Fed funds market, the market that it's seeking to influence. So a tiny imprint that was well transmitted into the economy, allowing for great control of interest rates.

Nelson: And when I left the Fed about six years ago, one of the first speaking opportunities I had was at a symposium at the New York Fed where I made the argument that the Fed should go back to doing it this way. I was surprised to find that the counterargument that was made was no, I mean, this was not by the New York Fed, but rather by another panelist. They can't do that because the TGA is the Treasury General Account is now so big and variable that the open market operations required to control that volatility would be too big.

Nelson: It would be unwieldy and for the reason... So maybe a little background in monetary policy here for a second. So monetary policy works by the Fed controlling the supply reserve balances that are out there. Reserve balances are a liability of the Fed, their deposits of commercial banks at the Fed and the Fed can control those. And they do so in a way to try to get interest rates where they want them to be. It's usually very easy. It sounds hard, but it wasn't that hard. And it was quite effective, but one thing that moves reserve balances around that the Fed can't really control is the size of the treasury's account at the Fed another Fed liability. So when the Fed learns of a move in the treasury general account, the Fed has to offset that with an open market operation.

Nelson: And the Fed used to handle that by making the treasury keep almost all of its money in the private sector at commercial banks. With only a small amount held at the Fed of about five billion dollars, it varied a bit and the Fed would offset it, but it was very small. What happened, happened in stages like so many of these things where a decision was made and then another decision was made, and then it ended up in a place where they hadn't really expected to be.

Nelson: So the first step towards the treasury general account getting much bigger occurred when the Fed opened its second and but quite large post-Lehman emergency credit facility. The commercial paper funding facility the Fed had been offsetting the increases in its balance sheet up to that point in the crisis caused by its interventions by running down its treasury securities, but it was out of shorter term treasury securities to run down at that point.

The first step towards the treasury general account getting much bigger occurred when the Fed opened its second and but quite large post-Lehman emergency credit facility.

Nelson: So the CPFF its expansion ran the risk of increasing reserves in a manner that the Fed was not going to be able to control. And although there had been hope that there would be treasury use of TARP funds as a way to fund that the CPFF at the start that didn't work out. And so what the treasury did was they just deposited $50 billion at the Fed as a way to help the Fed fund the growth in the CPFF. Once the CPFF expanded nevertheless, the treasury however, expanded its holding its deposit at the Fed. And then as the crisis continued and ended but the Feds kept interest rates at zero. It remained profitable for the treasury to continue to keep a large amount at the Fed for reasons that are a little difficult to explain.

Nelson: Bank deposit rates for the treasury, as well as for everyone else at that point were about zero. The Fed by contrast was paying banks 25 basis points on their deposits. So if the treasury moved its deposit from the banks to the Fed, that would mean that the Fed didn't have to accept those deposits from the banks paying 25 basis points. Instead, it had the deposit of the treasury, which got zero, even though the treasury was getting zero, the Fed remits all of its profits to the treasury.

Nelson: This sounds really complicated, but basically it was good for taxpayers for the Fed to keep its money at the Fed because the Fed was paying more to banks than was the market rate on the banks deposit to that time. But then that then became the norm in fact treasury just closed down its capacity to keep money at the private sector and just started keeping all of its money at the Fed. Then and I think about 2016, the treasury changed its account management practices.

Nelson: So it decided that it needed to keep a lot more cash on hand to cover contingencies. They don't say this, but I think it's largely contingencies having to do with the debt limit that they were worried about because instead of 50 billion, it rose to about 200 billion. Then that number exploded last year as the treasury thought they were going to be big expenditures and was getting big inflows and it rose to 1.8 trillion way out of the norm of what it expected. But it has since fallen to 500 billion which has... And it expects to keep that to be about the normal level.

Nelson: But of course, if we have another debt limit debacle it seems increasingly likely they're going to have to run that number down below 200 billion again. So you can see you now have this huge liability of the Fed that moves around by hundreds of billions of dollars, that has results in changes to the reserve balances that the Fed is provided the banking system. So for the Fed to make that something that is a matter of indifference to banks, it has to hold a vast quantity of reserve balances. So you're moving around the quantity and the reserve balances themselves aren't scarce.

