Transcript: What the Fed’s Big Balance Sheet Unwind Means for Markets

The Federal Reserve recently began shrinking its massive balance sheet, unwinding trillions of dollars worth of bond purchases that it started making during the depths effort to offset the effects of the Covid-19 pandemic. It’s not the first time that the Fed has undertaken ‘quantitative tightening,’ as the process is called. But this time around is different. The central bank is withdrawing stimulus at an unprecedented speed. The big question for markets now is what the impact of this liquidity withdrawal will actually be, and whether differences in the size and composition of the Fed’s more recent market operations make this bout of ‘QT’ different to previous episodes. Joseph Wang is a former trader on the Federal Reserve’s open markets desk and now blogs about the central bank as “Fed Guy.” In this episode, he walks us through the mechanics of the central bank’s big balance sheet unwind, explains how it might affect markets, and outlines all the uncertainties that still surround this huge operation. Transcripts have been lightly edited for clarity.

Points of interest in the pod:
How is post-Covid QE/QT different to post-GFC QE?QT? — 4:21
Swapping a Treasury versus swapping reserves? — 11:34
What are US Treasury yields telling us? — 14:45
Why is the Fed shrinking its balance sheet anyway? — 17:21
The role of the RRP — 18:59
The possibility of a backstop for US Treasuries — 26:54
What would force the Fed to pause QT? — 33:16
The end of the Fed-Treasury accord —  40:24

---

Tracy Alloway: (00:10)
Hello, and welcome to another episode of the Odd Lots podcast. I'm Tracy Alloway

Joe Weisenthal: (00:14)
And I'm Joe Weisenthal.

Tracy: (00:15)
Joe, what was the biggest thing that happened in markets in recent months, over the summer? It’s like a test. I think it's like a test of, you know, what people are looking at at the moment, what they find interesting.

Joe: (00:29)
I mean, God, I don't even… The Fed tightening obviously.

Tracy: (00:33)
Yes, this is the correct answer.

Joe: (00:35)
Okay. I'll stop there. I got it right. So I'm just going to stop there and you can go on.

Tracy: (00:40)
Okay. So the Fed started quantitative tightening -- we're recording this in late June -- and weirdly it kind of went by without that much fanfare. There were a few news articles about the Fed firing the starting gun on quantitative tightening and the unwind of its very, very large balance sheet. But there was so much going on at the same time. You know, there was that surprise 75 basis point interest rate hike, and then lots of talk about inflation and things like that, that it feels like it didn't get as much attention as it probably should have.

Joe: (01:14)
Most of the attention is paid to the rate, obviously. You know, QE when it was first unveiled or when Ben Bernanke did QE II, which was the real QE during the great financial crisis, it got so much attention, but there still seems to be a lot of ambiguity about A), how it works, what it does, what it accomplishes. And then in terms of like the degree to which unwinding the balance sheet is, or is not, an additional form of policy tightening is something that I just feel like is, at best, like still deeply misunderstood.

Tracy: (01:50)
It is kind of crazy that even after years and years of quantitative easing, there's still a discussion about what the impact is and how it actually works. I mean, I remember people still arguing about whether or not it pushes up asset prices and things like that. And there are people out there right now who are arguing that the reason markets have fallen might not actually have to do that much with inflation concerns and worries over looming recession, but could just be because liquidity is starting to exit the system.

Joe: (02:19)
No, no. I mean, it's totally valid. You know, it's worth noting that we have had, you know, 2014 through 2018 markets boomed, even though there was no longer a further expansion of the balance sheet. You know, we started to rally in early 2019 again, even as the balance sheet shrank for a while, going into some of the tensions, but it really is wild as you say, like how little we know and how little even, I mean, I think even the Fed knows about like measuring the effects of changes to the size of the balance sheet.

Tracy: (02:56)
Well, there is also an argument to be made that the QE that we've seen over the past couple of years is stylistically and quantitatively different than the ones we've seen prior. And so the exit is going to be different too. So we are going to dig into all of these very big and technical questions and I'm happy to say, we really do have the perfect person to discuss this. We're going to be speaking with someone who's been on the podcast before Joe, but I think you were actually away for that episode,

Joe: (03:24)
Thrilled to have him back.

Tracy: (03:25)
Yeah. So I'm thrilled to have him back. We're going to be speaking with Joseph Wang. He used to work at the New York Fed on the open markets desk, conducting repo operations, basically being deep in the weeds of money markets. And now he runs a blog called Fed Guy, which is really a must read if you're interested in monetary policy and in all of these big questions about how it actually works. So Joseph, thank you so much for coming back on Odd Lots.

