The market for US Treasuries is arguably one of the most important and liquid markets in the world. But it's been experiencing a number of hiccups in recent years, such as the sudden selloff of March 2020. And in more recent weeks, yields on US government debt have also spiked as the Federal Reserve raises interest rates. Some of that makes sense as the central bank makes big changes to its forecast for inflation and markets adjust to the new path. But the degree of the moves has also led some traders to conclude that there's a problem in the way this huge market is functioning. So why does a market that should be pretty boring keep experiencing all this drama? On this episode, we bring back Josh Younger, Managing Director and Global Head of ALM Research and Strategy at JPMorgan, to talk about why bonds keep going through all these shocks and what can be done to minimize them. Transcripts have been lightly edited for clarity.
Points of interest in the pod:
What is liquidity anyway? — 04:23
Market depth in the Treasury market — 8:11
What happened to Treasuries in March 2020? — 13:05
What are dealers actually doing? — 17:31
The role of HFT in Treasuries — 21:00
The growth of the overall Treasury market — 22:24
The role of sponsored repo — 26:35
Why not make the SLR exemption permanent? — 30:18
The possibility of cross-margining — 35:28
Does the Fed care? — 41:09
---
Tracy Alloway: (00:11)
Hello, and welcome to another episode of the Odd Lots podcast. I'm Tracy Alloway.
Joe Weisenthal: (00:15)
And I'm Joe Weisenthal.
Tracy: (00:17)
Joe, what's more volatile than cryptocurrencies at the moment?
Joe: (00:22)
Uh, I don't know. Wheat.
Tracy: (00:25)
This is a trick question, because I think there are a number of things. But one thing that most people didn't expect, up until recently, was bonds -- US government bonds.
Joe: (00:35)
I don't think they've been as volatile as cryptocurrencies, but they have been volatile.
Tracy: (00:39)
Oh now wait a second.
Joe: (00:41)
Oh, is that wrong?
Tracy: (00:42)
There was a day where they actually were more volatile than Bitcoin at least. And I think it was earlier this month after the Wall Street Journal released that article saying that the Fed might hike rates by 75 basis points. And suddenly we had this big move in bond yields and I think the 10 year US Treasury, the yield on that jumped something like 28 basis points in a day, which was, I know you love this, a four standard deviation move and the kind of thing that's only supposed to happen -- based on normal distributions -- like once in a century.
Joe: (01:14)
Have you ever seen that tweet about Nassim Taleb breaking through the wall like the Koolaid Man?
Tracy: (01:20)
<Laughing>
Joe: (01:20)
Have you ever seen that tweet?
Tracy: (01:22)
Every time someone says standard deviation? Yeah, I'm not surprised.
Joe: (01:26)
I'll find it for you. There's a really funny tweet about it, anyway,
Tracy: (01:29)
Thank you.
Joe: (01:30)
That was a wild day. I remember that Nick Timiraos at the Wall Street Journal said they're going to go 75 instead of 50 and everyone believed him. And of course he was right. And it was just this huge instant repricing of the entire curve. Like, really shocking.
Tracy: (01:43)
Right. And so a four standard deviation move -- make of that what you will -- but Bitcoin's move in the same period was something like 2.7 standard deviations. So we're talking about the market for US Treasury debt, which is one of the most, probably the most important market in the world. It's basically the risk-free asset against which all other assets are judged and it's supposed to be the most liquid market in the world as well. And yet it's having these big moves and this isn't the first time. You know, I said a move that's only supposed to happen once in a century, but actually we see these moves relatively often. And of course we had the big blow-up in yields in March of 2020. We had a big rally in yields in, I think it was 2014. We've had numerous other high-profile incidents where the market just seems to like, yeah, go nuts.
Joe: (02:35)
In extreme periods of volatility, we do see this stress and it's not supposed to happen, right? Like that's the whole idea here, which is that there are certain financial assets where yeah, it's supposed to happen kind of. With Treasuries because they're theoretically risk-free, and because say they can theoretically be, you know, theoretically from an economic standpoint, be like equivalent to reserves or almost like cash at some point, the ultimate safe haven asset, they're supposed to just be the one predictable thing, you know, the sun around which the financial universe revolves around. And every once in a while it doesn't happen and that's not good.
Tracy: (03:15)
Yes. Not good is a good way of putting it. The other thing that's happening at the moment is the Fed has started to unwind its massive balance sheet, right?
Joe: (03:23)
Yeah. So it's another factor here.
Tracy: (03:26)
Right. And that actually happened, I think they started unwinding in the same week where they raised interest rates by 75 basis points. So of course, no one actually noticed that this had started, but there's all these things going on in the market. It seems like we keep getting these dramatic events. So we really need to dig into it. All right. We do actually have the perfect person to discuss. We are going to be speaking with Josh Younger. He is currently a managing director and global head of asset liability management research and strategy at JP Morgan. He was previously doing sell-side research at JPM writing about interest rates and money markets, which I think was the last time we had him on and I didn't know this, but he was previously an astrophysicist studying things like black holes and what happens when large objects collide against each other -- so maybe the perfect analogy for the bond market. Perfect. All right, Josh, thank you so much for coming on Odd Lots.