Beckworth: So it's much harder to forecast and know where the Fed's balance sheets going to be from period to period where in the past, it was much simpler.

Nelson: That's right.

Beckworth: So it's like if my own balance sheet had this category that is swung around wildly, so I couldn't plan my expenditures, my activity very easily. So that's a problem and it has added further fuel to the fire, having a big balance sheet, because if you have these huge swings, you might also have extra reserves around. Okay, so that's problem number one, that's the treasury department. It's in part tied to the Fed's move to its floor system, but also the Treasury's made decisions as you outlined in terms of how it manages its account because of the debt ceiling issue.

Beckworth: That's the first problem that's made the Fed's balance sheet bigger, harder to manage. Second problem you mentioned, foreign official institutions have an increasingly hard time keeping cash at US banks and broker dealers in part because of the capital consequences associated with investing those deposits. So if they can't do it there, what are they doing Bill?

Difficulty Keeping Cash at US Banks

Nelson: The Fed solution was to remove the limits on the foreign repo pool where foreign institutions can deposit savings at the Fed and increase the interest rate that it pays on that pool. Now, it didn't increase the interest rate exactly, but it changed the formula that resulted in a higher rate. With the result of the pool has risen from less than 500 billion before the global financial crisis to about 250 billion in recent years.

Beckworth: So this is similar to the previous point that we just made. So the previous point is this other category in the Fed's balance sheet. And I'll call it a TGA for short Treasury General Account. And this is another one, so these are two items in the Fed's balance sheet that are exogenous or outside the Fed's control, and they can swing greatly, which again makes it hard to forecast and manage reserves. And it makes it hard for you and me to get to this dream world of a small lean balance sheet, which both of us want to get to. I'm going to come back to that later when we talk about your proposals and the recent talk of the Fed tapering. All right, next problem.

Nelson: David, could I jump in and make one quick observation?

Beckworth: Sure. Yes.

Nelson: Even though both of these two... So first of all, I want to say I'm roughly going in chronological order here, but these things are all a bit overlapping. But these two consequential matters as far as I can tell, there was nothing in the FOMC minutes or in some cases, the transcript to indicate that these were actually decisions that were made by the FOMC. Even though they're exhibit A in why the FOMC is now handcuffed from going back to the kind of monetary policy operations that it did before. So that's a pretty serious issue. And I think bears further investigations to the decision-making that went into these.

Beckworth: So do you think this is something Congress should take up then, like have a committee study why is treasury? Why is the foreign repo pool? Why are these changes happening and should they be returned back to an earlier form?

Nelson: Well, I think the Fed's decision to operate with a large balance sheet is absolutely, it's a very consequential decision that has impacts on financial markets on taxpayers and should absolutely be something that Congress should be asking hard questions about, or the GAO should consider. Sometimes people confuse the independence of the Fed with the independence of how they do their operations.

Nelson: The independence that one is worried about is that you don't want your central bank to be forced by the government to lower rates and goose the economy and cause inflation. But it's quite different to think about how the central bank goes about setting its interest rates. Moreover, even the Fed, the Fed will completely say, "We are responsible to Congress." So there's certainly nothing untoward about Congress asking tough questions about what the Fed is doing and this very consequential decision that they made back in 2018.

Beckworth: This goes back to the point I was making about the supplemental leverage ratio. You want to have the Fed have all the flexibility it can to implement optimal monetary policy. And if it has to worry about the TGA, it has to worry about the SLR. It has to worry about foreign repo pool. It's just an added noise to distract from the signal of what monetary policy is trying to do. So yeah, all the congressional staffers out there listening to the show, take this to your boss, GAO, get on it. Let's have a conversation about the TGA and why it swings as much as it does and whether that's the right response.

Beckworth: All right, let's go onto the next problem you highlight. QE3 produced a Fed securities portfolio so large that selling securities as originally planned could disrupt financial markets and generate huge losses for the Fed because the value of those securities declined as interest rates rose. So what was the Fed's solution Bill to that dilemma?