Joseph: (03:52)
Hey Tracy. Hey Joe, thanks so much for inviting me. It's a pleasure to be here.

Tracy: (03:56)
Yeah. So maybe just to begin with, could you give us the broad outline of how quantitative tightening or the QE that we've seen over the past couple of years, you know, as I alluded to, it's different to the QE that we've seen in the past. So I guess the question is how different is it this time? Like what makes this particular exit different to previous periods of quantitative tightening that we've seen?

Joseph: (04:21)
Sure. So I think this time QT is different, first in the level, and I think there's changes in the structure of the financial system that make a bit more difficult. So this time around QE looks like it's going to ramp up to about $95 billion a month. Now in contrast, the last time around when we did this, the maximum that we ever did was $50 billion a month. So in terms of pace, it's a much, much more aggressive pace. We're doing it at $95 billion a month, you can kind of contrast last time, the maximum we did was $50 billion a month.

So the way that this works through the system, I think broadly speaking, I think of QT as having two mechanisms. One is that it increases the supply of Treasuries into the market. That's one, and that's kind of how the Fed thinks of it, by increasing or increasing the supply of Treasuries, you are pushing the term premium higher. So it puts upward pressure on interest rates.

And the second mechanism has to do with draining liquidity out of the system. So these two mechanisms are related, but also operate in separate ways. And they also have a lot of moving parts into how they actually can play out. And these moving parts aren't completely within the Fed's control. So because of this, QT can play out in a range of outcomes. You can have very benign QT where it really is just watching paint dry, as Chair [Janet] Yellen mentioned before, or you can have QT that's more aggressive and very disruptive. Now, based on what I see in the current configuration of the financial system since there are so many moving parts, it seems what's happening right now is compared to last time, QT this time is going to be a lot more disruptive. I guess I can talk about why.

So I'll go by the first mechanism, the increase in Treasury supply, and then I'll talk about why draining liquidity this time will also be more disruptive. So when QT increases the supply of Treasuries into the private sector, the Fed doesn't actually get to decide what tenors reach the market. That's a decision by the US Treasury, overall what happens is that when the Fed is doing QT it's receiving repayments for the Treasuries that it owns. So the US Treasury issues new debt, and takes that money and repays the Fed. That's what happens. So it's the US Treasury that gets to decide what are the tenors of the new Treasuries that the market absorbs? Now you can do this in a way that's very market neutral. So let's say the US Treasury issues a lot of short-dated debt, Treasury bills.

Now the market can absorb these Treasury bills very easily. If you think back to the first quarter of 2020, the Treasury issued $2 trillion in bills and the market just lapped that up easily. So in a sense it's because bills are so cash-like, they don't really have any interest rate impact, but what the US Treasury is doing this time around, it's actually cutting bill issuance, because it received a lot of tax payments in April above its expectations. So all the QT increase in supply over the next few months is going to be in coupons and coupons are more difficult for the market to absorb. So it's probably going to place more upward pressure on interest rates. There are also a lot more mechanics behind this that make it this time around more disruptive. So for example, we are having a big change in who the marginal buyers are in this market.

Well, actually I'll sit back a bit and say, so the increase in supply this time around is much higher than it was last time around. I think it's useful to think about Treasury rates in terms of supply and demand. So in terms of supply, this time around, the amount taking into account of QT, the estimates for the increase in supply to the private sector, it's going to be about $1.5 trillion a year, for the next three years. Now, just for context, pre-Covid, the amount of supply that was going into the market was about $500 billion a year. And so the pace of the supply is just so, so much higher than it was the last time we did this. And that is happening in the context of, from the, I'll say demand side, from the buyer side, where the marginal buyer is changing and the market structure doesn't seem very strong.

The marginal buyer for Treasuries, before Covid, was actually the hedge fund, the hedge funds. So what the hedge funds were doing, they were buying, let's say hundreds of billions of dollars in Treasuries, several hundred billions in Treasuries. But they were buying it as part of a basis trade. So they actually didn't really care about things like growth and inflation. It was really about the spread between the cash Treasuries and the futures. So they were the marginal buyers. Post-Covid, it was all about the Fed and the commercial banks. The commercial banks, because of regulation, they have to own a lot of liquid assets and they were buying tremendously. So those players are not in the market anymore. And they were also players who are much more agnostic to things like where the interest rates were because they have to buy them as part of a pairs trade or for regulation.

Those people are out. And you're having a situation where we're looking for the new marginal buyer and that new marginal buyer is probably going to be more sensitive to things like inflation, rates, and, you know, as we see this happening in inflation, it's not clear what that is nor what the Fed's policy rate will be going forward. So there's going to be some volatility there.

