How The Treasury Decides Where To Borrow | Steve Hou on Maturity Issuance Policies of U.S. Treasury and Equity Index Construction
01:46:14 Share

The ECB is ready to do whatever it takes to preserve the euro. And believe me, it will be enough. Thank you. Let's close the door. So happy to welcome to Monetary Matters, Steve Ho, Senior Researcher in Equity and Multi-Asset Indices at Bloomberg. Steve, great to see you. Welcome to Monetary Matters. You as well, Jack. Nice to finally meet you on the podcast.

So, Steve, I want to talk about your work that you've done on the government bond market and how the U.S. Treasury funds itself by issuing debt and where issues on the curve impacts term premium. Because this is an issue that when you originally wrote your PhD thesis years ago, wasn't receiving that much attention. But this topic has really blown up as the U.S. debt has continued to balloon skyrocket exponentially.

mainly in 2020. So I was just reading it. Walk us through a few of your claims about the U.S. debt. I guess let's start with

How does the US Treasury think about where it wants to issue on the curve, whether it wants to issue with one month or a three month Treasury bill, or a 10 year Treasury note, a 20 year or a 30 year Treasury bond? How does it go about making that decision? I think that's something that's poorly understood by me, and probably a lot of people in the audience.

This was my PhD work at the University of Michigan. And as you said, back in 2017, the motivation at the time was the Fed was going to unwind its balance sheet. And there was worry about what this might do to SMR markets and that motivated to understanding so that effectively this will be unloading a bunch of treasure securities and how that might affect, you know,

So at the time, I was struggling to get people's attention on this topic because at the time, inflation was not anywhere on the horizon and government debt was large. People were worried about other sort of

top of mind issues, including the fight Trump was having with Powell. And in any event, the Fed did not proceed with its, you know, sort of unwinding. So that whole issue kind of went away a little bit. But my work at the time basically focused on two aspects.

In one aspect, the main chapter of my dissertation, the first chapter, I was asking when is the supply effect large in the government bond market.

And just very quickly, the intuition there is basically, the finding there is basically to the extent that one thinks about the Fed's, you know, summer portfolio being unwound and outright sold, you're talking about a bunch of debt being unloaded onto the private market, you know, and then it has to be absorbed by investors. And to the extent that you think about investors' consideration about these bonds,

they're thinking about not just the risk, thinking about correlations, right? How would the bond affect the overall riskiness of my portfolio? Which depends on how bonds co-vary with other assets, right? So when you think about it intuitively,

The findings should come as no surprise to anyone that the price for absorbing supply from the private market's perspective depends a lot on the extent to which bonds co-vary with stocks or other real assets for which stocks is a proxy. And stocks and bonds, if they are positively co-varying,

then bonds marginally were at riskiness to the overall asset, along with the portfolio, and you would expect investors, arbitrageurs, demand a higher fee for inventing the bond supply issuances. And to the extent that stock and bonds were negatively correlated between 2000 until just before the pandemic, you had this 20-year period

during which bonds were sort of essentially negatively correlated with stock returns and very attractive as a sought-after safe asset. And this was essentially spelling out a career of Bill Gross and Ray Dalio, the triumph of risk parity and bonds just 10 years ago in one direction.

In this environment, you would expect supply to have a more muted effect on bond yields. That's basically the main finding of my first chapter. I'll pause there to let you ask some questions before I talk about the second chapter, where I study the realistic rules with which the Treasury has issued debt, which you also alluded to.

So the Treasury aims to issue debt

and funded the government at the lowest cost to the taxpayer. And for a long time, I, and it's kind of intuitive, thought that, okay, an upward sloping yield curve, like overnight money is at 3% and the 10-year is at 5%, the Treasury would fund itself in overnight and in shorter-term bills. If we have an inverted yield curve, which is rare, but happens, such as we had in last year, let's say, the Treasury would want to fund itself longer-term. But actually, the

The academic work and what I've learned is actually true is the Treasury assumes that the yield curve is kind of correct. So when you have an upward sloping yield curve, issuing at 5% and the long end doesn't save money because ultimately what the curve is telling you is that short term interest rates are headed to 5%. So really what the Treasury cares about is the supply effect. Basically, how much is it going to

move yields up if it issues in that tenor, I guess. So if it issues 10-year treasury notes, how much increases the supply and holding demand constant, the price should go down and yield should go up, but how much? And that's called the supply effect. Your work found actually that how much the yields go up based on supply, in other words, the elasticity, I guess, of that

depends on stock bond correlation. And if stocks and bonds are negatively correlated, meaning that when stocks crash, bonds rally, bonds are a really good hedge for a portfolio. And actually the supply effect can be negative, meaning that the more the treasury issues, the higher the prices and the lower the yield is. And that's counterintuitive, but that is really interesting. So in a world like in 2020, where bonds, you know, in March 2020,

long-term treasury bonds, there was a time where they actually sold off and that was very scary. But ultimately, the 10-year I believe did get to below 1%. I think the intraday reading of 10-year was extremely, extremely low. They ultimately proved to be a good hedge.

And so yields would go down, the supply effect, your work would say that that was negative. But in a time where stocks and bonds sold off together or rose together, and that's what happened in 2022, which is quite scary for the 60/40 diversified long only beta portfolio, let alone quantitative strategies that are heavily leveraged and rely on exploiting these correlations such as risk parity, which of course includes commodities and other assets as well,

the supply effects can be very, very positive, meaning that the more you issue in the 10 year, the 30 year, the more those yields rise. Yeah. So, I mean, as with anything, you know, sort of, you know, you are talking about an equilibrium outcome, right? You know, there is the demand side and there's a supply side. I mean, the two chapters of my dissertation actually sort of dealt with, you know, the two angles a little bit, right? The first chapter,

sort of studies, you know, a way in which, you know, change variations in demand sensitivity will affect, right? How the result in the shift in the supply curve, right? That's, you know, the first chapter. In the second chapter, I document a bunch of stylized facts, right?

of what has historically summarized the way in which the Treasury has approached debt issuance, which is a very little studied and on the subject for the longest time is boring. Nobody really thought about this subject. It's a tiny little office manned by a few people in the Treasury Department and the Debt Management Office.

And you're talking about trillions and trillions of marketable securities out there. The Treasury has very limited capacity to be tactical in a meaningful way. It's not a sort of hedge fund. It doesn't really have the ability to respond that quickly to movement in prices. In any event, it is not really...

that sort of, I think, are opportunistic when it comes to short-term price fluctuations. So what I found was that the four stylized facts is that about half of the decision for them is just adjusting how much to issue in terms of with bills, right? There's the extreme short-term money instruments between up to one year of maturity and then the rest of it.

And then the second fact is essentially that they have a sort of strong desire to spread the issuance of long-term debt that's above one year across the tenors to maintain liquidity for different tenor points. And all of this is always constantly communicated with the private markets, which are represented on the TBAG or the Treasury Borrowing Advisory Committee.

And then there are two more facts is that the weighted average maturity of issuance, which is their marginal decision, tends to lengthen in response to rising funding needs, which comes in two sources. One is technical, one is more structural. The technical one is if they have a whole bunch of long-term debt that's maturing. So that tends to have a temporary impact on

issuance maturity tends to sort of push that out a little bit, but then it comes back down. It's a bit of a ripple. But second one is more structural where you have a big shock to the federal deficits, especially after a major recession where the government tends to run up deficits to stimulate the economy and perform countercyclical macroeconomic stabilization.

