Joshua Younger on the Treasury Market: Structure, Stressors, and Potential Reforms

Josh Younger is currently a managing director and global head of ALM research and strategy at JP Morgan, and previously spent over a decade as a senior market strategist focused on interest rate and money markets. Josh joins David on Macro Musings to discuss the current state of the Treasury market and various reforms that have recently been proposed for it. Specifically, Josh and David discuss the history and evolving structure of the Treasury market, the emergence of high frequency trading firms over the past decade, the factors behind the 2020 dash for cash, current stresses on the Treasury market, as well as potential reforms for the market going forward.

Read the full episode transcript:

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

David Beckworth: Josh, welcome to the show.

Joshua Younger: Thanks for having me. I'm a big fan, so it's great to be here.

Beckworth: Well, it's great to have you on. I have heard you actually on the Odd Lots Podcast several times. Anyone who has something to talk about Treasury markets, money markets, that is food for my soul and I've enjoyed your commentary. So, I'm delighted to get you on the show. I should give a shout out to Lev Menand, who's a common friend between us, and he kind of arranged for us to meet and get you on the program. Thank you, Lev. Josh, you've spent a long time in the Treasury market and money markets, and I'm really looking forward to discussing that with you today. Let's begin just with some basic overviews about the Treasury market. Tell us its size and just some basic descriptive statistics about it.

Younger: Yeah, sure. The Treasury market is the largest US dollar fixed income market. It's the largest fixed income market in the world, depending on how you cut those things. It's over $20 trillion, just about 25, in fact. So, that's close to 100% of GDP for the United States, and it really forms the backbone of the global financial system. The Treasury yield is the benchmark against which all other risk assets are measured. It's a store of value for foreign central banks. For example, anyone who has a current account surplus has been accumulating treasuries for some time. So, it's a source of not just assets for the United States, but for foreign governments, and it backs their money.

Younger: In some ways, it's an instrument of global dollar dominance, because the fact that those foreign central banks can source significant quantities of high-quality US dollar fixed income securities means that they're very comfortable using dollars for foreign exchange transactions, or for trade invoicing and other forms of cross border payments. And so, as we've learned over the past few months, that's a very powerful tool of state craft when wielded. And so, that system owes in large part to very broad based global support and faith in the full faith and credit of the United States as applies to the Treasury bond market. So, it's not just its size, but it's its liquidity, it's its depth and breadth of secondary markets and trading, and its availability to a very broad range of market participants.

Current State of the Treasury Market

Beckworth: All right. Thank you for that overview. Now, today, I was looking at interest rates on Treasury debt, ten-year treasuries. Last I checked it was just under 2.8%, 30-year around 3%, which is rather remarkable in my view, given that we added about $5 trillion in federal debt,  Treasury debt, during the pandemic. And yet, we have relatively low interest rates. There's been a lot of talk about, "Oh, this is the moment where the dam's going to break and we're going to see higher interest rates, runaway inflation." And yet, that hasn't happened. We've had some higher inflation, but it looks like we're on maybe the backside of it, and it's related to other factors. But we don't see this high interest rate concern materializing, and I'm wondering, from your perspective in the marketplace, why is that?

Younger: I think it's important to keep in mind that not everybody buys fixed income or yield, so to speak. They can buy it for a variety of reasons that it's not just to maximize the coupon that they earn on it. And so, as a really important example of that 5 trillion that was issued on a net basis over the past couple years, more than half of that went to the Fed, who's obviously not in this game to make money. Well, they're technically in this game to make money, but not that kind of making money. A big chunk of the remainder went to commercial banks, which are also, in a sense, printing the money with which they bought Treasury bonds. They don't do it the way the Fed does it, but they do create deposits.

Younger: When a bank buys a security, it creates a deposit, and that is a form of money creation. Where did the money come from to buy those assets of the 5 trillion? Something like four was simply materialized through that process. And then, of the remainder, there's a variety of participants in the market that are hedging interest rate risk through the purchase of high quality, long dated, long maturity fixed income assets. Pension funds are a great example of that, where they invest in equities and fixed income. And when equities rise, they are fully funded relative to their expected payouts to beneficiaries. And so, they try to lock that in. They defease it by buying things like Treasury bonds.

Younger: And so, rising equities can trigger that. I don't say that because that's the dominant flow. It's just an example of the risk management that can generate demand for an asset that, as you describe, it's not paying a particularly compelling coupon all else considered. But really the role of banks – the Federal Reserve, the central bank being the dominant participant there, but also commercial banks – and QE effect, quantitative easing when the Fed expands their balance sheet, creates demand for treasuries in another sense, which when we look at that commercial bank side – and this is my new day job – it creates deposits in the banks and in the banking system.

Younger: Those deposits are long dated in practice because people tend to keep their money at banks for very long periods of time. So, you can go get your money back any day you want, but you tend not to. That means it's a long-dated liability. That means you have to find long-dated assets to hedge it. So, that demand comes in part from deposit creation on the back of QE. I think it's hard to understate the impact that an expanding Fed balance sheet can have on fixed income markets when it's doing so at such a pace that it has for the past couple of years.

I think it's hard to understate the impact that an expanding Fed balance sheet can have on fixed income markets when it's doing so at such a pace that it has for the past couple of years.

Beckworth: So, you're a big believer in the portfolio balance channel of monetary policy. That's a fair assessment?

Younger: It's a combination of portfolio balance and simply conjuring demand when you create long-dated liabilities, then they need long-dated assets to support them. So, there's a liability-driven model of investment for places like banks and pension funds. There's a portfolio rebalancing channel, but there's also simply the money creation channel. And when that money behaves long dated, then it supports high quality money-like assets that can match the maturity.

Beckworth: This money view that you're describing, I think oftentimes academics will call, liquidity preference theory of Keynes. That's kind of the Central Bank and the money markets determine the interest rates. There's the expectation theory where the short term rates are embedded into longer rates. Maybe that at 2.8% is someone could argue, that's just what the Fed's going to do over the next 10 years.

Beckworth: But then there's also another theory, I want to throw this out too, that's kind of pejoratively called the loanable funds theory, I would call, the desired savings-desired investment theory, more based on fundamentals. So yes, the Fed does set the overnight rate. Yes, it can influence the path of expected interest rates, but it's subject to a mandate that has a lot of weight on price stability. So, it really cares about price stability. Over the long run, if the Fed ignores fundamentals, so real interest rate changes due to productivity, demographics, other factors, it's going to lose control over price stability. So, this theory would say that over the long run, fundamentals are ultimately shaping interest rates, given the Fed has this price stability mandate. What are your thoughts on that story?