Oh, and one more thing, and you can see Chair Powell mention this again. Treasury market liquidity is not very good. So when we have this tremendous increase in supply, changing demand, amidst very low Treasury market liquidity. So just, just for some context, so every day in the Treasury market, we do about $600 billion in cash transactions. And we have about a $23 trillion Treasury market for the private sector. So if you rewind the clock 20 years ago, we had about $7 trillion in net marketable Treasuries. Daily volumes are about $400 billion. So today the total Treasuries volumes have more than tripled to $23 trillion, but daily liquidity is only a little bit higher from $400 billion to $600 billion. So you can have, in a sense, you can think about, let's say the stadium getting a lot bigger, but the doors are not really increasing. And that's a big reason why we see these huge moves in Treasury yields. Recently, I think a few weeks ago we saw the 10-year just jump 25 basis points. We saw the Treasury market break in March, 2020, and we've had flash crashes in the past. So it's kind of, there's this storm brewing from my perspective where you have enormous issuance, you have a weak market structure, and you also have a demand side for Treasuries that's becoming a little uncertain.

Joe: (11:34)
So I want to explore the liquidity side a little bit more. And one of the things that we talked about in a recent episode, you know, like Treasuries and reserves are not that different from a sort of like economic perspective, right? So, okay. You talk about this like huge increase into the market of these Treasuries that the Fed will be getting rid of, reducing. But on the other hand, it's also diminishing the reserves, the liability side of the balance sheet. And so on some level there's an evenness to it and economically they're not radically different. They are somewhat different. Explain further the effect on liquidity from swapping two assets that are not that different.

Joseph: (12:20)
Yeah, that's a really good question. So I think there's a couple things to this. One is that reserves can only be held by commercial banks. So reserves are basically deposits at the Fed and only commercial banks, properly speaking, can have deposits at the Fed. So from a commercial bank's standpoint, Joe you're right. It's very equivalent. I mean, there is more interest rate risk in a Treasury, for example. But from a commercial bank’s standpoint, I can have reserves in my liquidity portfolio or I can have Treasuries. And what they've been doing over the past couple years is they're making that choice to say that I want to have Treasuries rather than reserves. Since Treasuries are yielding much more than interest on reserves, but that's not this decision faced by people who are not banks.

So for example, you and me, we have deposits at a commercial bank. We’re not eligible to hold reserves at the Fed. So when the Fed is doing QT from our perspective, the deposits in the system are declining. So when the Fed does QT, it reduces reserve assets at commercial banks, which are often backed by deposit liabilities. So it's this two-tiered monetary system we have, where non-banks have deposits at banks and banks have deposits at the Fed. So from our perspective, we're losing bank deposits, which are, you know, carry credit risk, and don't earn IOR. So the substitution is not perfect, but I think more broadly, the point though, is it seems like right now what's happening is that Treasuries are becoming less cash-like you can see this in the lack of flight to safety in market volatility. Bonds are selling off and stocks are selling off. So when we have high inflation and when we have a lot of rate volatility, it seems like the market is not rushing to Treasuries as safety, they're rushing to just cash. And so that makes it, I think, this asset swap that you talk about, which is broadly what QE is, is not a perfect substitute for each other.

Tracy: (14:45)
So I just want to touch on one consequence of the dynamic you just described before we go more into QT and the mechanics there, but we've spoken about this before, I think last year, but the implication that banks aren't necessarily buying Treasuries because they think that, you know, interest rates are going to go up or down, but they're buying them because they have to buy Treasuries to satisfy liquidity coverage ratios and regulatory requirements and things like that. And Treasuries are sort of the best option of the assets that are available to do that. So what does that actually mean when it comes to Treasury yields? Like when we look at a Treasury yield, now, how much information is that actually giving us about investors’ expectations for the future direction of the economy and things like that?

Joseph: (15:32)
When I look at Treasury yields, I don't actually think there's a lot of information content and I don't think so because, as you know Tracy, there's a lot of people who buy Treasuries for different reasons. So you do have investors who let's say, look at growth and inflation and look at yields and make a judgment. But Treasuries are very special in the financial system, in that they are considered a high-quality asset, a credit risk-free asset, and under a range of regulations, people have to buy them just because the regulations tell them to and banks, for example, they have to hold high quality liquid assets. As you mentioned, under things like the liquidity coverage ratio, what qualifies as high quality liquid assets, a very, very narrow range of assets, Treasuries being one of them. And so they have to buy some of that, but it's not just them.