Josh: (04:20)
Thanks very much for having me. It's great to be back.
Tracy: (04:23)
So maybe just as a beginning question, I think people have probably heard that there may be liquidity issues in the US Treasury market, or certainly there are these bouts of volatility, but could you maybe give us some color of what exactly we mean when we talk about illiquidity in the Treasury market and what exactly happens on a day? Like the one we saw recently where 10-year yields made this massive move?
Josh: (04:49)
Yeah. So I know it's always weird to see a four standard deviation event, but if people were around in 2014, we had a 15 standard deviation event, at least over a short time scale. So pulling in my old education, that's not supposed to happen in the lifetime of the universe. So standard deviations have their place, but you know, the normal assumption, the bell curve assumption, is not always necessarily the best one. And for Treasury markets in particular, you know, people like to say that there are lots of fat tails, like infrequent things happen at a much higher frequency than you would otherwise expect. That's true of all financial markets. It's definitely true of bond markets. But, the question when we bring up liquidity is if we look at big changes in price, is that because things have really changed or is it because there's some kind of market functioning or other issue that is causing the price change to be exacerbated for reasons other than fundamentals?
So, this is a tough environment because you know, the Fed is actually raising rates by 75 basis points in a meeting. So the outlook is truly changing rapidly. And so rapid price changes would be expected just because, you know, the fundamentals are shifting at the same pace, but we do want to figure out and watch if the market is unable to transfer risk. Because that's kind of what the market's supposed to do. They're supposed to line up buyers and sellers and take care of those bonds that don't have an immediate buyer, but have an immediate seller or vice versa. Like they're supposed to smooth out these fluctuations by having somebody in the middle. That's a market maker, who's the buyer to the sellers and the seller to the buyers and can hold onto stuff that doesn't have an immediate home.
It's just basically like running a store, right? Like you're running a candy store, you got inventory, you have orders coming in, you got customers coming in. You need shelf space, but you also need to make sure you have enough customers to take out the inventory you're getting in. So you need to line all those things up. And unfortunately, when you're a market maker, you can't really control your inventory, right? That's someone else's decision as opposed to the candy store, but there are different ways to look at this. Usually we talk about the depth of the market, which is, if you look at the screen, you're a trader on a desk at any given dealer, and you looked at the screen, the electronic markets between dealers, where they put up, you know, I'm willing to buy this much at this price or sell this much at that price. How much size could you move near the current price in the market? That's the market depth.
That's really low. So that's immediately concerning because if depth is low, that means if I go to sell something that's bigger than the size on the screen, presumably the price is going to change. And if that size on the screen is really small, then an incrementally larger trade, still small in the grand scheme of things, is going to move the price a lot. And the implication is the market can't transfer risk without a big price impact. So that's really important.
Joe: (07:37)
So just before moving on from that, on market depth, so, okay. Like Nick Timiraos comes out and says, they're going to go 75 instead of 50, of course you would expect to see a big move because there is fundamental news. There's a reason for the price of the curve to change as markets digest the sort of new thing about the Fed trajectory. But as you say, it's like, okay, that's going to happen. How do you measure depth? What would be sort of normal or healthy depth? Like how would you quantify that? And then what is the quantification of depth right now? When the dealers look at their screen.
Josh: (08:11)
So we're data limited here, because there's really two kinds of markets that exist in something like a hub and spoke or a network. The first is the end users, right? So the holders of Treasury bonds, it could be a foreign central bank. It could be an asset manager, a hedge fund, but the non-dealer, non-bank end users of bond debt. So they just hold it as an investment or a speculative asset. And then, and that market faces the dealers because when you go to sell that bond or buy another bond, you call up your dealer, right? So there's an interface there that is not particularly observable. We don't really know what's going on there. Usually people call that voice trading, which is really like in the institutional market. I would call up my dealer and say, ‘I need to sell you a billion of 10-year notes, like big trades, big institutional trades. And when retail gets smaller sizes usually, there's some institution standing between them and the dealers.
And then there's the market between the dealers themselves. And in this context dealer, we'll talk about them shortly, the HFT component of this, the high frequency component, but like the inter dealer market is the way that dealers pass risk around between each other. And in some sense, that's the more important market. Because if you think of yourself as buying whatever someone's selling and selling whatever someone's buying, the amount of size you're willing to show, the size trade you're willing to do, is informed by how easily and cheaply you can hedge that.
So the way I like to think about this, and I'll get to your question in a second, the way to think about this is you've got sort of risk coming in from the outside end users selling a bond. For example, say they're selling a billion of those 10-year notes. They're going to call one or two dealers, not going to call everybody. Those one or two dealers are going to buy that debt at some price, and they're going to try to hedge it. And what hedging really is, is socializing that slug of risk across the broader complex of dealers. So they're all going to keep a piece of it, but they're not going, the dealer's not going to hold a whole billion outright. That's too much risk to hold as a market maker and yields move a little bit, all of a sudden you've run through all of your capital and that's a problem. So you, the active hedging is the act of taking pieces of that and spreading it around. And that's what the inter dealer market does. That's the electronic market. And that trades much more frequently.