Younger: I think that's how a lot of market participants think about it fundamentally. When we look at, say, 5-year forward, 5-year interest rates, which is a good proxy for looking through all the noise of the near term dynamics in the economy and other factors, and saying, what is the long run growth potential? What is the interest rate which is consistent with the long run growth potential in the economy? It's like the Fischer identity. I've got trend inflation, and I've got trend growth, and if I know those things, then in equilibrium, interest rates, which is a 5-year forward, 5-year rate, it's well past the horizon over which I could have a realistic view on anything, should be equal to that.

Younger: Largely consistent with that set of assumptions, the inflation shock of the past year or so has not really moved 5-year forward, 5-year interest rates that much, certainly compared to, say, 1-year forward, 1-year interest rates. If you look at the near term path, that's very much expectations driven. How much is the Fed going to have to hike to regain its control over prices? But when you look out further, the view has generally been, at least of the market, that they are going to be successful in doing so, that the runaway inflationary processes of the 1970s are not necessarily an immediate risk, or not likely. And over time, maybe it's a year, maybe it's two years, maybe it's three, it's probably less than five, we'll get back to something resembling trend growth or potential growth for the economy and inflation consistent with that, and demographics reassert their control.

The inflation shock of the past year or so has not really moved 5-year forward, 5-year interest rates that much, certainly compared to, say, 1-year forward, 1-year interest rates. If you look at the near term path, that's very much expectations driven...But when you look out further, the view has generally been, at least of the market, that they are going to be successful in doing so, that the runaway inflationary processes of the 1970s are not necessarily an immediate risk, or not likely. And over time, maybe it's a year, maybe it's two years, maybe it's three, it's probably less than five, we'll get back to something resembling trend growth or potential growth for the economy and inflation consistent with that, and demographics reassert their control.

Younger: That is the market's view. There are certainly participants in the market who think that the market is wrong on that fact, and the move higher in volatility across all of these interest rates as priced into the options market tells you that the distribution of outcomes is widening considerably. So, we may think, on average, that the Fed will regain control of policy, the most likely outcome, but the distribution of possibilities around that outcome are much broader now.

Beckworth: There's more uncertainty, greater uncertainty, but I like that reconciliation. Short term, money view, liquidity preference view, long run. It's the fundamentals, and I'm glad to hear the market thinks that as well. But let me ask you this then. Going forward and you kind of answered it already, but over the longer horizon, you do think we'll be back in an era of low interest rates. You've mentioned demographics will kick back in, but we'll return to what we had pre-pandemic?

Younger: Well, the market thinks that. One of the benefits of being focused on asset and liability management is having a view on interest rates, but ultimately, your day-to-day is driven more by what the balance sheet gives you. And so, my time is more occupied by sort of having a good view on the risks that the firm has, and how best to hedge those risks and manage those risks, and maybe leave some residual exposure that reflects an expectation or how interest rates will go or how the economy's going to evolve or things like that. But most of the effort, most of the time is spent having a very clear picture of what risks the bank is exposed to as a general matter, and then, in certain levels of detail, and how those risks change as interest rates change.

Younger: That's not really answering your question. Part of my hesitance is that I've actually never really taken a class in economics. I've been in the markets for a while, but I'm certainly not a trained economist. I'm sort of better on those things than the pure economics angle, but as somebody who lists stuff a lot, it seems plausible that, at least over the long run, we're returned to something resembling normal. But in the meantime, the market can move considerably away from that. So, the risk of rates being higher or lower really depends less on what we think trend growth is going to be because the market rarely sticks to a view on that for very long.

Younger: But more sort of how the environment evolves over the near term, because in a recessionary scenario, for example, you could expect tenure interest rates to be much lower, simply based on the fact that the growth is slowing over the near term. It sort of brings up this issue that the market's a fairly poor predictor of the future. And so, it really reflects the current information we have and the risk tolerance of investors around that and hedgers around that information, rather than a really sort of informed view on the likely future outcome of future state of the world.

Beckworth: Josh, we'll get back to the discussion of the structure, and that's the reason I got you on this show. But one last question on this area, and that is after the dash for cash, which we'll talk about in a little bit, after that was done well into the pandemic and we're running big deficits and again, say in 2021, another $1.9 trillion was going to be allocated to the American Recovery Plan. You didn't see the Treasury market panicking, at least that was my impression. Correct me if I'm wrong. You didn't see this fear of, for example, fiscal dominance where the Fed would lose control of interest rates and inflation because it'd be runaway public finance issues. Is that a fair assessment of how the Treasury market took this in stride over the past two years outside of, again, the dash for cash?

Younger: Yeah. I think this sort of freak out scenario, which comes in different flavors, but the freak out scenario that some have called for really hasn't materialized. James Carville, who said, if he came back, he'd want to be reincarnated as the bond market so he could scare everybody. I mean that, bond market's not that scary right now. It's a bit of a conundrum to use another central banker's words, but it doesn't necessarily mean that the market's unaware of these risks. I think you mentioned runaway deficits. I mean, the CBO’s been projecting runaway deficits as long as I've been around.

Younger: So, the long run has been very uncertain for a very long period of time. It sort of gets the market's inability to really look to the likely future state of the world. It's more that these long term rates are a reflection of the current balance of risks and tolerance for duration exposure, exposure to moves in interest rates and how it affects people's bottom line. So, we look at the past couple years. The Fed has been a massive stabilizing influence. They've taken down most of the supply, more than half. Banks, commercial banks have been another significant stabilizing influence and they've taken down another 10, 15% of the supply, and that's been fairly price insensitive, because they just had risk to manage. And then, you had foreign banks were taken down some of it. There were pension funds which were doing risk management exercises.

Younger: If you look down the list of the buyers on a net basis of treasuries over the past two years, a big chunk of them are not really responding to price incentives. They're really responding to risk management incentives. That's more about the fact that they have the same and opposing exposure on the other side of their balance sheets. Higher rates are bad if you've got a liability that floats up with rates. And so, you want to hedge that. You've also got fixed rate liabilities, which increase in value when rates go down. So, we have to think about both sides of the balance sheet.