If you look at, let's say, government-sponsored enterprises like Fannie Mae or Freddie Mac, they also have similar regulations. They have to buy high-quality liquid assets. Or if you look abroad, if you are a foreign reserve manager, if you're managing the foreign reserves of let's say Japan or China or some other country, you can't really buy just equities or anything like that, you usually, other than this Swiss National Bank, usually you're very, very conservative. And so you can only buy things like Treasuries. So there's a lot of demand for Treasuries that's just not that driven by fundamentals. And of course you can have hedge funds who are just buying it as part of a basis trade, where they care about the spread between the Treasuries and something else rather than the absolute level of the Treasuries as measured by, let's say, economic fundamentals. So it's really hard to look at, from my perspective, to look at price and infer economic conditions.

Joe: (17:21)
So thinking about, you know, in terms of like the mechanics or the implications of quantitative tightening, why don't we start off with kind of a basic question, but it's like, why does the Fed feel an impulse to reduce the size of its balance sheet? Because it has the rate channel, it can hike rates. It has been hiking fairly aggressively, you know, 75 [basis points] at the last meeting. Where does the urgency, or just even the impulse, come from to decrease its holdings?

Joseph: (17:54)
I think from the Fed's perspective, it's a lot like you and Tracy suggested earlier in the show, the Fed doesn't really understand what exactly happens. And so they want to be with something that they think they understand well, like the overnight rate. So it seems from what I hear, they want to get out of this balance sheet stuff and go to something that they feel like they're more comfortable with, which is raising the overnight rate. And, you know, as you mentioned, you have disagreements within the Fed as to what exactly QE does, you have people who would feel like, ‘oh, you know, QE doesn't really do anything, just swapping one asset for another.’ And yet you have people in the market who look at QE and just, you know, max long because QE makes the market go higher. So I think it’s just not very clear what it actually does and they don't want to be doing things that they don't really understand.

Tracy: (18:42)
What do you think it does? Can you sort of describe for us what draining liquidity would look like in the current period versus draining liquidity from say 2018 or 2019? Because I think that might help us sort of understand the differences here and the difference in the mechanism.

Joseph: (18:59)
Sure. So when the Fed drains liquidity out of the financial system, it doesn't actually have control of where the liquidity comes out of. It can come out of the banking system, which she would drain reserves and deposits, or it can come out of the RRP, which would just decrease the RRP size. Now the RRP, as you see right now, it's very large. It's $2.2 trillion, the RRP you can think of as just the true excess liquidity in the financial system, there's all this money that people have nowhere else to invest in. And so they just leave it on deposit at the Fed and receive the RRP rate. So when you do QT, if money is coming out of the RRP, it's going to be a very benign because you're taking money out that really nobody wants, or you could take it out of the banking system, which conceivably someone somewhere is reliant upon that liquidity. The Fed beforehand doesn't actually know what will happen. If you listen to Fed presidents talk over the past few months, they just look at the RRP and they think that there's a lot of excess liquidity in the system and so, we can just do aggressive QT, no problem. But if you notice what's happening right now is that the RRP is not declining. It will probably go much higher in my view.

Tracy: (20:16)
Right. So I have to say, we're recording this right before quarter end. So right before the end of June, and there is a very high chance that it could shoot up. I think in May it went above something like $2 trillion, which was a record at the time, but we could get another record before this episode actually publishes.

Joseph: (20:38)
Yeah. When I used to run the RRP, we were very surprised for like $500 billion. Now that's too low. What happens? So the reason is that, you have all this liquidity, how it gets drained ultimately depends on who buys the newly-issued Treasuries and how they finance it. If the Treasuries are purchased by people who are levered investors, then it drains the RRP. For example, if you are a hedge fund and you buy the newly issued Treasuries with repo loan, then the cash from that repo loan ultimately comes from the RRP, a money market fund will withdraw money from the RRP and lend it in repo to the hedge fund investor. Money fund investors can only lend, can only make specific investments. Very narrow. One of them is repo. So that's basically the only, oh, that and increased bill issuance. But broadly speaking, that would be how you get the RRP lower.

On the other hand, if the people who buy the newly-issued Treasuries are, let's say cash investors who are buying it with deposits they held at commercial banks. Then what you will see is that liquidity will be drained out of the banking system. So that means that what's being drained is not necessarily liquidity that's held in the RRP that no one wants, that's excess, but liquidity in the banking system that maybe someone somewhere is relying on. Now, beforehand, it's hard to see where the liquidity will be drained. But the way that I look at this is that I just look at what's actually been happening the past few months. So the past few months, when the Treasury has been issuing coupon debt, the people who have been buying it have been people who are holding money at a commercial bank. So you can see that with the increases over the past few months, the amount of reserves in the commercial banking system is declining, but the amount in the RRP is not declining.