Balance Sheet Consequences of QE3

Nelson: Change its exit plan so that they no longer include assets stale instead just stay larger for longer. So yeah, so just some background here. If people will remember QE3 was the Feds last flow based asset purchase program. Whereas, QE1 and QE2 were big announced amounts, Q3 was assets that were acquired on a continuous basis each month. The original idea was that the program would be about six months long, but it went on for, I forget now, 18 months, something along those lines and vastly exceeded its original expected size.

Nelson: Now, the Fed had originally published exit principles that it expected to follow to normalize its balance sheet, to go back to the old of doing things after QE1 and QE2 those interested can go back and find my briefing of the FOMC on those original exit principles. But it became clear that QE3 was going to result in a balance sheet. Now, those original exit principles involved selling off the assets before raising rates in order to normalize the balance sheet. So that reserve balances were again scarce and interest rates could be raised and policy could be conducted in the manner that it was conducted before.

Nelson: But once QE3 got underway, it became clear that selling off the assets on the timeline, at least that was expected, would involve sales of such a large amount, that it would be disruptive. And there was also concern expressed in the FOMC transcripts and other publicly available sources about the potential for losses being made. If the Fed makes losses, it doesn't realize those losses and it doesn't record... I mean, it indicates them as unrealized losses on their books, but there was concern that I would actually have to realize losses, which was viewed as unappealing.

Once QE3 got underway, it became clear that selling off the assets on the timeline, at least that was expected, would involve sales of such a large amount, that it would be disruptive. And there was also concern expressed in the FOMC transcripts and other publicly available sources about the potential for losses being made.

Nelson: So in a very Groucho Marx move, Groucho Marx's famous line about I have principles and if you don't like them, I've got others. So if Feds change this exit principles to be ones where it would raise interest rates and then it would normalize its balance sheet later. And ultimately it decided that it was just going to hold its balance sheet pretty much where it was, it didn't sell anything in the end. It just gradually allowed it to slowly decline. And not that far in the end, as we all witnessed.

Beckworth: Yeah, now Bill was interesting and we'll come back to this. The taper talk seems to be following the outlines of that original plan. They're talking first about shrinking the balance sheet before raising interest rates. So that gets back to me the spirit of it. But let me ask this question, those original plans for returning to normal to shrinking the balance sheet, aren't they still on the books?

Beckworth: Was that ever officially changed? I know they didn't live up to them, but did they come out with a new updated version? Was it 2014? They had the original plans and then... I know that they changed course with quantitative tightening after things seem to get a little dicey in the economy, but can they still look to that and say, "Hey, this is our blueprint. We need to follow it."

Nelson: Yeah. And just to be clear, the current talk is consistent with the Feds revised exit principles.

Beckworth: Revised, okay.

Nelson: So they would taper purchases but they wouldn't-

Beckworth: Shrink the balance sheet, yeah.

Nelson: So they would expand to more slowly and chair Powell did indicate in the press conference that he wasn't taking off the table the possibility that they might raise rates before they finished tapering but they didn't anticipate doing so, but they weren't ruling it out if they needed to. Which I found a relief just because I really think that the model of the previous tightening as the timeline for what will need to be done going forward is incorrect.

Nelson: Just because an expansion coming off a financial crisis tends to be slow because you have to repair the financial system, but a recovery coming off of a global pandemic could be quite rapid. When the pandemic is done the growth returns and that is much more likely to require a tightening at a pace that's similar to past tightenings, which were much more rapid than the tightening after the global financial crisis.

Beckworth: So Bill, the other part of my question was do the original plans for exiting, for return to normalcy do they still stand, are they still on the books or have they officially come up with new ones? I know they've acted differently, but can they look back to the original ones as a plan for guidance?

Nelson: So they did officially switch to a plan after September 19th after the repo volatility in September 19th. So they officially switched to a plan in which they anticipated conducting policy with abundant reserve balances, not with scarce reserve balances. The FOMC adopted that, I think in January 2019. And then in September 2019 after the repo market volatility they indicated that they expected to keep reserve balances, roughly where they were when troubles began September 2019.