It's a decent chunk of the overall volume traded, but it's almost entirely concentrated in recently issued bonds. BrokerTec’s by far of the biggest venue for that. And that's where we have good data, because it's electronic. So we can see the screens so to speak. And we can capture that data and we can look at say the total number of bids and offers within say two to three levels of the best price. This is the central limit order book. People call them CLOBs, which is always kind of a fun acronym. But they can track that and say within three levels of the best price, there is X million dollars on the screen to trade on average during say the New York trading session. When you do that on average, something like $150 million over long periods of time. Over the past 10 years, that's averaged something like $150 to $200 million.
So that would be sort of typical and in bad times that can go down to $5 to $10 [million]. That's what we had in 2020. Now it's probably around $40 or $50 million on average. Now this fluctuates a lot over the course of the trading day, but this is like if you were to close your eyes and then, you know, pick a time and look at it, you get something like $40 to $50 million to trade within three levels of the best price. So that's low.
The reason why that's important again is because that's the liquidity available to the dealers themselves to hedge their risk. And so if that number gets higher, they're willing to make bigger markets facing clients. If that number gets smaller, they're going to make smaller markets because they don't have confidence that they can hedge out of the risk. So that's low along the lines of different, volatility episodes we've seen in the past. So if this was August, 2011, it would be probably a similar number. If this was June or July of 2013, the taper tantrum, you get a similar number to that. March of 2010 was much lower. In November of 2008, it was much lower. So you know, it's been worse, but it's definitely not great out there.
Tracy: (12:29)
Can we talk a little bit about March, 2020? Because I think that was when we first had you on and we were talking about this big sell off in the bond market, which was exactly what people didn't expect to happen when we had a huge risk off event, which was the global pandemic and equities sliding and all of that. What exactly is the consensus around what happened in March of 2020? What was the issue that whole incident exposed?
Josh: (12:57)
Yeah, so it's funny. I think if memory serves, I was supposed to talk about Libor reform.
Joe: (13:02)
I mean, I think we eventually did like a few months later.
Josh: (13:05)
We did. Yeah. More interesting events intervened. So this brings up a really important question, which is illiquidity versus market functioning, right? Like what was so much worse back then because the Fed is unwinding their balance sheet, like you said, they're not buying $3 trillion worth of assets over a few weeks, or $3 trillion rather. So there's clearly a lot less sense of urgency about the impact of this illiquid episode versus the events of 2020. And that's where I think we should be very specific about what we mean by liquidity, because what we really mean, we're always feeling around this is a different people filling different parts of the elephant, but like one’s a trunk, one’s a foot, but they don't really know the whole picture. I think if we were to have the full picture of the market, all of the participants and all of their risk tolerance and positioning and we would really be asking this question, when there's more sellers than buyers, how much does it move the price? And the implication there is can I turn my Treasuries into cash at a reasonable cost in a reasonable period of time at scale?
And that's not just part of the fact that the Treasury market is perceived as risk free and is quite large. It's built into a lot of the law. The liquidity coverage ratio does not distinguish between Treasuries in cash. I'm not a lawyer, but I believe there’s a part of the IRS code that says you can actually pay your taxes in Treasuries, which is the only financial asset you can directly pay your taxes in. So like a chartalist would say, well, then they're money, right? And so Treasuries have special status in lots of different ways. The reason they do is because they're supposed to be as close to cash as any financial asset can be. And that means I can move large size at low cost. Because when you turn your bank account into physical cash, they don't charge you a fee, right? So I need something that's convertible into real currency at low cost.
And back in 2020, not only was debt low, but the bid ask spread in the market was very wide. So what's going on now is the depth is low, but if you're able to trade, you're able to trade it. Roughly the bid ask spread, you were able to trade out a year ago, especially in sort of benchmark issues like the 10-year note back in 2020, it was three, four, five times wider. So there were no bids or offers within three levels of that market price. In fact, the market price, if you don't have bids or offers within one or two ticks of the market price, then it's not really the market price. We don't really know where the market price is.
And so back then the market was not functioning. It was not just illiquid and that's where the urgency comes in. So the question is what's different now versus then what's going on now is a revision to expectations. But more importantly, this is an environment that's sort of familiar in certain respects. You know, we've had inflation before, not recently, but we've had it before. We know what it is. We know why the Fed is raising rates. We have reasonable confidence around the range of potential outcomes. We'll argue about where the Fed's going to stop, but we know they're going to keep raising rates until inflation comes down and we can say 3%, 4%, 5%. But I don't think anyone's out there calling for a 30% funds rate. So these are within the bounds of things we've seen before back in 2020. It's so easy to forget just how crazy the set of uncertainties was. And like for example, there was speculation they would just close the market for some indeterminate period of time, right? So if the market's closed, the bid ask spread is infinite, because you can't transact. So if you need cash, you need to get it now, so like that creates a lot of urgency.