So just how the financial system is currently configured, there doesn't seem like there's going to be any increased demand for leveraged Treasury investing. So going forward, what you can actually see is that the draining from QT comes out of the commercial banking system. Whereas the RRP continues to increase. This, in a sense, is kind of like a double tightening effect because when the RRP goes higher, it's also draining liquidity out of the banking system. So this is why it seems on this side of the equation, from my perspective, draining liquidity can also be disruptive. You're not draining liquidity out of the RRP, which would be painless for the financial system. You're draining it out of the banking system. And the RRP is also sucking liquidity out of the banking system.

Joe: (23:21)
Let me ask you another kind of slightly bigger picture question, but you talked about the balance sheet remains a tool that the Fed is, you know, it's hard to quantify its effects, perhaps it's a little bit uncomfortable using it and so forth. And it seems to me that, you know, you think about the difference between post great financial crisis and post-Covid, that QE was sort of used differently. And so post, great financial crisis, the Fed had hit the zero lower bound and felt it needed to ease further. And so it bought assets. Whereas my sense of sort of March, 2020 was that there was a big element specifically of this liquidity effect and of this, you know, sort of credit easing and backstop credit markets specifically via asset purchases. I guess, you know, the question I've wondered is did they sort of backdoor themselves into using a tool that it actually never really wanted? That because it had this unusual situation, they didn't really want to have to go back to QE, but they sort of were forced to, and it stuck around a longer because they had this sort of different need when Covid hit?

Joseph: (24:33)
I think you're right that they used QE differently in Covid. So post-GFC, it was largely used as a tool to lower longer-dated interest rates. So the Fed hit the zero bound, they wanted to continue to ease by putting downward pressure on longer-dated interest rates. So in order to do that, it bought a lot of Treasuries. Now fast forward to March, 2020, it was a little bit different because the Treasury market broke. So what that meant was that people who wanted to sell their Treasuries for cash could not do that. So on a global scale, Treasuries are kind of where people keep their dollars. It's kind of like a huge bank, so to speak. So for example, if you and I, we will go to the bank and we want to get our cash out because we need cash.

We expect to be able to get that, but if the bank says, sorry, I don't have any cash. Then, you know, we panic. There's a run on that bank. And that's what happened in March, 2020 to the Treasury market. Everyone who wanted to sell their Treasuries for cash realized they could not actually sell their Treasuries for cash. In that sense, there was a run on the market and they started selling everything else they could to get cash. The Fed saw that, they wanted to help that by basically backstopping the Treasury market, becoming a liquidity provider of last resort. And they purchased, let's say about a trillion dollars of Treasuries in one month. That's how QE came back in 2020. But it stayed far, far, far beyond the liquidity event. And at that stage, I think it morphed back into easing financial conditions, as you suggested, which I take to mean the original QE motivation of putting downward pressure on interest rates. So that's how I think about that. I agree. It probably was not super necessary for the length of time that they did it.

Tracy: (26:19)
You know, you mentioned the Treasury market blow up in 2020. And again, this is something that we've been talking about quite a lot recently on other episodes, we also had the repo blow-up from 2019. And I think the response to that was the creation of the standing repo facility, the SRF, which basically allowed banks to exchange Treasuries for dollars. And so I'm wondering, does the existence of something like that, does it make it less possible that we're going to get some sort of major blow up or are there limits to what the SRF can do in the current environment?

Joseph: (26:54)
So last time around, QT basically contributed to the blow up of the repo market, as you noted. So I don't think that's going to happen this time round. So, I mean, we never have the same thing blow up, usually. I think stress will be somewhere else. And I think to understand why I think the stress will be somewhere else, it's helpful to revisit what actually happened. Why did QT cause the repo market to blow up in 2019? For some context in 2019, heading into, let's say September of 2019, when the repo market blew up, there was tremendous demand for repo financing. The amount of repo demand for repo financing increased by a few hundred billion in the months leading up to September, 2019. And those were all the hedge funds doing their basis trades and that pushed repo rates steadily higher and ultimately above interest on reserves.

So the banks saw that repo rates were above interest on reserves. And they note that lending in Treasury-backed repo from a regulatory standpoint is equivalent to holding reserves at the Fed. So they figured that they can earn some extra return by shifting the composition of their liquidity portfolio to fewer reserves and more repo. And so heading into September, 2019, the banks became the marginal lender in the repo market to the tune of hundreds of billions of dollars. So QT was playing in the background and what QT was doing, it was withdrawing the amount of excess cash the banks held, the reserves. So as we had from a demand side, continued demand for repo financing, and on the supply side, banks being the margin lenders in the market, their extra cash declining because of QT. Eventually the market hit an air pocket where repo rates spiked higher, uncontrollably.