And then the fourth one is actually sort of a little bit intriguing. It turns out because of the first three factors, especially the last one, there is actually historically a positive correlation between issuance maturity and the term spread where the government, you would think they would want to issue more long-term debt when the long-term yield is relatively low, but they have actually historically regularly issued exactly when the term spread or the term premium is highest.

Now, so make that what you will, but I think recently we had a little bit of political controversy that was brought up about how the Treasury has operated since Q3 of last year. And as far as my research is concerned, which was actually quoted by Steve Maran and Dr. Rivini in their paper, it's actually quite consistent with how historically they have conducted operations.

Okay, so...

When there's a recession, or let's call it macroeconomic shock, debt to GDP mechanically increases because of automatic stabilizers. People who get laid off, they go on unemployment, automatically the deficit increases. Their counter-cyclical tax cuts, automatically the deficit increases. That is typically in a Keynesian thinking viewed as a good thing to combat recessions. And so the treasury, when there's a recession, issues on the short end. Why? Because short-term treasury bills are kind of like money. So it's just...

not taxing the financial system that much as opposed to funding at the 10, 20, 30 year lend, which requires a lot more balance sheet? So generally it's really more just market demand, the catered to demand. Right. So out of usually in the middle upon impact of a recession, there is a lot of demand for liquidity in the economy. Right. And the Treasury tends to issue a lot of bills upon impact.

to cater to that demand and also sort of harvest some of that liquidity premium. But I think that's not the primary motivation. It's just, I think there's just a lot of demand and they need it.

So the treasury, like I wrote in my second chapter, this little simple model where the treasury is always balancing more or less two factors, right? Sort of the opportunity to earn some liquidity premium, but then at the risk of rollover risks, right? So how much are you going to sort of keep in your extreme short term short-term maturity that's just rolling over all the time?

If you think about all the company agencies in the US economy, of everybody that is out there borrowing, the US government probably has the highest capacity for bearing rollover risks. It's not some little company that has to go out there and borrow. It has that nominal credibility.

So what has to through a business cycle, what typically happens is that upon impact of a recession, economic shock, tragedy, issues, a lot of bills. And then as we exit from the recession, at which point the yield curve tends to only invert itself. Normalize. The yield curve steepens.

And then, you know, that's also when deficit to GDP or debt to GDP has run up a lot, right? During this period, the treasury tends to lengthen the maturity of his debt portfolio to bring it towards this, you know, sort of target historical weight average maturity of the entire debt portfolio, which is why you see this

interesting pattern. And indeed, you saw this again, by the way, after the COVID recession, where the weight average maturity plunged momentarily, and then it bounced right back. And in fact, peaking is, I think,

sort of 40 year high at the end of the third quarter last year. If you think about the treasury as just, if I put you Jack in that position of say, so, hey, we have a huge amount of borrowing, manage this carefully, we want to sort of borrow cheaply, but also don't screw up, right? Don't go crazy.

you would do something along this line where you maintain roughly a reasonably long average maturity so that not all of your obligations are maturing all at once. So they're a little bit spread out.

And you try to do it in a moment that is not causing too much market stress because demand curves are downward sloping. And there is also a market demand for liquidity across the different tenors, especially at the key sort of maturity tenors that markets like to trade. 10 years is a key tenor point, of course, 257.

So that being the case, what you end up doing is more or less where you sort of

tactically in the moment of shock may issue a ton of bills that will raise your rollover risk at least momentarily. And then as the crisis pass, you will basically start migrating more towards the long end. So as the short-term bill matures, you will try to issue more and more long-term debt going on the curve.

And what is typically fortuitously the case is that upon impact of a bad recession, especially the entire curve level tends to be lower. So the slope doesn't really matter. It's kind of second order. So the whole level shifts down and then you move to initially move to the short end and then it is the curve.

steepens, but still generally at a relatively low level, you can move further out the curve. So your entire net portfolio is actually spread out, sprinkled out, along the entire maturity curve, but still at the generally affordable level. That's what happened after COVID, exactly what happened, right?

The recent controversy is only that sort of, you know, the entire curve also had a re-rating because of the whole sort of, you know, regime shift with inflation that we're now facing a very,

high curve, right? I think that's part of the problem that we're facing now, right? Thank you, Steve. So we'll get into the controversy about activist treasury issuance, whether the Biden-Yellen treasury is manipulating the treasury market, not my words, and that the US debt was downgraded from AAA to AA, I believe. But just to put a point on it, so rollover risk is you borrow money in short term, you have to pay it back, and you

you struggle to refinance that debt. So in the private market, that happens very frequently. For real estate developers, for example, they borrow short term, they have to pay the money back and interest rates shut up. They may have a hard time with that. But the US and other governments around the world may have a hard time with that. But Steve, the US government, does the US government effectively face rollover risk? I feel like

Going back to a time where the US government, like can the US government borrow money? I mean, you'd have to go back to like the Civil War, I feel like, you know, like the US had credit for well over 100 years. Am I wrong? Yeah, exactly. Which is why I actually in my paper, which I will actually put up finally, the second chapter, I put rollover risk, like I put quotes around rollover. I mean, the US government fundamentally as an entity doesn't really have a very meaningful amount of rollover risk.

It's just, I think psychologically and maybe institutionally unpleasant to be saying we're going to be issuing 20 trillion or 30 trillions of T-bills all at once every few months. I'm making up some numbers here, but you get the point. It's more just institutional prudence. I would say more than anything,

there is a market demand for the treasury to be issuing debt further out of the curve more than the treasury itself really needs to be sort of doing so much

issuance, you know, all these random tenets. In fact, I've always been pretty critical of the decision of issuing tips. It's quite expensive. And it seems to me like the entire motivation seems to have been motivated by a desire to observe market implied inflation expectations, which seems like a very expensive way to do it by issuing real commitments. But in any event, yeah, I mean, to your point,

I think the role of risk is kind of a, in quotes, soft one. Generally, you know, is the idea is those four facts that I pointed out. Right. And in fact, you find that to that exact point.

This is why Treasurer has actually felt quite comfortable historically having a fairly large amount of debt portfolio at very short maturities because the liquidity and relative safety of his confidence in being able to renew his debt. So, if we put rollover risk aside as kind of not the real reason,

What are the reasons that the US Treasury wants to borrow long term if not for rollover risk? You said that the market can want it, but if that is the really only reason, isn't that kind of not a good reason? Because just because hedge fund managers or the private markets want something doesn't mean the government has to give it to them, right?

Of course, this whole thing comes with historical evolution and the market has matured. But this came as a long journey. After World War II, when we transitioned away from this one-off war bond auctions and so on and became much more of this regular, predictable market where debt is constantly issued and rolled over in

into a similar tenor, there's just, I think, a certain institutional momentum for maintaining that. And generally, it's not a bad idea to have a good chunk of your debt that is not maturing so much all the time. Then from an operational point of view, you avoid accidents.