Joe: (16:29)
So just going back to March, 2020 for a second, and as you laid out nicely, like the law basically says Treasuries are cash or very close to cash or cash-like and should be treated as such. But if I recall, like with March 2020, part of the issue, and part of the reason we saw these sort of extreme widening of the spreads, is that even though there's theoretically free money on the table for someone to come in and close these spreads, that the demand for pure cash was so strong that nobody, no actors in any part of the ecosystem, whether it's hedge funds or banks or whatever, wanted to deploy capital, deploy balance sheet, to close these spreads to take advantage of these opportunities at scale. And so is there sort of like a core problem with, yes, we want to treat Treasuries as cash, but the way we have the system set up is such that we still depend on the risk taking appetite of private investors to actually do that job in the end?
Josh: (17:31)
Yeah. So this gets to, you know, what are these dealers actually doing? And I was referring to, you know, matching buyers and sellers is one function and holding inventory is the other. I'm going to reference the SLR, the supplementary leverage ratio, because it's one example of a regulation that creates a very different set of incentives for dealers that are housed within banks. And what the SLR does is it makes all balance sheet count the same towards your capital requirements. It makes balance sheet a scarce commodity, right? And that's not necessarily a problem if it's not strictly binding or there's relatively two-way flow, meaning as many buyers or sellers, but in a one-sided market where there's most sellers -- and that's what you're referring to, that most people want cash, they don't want securities -- it makes it very difficult for a dealer that's housed within a bank to hold that inventory.
Because what balance sheet constraints do, what leverage constraints do, is they turn that allocation of balance sheet into a zero sum game. So when the Treasury desk, for example, wants to buy that billion dollars worth of 10-year notes and they are near their allocation limit because if you want to limit the amount of balance sheet a business uses, you just give everyone a limit. Like, that's a plausible way to do it. So that creates rigidities and let's say they call up their manager and they call their manager and it goes all the way up the chain. They say, I need another $5 billion of balance sheet because there's so many sellers and I'm the Treasury desk. I'm processing the most important market in the world, right? So I need that extra balance sheet. The reaction is well, so does the prime business, so does the credit business, so does the front-end business, the short-term credit desk? So does the corporate lending business where you're drawing revolvers.
So everyone's got their hand out and that's a zero sum game in the sense that there is a balance sheet allocation that the firm can use and it makes it very hard in practice, especially at a high frequency, like a high pace, things are moving quickly, to get those allocations where they need to be when they need to be there. If you think back to that period of time, it's easy with hindsight to say, well, you know, there wasn't that much selling, but you don't know that in the moment, right? It’s like driving a car 300 miles an hour towards a wall with your eyes closed. And like, it might work out, but it might not. And it creates a lot of hesitation. So you have in 2020, what's referred to as the dash for cash, but it's really about a relative urgency to find liquid assets, not securities, not long-term securities.
So due to this concern, that one, you might need that cash for something and two, they might close the market tomorrow. That's the background. I think the important thing to keep in mind as well, and I think we talked about this back in 2020, there was this push over several years to take those market making functions and in part split them into two parts and push them out of the banking sector. Now this wasn't purposeful, but that was the set of incentives that were created. So that when you think about matching buyers and sellers, high frequency traders took on a lot of that responsibility. And they did that by doing very, very high frequency, fast arbitrage trades to try to spread the risk around really quickly and go home with no inventory. So high frequency traders generally don't have inventory at the end of the day. So they're doing only matching like that's most of their function and they were 70%, 80% of that onscreen depth. So they were the majority of the market, but they can't operate in a high volatility environment.
Tracy: (20:46)
Right. The thing about HFT is it tends to actually back away if something big is happening, right? Like if there's a big data release, like often the machines will kind of back away from making that market until things settle down a bit.
Josh: (21:00)
Yeah. Like, at the time, if you were to look at the way the market trades in the three seconds before the payrolls number drops on monthly Fridays and look at any day, any period of that two-week span and just how the market was trading, like it felt like payroll Friday at 8:30, all day long for two weeks. And the reason is because if you're a high frequency trader, you're trying to make those bid ask spreads, you need to sell for a dollar, and buy it back for a little less. And so if market markets are moving around a lot, it's really hard to monetize those spreads. So there's a reason for this, but in making it more complicated and costly for bank dealers to perform that market-making function in businesses like Treasuries, the HFT stepped in to fill that need -- like a ‘life finds a way’ moment. So they step in and fill the need. They do a great job until things get dicey. And then that business model does not operate well when vol is very high. So they tend to jump and drop the size they're willing to show.