Nelson: Plus a buffer to control for swings, leaving them in a position where they didn't anticipate having to move things around a lot. Which I guess is something like 1.6 to $1.7 trillion in reserves. Now, we're now at four trillion dollars in reserves. So it seems likely to think that even with that very elevated level before they've got some shrinking to do. But that's as you and I have discussed before, I think they could get a lot smaller and we'll discuss that at the end.

Beckworth: Yeah, no, I think that's a good point. And yeah, that seems like a good way to answer my question is that they explicitly came out and said, "We're going to stick with the abundance reserve system or a floor system." So that would nullify the pre-existing plans just by definition. We're going to keep a large balance sheet, therefore we're not going to be shrinking it to what it was before, but you know what? Just to touch on this a little bit and we'll come back to this. But there are thinking is that they did too much quantitative tightening.

Beckworth: They got too close to the demand for reserves that the banking system wanted and we'll come back to this, but you've highlighted many times, I think, properly that there's a difference between short run demand for reserves and long run demand. So it's true. They accidentally backed in too far, Darrell Duffie has this paper saying reserves in fact, weren't abundant during his period because they had hit the demand for them. But in any event, let's move on to the next problem that you note.

Beckworth: So your next problem, Bill that you highlight says the giant balance sheet caused by QE3, created so much liquidity that the Fed wasn't sure in 2014 if it could increase the Fed funds rate just by raising the interest rate, if it pays to banks on their deposits. So what was the solution Bill?

Nelson: So the solution was to create a temporary in quotes overnight reverse repurchase facility at which the Fed was able to pay interest to money market mutual funds and GSEs by borrowing money from them. Expanding the Fed set of counter parties to include entities about which policymakers otherwise have deep concern. So just to provide a little more color there because the Fed had created such a large quantity of reserve balances after QE3 it was concerned that... So the way the Fed thinks of itself conducting policy in its floor system and its abundant reserve system is you oversupply reserves so that everyone is stufFed full of reserves. And then you pay an interest rate on those reserves roughly equal to the interest rate that you would like to prevail in money market rates generally.

Nelson: And since all the banks and thrifts and credit unions have a lot of reserves that they're earning the interest rate that the Fed is paying on them, the interest rate on reserve balances or IORB rate. Those institutions shouldn't lend out money at anything less than that rate. And they wouldn't borrow at anything more than that rate because they don't need to borrow. So basically you would expect interest rates to prevail at about that level.

Nelson: What the Fed found happened however, for a few reasons is that when it produced a huge quantity of reserves, that market rates actually fell well below the interest rate on reserve balances. And that was partly because there were entities out there that held reserve balances GSEs, or at least held deposits in the Fed that didn't earn interest on them. So those entities were lending their money to banks, and the banks were borrowing that money at somewhere in between zero and the interest rate that the Fed was paying.

Nelson: And then there were a lot of other entities that were lending into money markets in particular money funds that didn't have access to any of this. So what the Fed did it's supposedly temporarily was to open overnight reverse repurchase facility where money funds and GSEs could come in and basically deposit money at the Fed. Where the Fed was essentially paying interest on this broader set of counterparties, that it didn't have authority to pay interest on the deposits of them.

Nelson: But so instead it had a standing open market operation. Now, a reverse repurchase agreement is basically like a deposit. So first of all, when the Fed describes operations, it does so from the counterparties point of view, which is always confusing. So when the Fed says, this is a reverse repurchase facility, it means that the money funds can conduct a reverse repurchase agreement with the Fed. The money fund comes in. It gives the Fed money. Fed the money funds treasuries as collateral.

Nelson: Not that they need to, but that just structures it as a repo. And then the next day, the money fund gets the money back and gives the treasuries back. So effectively the money fund has made an overnight deposit and earned interest on that deposit. So now you have a broader set of counterparties earning interest on that floor. Now, the overnight RP facility is an interest rate that's lower than the interest that the banks were getting, but nevertheless, you have this two-pronged approach providing interest that provides a more complete floor and allows the Fed then to raise interest rates when it wanted to. By lifting both of those rates, the rates paid to banks, and the rates paid the money funds to pull up market rates along with those policy rates, which worked.