So what do you make of the accusation against the Yellen Treasury Administration that she was involved in, quote unquote, ATI, activist treasury issuance, i.e. issuing a lot of treasury bills in order to stimulate markets? And I guess, Steve, the accusations of ATI have two components to them, really, as I perceive them. Number one is that issuing treasury bills is stimulative to markets and the economy. Issue two is that it was the intention of...

the Treasury to do that. I think issue one, I mean, definitely could be true. I mean, it's kind of a fact of monetary plumbing that a one-month Treasury bill is a lot more stimulative. It's basically money, whereas a 30-year Treasury bond is an investment. It's a security. I mean, one-month Treasury bill is a security too. But I think that's just a matter of fact. But

I think that the Treasury was fulfilling their own job of tax cost to the taxpayer. The markets wanted short term Treasury bills. Demand was high rather than manipulating the markets. I think intention is very hard to prove. So what do you make of that claim? Just your general thoughts on it.

So, first of all, Steve Moran is a good friend. I like him very much and I respect his work very much. So, and, you know, sort of, he actually cited my work, referred to it in his white paper.

From a quantitative perspective, to the extent he was estimating the impact of these immaturity, that maturity structure changes and whatever short-term impact it might have had on bond yields, I don't really particularly disagree that much. What I do find a little bit curious is this whole sort of controversy, quote unquote, because I don't really see this as...

deviating that much from how territory has historically conducted itself. And in any event, I don't really think it's that big of a deal. So let me tell you why. So first of all,

We already mentioned, historically, the Treasury's issuance decision around the curve has always been that they tend to go very short maturity upon impact of a big recession. And COVID was as big as it comes. And weight average maturity did plunge. If you go to the Treasury's presentation to the T-BAC committee, which they put out every quarter, you will see this chart showing

that plots the weighted average maturity of the market, what market was that outstanding. And it did plunge right in 2020, reflecting, you know, the issuance of T-bills and then it bounced right back. And in fact, it continued rising from the bottom of COVID all the way through the Q3 of 2023.

And I think by Charger's own calculation, that overall that portfolio had a weight average maturity about 70 months, which is substantially higher than its historical average since 1980, based on the chart. I think it's higher than any level since the 1970s or maybe even the 1960s. Correct.

So this argument that the Treasury did not term out the debt is just very bizarre to me because it did, right? It very much did, you know? And the Treasury kind of, as a debt borrowing, you know, agent acts kind of, you know, opportunistically in that regard, right? So in fact,

the idea that they will counteract the Fed is not the first time either. Like during QE, the Treasury will try to issue long-term debt. So, hey, wait, there's a new buyer in the market. We're going to issue a ton of new long-term debt. So I think they did exactly what you would expect them to do.

Of course, if you're the treasury, you can't really just go out there and borrow $500 trillion and let it sit on the balance sheet just in case interest rates go up in the future. You can only have the ability to borrow up to the debt limit and up to the spending needs. So after that first wave went away, when we continue to have more borrowing needs, especially

due to interest rate cost itself and debt service cost itself, which has become a substantially part of the overall federal deficit, then you really have no choice, at which point if you are sort of ignoring politics entirely, which I don't think is a factor here at all, personally,

I think they just perceive themselves as having some wiggle room here to not have to lock in long-term borrowing at historically high 10-year yield and believe, the market broadly believed back in 2023 that long-term interest rates will come down gradually and they will be able to borrow and turn out whatever additional borrowing

hopefully lower interest rates, as it turns out, 10-year yield did come down a lot and then you shot up again. But I think this whole... I mean, but the thing is, if you look at 10-year yield where it sits today relative to say the peak of Q3,

you are still not higher than what it was back then. In other words, you have had this entire time to still do what you need to do to term out the debt, if that's the consideration here, and not rather than pushing a whole bunch of long-term debt, just as the market itself is panicking a little bit about the whole supply and whatever issue back in Q3, 2023. So that's kind of how I see this whole issue

I think the significance has been

I think the whole motivation, I don't know where it comes from. I don't see evidence of it. And I think also, even I think the economic significance has also been overstated. We will see how the new treasury under Scott Bassett and in fact, our good friend Steve Moran, who is becoming the incoming chair of CEA will handle this situation. But I suspect a continuation of historical conduct will resume.

Right. So in March 2020, under the Trump-Munichian Treasury and where Steve Brown worked, the Treasury issued a lot of Treasury bills, which is in accordance with mainstream policy of when you have a financial crisis, an economic crisis, a recession, a depression, you issue a lot of short-term Treasury bills because that's what the market wants. And it also kind of provides a much-needed stimulus. So then Treasury was

The bill share there was very, very high. And it spiked bill share as a percentage of marketable debt outstanding as bill share also as a percentage of net issuance.

And so there the Stan Druckenmiller critique is accurate, I think, of, well, when interest rates were so low in April 2020, March 2020, they should have issued a bunch of long-term debt. But I think that's a conflict of a private markets hedge fund manager's view versus a treasury view. Like the most, if the treasury wanted to make money, yes, it should have issued a ton of short-term, you know, I mean, I think the 10-year like bottomed at 50 basis points. That would have been very good for the public purses of the United States, but it probably would have, you know,

caused a disruption in markets and it could have exacerbated that recession and inhibited the recovery. Your point is that later on in 2021, which just so happened again, I think it's a pure coincidence, but Yellen and the Biden treasury extended the debt and so did the opposite of what the ATI accusation is. And that was because you're emerging from a recession and the market now wants longer duration coupons and you want to move the

the weighted average maturity, the WAM, to more historical levels. So that's, I just think that, yeah, like the weight of average maturity right now is very, very high. It is slightly lower than it was during the early years of the Biden administration. But during the early years of Biden administration, it was the highest it's been since like the late 1960s. I mean, yeah. So I think from a

so to speak, stock, right, in terms of the decisions that have already been made, perspective,

I don't think there is really too much to be said about what they have done, you know, to, to, to a certain extent, you can't really pre-commit you all of your future selves, good decisions today. Right. You can't be like, oh, you know, I'm going to commit to a healthy diet by not eating today forever, you know, just so that, because at some point you still have to revert to eating something every day. So, um,

I think that's part of the tension. This idea that the Treasury sitting on top of $20-30 trillion of debt portfolio can act very opportunistically, very quickly is not realistic. They are not always going to be acting

on the dot. But over time, I think the average is okay. Just like when people were freaking out about Biden releasing the oil strategic reserve to help with oil prices. And people were like, "Oh my God, he's releasing all these oils and he's not refilling it, and prices are slow." But then you look at the price, it came down a lot

didn't do very much or did a little bit and then price went up and then it came down again so now okay fill it and they are filling it so i i think that's kind of how the government has typically operated they operate you know not that uh quickly but i think if you get the main forces right on average you're gonna be okay but i mean right obviously right now um

I think that interest rate expense and debt service costs is going to be very much for first order and whoever has to tackle and deal with it is going to have a big problem, a tough problem they have to solve. And I don't think locking in whatever debt we had incurred a few years ago, historically, a lot of rates really would have helped in terms of the new debt that's being incurred in terms of debt service costs

interest expense and new spending and so on. Yes. When we talk about the Treasury, the Treasury has zero power whatsoever to influence the deficit, how much the government has to borrow. Its only job is to fund the government and refinance the old debt as well as borrow new money.

for the new deficit. Steve, so if the Trump-invested Treasury tries to extend duration and borrow more long term, do you think that will have a supply effect where 10-year yields, 30-year yields will go up because of the supply of bonds? And how much of that depends on your view of the stock bond correlation, which I believe is now negative once again, I believe?