Tracy: (21:56)
Can you talk a little bit about the overall size of the Treasury market as well? Because we have seen the US deficit growing. I think the amount of Treasuries outstanding is at something like $23 trillion at the moment, and it's gotten a lot larger in recent years, even though the Fed has been purchasing US Treasuries up until recently, how does that impact ease of trading or the way the market functions?
Josh: (22:24)
Well, the Fed helps a lot when they buy half of the net supply for the past two years. So there's been sort of two things going on there. The first is the Fed is really there to absorb the supply. And the second is the commercial banking sector, which I believe is the sixth or seventh, the US commercial banking sector, which I think is the sixth or seventh largest holder of Treasuries, has been the second largest buyer over the past two years. And that's because when you grow the size of the banking system, which is what QE -- quantitative easing -- does, there's this natural need to find assets to support those new deposits, those new liabilities. And so, you know, banks have stepped in to absorb a lot of that supply and there's a bunch of other components in the market that do sort of take on a decent share of it.
But the question is more, is the market too big for the intermediation capacity offered? I would argue that's probably still the case. It's certainly grown if anything, and bank appetite to market making Treasuries has not increased, but that has vulnerabilities. So just because that ratio is a little off doesn't mean the whole world is going to fall apart at any moment. But when the stress hits, which is what happened in 2020, it really struggles to serve that function. And that sort of brings up the second thing that non-bank dealers were doing, which is on the hedge fund side. We probably talked at the time about basis trades and specifically holding securities levered with repo. So levered longs in Treasury securities that were hedged with futures positions. There was a big position that built up over time because if balance sheet again is a scarce commodity, it's a zero sum game.
Then you have to sort of use it or lose it, as a client. And one way to do that is to have positions that utilize a repo line. You're sort of using your line of credit, but you're hedging the risk in the futures market. And so you're not actually taking any market risk, but you're using your allocation, which looks a lot like holding inventory. What dealers tend to do is they tend to be long off the run securities. Off the run securities are bonds that were issued a little while ago. They're not the most recently issued security. That's what long term holders tend to have -- stuff that was issued a while ago when they sell it, dealers hold that inventory. And they hedge in the futures market. On the hedge fund side, there was a position that built up, which was pretty passive, mostly designed to, you know, use it rather than lose it on balance sheet. And that looked a lot like inventory.
And so you have these two functions of matching trades performed by HFTs and holding inventory performed by hedge funds. That was a brittle arrangement, because it was basically taking the market making capacity and allocating it away from banks to relatively unregulated institutions who are not subject to the SLR for example. And what happens in 2020 is that whole system comes crashing down because of a variety of operational and market-related concerns that basically make it untenable to hold those basis positions and very difficult to operate a high frequency trading operation. And so you have all of this intermediation capacity that's in principle provided by non-banks, all of a sudden disappears. And when the banks are asked to take on the slack, they simply can't do it for reasonable price, which is why this whole arrangement came to be in the first place. And so the market stops functioning.
Tracy: (25:39)
Can I just ask, is this sponsored repo?
Josh: (25:42)
So this is just repo in general. Some of it's sponsored. Sponsored repo is definitely worth talking about in terms of some of the solutions that are proposed to this issue of intermediation capacity, but this is just repo-funded positions in general.
Tracy: (25:56)
Okay. Should we talk a bit about sponsored repo in that case? Because this is, we're talking about dealers’ ability to intermediate the market being perhaps too small or too constrained versus what's going on there. And as you say, one of the solutions to this issue has been the idea of sponsored repo, which I think basically I wrote about this years ago, so I can't remember everything, but I think it basically allows banks to transact with counterparties like hedge funds without necessarily bumping up against balance sheet constraints. And I think they get it from the FDIC or no, sorry from the FICC.
Josh: (26:35)
Yeah. So if the issue is balance sheet is a zero sum game, then if you can increase the amount of balance sheet that's being passed around -- the balance sheet capacity of banks -- then you've addressed the problem in part. So there's a couple ways to do that. The first is you can just change the rules, right? And that's what the Fed actually did on a temporary basis. In 2020, they made a temporary change to the supplementary leverage ratio that said, if it's cash or it's Treasuries on balance sheet, which is an important distinction, but if it's cash or Treasuries, it doesn't count anymore or for at least a year. And the idea of that was to create capacity, right? Because now the size of my balance sheet I use for that supplementary leverage ratio is just smaller and certain things don't contribute to that number.
So I can do more of them in principle. That was the theory. It doesn't necessarily work that way in practice. But you know, it's a separate thing. So one way to do this is to change the rules, change the game. The other way you can do it is play it a little differently. And so, you know, one thing that was proposed very quickly after, after the crisis in 2020 was to introduce a broad clearing mandate for the Treasury market and the logic for that was on the one hand, it reduced settlement risk by having all of these transactions go through a central counterparty. So it would sort of be less likely that securities, for example, could not be found in time -- less failures to deliver, which in principle has an impact on certain capital requirements and so forth.