Younger: There's the asset and liability side of things. And so, that makes you a little less sensitive to the headline level of yield. It's more about changes. Does that mean things are going to be stable forever and we've got nothing to worry about? I mean, that's certainly not the case. That's never the case for anything. You can end up in a situation where instability is build, and we'll talk about 2020 in a moment, but the outcome in 2020, this dash for cash, that volatility episode, which really was the opposite of what most people expected when you had a massive demand shock and economic contraction. That was a consequence of the appearance of stability. But the reality was that there was a freak out. It's just no one saw it because it was being held off stage. And when things actually hit the fan, all of a sudden you can see through all the set pieces and you got a problem. That's the more likely, I think, path for disruption, simply because things don't really happen in his dramatic fashion as we might sort of want them to.

Beckworth: All right. Well, let's get back into the structure of the Treasury market. I read a report that came out, I believe late last year, the interagency working group on the US Treasury market. They broke it down into an outright purchase or sales market or what they would call cash market. They also had a repo market part of the Treasury market, and then the future. So, cash, repo, and futures. Is that a good all-encompassing perspective of the Treasury market?

If you look down the list of the buyers on a net basis of treasuries over the past two years, a big chunk of them are not really responding to price incentives. They're really responding to risk management incentives. That's more about the fact that they have the same and opposing exposure on the other side of their balance sheets. Higher rates are bad if you've got a liability that floats up with rates. And so, you want to hedge that. You've also got fixed rate liabilities, which increase in value when rates go down.

Structure of the Treasury Market

Younger: Yeah. I think that there's sort of a higher level of categorization, if you think of it in trees. You say, on the one hand, I have end investors. They are term funded. Meaning, I either have cash to invest or I have very sticky or long-dated liabilities, that means there's no risk that I need to liquidate my position because my funding is running. So, that is pension funds, that is people, individuals, me. I'm not short-term funded. You're not short-term funded. We have cash to invest. It's a variety of other asset managers and things that tend to have very stable funding. It includes banks because deposits tend to be very sticky. They don't tend to run unless there's some kind of really catastrophic event. That's the end-user market.

Younger: And then, you have the lever market. Lever marketers are short-term funded. They’re just runnable in principle. So, that's repo and futures. Repo is a form of leverage on securities, which means I need to borrow $100 to buy my $100 bond, and I'll pay interest on that 100 bucks. Futures are derivative, which means I don't necessarily have to come up with $100, but I'm getting leverage because I'm getting exposure to a much larger dollar value of asset. My mark to market is going reflect a much larger exposure than the amount of money I have to put, quote, unquote, down. And so, futures are a form of leverage. Swaps are a form of leverage. So within futures, there are swaps on Treasury yields. There are total return swaps like that. There are Treasury bond forwards that are traded over the counter, if not listed. Features market is one component of that broader derivative complex that references Treasury yields.

Younger: But those are the primary kind of venues for trading or the primary types of investors in the market. Their stability is interconnected. That's what we learned in 2020. So, end users buy treasuries in part because they expect to have liquidity. We'll talk about what that means, but they expect to have a deep and liquid market with which to sell them if they have to. This is a high quality liquid asset. And that market, that secondary market, relies on leverage because in the middle or intermediaries who are highly levered on the other side of the trade, if you can't find another end user to buy the bond you're selling, it's probably going to go to a levered participant. So, there's this sort of holding pen, this purgatory, that represents the levered community. And so, that's a critical component of the market, because it facilitates the end users feeling comfortable taking on very large positions.

Beckworth: Now in the cash market, I want to go focus in on that a bit, because what I understand is there's been some interesting changes over the past decade, where dealers used to be the most important part. Now, these primary trading firms or PTFs doing electronic trading, almost like algorithms, is taking over the market. Is that a fair assessment?

High Frequency Trading

Younger: Yeah. Let's maybe take a broader view. There's an end user. The most sort of stable form of end user is arguably a foreign central bank, because they have issued liabilities. They've issued money on the other side of their balance sheet. So, they've bought a Treasury, and they've issued local currency, and that's because they have a current account surplus. They're taking in foreign currency. They need to sterilize the interaction or something like that. So, they have very sticky funding because they have the government, and they're insensitive to returns, and they are funded by the issuance of money. So, they want to sell that Treasury bond.

Younger: Let's say they want a billion dollars’ worth of five-year Treasury notes. What do they do? They call probably a couple of dealers, but not more than one or two, maybe three. Dealers being largely US bank affiliated dealers. They sit within the largest US banks. They literally call. They call them on the phone, voice trading, and they say, "I need to sell you a billion dollars’ worth of five-year notes. What's the price for that?" And so, they get a price and they say done, and now there's a transaction. All of a sudden, this dealer, one dealer, is sitting on a billion dollars’ worth of five-year treasuries. And for every hundredth of a percentage point that goes up and down, they're going to make or lose $50 million. That's not a good situation.

Younger: So, they need to find a way to get out of that position. They need to hedge it. Actually, it's $500,000. That's a lot of money. You can't just sit on that for no reason. Nobody who manages that desk is going to be happy with that level of volatility in their earnings. And so, hedging is not necessarily buying and selling the same security. Ideally, what they would do is buy the billion dollars for the five year notes and sell it from one central bank and sell it to another. And now, they're really acting as an agent, not as principal, and they can make a little spread off that and everybody's happy. They take no risk.

Younger: That almost never happens for a trade that size, especially because a trade that size is probably not the most liquid issue. It's probably a bond that was bought a year ago, so it's not the most recent five year note. It's an old five year note. They're not exactly sure where it's supposed to trade. And so, it's hard to find an instantaneous other buyer for that. So, they hedge with, let's say, the current issue or the future's market. So, they go short. And they go short, not facing one counterparty but several counterparties, because you're not going to just offer to go short a billion dollars’ worth of five-year notes. You're going to do it piecemeal. And so, the other side of that short is probably another dealer. And if you're facing another dealer, you're basically saying I'm socializing the risk. This five year note has a certain amount of interest rate risk. I can't hold it all just me. So maybe, the 12 of us should get together and take a little piece. The way that actually happens through transactions is through hedging.