Yeah, so depending on how you estimate the stock bond correlation, which I had earlier suggested, essentially is proxies for the price of duration risk, or how much one charges for absorbing statutory supply. It has been falling pretty rapidly over the last few months.

So that's good. So in terms of market price for supply is lower than it was before. I don't know what the new administration is going to do in terms of that management, but I suspect it's going to become of higher importance than it was in the previous 20 years prior to COVID.

I expect them to move judiciously and generally extending maturity just because that's kind of how historical literature has always done is that we said, you know, sort of the four facts that I laid out in my dissertation. The third one is that, you know, if there is rising deficit, right,

they are going to want to respond by extending maturity, even though the entire curve has moved higher. You don't really have that much of a choice, but I suspect they would do it in such a way that you will observe them doing similar things where they don't try to overwhelm the market by issuing everything all at once. And the one thing that I did in my paper studying

the impact of supply which i'm a little bit proud of is that uh you know people i often look at this they look at some sort of just simple correlation of uh treasury supply and then the interest rate but what i was

figuring out was that really what matters is not the entire stock, but the treasury issuance. And not only is the treasury issuance, the new issuance that matters, there is actually anticipation. So when I calculated the weighted average maturity, weighted average duration of issuance, I was actually going six or 12 months ahead. So I was blending trading six months and a future six months, or looking at

outright future six to 12 months because the market anticipates supply that's coming and is being very painstakingly communicated to them. And in fact, taken as advice from the market participants, primary dealers, right? So it's not like, it's very rarely the case that market would be shocked

shocked by supply. And I think the only time I remember most recently was middle of last year, Q3, right? You know, from August through November, I think the market was just kind of a bit asleep on the whole issue because he hadn't thought about it in such a long time that when he suddenly came up, even though it had been sort of repeatedly told about it, he suddenly, you know, was a little bit surprised by it.

But after that, I think very quickly you realize that QRAs stopped being a market event informationally. Like that day doesn't carry particular significance before and after if you do some sort of event study, right?

because the market generally anticipates and then smoothly adjusts towards it. So I would expect something similar where generally the fiscal picture is not pretty and you would expect supply to have that impact on bond yields, driving bond yields higher, at least sort of always keeping

being a force, sort of pushing bond yields higher while other forces may want to push it down, right? If there is a recession demand or whatever. But I think fortunately supply is actually a choice variable. So I think they would do it judiciously so that the market doesn't go crazy.

Yeah. Why do you think it was that the quarterly funding announcement mattered so much in 2023? That's the QRA commonly known as just simply the refunding. Because, yeah, I think in that second quarter refunding, as you said, we're going to borrow a lot more in the longer term. That long-term interest rates went up as high as 5%. I think they probably peaked October 30th. And then November 1st,

first, the Treasury came out and said, we're actually not borrowing as many, we're still borrowing tons of long term money, but just way less than the market thinks. And then interest rates shot down. And the tenure went down by over 100 basis points in like 30 trading days. So one and a half months, basically. And I mean, you could say that's just a coincidence, but it sure didn't seem like a coincidence. And I know many people were paying attention to it. So why do you think the QRA mattered in 2023, given that the year where you had the stock bond sell off was 2022?

And actually in 2023, I mean, bonds sold off as stocks rallied a lot, which means which is a negative correlation, which is, you know, they are a hedge for each other.

Yeah, I think part of it, you know, economics, there's this thing called the peso problem, right? You know, sort of, you know, like the thing that goes horizontally and then suddenly drops, you know, instead of adjusting continuously towards an event that you know is there, right? So the market has always known with quite a bit of foresight, the borrowing needs was going to be there, right? And whatever borrowing maturity structure was going to look like, you know, as of end of Q2, right?

I think the market as a whole just wasn't really paying a whole lot of attention to the issue because it had just dealt with SVB and all these other things. Bob Shiller has this great book called Narrative Economics, and market sort of goes from narrative to narrative over the long run. I think things...

price out correctly in equilibrium. But in the short run, from week to week, month to month, there's always a dominating narrative. And at that moment, suddenly, this became the dominating narrative. And I remember, the 10-year yield went up 100 bps from that eventful QRA at the end of Q2.

And then, if you take a different, your favorite term premium estimate, your term premium went up about 60-70% of that. In other words, a lot of it was actually driven not necessarily by a change of view about the long-term interest rate short-track path, but rather by risk premium compensation for absorbing bond supply.

which I think interestingly, Paul was asked about it and he actually affirmed this viewpoint as well, that you can see how much long-term yield went up and measures of term premium went up. So I think this really came as a result of, depending on the quarter of the year, whether inflation was or was not a major consideration, right? Because you remember when...

The 10-year yield then soon after that third quarter thing plunged very quickly because that issue got narrative, got taken away a little bit. And most recently, I think last year, when the 10-year yield suddenly plunged from 4.5 to 3.5 or 3.6 or whatever, because we had a growth scare there.

Nobody seemed to suddenly during that moment, it's like nobody was talking to supply anymore or QRA because it doesn't seem to take a whole lot for recession fears and demand for safety, demand for a safe access to come right back.

So I guess the long short of it is I think the whole event was more of a narrative having caught on. The financial world definitely does run on narratives. Steve, the final thing I'd like to ask you about this is you got this fascinating chart going all the way back to 1920 showing...

But the weighted average maturity of the U.S. debt in different colors, orange for short-term bills, green for one to five year, pink, red for five to 10 years, and blue for 10 and above. And then you've got a red line of just the weighted average maturity overall. And it's fascinating how since 1980, or probably 1977, I guess, basically the weighted average maturity has been going up.

And then from 1920 to late 1970s, it was going down. So it really is like two different regimes. And we are in a regime where the weighted average maturity is going up. So although debt to GDP has been going up and that chart looks a little bit scary and has many, many people saying a U.S. debt crisis is coming or something, it is notable that the term structure of our debt

does not look like Argentina or an emerging market. It actually is quite, quite responsible. Yeah. If you look at chart 2.1 of that paper, you have just mentioned, which I will put up right after. I would refer to those regimes, the transition as being financial development or maturity, right? The US had gone from a relatively financially immature country or

or less developed country to one that's really mature and institutional, right?

And the first half in the shaded area where you see a sharp and steady decline in maturity is exactly that the market is able to handle liquidity better. It's able to handle sort of transactions of large volumes of T-bills and short-term maturity. And generally, I think you see this in the entire economy, right?

the US financial sector, financial services being the world's most sophisticated is going to be able to handle, you know, sort of better role of risk better, right? I mean, it's all about risk management, ultimately, right? So if you have the mechanism, you can you can take more risks

And since I think the modern era, you've basically seen the weight average maturity to be what is statistically known as a stationary process where it sort of goes up sometimes and comes down. Sometimes it looks like a sinusoidal curve. And that evolution has more to do with business cycle evolution.

various other structural forces. Sometimes it's a little bit higher, sometimes it's a bit lower. Um, and right now we're actually towards the longer end of, uh, the historical range, which is good, right? Because you happen to be dealing with a lot of, uh, you know, interest rate risks. So, you know, we have a, that we seem to be, you know, looking at a debt portfolio that's relatively well, you know, positioned that is not sort of, uh,

stock with having to go to whatever the highest sort of short-term interest rate is every three months. That makes sense. One thing that does give me a little bit of worry is just I think there's a lot of leverage in the treasury market and a lot of

Treasury is owned by banks, which are leveraged, but also hedge funds, which are borrowing heavily in the repo market, as well as involved in the basis trade. So this issue is a little more complicated. Steve, let's now take a detour. Tell us about the index work that you do for Bloomberg, as well as with the products that you've got out. What are you working on now? So my work currently is actually

quite different from my work on the treasury market during my PhD. And I work more on investable indices, especially systematic equities. And we've worked on various factor ideas that leverages Bloomberg's proprietary accounting data.