But you know, the more interesting thing from my perspective, [the] impact of that was that in the repo market, if you were to clear all those trades, the impact of central clearing is, say ,everyone's facing the same counterparty. They they're all facing FICC, which is the clearinghouse. It's the same way it works in derivatives markets where everyone faces the CME, right? Because all their derivative exposures, all those contracts are novated or transferred to a central counterparty that serves as the other side to every trade. And so they're naturally offsetting because derivative markets are, you know, definitionally zero sum. And so they have all the positions, they can match off trades and reduce the overall sort of credit risk embedded in derivative contracts. So in the Treasury market, the idea was, well, what if everybody just faced FICC? And that has a couple of implications, but one of the most important is that when the banks measure their balance sheet, one of the ways they do that in repo markets is they say, am I facing the same counterparty?
So if you borrowed with the left and lent with the right, you're kind of economically neutral, but if you do that facing different counterparties, then they both contribute. Or one of them contributes to the leverage. And that's because those two trades can't see each other, in, in principle. But if everyone's facing the same counterparty, then they can. And so the idea there was, well, banks will get balance sheet relief and elasticity, meaning they can grow their exposures as needed because they'll be lots of offsets in the way you measure leverage. And if you can count all your economic offsets in your regulatory exposure, then that ratio is less binding. Which is like a lot of technical ways to say, you know, a central counterparty will reduce the amount of letters you have to show for your regulatory requirements.
Joe: 29:58
Why can't they just make those rules -- the one-year rule where they said Treasuries are equal to cash -- why not just make that permanent, especially since other parts of the law indicate that? And as you mentioned, like, you can pay your taxes with Treasuries. If so many parts of the law say Treasuries are cash, why not just have that be a permanent part of bank regulations?
Josh: (30:18)
Well, people have made that argument. It's an ongoing debate as to whether Treasuries should be excluded. It would be a little difficult in the context of international standards to exclude Treasuries and not be a little out of the mainstream, but certainly cash at the Fed is one of those things that people argue shouldn't be included. And one way to think about that is if you get a home equity line of credit, right? And you take a hundred thousand dollars out and you take that hundred thousand dollars, you put it in a bank account and keep it there, don't do anything with it. So you always have cash in the bank to pay down the loan, but you still have the loan. Has your credit gone down? Are you less creditworthy now as a consequence of doing that?
Because if you want, if you have a new liability and you have cash to back, you're sort of not really necessarily reducing your credit quality as a borrower, right? And so should you have to hold more capital against that position is kind of the question, and a lot of jurisdictions have, or some jurisdictions have said, well, you shouldn't. You're not less safe for sound as a consequence of having more cash. On the Treasury side, you know, people have argued well that that has interest rate risk associated with it. It's not purely fungible. It's not purely cashless. Look what happened in 2020, you might have a market functioning issue. But this is like an ongoing debate, which is what's the proper measure of size for a bank, because the need to think about size constraints in general is something that the Basel committee is very focused on and regulators are very focused on after the 2008 crisis.
But you know, how you measure that size, what actually should contribute to that size measurement? How big are you really? And how risky are you really? Is an ongoing source of debate. But the clearing question, what's interesting there, I think, and important to note is it's not a matter of direction, meaning, you know, if you were to introduce a clearing mandate, the amount of balance sheet allocated to repo, like the amount of balance sheet size you'd show would go down. The question is how much, and that's where sponsored repo comes in, which is clearing is available currently. And there's a decent chunk of the market that's actually cleared already. And so, you know, it's not obvious. And there's lots of reasons to think that it would likely not meaningfully improve the leverage constraints that banks are facing. That is a very, very mild self for what is otherwise a much more acute problem.
Tracy: (32:35)
So we don't necessarily have a clearing mandate in US Treasury trading, the way we do for derivatives, where trades have to go through a central counterparty, but we do have sort of some clearing going on in the market through for instance FICC. Rather than just announce a total clearing mandate, is there a way to incentivize market participants to do more clearing voluntarily?
Josh: (33:05)
Well, those incentives are there in the price. So at times, for example, if you're a cash lender, it's been advantageous to do that in a cleared format, to do that via sponsored, like banks will essentially pay up a little bit to incentivize you to do that. And the same is true on the borrower side. And so, this kind of balance sheet optimization work is something that the larger institutions have gotten quite good at. And they know how to price trades to sort of push people or nudge them in the direction that benefits their regulatory constraints. So in that sense, people are acting economically. The issue is not so much, you know, can we push people in that direction? It's more, you know, what direction do they want to face? And by that, I mean, when rates are rising, for example, now the way that hedge funds position for that is they don't do repos, which is a way to buy Treasuries when using leverage, they do reverse repos, which is you're going short the market, right? That's how you position for rising rates.
And so that creates netting inefficiencies, which basically means you can't net off trades, on the borrower versus the lender side, because there are simply less borrowers out there. So, you know, the amount of netting you can do in your book goes down as a consequence of the positioning of the hedge fund and speculative industry, those who are trying to short Treasuries. And so in practice, that actually makes a bigger difference than the availability of sponsored trades. So it's not about access to clearing as much as this is about, you know, do people think rates are going up or down, which is something you obviously can't control.