Another way to think about this is that the markets that benefited from QE cash were hurt by QT and the major beneficiary in QE cash last time around was repo. We don't have that problem at all this time because repo rates are much lower than IOR, banks are not lending in repo. What they have been lending in, as I mentioned earlier, is in Treasuries and agency MBS, to the tune of $0.5 trillion the past couple years. So we have this dynamic, we have a similar dynamic playing out, but just not in the repo market. This time around tremendous continued demand for financing by the US Treasury met by the marginal lender in the market, the commercial banks having less cash to lend. So if there is another blow up because of QT, it's very likely to be, in my view, in the Treasury market, since the same dynamic is playing out.

And what that could eventually mean is some kind of, let's say liquidity backstop for the Treasury market rather than for the repo, which I think is probably very logical given what the Fed is already doing. If you recall, as you noted Tracy, when the repo market blew up, Fed stepped in with an emergency liquidity facility for repo, when the FX swap lines blow up, the Fed has their FX swap lines. When the commercial paper market blows up, they have their, you know, they have their tools for that. In the past when the Treasury market blew up, they just did QE, which is a very blunt instrument. A more calibrated instrument would probably be some kind of emergency backstop willing to buy Treasuries at, you know, a set interest rate, set above the market as a liquidity backstop.

Joe: (30:31)
There is a standing repo facility, right?

Joseph: (30:34)
Exactly. That provides emergency liquidity. If you have Treasuries, you can repo that for cash. So it provides emergency cash. It doesn't put a ceiling on rates in case the Treasury market blows up because of selling.

Joe: (30:49)
In theory. Why is that not sufficient to avoid a blowup? If a holder of Treasuries can know that there is this window or this desk out there that will swap at any time Treasuries for cash, why doesn't that short circuit the sort of run dynamics in the first place?

Joseph: (31:09)
Exactly. So that has to do with balance sheet constraints. And I'll explain that a little bit more. So the people who have access to the Fed's repo facility are the primary dealers. So if you want to have liquidity flow from the standing repo facility to the market, it has to go through the primary dealers. And how that would play out is the primary dealer would borrow from the Fed, let's say a hundred dollars from the Fed. And then let's say on the asset side, lend out that a hundred dollars. So it expands the balance sheet of a primary dealer. Primary dealers are basically like the pipes to which money flows from the Fed or money market fund cash investors into the broader market. Pre-GFC, there is not a lot of limit to how wide these pipes could be. Post-GFC, because of a number of regulations, the pipes actually have kind of a fixed size. So for example, if there's a tremendous need for liquidity, a primary dealer cannot borrow like a hundred billion from the Fed and just lend it out to the market because they would hit these regulatory constraints.

From a high level, pre-GFC, the dealers were doing about $3 trillion in repo. Now that's, you know, pre-GFC, that’s in 200. Today, they're doing about $1.5 trillion. So, you know, everything in the market has gotten much bigger, yet the amount of repo primarily dealers do has gotten smaller by half. And those are the pipes of the financial system becoming more constrained. And also why we had these blowups in March, 2020. Dealers, even though at the time, they also had access to this repo facility the Fed had, their balance sheets, the pipes were simply not wide enough to accommodate all that. So there are regulatory things they can do to tweak that. And I think there's work being done on that side. But at the moment it's not complete yet.

Tracy: (33:16)
So what would cause the Fed to pause QT? Like what would be the catalyst for it to step back and go, wait a second, we're doing this at too rapid a pace, or we're doing too much too soon and basically reconsider?

Joseph: (33:30)
So I think the thinking is that eventually the Fed will hike or do QT and eventually something will blow up and they'll have to reverse. I actually think that's totally accurate in what happened in the past, but I think what's happening now is that the Fed actually has enough tools so that they don't have to stop. So if you think about, broadly speaking, the Fed has basically become a one mandate bank for the moment. Powell was telling you that its commitment to price stability is unconditional. He's telling you that full employment is conditional on price stability. So the only thing that matters for him right now is inflation. And so he's going to be very aggressive in his monetary tightening. And that means, of course not stopping QT and not stopping rate hikes.

He can do that now because the Fed has rolled out so many new facilities such that wide sectors of the economy can be supported, even if something breaks. The Fed during March, 2020 pioneered facilities, to make them lender of last resort for a wide range of markets and for a wide range of sectors. For example, they were backstopping the municipal bond market through their municipal liquidity facility last time around and the corporate bond market through the corporate credit facility and conceivably, they could also have new Treasury facilities as well. So I think eventually something will break because it always breaks, but it doesn't mean that they'll stop. It just means that they could use their facilities to further extend their tightening. These facilities, in my view, greatly extend the possibility of how restrictive monetary policy can be simply because they remove liquidity risk.