Younger: Now, that's the old world. That's the pre-2008 world. Now, you get to the new world and that inventory management becomes very expensive, for reasons that we'll talk about. It's hard to hold large inventories at Treasury bonds. And so, there's another component of the market that grows up to facilitate those transfers, to speed them up. That's the principal trading firms, that's the high frequency traders. And so, what they do is they do very, very fast, small transactions all day long, usually enforcing relationships between securities. But in doing so, they're spreading the risk around faster. The socialization time scale for that risk is much shorter today than it was.

Younger: A while ago, those principal trading firms go home with no inventory. They're just sort of doing the trade matching process. They're not doing any of the inventory management process that dealers used to do both. They used to find buyers and sellers, used to hold the bonds in inventory they couldn't sell. So those on a volume weighted basis, those high frequency trading firms are 70 or 80% of the depth of the market on the screens. The amount you can transact, if you were to look at the screen and say, "How much can I sell at the market price?" 80% of those bids are high frequency traders on average on a typical day.

Younger: The market has become heavily reliant on those [high frequency trading] firms to socialize risk among dealers and other levered participants, other high frequency traders in a way that facilitates the broader optical stability of the market. So, when they have problems, we all have problems basically. That is the grease that keeps the wheel turning in the current market structure. And then, on the other hand, if they can't hold an inventory and there's too much risk in the dealer complex, let's say it's a $50 billion trade, not at $1 billion trade, who's on the other side of that? If it's not another end user or not in totality, hedge funds are another outlet. Their time scale's a lot longer than high frequency, but a hedge fund can hold a bond for a day, a week, a month, something like that. They use repo leverage to do so.

The market has become heavily reliant on those [high frequency trading] firms to socialize risk among dealers and other levered participants, other high frequency traders in a way that facilitates the broader optical stability of the market. So, when they have problems, we all have problems basically. That is the grease that keeps the wheel turning in the current market structure.

Younger: So, that's the process by which transactions are facilitated. And ultimately through this exercise, the price will adjust and some other end user will come in and say, "That looks like an attractive buy." They'll buy it, not using repo leverage, and take it out of that complex. So, risk comes in from, say, a foreign central bank. It gets socialized among dealers and hedge funds. And then, eventually, it finds its way back out to some end user. And so, the market represents the steady state inflow and outflow of that process.

Beckworth: In reading and preparing for this show, one of the observations I made is there's a trade-off here between efficiency and resiliency with high frequency trading. So they come in, they do a better job socializing that risk, as you said. They're instantaneous, they're quick, but they don't have the relationships that dealers had, for example. You can't call up your friend. Or if something goes wrong, you can't call up someone and say, "Hey, let's work this out." Now, that means you're less efficient if you're calling someone and talking, you're less efficient, but maybe you've given up a little bit of resiliency. There's a trade-off there that's being made. But this is where we are. This is a trade-off that potentially can create problems. I think we're going to maybe talk about this in a minute with the dash for cash in 2020. But one question I had about this is, was this move to increase use of high frequency trading inevitable? Or was it, in part, a result of the new regulations coming out of the Great Financial Crisis?

Younger: Yeah. This is the market's equivalent of just-in-time supply chains, which function really well under normal circumstances and really struggle under stress. That's what we're finding from the most recent few months. We found it in 2020 with the Treasury market. And so, was this by design? I don't think so. Was it inevitable? Nothing's really inevitable. Fundamentally, isn't it the logical outcome of a series of decisions that were made that significantly changed? The incentive structure for dealers? I think the answer is yes.

Younger: So, what happened in 2008, there were no more independent dealers of significant size. They were all bought by bank holding companies, or they convert it into bank holding companies, or they went out of business. So, Bear Stearns is acquired. Merrill Lynch is acquired. Lehman goes out of business. Goldman's now a bank. Morgan Stanley's now a bank. That doesn't mean there are no independent dealers left. There certainly are, but that was the vast bulk of the market share in Treasury trading among dealers.

Younger: Then the question was, do we want banks to do Treasury trading? And if we want banks to do Treasury trading, and this is an open question, this is something people are still wrestling with. But if we want Treasury trading to happen principally within banks, then we have to make it economical to do so. I mean, this is part of the lesson in the 1950s was if dealers aren't making money, they're not here to do it for free. And so, they need a business model that works, generates profits and does so in a stable and predictable fashion. And so, for banks, there was a trade-off in the regulatory reform that happened.

Younger: On the one hand, the lesson of 2008 is “big as bad,” generically. That was the lesson that regulators learned. It's not just because you have a lot of credit risk. It's that maybe that credit rating isn't what it's supposed to be. So, you had a AAA asset, doesn't look like a AAA asset anymore. Now, treasuries are obviously different than that. But the conclusion was that size, as a general matter, increases the cost of your failure for whatever reason, and those costs should be incorporated into your capital requirement. So, you need to be even more safe and even more sound as a consequence of being big.

Younger: The second issue with that is the stuff that you can't do as easily when that happens is low risk, high leverage activity. That's precisely what Treasury and repo trading is. It's very low credit risk, if any, and it's very high leverage because you're expecting to be able to transact at very low costs. So in a low transaction cost market, it's only economical to be a dealer if you can lever that up many, many, many, many times. Otherwise, transaction costs have to be bigger. I mean, you got to pay the rent somehow. And so, the decision was made to preference size constraints over consolidating that kind of activity within the banking perimeter. That's an ongoing debate. I think that's something that the interagency working group is considering maybe not necessarily in those terms, but the logical outcome of that is scoping in non-banks to provide, in principle if not sort of by design, a lot of the intermediation capacity that the market generically needs.

Younger: I mean, this is a life finds a way moment. To mix a bunch of different metaphors, we could use Jurassic Park, I guess. But there's a natural process here by which an opportunity arises to as a non-bank provide intermediation service and capacity to a market which needs it simply due to its size. That's high frequency trading firms, and that's relative value hedge funds. They weren't doing this on purpose. They were not dealers, but they were responding to price incentives and economic opportunity, and in doing so, the market structure simply changes.

Beckworth: Okay. Let's segue into the dash for cash. We've been talking about it. Let's actually jump into the deep end of the pool on this particular event that occurred in early 2020. The name itself, dash for cash, let me ask this question. What does that mean? How was the market dashing for cash? Was it going for T-bills? Was it going for some other type of asset? What exactly happened that led to that naming?