Joe: (34:39)
Going back to the present tense. And, you know, obviously March 2020 was an extraordinary situation. I mean, truly, you know, maybe hopefully once in a lifetime or once in a century. What we've seen more recently should not be that rare. And, you know, it's going to happen multiple times, probably in our careers, that people are surprised in a significant way by the direction of the market, because that's what markets do. What are the lessons here and how bad is it that, you know, we have this lack of depth? That we have this gap between Treasuries that are, you know, on the run versus off the run Treasuries and so forth. And is it something that needs to be fixed? Is it something that, you know, when you look at this lack of depth needs some sort of like policy or architectural solution?
Josh: (35:28)
Yeah. I think it's important not to over solve for the problem. So like, as an example, if you have a plate and you think it's cracked and then you smash it on the ground and it breaks along the crack, you've confirmed that it was cracked, but that doesn't mean you should only have plastic plates. So, okay. That's not the right level of stress to solve for. And in a lot of ways, 2020 is the smashing of the plate in that it's usually going to break it. And that's because it's extremely unusual. It is unique in lots of respects. And yes, you could probably say that about any crisis, but it's important to recognize that we should not be creating or designing a Treasury market that is specifically calibrated to survive a once century pandemic liquidity squeeze.
That is not the right level of rigor. And I think it's likely we can't do that within reasonable constraints. Like there's a lot of reasons to think that even in the absence of leverage constraints, even in the absence of other issues, 2020 would've been a mess in lots of ways regardless. So it's unclear that we could actually have avoided that fate, but the question is, do we want a more resilient Treasury market? I think there, the answer is very much, yes, it's important. When we think about that, if we go through the list of reforms that have been proposed to think more in terms of what is the quantitative impact? Like how much is this going to help? Not, will it help, but how much will it help to reduce the fragility in a system that under much less extreme circumstances has exhibited a lot of frailty.
And that comes back to this issue of disincentivizing banks from taking on high leverage, low risk positions -- among those are repo and Treasury intermediation. And to say that there's a lot of reasons, and this goes all the way back to the Fed Treasury Accord in the 1950s, where the proper function of the Treasury market is a matter of national importance, like the economy requires it. And so it needs to be robust to modest shocks or even significant ones. And so, you know, the question is when we, for example, include reserves in the leverage ratio, and the reason why I'm talking about cash and not Treasuries is that, you know, capital's fungible, right? You have some capital requirement. And if you create more space relative to minimums, if you make sure the supplementary leverage ratio is really a backstop rather than a binding constraint, then you have space to do other activities that would otherwise consume balance sheet. And one of those is Treasury trading.
You know, we have to think about, and the Fed has been quite clear that they are thinking about this, whether, you know, the banking system is less safe and sound is a consequence of a higher reserve balance. And whether that's the right way to think about capital requirements and safety and sound is so that's on the regulatory side. And then on the market structure side, you know, there are certain ways in which you could reduce these procyclical tendencies. And one of the things I like to highlight is cross margining. Now, what is cross margining? Back in 2020, vol was very high. And in those basis positions, the two legs of the trade were margined separately. And by margined separately, I mean that the margin you had to post against the futures leg was based on an outright exposure.
The margin you had to post against the cash leg, the securities leg, was based on outright exposure. They couldn't see each other. So you could be sort of well hedged, but still have to post margin on both sides of your hedge. Because those two hedges were margined independently. That sounds like a technical nuance. Why am I talking about that in the context of like changes to the overall structure of the banking system? It’s because that creates post procyclical. Back at that time, the futures margins increased by several times very quickly. And that naturally delevers the entire financial system, because if you have to post more cash against a position, you get less leverage and you know, that might be desirable under normal times -- I'm not going to say it's necessarily desirable. You might want less leverage in the system. You might want more, but reducing it rapidly during a period of stress is sort of de facto problematic both in practical terms, because you’ve got to find that cash, but also like the signaling value that it's important to keep in mind if margins get tripled and no liquidations actually come as a result of that, no one liquidates positions, they find the cash elsewhere.
You could still trade the market. Like there might be liquidations and then it sort of has the same impact. And you could see that in part from the commodities experience in the past few months, right? There was this concern, that margin requirements were going to force liquidations of positions and that just creates a whole mess. So, you know, one of the things you can do is you can say, look, if you're well hedged on an economic basis, you shouldn't have to post as much margin. And that just reduces the pro cyclical dynamic that the margin cycle introduces, and makes the market sort of more resilient to volatility shocks.
Tracy: (40:31)
You mentioned how the Fed is thinking about these issues. And I, I think they published a couple papers on this topic, but why doesn't the Fed feel compelled to intervene? Because clearly it's comfortable unwinding its balance sheet. This is just to play a devil's advocate by the way, but it's comfortable unwinding its balance sheet. It's comfortable, I don't want to say abandoning forward guidance, but certainly surprising the market as it did recently. What do they see here that maybe market participants don't or where does the difference in opinion come from?