And I've tackled a few issues that are, I think, pertinent to the market over the last few years. Most recently, I put out a new white paper on the innovation factor, where I think about how to systematically capture innovation. People love talking about innovation, you know, but it's actually not often, I think, clearly spelled out what innovation actually means and how you actually go about systematically capturing it.

So this white paper that I put out, which has a linked index that has been licensed by First Class as an ETF, essentially looks at the R&D expenditures of public companies, US public companies, and see which companies are genuinely innovative. And one core finding that we have

is that contrary to the past, where people have looked at R&D spending as a source of innovation, we find that it's not so much the level that matters as much as the persistence. As anyone who has actually tried to do innovative work probably knows, or even just work of any kind, that sometimes what really matters in the long run is having the persistence of doing it again and again.

So if you screen companies that have, say, three consecutive years of R&D growth, regardless of anything else, right, just do nothing else at all, and you take top 100 companies in the US large mid cap, right?

And you sort of take those companies and you market weight them. And then I guess you apply a little bit of a weight cap so that individual companies don't exceed 8% or something like that. That gets you close to the benchmark weights. You recover a line. And that line, I'll let you guess what that line looks like. It looks like an index that everybody knows of. S&P? Yeah.

The other one. NASDAQ. Correct. Okay. It overlaps the NASDAQ, which is kind of interesting. So there's a blog that we can link to

So basically the NASDAQ index, of course, we know as we know, is just all the companies that happen to be listed on the NASDAQ exchange. And here we have constructed mechanically through accounting variables without referencing any stock exchange, something

that looks just like and in fact, as most recently, essentially outperformed it because it contains companies that are not

listed on the nasa exchange uh you know there are innovative companies in the us that don't list on nasa exchange with you know without taking a dig in that exchange at all like you know we we know that for example eli lilly you know which is a major pharmaceutical innovator and and one of the inventors of uh the glp wonders drug happens to be listed on the new york stock exchange

So this is actually something we found that basically persistence works better than intensity. You can screen for companies with most intense R&D spending relative to say normalized by sales, but persistence just has historically done a lot better.

And this is actually true, not just in the US large mid cap, but in the US, you know, different market segments as well, mid small cap and overseas, right? So that's an interesting fact. But then we also go on to sort of, you know, actually concretely verify why this R&D actually is innovation and not just some sort of luck, you know?

When did you publish this index? This index was published, I think, early last year. I think late 2003, actually, and then became licensed by First Trust.

Got it. So the ETF has outperformed the NASDAQ since it was created. It was created in May of 2024, so it's been out for less than a year. What is the index? The index existed before then. Yeah, but the index has existed before. Tell us about what the index has done since you created it, as well as what the index did historically, although that is an out-of-sample thing, and you created it before you did the outperformance.

Yeah, so as I was saying earlier, right, you know, I think historically, you know, this index has looked a lot like the NASDAQ, you know, index because it captures a lot of the similar names. And we all know how tech and innovative companies have done, you know, over the last, you know, 20 years, especially, you know,

the most recent 10. And part of the reason we have seen this dramatic outperformance is actually not because, oh, market has just gone into a bubble and mania and so on. If you look at the fact that footprint of the companies that are in this index, you'll find that you overwhelmingly load on quality and growth. Obviously they're expensive, but these are companies that are highly profitable, right?

I think what we are seeing is essentially the capitalization of successful innovation in America. And I find it quite fascinating that there was this book in the 90s called The Innovator's Dilemma, right?

And the author, Christensen, he wrote that over time, as companies get larger and larger, they find it difficult to innovate. Why is that? It's because whatever new line of business they discover, even if it succeeds, it doesn't make an immediate first-order difference to their overall revenue, because this has to be a very small business. And if anything, it may actually capitalize their current business.

And then you have other factors such as bureaucracy and other issues that tends to prevent companies from innovating. But what we saw over the last 20 years in the US is that not only has the US managed to continue innovating, but innovation has actually occurred a lot at very large tech companies. So much so that today we're talking increasingly about Magnificent Seven, and so on and so forth. These companies

On the one hand, they are dominating the markets with very expensive valuation. On the other hand, they've also just become indispensable in society because they've come up with genuinely valuable innovations that are very, very profitable and customers pay for those things.

And to the extent that the index that we published has actually deviated and outperformed, say, the NASDAQ index, it's because this is actually done in a systematic fashion. So we actually pick up companies that may or may not list on the NASDAQ Exchange, but qualify based on the signal.

Oh yeah, three, two, one. So you're basically caveat companies that spend a lot on R&D but are profitable and been having spending money on R&D for a while. So it's probably not going to get stocks that just IPO'd

So what would an example be of a company that is in the NASDAQ but is not in your R&D index? And what's a company of example of a company, you know, you can just say a generic example of a company that would be in your index that's not in the NASDAQ. So I gave an example earlier that's, I think, a pretty, you know, you know, famous one. That's the the.

Eli Lilly happens to be listed on the New York Stock Exchange. There is also Caterpillar. That's another one. People don't think about industrial companies and they're not sexy, but Caterpillar is actually a very innovative company. If you go further down the

the market cap, you start looking at companies like Snowflake, Block, Square. So there are companies that are not necessarily listed on that exchange credit to NASDAQ. They sort of really were very early on capturing, became the home of tech companies, innovative companies, and there was a snowballing effect.

But the broader point is that it's not necessarily exclusive. Good ideas, innovative ideas, it's not actually exclusive to a single stock exchange. And so I guess the NASDAQ, I think NASDAQ 100 is 60% tech as we record, whereas your index is much less tech. It still has a lot of tech. So you're capturing innovation that's not in the technology world as well. Tell us about sectors because sectors are kind of...

They're kind of nonsense. I mean, like Amazon is a tech company, but it's in consumer discretionary. Let's see, Google is a tech company, but it's in communication services. Like these, I guess, telecommunications. Do these names, I mean, do these sectors, do they really help? Do we think we need a...

Either get rid of all the sectors, get rid of all the labels or generate new labels because often they say, "Oh, well, we're diversified. We don't even have... We're only 40% tech." But then the thing that's not tech is Amazon, which is not tech technically, and Google, which is not tech technically. Yeah, you're exactly right. I mean, this goes beyond the scope of this index itself. But yeah, I mean, I think...

Over time, we come up with new ways of modeling the world. Sectors are really just a broad type of categorization. Once you start thinking about it, why is one company designated tech and not the other one? And then over time, companies move too across sectors because

If you get into the weeds, in the sausage factory, you can talk to our colleagues in the BICS team at Bloomberg, Bloomberg Industrial Classification System, that has a lot of judgment calls about what percentage of revenue is derived from what source,

qualifies you for what sector. So, yeah, it's quite right. And of course, there's emerging patterns in new industries, right? You know, EVs and drones and new things coming out all the time. And companies may try to get into new sectors.