Joe: (35:14)
Wait, sorry. Can you just explain that a little further? You're saying these new facilities

Joseph: (35:18)
Okay. Sorry. The facilities that were pioneered back in March….

Joe: (35:22)
So these new facilities that were pioneered in March, 2020, the explicit backstopping of the credit markets, the muni facility, which was obviously extraordinary and sort of, you know, this brand new thing, explain how do you see them potentially being used? Because I feel like these facilities have largely been forgotten about, no one ever talks about either one of those these days.

Joseph: (35:40)
Exactly, exactly. So they're all forgotten and they're not commissioned right now. So what I'm saying is that if QT, or if Fed rate hikes actually break something in the market, the Fed does not have to stop. It does not have to stop because it can continue to keep the financial markets humming along.

Joe: (35:58)
So it can continue to tighten in pursuit of its inflation goal while sort of more strategically repairing potential breaks in the financial markets?

Joseph: (36:08)
Exactly. Joe.

Joe: (36:09)
Okay. Interesting.

Joseph: (36:10)
Because again, the Fed has become lender of last resort to such a wide range of market participants from the corporations, to the municipals and indirectly to small businesses through the banking system, through their mainstream lending facility. So it's basically so significantly expanded their footprint, that there's less reliance on the transmission of monetary policy through the market. And you can kind of, well, it's not ideal, but you can kind of indirectly reach it through these programs, such that even if something breaks, it doesn't actually mean they have to back down, lower rates and continue QE, especially if inflation is too high.

Tracy: (36:50)
What's your bet on what could break if you had to wager something right now, like what would it be? Where's the biggest area of weakness?

Joseph: (37:00)
I still believe the Treasury market is the highest risk, first because as I mentioned, the repo breakup dynamics that we saw in the prior QT are playing out in the Treasury market, have tremendous supply coming up a few years. The demand, it doesn't seem like it's there because the marginal investor is disappearing and we have very weak, low liquidity, weak market structure. And just watching the Treasury market over the past few weeks seems like it's becoming more volatile. So I think that's probably the place that is most likely to break this time around.

Joe: (37:35)
So one of the things, you know, as you were talking about in the beginning, it can be sort of hard to, well, it can be hard to predict what's going to break. It can also be hard to predict where the liquidity is going to get drained out of the system first, like all of these things, it doesn't seem like it's a hard science anticipating it. Is this sort of fundamentally why the Fed is so uncomfortable quantifying the tightening effect of balance sheet policy? Because, you know, it's sort of easy to see like the transmission mechanism of a rate increase. It's like you hike rates and rates go up and then you see it in mortgages and car loans and that tightens the housing market, it's somewhat straightforward, I think. Whereas if there's so much uncertainty about where is the liquidity going to come from or be pulled from, it seems that makes it inherently a much tougher tool to quantify and calibrate.

Joseph: (38:31)
I agree completely with that. And I don't actually know if the Fed understands this. If you hear, I think there's work from the Fed that's also been mentioned by Governor [Christopher] Waller that, you know, let's say $2 trillion of QT is equal to like 50 basis points. I suspect that's probably not true. And it will be something that they wish they didn't say because the thing is, there's so many moving parts to this. There's so many ways as it can go, it's not something that can fit it in an equation. Now, if you want to approach the world as if it were a giant equation, you need to have relationships that are consistent and don't change. And this is very much true in physics. If I drop a rock here, you know, 9.8 meters per second square, it goes down. Same if I dropped it in London or if I dropped it a hundred years ago, that works well for things that don't change. But if you're looking at the financial markets, the relationships are always changing. There are different regimes and there are different actors and different regulatory changes. So you just really can't know what will happen. Any estimate, I think, is just not very useful. And if so, in that sense, it's kind of good that they get out of this.

Tracy: (39:38)
This is a related question, but what's the future of the Fed and its relationship with financial markets in the sense that, you know, as you've been describing now for the past, at least the past 10 years, you know, more than a decade since 2008, whenever something goes wrong, the Fed comes up with some sort of new program to enable it to keep pursuing its policy goals or keep doing what it was doing. Is that just how it's going to be for the foreseeable future? You know, something goes wrong. The Fed comes up with a new program, it gets added. Eventually it becomes the new normal, eventually something else goes wrong and there's a new program and so on and so forth. Or is there going to be a larger shift or break in this pattern at some point?