Nelson: But when you read the transcripts and I encourage everyone to read the transcripts associated with the Fed's decision to open the ONRP facility, they were very concerned about this facility. They were concerned that it was brand new, that it would contribute to quality that it would expand the Fed's footprint in the financial markets. But they did it because they thought it was going to be temporary.

Nelson: Governor Tarullo was particularly articulate about the concerns about becoming permanent. And there were caps on usage. Now, we started by talking about end of the exclusion and the SLR. So back just a couple days before the Fed ended the SLR exclusions, the supplementary leverage ratio exclusions the Fed announced that it was lifting the caps counterparty on the ONRP facility from 30 billion to 80 billion.

Nelson: And at the time I raised the possibility that maybe they're getting ready to end the exclusion, and they did so a couple of days later. Now importantly, by ending the exclusion, it made the banks deposits to the Fed much less attractive to banks. Meaning that the Fed was going to have to shift to the overnight RP facility to fund itself, which is exactly what happened as you noted at the outset. So since then the overnight RP facility has grown from zero to a trillion dollars.

Nelson: And deposits of banks they've continued to rise, but at a much slower pace now. So the overnight RP... But now interestingly when chair Powell was asked recently, is he worried about this? He said, "No, it's working just as we expected it to work." So there's this familiarity breeds contempt problem here that's... Well, originally we were worried about it and was going to be temporary, but now it's part of the infrastructure and we're not worried about it anymore. It's just working the way it should.

Beckworth: It's not temporary anymore. It's lost its temporary status.

Nelson: It appears not to be.

Beckworth: And it seems as you mentioned that they made the tweaks to the overnight repurchase facility right before the SLR exemption ended, because they knew all this excess liquidity had to go somewhere and it was driving rates down. This actually touches on several of your other problems you highlight here. So maybe we'll tie this all together right now. I mean, the Feds ongoing QE purchases, it's just flooding the system money markets with reserves. And there's only so much balance sheet capacity due to the regulations coming out of the last crisis. Which also is tied into SLR to some extent as well. We talked about earlier that it's going from banks into money market funds and these other non-bank entities have to find some place to park.

Beckworth: And so the Fed's opening up its facility to do that. Now, let me ask another question related to this. So one of the stories that's been given for why the 10-year treasury yield is so low, despite everything that's happened is because all of these reserves have flooded into the market.

Beckworth: This story was told more often before the Fed really opened up the overnight repurchase facility, but the story is the Feds flooding the market with so much liquidity that it's driving down yields farther out on the yield curve. So people are willing to take a little more risk rather than earn zero, maybe even a negative return by sitting in money market fund. And so the Fed maybe took some of the pressure off of that, but do you think the drop in long-term yields in any way is tied to this story of people working their way up the yield curve, because there's just so many reserves and it's pushed down yields at the short end?

Nelson: So I guess not exactly. And so reserves are an asset of the bank. So if you push reserves into the banking system, the banks have to hold the reserves. They don't have excess capacity that doesn't lead to a need for the institutions to buy other things. In fact, they have too many things already because the reserves are an asset. On the other hand I do feel that the promises of the Fed to keep rates at zero for very long periods of time. They made an unequivocal promise to keep them at zero until both parts of their mandate are met. They backpedaled on that, but nevertheless, it's very strong forward guidance has been pushing people out the yield curve to get yield.

Nelson: And you might think that the Fed's asset purchases themselves are pushing down yields. That's the original idea is that you buy longer term assets to push down term premium by just a supply and demand effect. You take long-term assets out of the market, and that increases their price lowers term premium. There's something funny about the Fed's current program and that they're buying short term assets, as well as long-term assets.

Nelson: Which isn't really in accord with any of the stories the Fed tells itself about why QE works. So that made sense maybe for the original, they launched their asset purchases initially to address financial market functioning concerns. And then they backed into it and changed their story as being for stimulus concerns. But it just raises this big question of why they're buying short-term securities, but they are buying a lot of long-term securities, and that could be contributing to the downward pressure in long-term rates, too.