So it is hard to say, but I think as a broad categorization, you know, sectors, you know, categorization kind of gives you a sense, right? I mean, just like, you know, we have different style factors now, value, momentum, quality, and people can debate whether a company, you know, is necessarily one versus the other. And depending on the intensity, I think they're just sort of approximations.

And Steve, because you're, you know, spend so much time talking about indices, I want to ask you,

What do you think about the S&P 500? What do you think about these passive indices? Why are they so hard for active managers to beat? Why is it that these seemingly kind of arbitrary rules of, okay, you take... I don't even know what the S&P 500 rules are, but I guess they have to, I guess, be kind of profitable sometimes. They have to be a certain size. And then you kick out the people who are smaller. So that's how Tesla gets in, how Supermicro gets in, and you get things reweighted. But ultimately, the principle is...

The better the stock performs, the bigger it is as a percentage of the index. And you don't do rebalancing. You'll say, oh, Apple's gone up twice as my Intel, so I'm going to sell half my Apple and rebalance Intel. You don't do that. It's market cap weighted. So tell us just, you know, and that has done very, very well in the US. Part of that is we've had a bull market in US stocks. We've had a bull market in large cap tech stocks. So that has worked. But just tell us your overall thoughts on that S&P.

Well, let me actually stick to Bloomberg indices. So Bloomberg, we have a B500 index, which is basically the largest 500 companies by market capitalization. And there are some buffer rules around what gets in and out on the quarterly rebalance frequency and review.

But I mean, behavior, like if you plot the two lines, you will not be able to tell one from the other. They are very similar. So there is some sort of committee based discussion, I think, S&P practices. And because it's the most well-known and the most tracked trading index, there's sort of front running around those decisions. But by and large, what really matters is what's at the top and with the market being so top heavy,

you know, like people sort of lose their minds over like, oh, SMCI got added and you lost a bunch of, you know, and it seems to happen regularly. The thing is, it doesn't really make a first world difference at all to the level of the index, right? Because such as a such a tiny way as a new new addition.

I think regarding your question about why it is that passive index has had such a strong track record, I think it's a fascinating question. I've pondered about that myself a lot. I think you've alluded to it, and I suspect that's a really big reason, is that

you just don't really have as many opportunities for the active manager to come in and meddle with and make a bunch of decisions that more often than not actually hurts. So you don't have the kind of risk-based

rules where you have to cut some weights, if the winner is running and you just have this passive set of weights that's based on market cap. And so mechanically you are loading on momentum, right? And you're also going to, I think, have some, based on rebalancing some means of mean reversion.

But yeah, I mean, it's a fascinating question, you know, why that is. And I don't think I can give you a full answer on the spot right here, but I think you've alluded to a lot of the reasons. Okay, so we won't talk about the S&P. We'll talk about the Bloomberg 500, which sounds like it's just the biggest companies. And I guess, first of all, what's the name of that? And what's the difference between that and the S&P? I guess there are some companies that get big, but they're not allowed in the S&P until certain things happen. Like I remember Tesla wasn't allowed until...

the end of, was it 2020 or 2021? I forget. But there was a time where it probably was a member of the Bloomberg index. It got added immediately once it eclipsed the 500th biggest company. But then it took a while to get into the S&P. So talk about that. And yeah, I just, you know, I think we can mine this vein a little bit more because there's a lot there. Sure. I think that's an interesting thing, right? So, you know, as you know, S&P 500,

index is an old index that's committee-based, right? And there is a certain degree of discussion as to which companies can edit and delete it.

Obviously, the reason they do it that way is because of legacy that's sort of been that way. But also, I think part of this because it has become such a market event, which companies get added because of the passive holdings on demand for these shares, that there is front running and trading around the index reconstitutions.

Bloomberg B500 index is fully rules-based. There's no committee, right? And there's no... And it's just one rule. It's the biggest companies in the US or global? US, yeah. And there is no profitability rules. For example, I think S&P requires that a company to be profitable consistently for, I think, three, four quarters in a row, right? Before it's eligible to be selected.

We don't have that. But beyond that, I mean, there's sort of... And then there's, of course, some buffer so that companies don't immediately qualify upon being momentarily the top 500 for an instant. But beyond that, it's entirely rules-based. I think this is actually, in my opinion, the more intellectually consistent way to do it. Because

to the extent that you just want a benchmark that captures the market, I really should not have to want to put another quality or profitability screen on it. Why do I care if it's profitable or not profitable? Well, I mean, you put a quality screen on your R&D thing. I think you're all about

quality screens. I think it does matter. What surprises me is that the Bloomberg no quality screen whatsoever, I mean, according to you, it has the same line as the S&P. That surprises me. And that tells me that maybe the added quality screen of adding the profitability is accounted for because once a company leaves the biggest 500, that is the quality company. And so maybe there's a lesson there that you don't want to own the company that just exited the S&P 500. That's not a good thing to own.

Yeah, I mean, the thing is that it's not like a consistent quality screen either, right? Because it's not like, oh, you know, Tesla, you know, only becomes eligible if it's profitable for a number of quarters. I know if it stops being profitable or if it loses money, then it gets kicked out. It does get kicked out. It's a quality screen to get in, but it's not a quality screen to stay is what you're saying. That's right. So the impact is actually pretty second order if you

If any at all, it's very minimal because of that. So at least that's sort of how I see it. And the R&D index, by the way, this is a different set of animals. I think AeroBeltrunas will call them hot sauce indices. These are more akin to your sort of factor indices that are like value, momentum, quality, and so on and so forth.

And I find this whole concept of passive indices rather intellectually unsatisfying. What is actually passive about an index that's constructed based on factor rules that gets rebalanced? Fundamentally, how different is that from, say, a factor fund?

Intellectually, I find the line between active and passive to be quite blurry. And I think of it as being more a continuum than a clear cut. True passive would be owning 500 stocks in some weight and then just never changing that ever. And I actually saw this on Twitter. So I don't know, a year or two ago, I don't know if it's true, but actually like the S&P 500, if you bought it in, let's say 1965 and never reweighted it,

it actually that performance would be even better than the reweighting of the S&P over then, which that is interesting.

Well, you would have suffered mightily from survivorship bias, right? A lot of the companies would have gone bankrupt. I mean, part of the reason why index passive index works so well is because, you know, sort of autonomously gets rid of companies that disappear and pick up new companies. Yes, I guess. But I guess that is pure passive. So again, it's something I saw on Twitter. I guess I wonder why that would have...

I guess, does the S&P ever reduce its weighting on something because it's gotten too big? Does it ever do that or no? I mean, increasingly, recently there is a debate about capping the weights of these biggest companies in the indices. And that's actually a discussion that's been happening for S&P and Russell and so on.