The lesson of 2008 is “big as bad,” generically. That was the lesson that regulators learned. It's not just because you have a lot of credit risk. It's that maybe that credit rating isn't what it's supposed to be. So, you had a AAA asset, doesn't look like a AAA asset anymore. Now, treasuries are obviously different than that. But the conclusion was that size, as a general matter, increases the cost of your failure for whatever reason, and those costs should be incorporated into your capital requirement. So, you need to be even more safe and even more sound as a consequence of being big.

Explaining the 2020 Dash for Cash

Younger: Yeah. There was this cognitive dissonance in the regulatory framework, which was not particularly relevant most of the time and became really problematic in 2020. When regulators think about different types of financial assets, cash is sort of a poorly defined thing. We talk about cash a lot. But cash means something that has essentially no interest rate risk, no credit risk, and is generally fungible. A dollar's a dollar's a dollar, and it's not going to change in value. This is a Gary Gordon's no questions asked thing.

Younger: There are different vehicles for that, different things that approximate that. They've got reserves at the Fed and paper currency. And outside of that, it's a spectrum. Some things are closer to cash than others. When the regulators were trying to decide how to maintain liquidity in the banking and financial system, that means having a sufficient store of cash and cash-like things, and you have to draw a line somewhere. So, the line was drawn, including treasuries, for a lot of practical reasons, in that category of cash and cash-like things, like the liquidity coverage ratio, which tells banks they have to hold a certain stock of high quality liquid assets which they can liquidate in the event of a run.

Younger: Treasuries are given the same weight as reserves at the Fed is cash. Reserves of the Fed are basically paper money. You don't have to put it somewhere. So, treasuries are treated the same for the ratio, meaning they are given equal weight, which means they're viewed as functionally the same level of liquidity and credit quality. That is reasonable if you can sell those treasuries at a reasonable cost at scale and quickly, and turn them into cash, if they are functionally exchangeable for cash, because it's kind of a bank deposit. The bank deposit's valuable if you can, in principle, go back and get your money. So, the federal government explicitly stands by bank deposits. They don't explicitly stand behind the Treasury market.

Younger: And so, in 2020, there were a bunch of different demands on liquidity. Essentially, everyone wanted to sell their treasuries at the same time. They all wanted cash. Because they all wanted cash, the market had to provide that transfer, that transformation, one direction not the other. So, that was the problem. This works if there's a buyer and a seller. It doesn't work if they're all sellers. Or rather it does work if there are a ton of sellers, and dealers have the capacity to provide that transformation and warehouse the risk themselves. But in the post leverage constraints, post Dodd–Frank world, their ability to grow their balance sheets was much more limited. And so, that transfer, that transformation could be done but only at high cost, not at scale, and relatively slowly. That's the market functioning issue that the Federal Reserve was so worried about. Because it called that cognitive dissonance into very sharp relief and presented in many ways an existential threat to the financial system, because it had been oriented around this assumption that treasuries were cash-like. And if they suddenly weren't, that has a lot of similarities to a traditional bank run.

Beckworth: My understanding is investors sold long-term treasuries, off the run treasuries, and they fled into money market funds into Treasury bills. I guess my question would be that they also try to run into bank deposits. I mean, there's a limit to what they can do. There are bank reserves. I guess the feds stepping in provided a lot of reserves would be maybe one way to look at that. But literally, they were transferring one former government liability for another, and it just happened to be, they wanted the short end ones, the really liquid ones.

Younger: Yeah. They want the more money-like things. So in that spectrum of money-like assets, where long term treasuries were considered functionally equivalent to short-term treasuries and reserves at the Fed, for a lot of reasons. This is not just banks, this is everybody. I mean, the standards that we apply to the banking system tend to be adopted by broader segments of the financial market and system more generally. So, they're just trying to get closer to money. Closer to money means something closer to literally the paper money in their wallet. So, how close can I get to that? And that process broke down.

Younger: Now, it's important to also keep in mind this wasn't just a dash for cash driven by need for liquidity. There were a lot of rumors flying around the market was going to close, just to actually shut down for some indeterminate period of time. A lot of folks remembered the experience of September 11th, 2001, when the Fed wire shut down for a period of time. Fed wire's the transfer system for reserves. So, you couldn't transfer your cash, or you couldn't actually call your repo dealer to unwind your position. So, there are a lot of operational concerns in early March that significantly accelerated the process that would otherwise have been happening, but was much more acute based on the nature of the event.

In 2020, there were a bunch of different demands on liquidity. Essentially, everyone wanted to sell their treasuries at the same time. They all wanted cash. Because they all wanted cash, the market had to provide that transfer, that transformation, one direction not the other. So, that was the problem...it does work if there are a ton of sellers, and dealers have the capacity to provide that transformation and warehouse the risk themselves. But in the post leverage constraints, post Dodd–Frank world, their ability to grow their balance sheets was much more limited. And so, that transfer, that transformation could be done but only at high cost, not at scale, and relatively slowly.

Younger: Because we've seen a lot of dashes for liquidity before. We've seen a lot of dashes for cash before. It certainly happened in 2008, but this was different in the sense that it was very accelerated. Everything about this pandemic has happened really quickly. This is exponential growth, again, that people are just really bad at intuition around exponential growth. So, it happened again, and the market simply could not provide that transformation capacity at that speed and at that scale.

Beckworth: Let's say for the sake of argument, in March 2020, that all the reformers we'll talk about in a minute, tweaking the supplemental leverage ratio, standing repo facility, expanding the capacity of these firms to intermediate in such a setting. Even in an ideal world like that, couldn't it still be the case that the pandemic, this once in 100 year event, the accelerated timeframe, the scale of the liquidity demand, would've stressed even an ideal system?

Younger: Yeah. I mean, it should. It's not the thing to solve for, because like you said, this is a once in 100 year crazy event. We shouldn't have a market that can withstand widespread speculation that will shut down tomorrow for an indeterminate period of time. That's not a reasonable expectation of a market structure. Now, I think of 2020 as not presenting a problem to be solved as much as highlighting concerns that we're already there, but much easier to kind of ignore. So in 2014, we have the flash rally where Treasury yields go down 40 basis points in 15 minutes to come back. I actually was going to a meeting and missed the whole thing, because it was only an hour. I came back and I was like, "What happened?" You missed the craziest thing I've ever seen.