Beckworth: Bill, you have a litany of problems on this list, and we've got to run through them quickly because we're running out of time. So I'm just going to fly through these and maybe not spend a lot of time on the Fed solution, but I do want to give you time at the end, have the flesh out your solution. So let's go at it here.

Beckworth: So next problem, banking examiners, and bank investors got used to the liquidity from the Fed’s inflated balance sheet. As the Fed tried to shrink back down to a reasonable size in September 2019, those elevated demands came up against the variability and liquidity from swings in the treasury account. This caused money markets to spike alarmingly. Bill, what was the solution?

Dependence on the Fed’s Inflated Balance Sheet

Nelson: Reverse course and get bigger beginning by lending primary dealers, $200 billion.

Beckworth: All right. So the analogy at the beginning of the woman swallowing the fly just she's swallowing the cat wow the dog-

Nelson: We're at the cat.

Beckworth: We're at the cat now yeah, absolutely. All right. Next problem. The giant balance sheet produced so much liquidity that the Fed funds market where banks, short of liquidity at the end of the day, borrow from banks with excess liquidity, evaporated. The only remaining lenders are GSEs, which have Fed accounts, but earn no interest. What was the solution?

Drying-up of the Fed Funds Market

Nelson: Keep oversupplying reserves so banks don't need to borrow from each other, make the temporary overnight RP facility semi-permanent so that the Fed funds rate remains between the rate GSEs can get at the facility and the rate banks can get on reserve balances.

Beckworth: All right. So now that's the dog the woman swallowed the dog at this point. Okay. Let's go on to the next problem briefly. Bank regulations hostile to capital market intermediation succeeded in reducing broker dealers capacity to intermediate in treasury and corporate bond markets. What was the solution?

Bank Regulations Hostile to Capital Market Intermediation

Nelson: Purchase $1 trillion in treasuries in three weeks and backstop the corporate bond market when the COVID crisis hits. Keep buying $120 billion in treasury and agency securities each month. Open a permanent standing repo facility to lend to large broker dealers to finance their holdings of treasury securities.

Beckworth: All right. So now we've got what the horse in the-

Nelson: No, I think that was the goat.

Beckworth: That's the goat. All right. Maybe the next one or two next problem, the market got used to Fed QE4 purchases, even though market functioning returned to normal by the second half of 2020, solution?

Nelson: Keep buying.

Beckworth: Okay. Now, we're at the horse. Is that right?

Nelson: That was the cow. I don't think we got to the horse. Okay. This is the horse.

Beckworth: Okay. Next one. Next and final problem you highlight. So we have a long list here, leverage ratio requirements, which treat all assets as equally risky calibrated by the Fed when reserve balance is projected to be 25 billion became binding on banks with reserve balances at four trillion. The Fed would have to pay banks a high interest rate to get them to fund further growth in balance sheets solution?

Nelson: Borrow instead from money market mutual funds, which are exempt from the leverage ratio requirement at the ONRP facility.

Beckworth: All right. So that would be the cow.

Nelson: I think so.

Beckworth: Okay.

Nelson: Hopefully, we don't get to the horse.

Beckworth: All right. So we have this huge list of problems. And again, we'll put the link up so listeners read through it, follow his links if you want more information, more data that supports this. So what is your solution to all this Bill? So your whole point of this essay is, look, they're just making this problem bigger and bigger. The reach of the Fed's growing more so what is your way out of it?

How to Scale Back the Fed

Nelson: So, I suggest a three-part process through which the Fed can reverse this and get back to being a reasonably sized institution, conducting policy with a light imprint on financial markets. So first, clearly they have to taper and stop their purchases as soon as possible. And it looks like that's increasingly the course that they're on, which is good to see. Second, they need to comprehensively review the regulations that have diminished the US financial system capacity to handle flows of securities without government intervention. So that means in particular, say the supplementary leverage ratio that we've been discussing, a requirement that treats all assets as equally risky, but as the binding requirement for some institutions now. But also the G-SIB surcharge, which is an extra capital amount that large institutions hold.