Joseph: (40:24)
I think going forward, I think the inevitable outcome is probably a reversal of the Fed-Treasury accord simply because the Fed itself is becoming so much more involved in the markets. It's going to need to have more accountability. It's essentially becoming lender of last resort to everyone in the system, but also because of changes in the structure of the economy, such that the Fed probably can't carry out its task the same way that it was able to say at its inception. And this has to do with how the public sector is just a bigger part of the economy. For example, if you think back a hundred years ago, the government was a very small part of the economy and the Fed with its mandate of controlling inflation, it can simply adjust interest rates and private actors respond to that. It works much better.

If you are a private sector actor, you care about the price of money. And if interest rates are higher, you moderate your economic activity. And if interest rates are lower, you know, maybe you spend more. But the structure of the economy has changed so much over the past a hundred years, such that there's a greater part of the economy that's basically the public sector and the public sector doesn't really care about interest rates. So when the Fed hikes, when the Fed cuts, that doesn't really affect their economic activity. Their economic activity has to be affected through the legislative process. So as this trend continues, a greater part of the economy becomes insensitive to the Fed's interest rates, thus making the Fed less effective in controlling rates. And two, as the Fed becomes much more involved as lender of last resort to a wide range of the market, essentially becoming more in the allocation of credit business, which I think is probably something that more probably belongs to, if not the private sector, at least someone that has public mandate. So it seems we're heading towards the world [where] it will make more sense for more coordination between Fed and Treasury to achieve these goals simply because the Fed is doing more stuff that is fiscal policy-like, and also it has less ability to influence economic outcomes.

Tracy: (42:34)
All right, Joseph, it was so good having you back on Odd Lots. Thank you so much. Really appreciate it.

Joseph: (42:39)
Thank you so much for inviting me. I love Odd Lots and I really appreciate the opportunity. Thanks Tracy. And thanks Joe.

Joe: (42:45)
Thank you. It was great chatting with you.

Tracy: (43:00)
So Joe, I thought that was incredibly interesting and really good to get into the weeds of some of this. And also, I mean, one thing that is becoming clear from recent episodes is that lots of people seem to be saying that liquidity in the Treasury market has deteriorated for various reasons and that there are some vulnerabilities there, but Joseph's mention of the idea of Treasuries becoming less like cash, or less cash-like in the way they are traded and in their position in the financial system, that would actually be a sea change for markets, I think.

Joe: (43:34)
It's weird because I mean, clearly with the existence of the standing repo facility, the Fed's goal is to make it more explicitly cash-like, I mean, that's the idea, right? They're similar, they've always been somewhat money-like and similar to cash and, okay. And now they have this formal standing repo facility so that at least the primary dealers can swap them into cash at any time, even when it's not an emergency, so the fact that like liquidity is still deteriorating, the fact that, you know, we have had all these issues, you know, to his point, there's clearly still a lot of unfinished business.

Tracy: (44:14)
Yeah. The other thing that kind of struck me from that conversation was his description of how, when the RRP, I don't think we ever actually said what the RRP stands for, but it's the reverse repurchase facility, but when the RRP goes up, it doesn't necessarily mean that liquidity in the overall system is going up, which I think there are still a lot of people out there that look at the RRP at $2 trillion or whatever. And they go like, ‘oh, liquidity is sloshing around the system, buy everything.’ I'm thinking I'm thinking in particular of a certain subreddit where the RRP is a really big talking point, but Joseph's point that actually the RRP going up means liquidity might not be going out of the banking system and so, you know, financial conditions are tightening, that's worth remembering.

Joe: (44:59)
And just to his broader point, which he hit a few different ways, like the sort of relationship between quantitative tightening and where liquidity can come out of the market at any given time. And to some degree, you know, the unpredictability of it, it's different under different regimes. I think that may be like the clearest explanation of like, why the Fed and nobody else, really even talks about the tightening effects of QT in part, because it's just not nearly as straightforward or predictable. So the idea of like putting a number on it or saying like, okay, QT is like worth this many rate hikes or whatever, it seems way harder to judge.

Tracy: (45:42)
Yeah.

Joe: (45:44)
They should just keep the balance sheet big and forget about it. That's my solution. If I were there, I would just be like…

Tracy: (45:50)
It’s just too much of a headache to figure out?

Joe: (45:52)
Keep it. Just keep it there. If I were on the FOMC, that would be my vote.

Tracy: (45:58)
You know what everyone, campaign for Joe for Fed chair. You know, a simplified Fed. Simplified, open market operations. That's Joe's campaign platform.

Joe: (46:09)
Balance sheet only goes in one direction when I'm on the bench, it only gets bigger. We never do the opposite.

Tracy: (46:14)
You know what? I think that might actually be a very successful talking point. Okay. Let's leave it there before we say anything else.

Joe: (46:20)
Sounds good. Let's leave it there.

You can follow Joseph Wang on Twitter at @FedGuy12.