Younger: Then, we have 2019 with the issues in the repo market, which is fundamentally different than what happened in 2020, but still relevant. More generally, people felt that there were these air pockets in trading that seemed to open wider when things were more stressed. But it was easy to brush aside and say, hey, that was a weird open on Tuesday, or this 2019 repo thing was all about intraday liquidity and we'll just grow the reserve balance. And now, the reserves are ample again and tax day isn't for another year. There a laundry list of reasons why now was not the time or this was not a sufficiently stressful event. And so, what 2020 does is it really reveals the flaws in the system in a very naked way. They're worth fixing, even if we aren't hardening the system against that sort of event.

Beckworth: Okay. Let's talk about current stresses, if any, on the Treasury market. One thing that comes to mind would be quantitative tightening. At least one potential candidate would be the Fed shrinking its balance sheet and not buying up additional rounds of treasuries. Any thoughts there? Has that provided any stress on the market or any other source of stress currently on the Treasury market?

Current Stresses on the Treasury Market

Younger: Well, I think it's an interesting contrast. The market is quite illiquid by any measure and relative to history. So, you look at the depth of the market. You can look at the breadth of the market. You can look at turnover versus depth. You can look at any sort of ratio that people tend to look at. You'll find that it's harder to trade now at scale and at reasonable cost and quickly than it was before. But there's a big difference between illiquidity and market functioning, because the market is functioning, it's just illiquid. One would expect, under these circumstances, the market to be illiquid.

Younger: There've been periods of illiquidity before, which have not caused the Fed to buy $2 trillion worth of assets in a few weeks in the summer of 2011 and in 2013 and 2015. There are these recessionary scares, the devaluation of the yuan. You have these whole range of things. There are stress events in markets. That's fine. That's what they're for. I mean, they're there to facilitate price discovery. Prices are not always known very specifically. And volatility is the process of finding the new clearing level. Sometimes, that's hard to do without a lot of bumps along the way. This environment is somewhat different because, as we were talking about before, the fundamentals are in play. It is still an open question as to whether long-term inflation expectations will remain well anchored. Now, some people would argue they will for a variety of reasons. Some people argue they won't, but it's certainly an open question now. Whereas five years ago, there was no debate as to whether that would be the case. So, the Fed's credibility is to a greater extent on the line now than it was before. That should generate the volatility in long-term interest rates.

Younger: But the market is functioning in the sense that transaction costs for trades that actually clear are reasonably small, consistent with norms. Dealer inventories are not particularly elevated, which is really important because it means you're not seeing those end users liquidate. You're seeing a bunch of hedge funds compete against each other to see who can call the top end yields as opposed to foreign central banks liquidating large blocks. No Treasury securities. So, that's going to generate headline volatility but it's not going to break the market necessarily.

Younger: And if anything, the treasuries is looking overfunded for the next bit. I mean, that's what the refunding suggests. So, the deficit outlook is improving. It's still wide, but it's improving. Now, net of Fed runoff, net of the impact of QT, the story's a bit different, but the market looks like it's capable of absorbing that risk transfer from the federal government and the federal reserve to the private market. That interest rate risk transfer without, quote, unquote, breaking.

It is still an open question as to whether long-term inflation expectations will remain well anchored. Now, some people would argue they will for a variety of reasons. Some people argue they won't, but it's certainly an open question now. Whereas five years ago, there was no debate as to whether that would be the case. So, the Fed's credibility is to a greater extent on the line now than it was before. That should generate the volatility in long-term interest rates.

Beckworth: Okay. One related area to this is the liability side of the Fed's balance sheet. I bring this up just to speak about the Fed's footprint and where it is. If you look at the reverse repo facility, over $2 trillion there, the Fed's becoming a big counterparty in the repo market. Any concerns there? Is the Fed becoming too large of a presence, or is it making the system less robust, resilient because people come to expect the Fed to be the counterparty?

Younger: Well, there's a couple different ways to answer that question. The reverse repo facility generates a lot of controversy, which is sort of hilarious to anyone who's been in the market for a long time, because it was such an idiosyncratic backwater for so long. And now, it's on Reddit, and they're talking about GameStop and the RRP. But everything's cool eventually, I guess. The RRP, some people view the RRP as excess liquidity in the system, which is fair, but they go a step further and they say, "Well, this is a rolling report card on the efficacy of QE, because the bigger the RRP is, the more the Fed has overdone it, because this is liquidity that shouldn't be in the system in the first place. This is the drain. This is stuff that doesn't want to stay in the banking system that's rolling into the repo markets. And so, it's definitionally distorting market functioning in the money markets." Bigger is bad for the RRP.

Younger: Another school of thought is, this is sort of the Fed's facilities performing as they should, which is on the one hand, we have no good model for precisely how much liquidity the system needs. We need a good outlet. The quantitative tightening process is generically the reverse of quantitative easing, but it doesn't mean it's sequentially reversed. So, when quantitative easing was happening, first bank reserves went up. Then, the RRP went up. That doesn't mean with QT, the RRP should go down first and then bank reserves go down. I mean, that depends on who, at the end of the day, is buying those treasuries. So, the RRP's doing what it's supposed to be doing because it's keeping a floor on short term rates.

Younger: I think that's critical in this environment, because what the Fed's trying to do fundamentally is tighten financial conditions to fight inflation. The base rate for most lending now is tied to the repo markets, the secured overnight financing rate. So far, that's after LIBOR went away. That's the replacement we got. That's very closely tied to the Fed's policy rate, the Fed funds rate. And so, if the RRP wasn't there and the RRP keeps the foreign repo rates, then the Fed could be hiking 75 basis points a meeting, and interest rates that really feed through to the economy could be moving much less than that because of the excess liquidity in the system.

Younger: So, you need the RRP to implement monetary policy and to get that kind of tightening and financial conditions that can actually bring inflation under control. I think it's critical to what the Fed is doing, and it's largely doing what it's supposed to be doing. That at large simply reflects the fact that there is a lot of liquidity in the system. It's being drained at a reasonable pace. And in the meantime, we need money markets to behave relative to the Fed's policy expectations.

Beckworth: All right. Let's move on to proposals to reform the Treasury market. Let me start with the standing repo facility. What difference would it make? It's already been established, they're onboarding participants, they have primary dealers, they're getting depository institutions, but what difference would it have made, say in 2020 or 2019, in those crises?

You need the RRP to implement monetary policy and to get that kind of tightening and financial conditions that can actually bring inflation under control. I think it's critical to what the Fed is doing, and it's largely doing what it's supposed to be doing. That at large simply reflects the fact that there is a lot of liquidity in the system. It's being drained at a reasonable pace. And in the meantime, we need money markets to behave relative to the Fed's policy expectations.