Clearly they have to taper and stop their purchases as soon as possible. And it looks like that's increasingly the course that they're on...they need to comprehensively review the regulations that have diminished the US financial system capacity to handle flows of securities without government intervention.

Nelson: And it depends as well on reserve balances and other things. So that needs to be reconsidered. It also has some strange ways that it's calculated, it's calculated at period end, which caused period end problems as opposed to using a period average which might make more sense. So third, this is the important bit, it needs to contract its balance sheet like it did in 2018 and 2019 by letting assets run off, could even sell assets gradually. But importantly, it needs to remember how it used to conduct monetary policy. And as reserve balances get back down to the levels where people are used to them, it then needs to start controlling variability and reserve balances. This is what it forgot to do in September 2019, even though we'd all been talking about tautness and the demand for reserves showing up over the previous year.

Nelson: Once you get down to that level, you need to start controlling the variability in reserve balances and then gradually nudge those balances down so that the market rates rise up above the interest rate that the Fed is paying on its reserve balances. Once you do that, market forces are going to help you because financial institutions, banks will find ways to economize on reserve balances. Because they're now costly to hold rather than profitable to hold, because market rates are a bit above them rather than below them. And then gradually as that takes place banks will hold smaller and smaller levels of reserve balances, maybe not getting back to where they were, but ultimately getting back to a much smaller balance sheet. So anyway, that's the process that I'd encourage the Fed to take and I think they can take.

Beckworth: And now's a good time to start talking about this. Because the Fed is talking about tapering maybe by the end of the year. So we want to start thinking long-term, what's our destination? Even if we don't get there six months a year from now, but where are we headed? We need some kind of goal, some target to set up for the Fed's balance sheet. Now, in the past Bill, you've also stressed the importance of regulators and how they interpret the regulations they apply to banks. And in particular, this notion of how they view high quality liquid assets and how to interpret it. So maybe speak to that the important role they can play as well in making this happen.

It needs to contract its balance sheet like it did in 2018 and 2019 by letting assets run off, could even sell assets gradually. But importantly, it needs to remember how it used to conduct monetary policy. And as reserve balances get back down to the levels where people are used to them, it then needs to start controlling variability and reserve balances. This is what it forgot to do in September 2019.

Nelson: That's an important point. So the process through which the Fed ends up getting so big is that by providing a large quantity of reserves, providing it as bit of a subsidy rate is that everybody gets used to solving their liquidity problems with those reserve balances. Supervisors get used to banks holding high quantities of reserve balances. Banks get used to holding high quantities of reserve balances. And so once you start reversing that process, you really need to prevent the friction that comes into play when everybody is used to something. So one story that a banker told me was they had been holding a week's worth of reserve balances of contingency funding. The requirement was three days worth, and they were thinking of moving down to three days worth holding treasury securities for the remainder that you can repo it rapidly.

Nelson: But they realized that their supervisor would want to know why they made that change. And then when they thought about just going through the conversation, they just decided not to do it. So there's just frictions that are interjected. You also saw that during the explanation of the aftermath of September 2019, where banks indicated that they didn't really want to run the risk of running even a daylight overdraft. Now, because if they did, it'd be harder for them to explain to their supervisors how they wouldn't run a daylight overdraft under pressure.

Nelson: So under resolution, for example, so there's all kinds of different ways that people... I thought it was very interesting. And I often quote the Norges Bank the Central Bank of Norway in explaining why they went from an abundant reserve framework to a scarce reserve framework. They said that when they provided a lot of reserves, they found out that people got used to holding a lot of reserves and they stopped using the interbank market. And it wasn't the Norges Bank's job to solve liquidity problems for banks, and they wanted banks to use the marketplace. So they decided to go in the opposite route, seeing the same dynamic that we see here.

Beckworth: So there's a role for the regulators, the vice chair supervision at the Fed, everyone to communicate and signal that, look, you can't hold treasuries in lieu of reserves to meet your requirements. So along those lines, and I think this will be our last topic because we're running out of time here Bill. Is the Fed recently introduced as a permanent... Well, actually two permanent standing repo facilities for the domestic, and then for foreign entities, it already had temporary ad hoc versions of them.