I think it's in response to the fact that you have such increasingly high concentration of market weights because of the high capitalization of the biggest handful of companies. I think this morning I read that 26 companies account for half of the index weight, making basically companies that are towards the second half of the 500 named companies

economically relevant. They can go up 10 times and you wouldn't even notice it at the index level. Well, one company, the 499th biggest company, if it went up 10 times, you wouldn't notice it. But if 100 companies, 400 to 500, you would notice that and then they become economically more significant. And the point is, it's because Apple has been such a great company and it has been such a great company that they've earned the spot. They weren't given the spot by the geeks on the

S&P index deck. They've earned that spot. So, but Steve, I mean, I did a small amount of research into this, but I don't know how much you're familiar with like the telecom bubble in Canada. Like Canada famously has tons of oil and gas stocks and few tech stocks. So when they had this one tech stock, Nortel Networks, it was like 30% of the index. And it ultimately had severe accounting issues of fraud that the

CEO had to resign. And so when you have a stock that's 30% in an index and it goes down a lot and has severe issues, that can suffer. But I don't think that the current S&P is as vulnerable as that because Nvidia is 8%, Apple is 6%, Microsoft is 6%, 7%. These are high quality companies and they definitely can go down a lot. But

I do think that those companies are generally higher quality than the average company in the S&P 500 and averagely the Russell. So it's like if it's dominated by bubble stocks, then yes. But I'm not of the opinion that those stocks are bubble stocks yet. But I mean, what do I know? What are your thoughts on this matter? I mean,

Well, two issues. I mean, it's interesting you brought up Canada, right? I think, first of all, I think country boundaries are a little bit arbitrary, right? You know, like if you pick Saudi Arabia, it's going to necessarily be dominated by oil and gas, and Aramco is going to be by far the biggest company. And why do we actually draw national boundaries, you know, and construct investable indices that way, apart from the fact that, I mean, obviously you do have, you know, sort of national regulations and so on.

Generally, I like to think of companies along as much as possible universal characteristics. And to that extent, US companies have just been over the last however long, growthier, higher quality, more profitable, and longer duration.

I find these charts where they compare US stock market performance with the rest of the world. Often I roll my eyes very hard because you're not really comparing apples to apples exactly as you alluded to earlier. The sectoral compositions are totally different between US and Europe and Asia, right? And you don't compare growth stocks with value stocks. And for that reason, you shouldn't just compare directly. And if you adjust for that difference,

equity duration or growth profile, you'll find that Europe and American stocks, the difference, the delta while there is still some because the valuation premium in the US is substantially smaller, much smaller. So that's sort of my first point.

Regarding your second point, I think this goes back to the nature of the US economy. It's a very large, complex, diverse economy that's almost a world onto its own. The joke people often make about Americans calling baseball, National League as the World Series,

There is something there, right? This is a country that is just really, really big with lots of stuff going on and with all the sectors, with all the things that you want. So

and it has the most developed financial sector. So the bubble, to the extent one worries about concentration of equity in the indices, passive indices, the B500, you see a similar high degree of concentration towards the top, really, in my opinion, is mirroring a bubble, quote unquote, in the real economy, that you have an increasingly winner-takes-all economy.

instead of a mean reversion economy. The sectors that are dominating are sectors that tend to be extremely high positive networks, externality, social networks type of economy where

Apple is sitting on top of the most valuable set of users in the world to which you can deploy the latest technology at the snap of a finger. So I think that is really the more interesting question. So rather than focusing on how much market cap or market index weight a certain company has, I think the more interesting question is

should we expect this evolution

of network externality to continue? Can we keep expecting Costco to kill off local mom and pop grocery stores and everyone just shop at Costco and use an iPhone and do everything with, you know, Nvidia, you know, chip driven AI applications? Like, is that the world we're going to live in? Or should we expect at some point some kind of a mean reversion and the small guy strikes back, you know?

Right. And I think that so much of the earnings growth in the S&P 500, it may be literally over 100%, has been from the Magnificent Seven. So when people say, I mean, have you seen what the Magnificent Seven stocks are doing? The Russell hasn't even moved. It's like, yeah, the small cap stocks, their earnings are contracting. And the large cap stocks, their earnings are growing. So-

stocks that have growing earnings should be going up and stocks that earnings are going down generally should be going down. That's how it kind of works. That being said, you did mention Costco. I think the earnings there have been growing a ton. They have been taking market share away from mom and pop, but I think the price to earnings ratio there is...

nosebleed levels I mean 50 PE for a grocery is um I've been telling people we will witnessing is a quality bubble right it is uh you know you are seeing for the simultaneous expansion of earnings of very quality companies being quality of being a factor right so like profitable low debt low leverage and low earnings variability price strong pricing power companies these companies are being bid

Right. And they have been bid in the low rates era. They were being bid as though they were quasi bonds. Right. You know, because you have low interest rates and everyone wants to earn a bit of earnings yield.

And now I think it's just, I don't know what's driving it all, but you're sort of seeing the evaluation really go up a lot. And going back to what we were saying before, this does not feel very sustainable, but if let's say we get a major correction, what is the next thing you'll buy? I don't know if you wouldn't just run straight back to these things that's at a lower valuation.

Right. Yeah. I think if 2020 and 2021 was a bubble in junk speculative growth stocks, 2008, the bubble was mainly outside of the stock market. 2099, bubble in dot-com stocks, growth stocks, but they really didn't have any revenue, extremely speculative. Now it's just good companies are priced extremely richly. But I think a correlate to that is that there are high quality companies that are growing that are priced... I think...

The worst saying of all time is it's a stock picker's market because it means nothing. But there are opportunities where I feel like in 2020 and 2021, all stocks were overvalued, where I feel like now it's perceived quality stocks are overvalued. Steve, any response to that? Or just what else are you working on at Bloomberg Indices? What are you cooking?

Yeah, so related to what we were just talking about. So we have this innovation index, but we also previously have created two

Other indices that I want to mention quickly, one of them is called the Pricing Power Index, related to what we were just talking about. Costco is a consistent member of that index and we wanted to systematically capture companies with pricing power and we found that it turns out you can find companies with strong pricing power by looking at the stability of their gross margin over the trailing five years. Companies that maintain extremely stable gross margin

tends to be companies that have very strong pricing power. And interestingly, they tend to have a certain profile to them too. So economic logic wise, why is that the case? It's because the gross margin is basically the closest thing to the top line. Everything is basically top line revenue minus variable costs, which includes all your input as well as hourly wages of your tech workers.

And a company that has very stable gross margins comes in on two ways. Either they see that cost increase and they pass on the cost by raising prices and maintaining gross margin, or if they see a surge in demand for their products, right, and gross margins going up, then they will supply into that additional demand, right? So what you find by doing such a screen is that you see companies that tend to be niche suppliers in niche... Sorry.

suppliers in niche industries of which there are a handful of them. So interestingly, you pick up recently a lot of the AI stuff, so the cadence design is a company that gets picked up because they design software for designing microchips. So things that is not very... Sure. Yeah. That pricing power index was interesting. I expected to see a lot of the top names. I mean, Apple, Microsoft, and

Nvidia, I mean, what company has more pricing power than Nvidia? I mean, if Apple wanted to charge, you know, $1,500 for their phones, a lot of people would buy it. But those names are absent. Why are those kind of Mag7 names absent from the pricing power index, whereas for the...

R&D one, as I said, it looks a lot like NASDAQ. Was it intentional or is it just that your weightings dictated that, "Nope, these don't qualify because it doesn't meet XYZ weighting." Yeah, that's an excellent question. So these are systematically constructed indices based on

Systematic products, in the case of pricing power, is based on the stability of gross margin. In the case of R&D, it's based on the persistent growth in R&D expenditures. So whatever comes up, comes up. So why didn't Nvidia come up? And why isn't Apple there?