Reforms: SRF, SLR, and Increased Central Clearing

Younger: Yeah. I think the standing request facility, the SRF, is really interesting tool. It's effective as it varies with the environment. So, in an environment where banks are size constrained, and they're typically size constrained under periods of stress when the Fed is growing their balance sheet, when QE is happening. 2020, that's a great example of that. Under those circumstances, it's not particularly effective because when you go to borrow money from the SRF, you're guaranteed the rate at which you can borrow but it generates new capital requirements which affect your cost of funds. So, you have to support those new borrowings with new capital, or you have to allocate capital with other UIs use for other activities.

Younger: The thing to remember about leverage constraints is it just turns the whole business into a zero sum game. So, I'm the Treasury desk. I need funding at the SRF. The repo market's too rich. I can't afford to borrow in the private market. I'm going to go to the Fed. And I also have all these new trades I need to fund. And then, you got to go to the equities desk and say, "Hey, you got to do less." Or you go to the credit desk, "Hey, you got to do less because this is zero sum game. There's only so much to go around and I'm more important on treasuries." And then, the prime brokerage business says, "Well, I make more money and I need balance sheet."

Younger: So, you generate those kinds of arguments in a leveraged constrained world. That means that going to the SRF, especially for new business, is difficult. It provides some ceiling on repo rates, because if it's not new business, it's all business. At least I have a uneconomic participant who's giving me an offered rate. It's administered and not market determined. And so, that provides some bit of a ceiling, but if it's new business, maybe I can't do it. Maybe I can't do that business. It's got to go somewhere else.

Younger: If you're risk constrained, then these are riskless transactions. It's short term lending secured by Treasury bonds, overcollateralized with Treasury bonds. And so, that doesn't contribute to my risk based capital. And so in a risk based capital constrained, we're all like 2019, then I'm a lot less affected by that consideration. It would've helped in 2019. I don't know if it would've completely solved the problem, but it certainly would've helped. So in that sense, it's productive. You're putting a ceiling on short-term rates in addition to a floor. The reverse repo facility is the floor. The standing repo facility is the ceiling.

Younger: And as we have moved into a world where repo rates are central to all kinds of lending through benchmark reform, LIBOR is gone. Now, we're all exposed to repo. I mean, mortgages have repo exposure. Commercial term lending has repo exposure. Credit cards are probably going to have repo exposure to some extent. Auto loans. The new benchmark rate is a repo rate. And so, controlling its behavior within the corridor rates that the bed is comfortable with is really important. So, I think it's a good tool. I just wouldn't think of it as a completely perfect solution to every problem. It's useful in that sense, but not necessarily sufficient.

Beckworth: Well, one takeaway from what you just said is the Fed needs to drop the federal funds rate as its target and have that repo rate as a target for monetary policy instead, because it is such an essential part of the money market. This also sounds like what you're suggesting is that the standing repo facility would work best if we also had another reform put in place, and that is tweaking a supplemental leverage ratio so that reserves are removed from the denominator, maybe even treasuries. But that sounds to me like a better mix right there, having both of those in place. If that had been the case, would that have played a role in 2020?

Younger: Well, excluding reserves is I think the most straightforward argument to make, which is conceptually they're riskless, fungible central bank liabilities. I mean, we always talk about things in dollars. Reserves are dollars. So, it's not just the unit of account. It's the thing that generates the unit of account. And so, in that sense, supporting that with capital seems a little funny. You could still come back and say, "Look, big is bad. I don't care what it is. You need to support all of your size with capital," but just simply because being a bigger institution means you're more systemic and therefore the cost of your failure higher, and so, that should be capitalized. But from a fundamental standpoint, as a thing that generates the risk of failure, reserves are not on that list.

Younger: And so, the most important part of that is the reserve balance. The amount of reserves in the system is also not up to banks. That is not the bank's decision. That is the Fed's decision. That is a monetary policy decision. And so, when the Fed grows their balance sheet, they don't necessarily want to increase the capital requirements of the banking system because that forces one of two things. One is you have to raise a bunch of capital, which increases your cost of funding, because you're getting low-yielding riskless assets. And as to issue expensive preferred or preferreds or equity or some form of new capital, or you can stop doing stuff you would otherwise do. That's not appealing either from the standpoint of easing monetary conditions in financial conditions.

Younger: So when the Fed grows their balance sheet, they want that to pass through to the real economy effectively. That means not necessarily tying bank capital requirements to the size of their balance sheet. So, I think there's a good conceptual argument there. Treasuries are potentially a tougher call because they do generate interest rate risk. They do have liquidity risk within that in 2020, they are credit risk free and there may be compelling public policy reasons to preference them in bank regulations. We already do that in lots of different ways. When Glass–Steagall was passed in 1933, they specifically allowed for dealing in Treasury securities within banks because they wanted to preference the government securities market. So, there's lots of ways in which government bonds get special treatment. Maybe leverage is one of those places where it's appropriate.

Younger: But all of this is an ongoing debate, and it's all happening in the context of international standards. That's where the leverage ratio comes from, which are somewhat restraining. I mean, the Basel Committee says you can carve out reserves on a temporary basis for the purposes of implementing monetary policy, but you have to raise your minimums at the same time. So, you can't release capital through changes to the way you measure leverage. That's the international standard. Now, obviously, that's not always followed, but it does create sort of political noise around trying to deviate from those standards. And at the same time, leverage after 2008 is a tough thing to deal with in a public forum. That is something that is a very complicated issue for a lot of policy makers. I think that's why the Fed has said they want to recalibrate this thing. It's just not happened, because it's very difficult to come to a consensus on precisely what to do, that balances all of these interests.

Beckworth: Yeah. I will note the Bank of England recalibrated its supplemental leverage ratio and the Fed, I think, should follow suit. I'm very sympathetic to this proposal here. I've mentioned many times on the show before, but I'll repeat it, and that is when the Fed introduced this at 2014, 2015, the expectation was that its balance sheet would be smaller, gets very small. The Fed wasn't expecting the supplemental leverage ratio to be the constraint that it was during 2020 when its balance sheet got really large, as you mentioned, and it required capital. So, it did temporarily release that ratio. The other one I want to bring up is increase central clearing for the Treasury market. That's received a lot of attention and a lot of buzz. What are your thoughts on that?