Josh: (41:09)
Well, I think this comes back to the question of is it illiquid or is it not functioning? And this is where I think the somewhat hyperventilating language that traders often use is not terribly helpful. So like the number of times anyone's been told their ‘face has been ripped off’ is like probably not the right number of times relative to how like intense that image is. And so things are not, I guess, in that sense, that bad because you know, transaction costs are manageable. Like if risk is clearing, if you can get the trade done, you can do it at a relatively tight bid offer, at least in the current issue, the most recently issued bond. So like in that sense, the market is functioning. The high frequency traders have stayed involved much more so than prior volatility episodes.
There have been some structural changes to the way these markets operate. And one of the big advancements since 2020 has been the rise of what are sort of called bilateral streams or private central limit order books where, you know, the issue with BrokerTec is when you put it in an order, everyone can see it, even if they don't know who put it in. Whereas these bilateral streams and private order books, allow you to stream prices directly to specific clients. So you don't have to tip your hand in the same way. And that sounds like a small change, but it makes, it makes the showing of bigger size in general much easier. So markets are more resilient in this consequence. That's become a much more common way to trade interdealer that than was the case even a year ago. The question is why are yields actually going up and is that bad?
And if you're the Fed, you want to tighten financial conditions, presumably, so rising yields are not prima facie a problem. If you look at dealer inventories, they're pretty low. If you look at the way the repo market is trading, it suggests a scarcity of collateral. It suggests there's not enough bonds in the system rather than too many. What's probably happening here is speculative investors are going short the market in the way I described, through reverse repos. More so than real money end users of securities, of Treasury bonds, are actually selling them. And so that's a lot less concerning, right? So on the one hand, you're kind of getting the macroeconomic outcome you want. And on the other, you don't have the outflows from the real firm hands in the market at the nearly the same scale you had in 2020. So the need of the Fed to take those bonds out of the system is a lot less because they're floating around in the levered complex anyways.
It's not like you have long term holders looking to liquidate and no one on the other side. You take that all together and if I was at the Fed, would I be concerned? Absolutely. You know, low levels of depth, high levels of vol, a very uncertain macroeconomic environment and policy environment are all reasons for concern. But would I be convinced that not only are markets functioning poorly, but that they are not functioning to the point where this could create other contagion effects in the financial system and necessitates basically sucking the venom out, which is what those market functioning purchases were, we’re definitely not there. I mean, we're very far away from that kind of thing. So, you know, in that sense, it's a familiar kind of stress, which is still no fun, but it is not the kind of existential angst created by the environment of 2020.
Tracy: (44:21)
Existential angst. That’s a good way of describing it. Josh, we're going to have to leave it there. I feel like we could talk to you for another hour on this topic, but thank you so much for coming on Odd Lots. That was great.
Josh: (44:32)
It was great to be back. Thanks for having me.
Joe: (44:33)
Yeah, thanks Josh.
Tracy: (44:51)
So Joe, I feel like existential angst is a really good way of putting the experience of 2020. And that's when we did see some momentum towards actually addressing the problems that Josh was outlining in the world's most important market, but it does feel like absent a major crisis, these issues just kind of like limp along.
Joe: (45:12)
Yeah, I think that's right. And you know, some of this stuff, it's not the end of the world per se. It's not an imminent crisis, but because it is the world's most important market, there are reasons to be concerned when you see some of these lack of liquidity or the fragmentation that he discussed. You know, the thing that gets me is it feels like there are so many different moving parts that work at cross purposes. So you have some laws that say Treasuries should be cash, right? You have other laws that say banks can't get too big. You have other situations in which expansive Fed balance sheet helps liquidity. On the other hand, the Fed feels it needs to shrink its balance sheet for its monetary policy purposes, so you have all these different things that each individually may have some logic, but it's the confluence of all of them that seems to create these problems.
Tracy: (46:00)
I think that's exactly right. Also something I didn't know up until recently, do you know, repo contracts are apparently exempt from the automatic stay of bankruptcy? That was something Paul Volker did.
Joe: (46:10)
What does that mean?
Tracy: (46:11)
So it means if a company goes bankrupt, their repo assets don't automatically get frozen and then distributed to creditors. So, you know, if you want to treat Treasuries like cash, but on the other hand, if someone has a Treasury that they've used as repo collateral, in a bankruptcy, it wouldn't necessarily be distributed. Sorry. I'm getting really into the weeds.
Joe: (46:32)
No, but it sort of speaks to the issues.
Tracy: (46:34)
Right. Do we want to treat them like cash or do we not? And should we actually have a holistic approach towards the market?
Joe: (46:44)
All the regulators and policy authorities just have to get in the same room and it's like, let's just get on the same page with this stuff. That’s the trick.
Tracy: (46:51)
Yeah. Okay. We’ll make that happen, I'm sure. Should we leave it there?
Joe: (46:54)
Let's leave it there.