Beckworth: But now it's official, it's permanent. So one of the arguments or motivations for doing this, and I'm not sure this is the Feds, but others have made this argument. And I found that convincing and I think you find it less convincing is that having these standing repo facilities, particularly for the domestic banks, is that it would make it easier for them to hold treasuries instead of reserves, because they know if push comes to shove and they got to quickly turn those treasuries into reserves, they got a standing repo facility just waiting to do that. So do you think having this standing repo facility will make that transition to a smaller Fed balance sheet more likely?

Nelson: Sure. Good point and a bit of a challenge for me to respond to of course, because the repo facility did in some sense, expand the Fed's counterparties. So the repo facility that the Fed announced is one that they were going to be providing for primary dealers or broker dealers. Many of them are subsidiaries of banks. The primary dealers don't themselves hold reserve balances.

Nelson: So in order for a standing repo facility to make it more possible for banks to hold smaller quantities of reserve balances, it has to be a standing repo facility for banks. And the Fed has indicated that they will consider, they will gradually open this up to banks. But again, I think I'll believe that when I see it pass, like we were discussing earlier about past promises. But it's important to note that a standing repo facility... The banks already have a repo facility that takes treasury securities as collateral.

In order for a standing repo facility to make it more possible for banks to hold smaller quantities of reserve balances, it has to be a standing repo facility for banks. And the Fed has indicated that they will consider, they will gradually open this up to banks. But again, I think I'll believe that when I see it.

Nelson: It's called the discount window and banks can pledge treasuries to the discount window. It's supposed to be a no questions asked facility. It's currently charging the same rate as the standing repo facility and convert those into reserve balances. The important point is that there's a lot of stigma associated with the discount window. So really it's all just about packaging. So then the question is can... The discount window is extended under certain Fed authorities, section 10B a repo is done under section 14, it's a repo.

Nelson: So the question is if you lend banks money against treasury securities and call it a standing repo facility, will that make banks more comfortable than if you lend money against treasury securities and call it the discount window? Maybe, maybe not. I definitely agree. And I just published a post on this point that getting rid of the stigma associated with the discount window would be extremely helpful. And one way that it would be extremely helpful would be enabling the Fed to get much smaller, because if banks know they can always turn to the discount window for emergency liquidity, then they don't need to have such large deposits up front.

Beckworth: Yeah. So the standing repo facility, as it currently stands, would need to expand who has access to it to banks and it currently does. And then your next question is, well, will it overcome the stigma factor that really impairs the efficacy of the discount window, but hopefully it would. Now Bill, is this a two edged sword?

Nelson: I just want to emphasize that it's not an expansion of the Fed's reach into financial markets, it's just a repackaging.

Beckworth: Okay. That was my next question is-

Nelson: They're making loans to banks. And then they're making loans to banks in terms of having a standing repo facility for primary dealers. The purpose of that is to enhance treasury market functioning and treasury market functioning has been impaired partly. But to be clear, no treasury market was going to be able to handle what happened in March 2020. That was the tsunami to end all tsunamis. But I think that the concern is that broker dealer capacity has been declining while treasuries have been expanding. That decline owes importantly to an increase in leverage ratio requirements, which force banks to hold capital, which is expensive against riskless assets, treasuries. And really repo is against treasuries, which are also effectively riskless.

Beckworth: So the standing repo facility could be a great thing if it is developed along the lines that you've outlined. And if we get past the stigma, so let's hope that this happens and that we see a smaller Fed balance sheet in the future Bill. Well, with that, our time is up, our guest today has been Bill Nelson. Bill, thank you so much for coming on the show again.

Nelson: Always a pleasure. Thanks for having me, David.

Photo by Mark Wilson via Getty Images

About Macro Musings

Hosted by Senior Research Fellow David Beckworth, the Macro Musings podcast pulls back the curtain on the important macroeconomic issues of the past, present, and future.