Younger: I think it's an interesting proposal. It has a lot of potential benefits and not all of them are related to leverage and binding constraints on bank activity, but a lot of them are. So, I think there's a lot of operational gains from things from central clearing. There's a lot of settlement risk that could be reduced in the market. So, the risk of failures to deliver, for example, would go down. But ultimately, the way I like to think about the question is, would this have meaningfully addressed the, quote, unquote, problem in 2020? Just because something has other benefits doesn't mean that it is a broad solution to a very acute problem. And so, if we want to introduce central clearing among the reasons to do so, and from my perspective at the top of that list, is a meaningful reduction in the way we measure leverage in the banking system.

Younger: Because what central clearing does is it said there's functionally one counterparty in the repo market. And when there's one counterparty in the repo market, from an accounting standpoint, you can net otherwise unnettable positions more efficiently. What does that mean? It means if I'm long with the left hand and short with the right hand and I'm facing two different counterparties, I'm economically flat. But when it comes to reporting my leverage, one of those sides contributes because it's different counterparties. When you have essential clearing model, there's only one counterparty at the end of the day. So, those trades would net and there'd be no contribution to my reported size.

Younger: And so, you can do a lot, in principle or in theory, more business in the repo desk without actually increasing your footprint from a leverage exposure standpoint. The way that exposure measures is calculated. So, that's directionally really interesting. The rub there is we're already kind of doing this, and actually we're doing it a fairly efficient way because clearing is open to non-banks, and non-bank can participate in central clearing of treasuries in repo. They do it by sponsorship. There's a sponsored service run by the largest clearing house where clearing members, banks, can sponsor non-banks, like hedge funds and money market funds and others, for membership. That comes with a certain set of obligations.

Younger: But at the end of the day, that means their trades ultimately get novated to the central counterparty and they get all the benefits of this clearing mandate that we're describing. The incremental gain from a broad clearing mandate that scopes in everybody, relative to the current market structure, which has many participants already clearing their positions is pretty modest, unless you make really invasive changes to the way the market trades. And so, while directionally beneficial, our view is that this is not the place to stop. There are certain aspects of it which are kind of low hanging fruit, but we should give credit where credit is due, which is the sponsored model. A three component model for the repo market where you have bilateral trades, you have tri-party trades, and you have sponsored trades. I mean, that's actually serving the market quite well.

Younger: And so, in that context, and that was all in place in 2020, clearing would not have avoided the most disruptive elements of the dash for cash. We still would've had market functioning issues. We still would've had difficulty maintaining stability in the repo and Treasury markets, and the Fed would've still likely had to intervene at something like the scale that it did. So, that doesn't mean it's not worth doing or considering, but it does mean that it is not the way to ‘solve the problem’ in a vacuum.

Clearing would not have avoided the most disruptive elements of the dash for cash. We still would've had market functioning issues. We still would've had difficulty maintaining stability in the repo and Treasury markets, and the Fed would've still likely had to intervene at something like the scale that it did. So, that doesn't mean it's not worth doing or considering, but it does mean that it is not the way to ‘solve the problem’ in a vacuum.

Beckworth: Of the three that we've mentioned, the supplemental leverage ratio is being examined under review right now, the standing repo facility is actually implemented, and this central clearing proposal, is there any work being done on it or is it just an idea somewhere?

Younger: Yeah. The interagency report calls for study. I think that the Fed has been very clear that they're considering this as a possibility. Like many good ideas, it's worth considering. The way I approach it is, this is a directionally a good idea. Does it go in the right direction of solving the problem? The answer there is clear. Yes. But on a magnitude basis, quantitatively, what is the benefit? How much balance sheet is actually going to get released? How much elasticity are we actually going to generate for the Treasury market? Is that commensurate with the demands of not necessarily 2020, but a reasonably significant shock that we think should be handled well by the market structure? That's very much not clear.

Beckworth: But it sounds like there's just a study being proposed on it as opposed to actual policy action being done. The Fed is actually reviewing the supplemental leverage ratio. We actually have a standing repo facility. But, at best, we're just doing another study on central clearing. Is that what I'm hearing?

Younger: I'm not aware of nor could I speak to the details of what they have on the table. However, I think studies don't go on forever, I guess is a good assumption. We've been talking about this for a while, so presumably there's concrete work being done.

Beckworth: Yeah. Okay. Now, there's a few other things that have happened in terms of reforming the Treasury market that the SEC has new suggested guidelines. And then, the Treasury Department has proposed also new transparency and disclosure rules. Any thoughts on those recent proposals?

Younger: Regulators having a better view of what's going on in the market is, generally speaking, a good thing. So, transparency facing the regulators, I think, is something we can all to some extent support. Transparency as a general rule is not necessarily bad. I think the key is, who is the transparency benefiting? In this case, we have a really interesting case study from the derivatives market, where in Dodd-Frank, there was a move to have public disclosure of derivatives transactions. The view was the market will benefit from public dissemination of pricing and transaction activity and details, because everyone will know you paid a dollar for this. I paid a dollar 20, what gives kind of thing. In the end, that didn't necessarily benefit the clients. It likely benefited the intermediaries, because if there was a transaction that say you were called about and didn't actually print, and then you see it hit the public disclosure, you can get a good sense of where that price is.

Younger: And so, you get a lot of market color on pricing as a dealer, because in the end, it was really about knowing what you were looking for, rather than sort of broadly scraping and doing data science magic on the data repository that was created through this process. So, I think at the end of the day that the information is always valuable. It's not necessarily valuable to the public or the clients. It could be valuable to the dealers. And in the case of the Treasury market, there's a lot of trading, particularly among the end users who are so critical to its stability in relatively off their own things that don't trade very often are pretty illiquid.

Younger: The ability to transact in those at a reasonable cost relies, to some extent, on the ability to do it quietly. And if there were to be real-time dissemination of that information, it would make it much more difficult to trade those less liquid securities. So, is that benefiting the market as a whole? Not necessarily. I think that's the key is, again, the information is always valuable. The question is whether or not it's valuable in the public interest as opposed to valuable to some component of the market that the government may not want to support.

Beckworth: Well, with that, our time is up. Our guest today has been Josh Younger. Josh, thank you so much for coming on the show.

Younger: Well, thanks so much for having me. It was a lot of fun.

Photo by Karen Bleier via Getty Images