First of all, Apple will be there if you expand the list to a bigger number of holdings. We happen to be picking the top 50 by pricing power. Apple, if you think about it, does Apple really have strong pricing power in terms of its iPhone? The iPhone flagship price has actually not changed very much over the last few years. If you look up the iPhone Pro or whatever, right? Yeah. Okay. I guess I'm out of date. I thought that the iPhone now costs like $1,000 and I remember it costing...

you know, one or $200, but maybe, you know, the price went up from 2008 to 2020, but over the past four years or now five years, you're, you're correct in saying that, that, I mean, I know Warren Buffett has said that Apple has pricing power. And basically he's saying, you know, if people charge a thousand dollars for this, it, it adds a lot more utility to their lives than a thousand dollars. And, you know, perhaps he's right about that.

Apple absolutely does have pricing power. But if you think about the last few years, when you just said that the flagship phones price has not changed very much, that was also during the year in which costs went up a lot because of inflation. So Apple's margin actually suffered while they kept their price the same in order to maintain market share. So Apple is also trying to pivot towards services and so on. So there is a reason why mechanically Apple has not picked up.

NVIDIA is another interesting example in which the gross margin stability was not there. In the case of Apple, the gross margin went up, went down, right?

and therefore not stable. Nvidia's growth margin went up a lot, right? And stayed up, right? And it kept inching upward. - I mean, it's ridiculously high growth margins for Nvidia. - Correct, correct. - 7%. - 80% growth margin or whatever, like half the cost if you're running an AI company is going towards Nvidia. So why did it not get picked up? I think this is one of those cases where it's a false positive, right? So I...

we want to construct a simple signal that has sufficient statistic that picks up most of the correct names and you may miss out some names that should be there. In this case, I will struggle to find companies that have a sustainable short-term surge in gross margin that don't really have pricing power. And NVIDIA's case is really more of a unique

set of circumstances where the surge of gross margin has shot up and stayed up. Let's see what happens in a few years. If let's say they keep this up for five years, it will show up in the index. But for now, it just has not shown that level of stability. Who knows? This is one of those cases where you

you do sometimes miss out on a couple of things because you're not able to quite perfectly differentiate the true and the sort of zero and the ones. - Yeah, I do not expect Nvidia's gross margins to stay at 76%. I don't think anyone should expect that. If they expect that, they'll be disappointed. And if gross margins for Nvidia do stay at 76% for the next 10 years, then no.

No one owns enough Nvidia, but that's not my expectation. Steve, we could talk more about what you're working on, but I'm sure you see a lot of indices created by other people. What is kind of your pet peeves in terms of indices that don't make sense? I'll give you a few examples. One could be the Russell, where it's all small cap stocks. So basically the good stocks in the Russell that go up leave the Russell. That's kind of a problem.

And then another example would be the Dow, where it's price weighted, not market cap weighted. So the biggest stock in the Dow was United Health because it didn't do a lot of stock splits. It was bigger than Apple. And that stock being down hurt the Dow, which I think is why we had nine or 10 days in the Dow being down in December.

So, I mean, those are two examples that come to mind for me, but I'm sure you have much more high level examples. Like what are some things you see in the index world where it's like, oh, this is wrong? Because I know on Twitter, Steve, you often, you'd like to be a little bit critical and critical is not a bad thing, you know, like art criticism, movie criticism. It's how we improve. But what are some criticism that, you know, what the index criticism, what is Steve Ho has a weekly column called index criticism. What's he writing about?

So, yeah, great question. I mean, first of all, to address the two indices that you mentioned, I actually think the Russell one gets, I think, an unfair rap a little bit because going back to what we were saying before, there is actually a strong correlation between size and quality.

Historically, it has always been the case. There is this very famous or infamous paper by my former employer AQRs, folks, that's titled "Science Matters If You Control Your Junk." I won't explain why, it's controversial. But basically what they find is that whereas historically there was like at the beginning, dawn days of factory investing in the 80s, bands found that small cap stocks tend to outperform.

you sort of that went away, right? And you actually see a strong correlation of junk quality and small cap, right? Small cap companies tend to be less profitable, often unprofitable, and bit large cap companies tend to be sort of the opposite.

And I think that's only worsened based on what we just were discussing. There's a structural advantage of being bigger increasingly. That's so much so that the biggest companies almost are no longer constrained by macroeconomic forces because they generate so much profits and cash flow they don't need to borrow. Whereas small companies are often highly levered and they get hurt by high interest rates, they get hurt by everything.

But the thing is, what's curious about it is that you would think that, well, first of all, small companies

become big companies that's how big companies came about it's not small this is a startup right so if you think about in the rational perspective you expect small companies being junkier be priced cheap enough in relative to earnings or whatever you want to measure such that the expected returns should be high now on average that they should earn higher returns to compensate investors for holding these you know sort of companies that are awful to look at

But that has not been the case. In fact, we've seen quality kill it. Quality has earned. So you have this situation where over the last 20 years or maybe longer,

easier to hold companies, companies that have low volatility in earnings, low volatilities in stock returns, high profitability have done better. Whereas the companies, junkier companies that are not profitable, small cap companies have absolutely languished over the last however many years, say seven, eight years. If you plot small cap index versus the large cap,

you will see that yawning difference. But if you take a longer horizon, you do see periodically, right, sort of like small cap outperform, right? So I think some of this is also recency bias and structural changes in the macro economy. So I don't really think there's anything particularly wrong with, you know, constructing indices based on, you know, a small mid cap segment. You just have to sort of, I think, be careful about how you control for quality.

And in fact, I actually show that

in the R&D index white paper, that small mid-cap segment until basically 2021 or so, innovative companies in the small mid-cap segment have actually outperformed the large mid-cap version. So small cap is not necessarily a problem. The DAOs, I think that's more historical legacy and people like to cite it, but there's not

that there's not a meaningful amount of asset that's linked to it anymore, right? So it's just whatever, you know? Yes, but when on CNN or PBS NewsHour, any non-specialist financial program, basically any news that's not CNBC or Bloomberg or like a financial program like itself quotes the Dow and...

They say what the NASDAQ did, they say what the Dow did, they say what the S&P did. Often they cite the Dow first. So most Americans and most people, I think, think that the Dow matters, even though you're probably correctly saying that it doesn't matter. And we both agree that it's kind of a silly index. Well, what is interesting though about the Dow, and I hope soon we'll actually start quoting the B500 index. Yeah, it's the B500. B500. B500. B for Bloomberg. Got it.

I think the Dow interestingly has a bit of a value tilt. That's why on any given day you will see that if NASDAQ is doing well and Dow is not doing well, it's probably the case that growth did very well and value didn't do very well and then vice versa. On the days when Dow is doing really well and NASDAQ is down, it's usually the case that value has had a good day. So as a proxy,

I think that's probably a tangential purpose he still serves apart from the fact that it's just based on familiarity. I've got a question, Steve.

This may be before both of our time. It actually definitely is before both of our time. But I know that part of the reason why the Russell has been performing so badly, in my opinion, is because there are so many junk companies that went into the Russell because there was an IPO boom in 2020 and 2021. I just looked at the IWM and the Russell in 2000, the recession and the dot-com bust.

The Russell was only down like 25%, whereas the S&P and the NASDAQ was down way more. So is there some kind of quality control on the Russell? How come all these bubble stocks like Pet.com didn't go into the Russell? I'm assuming they didn't go in because if they did, it would be down way more than 25%.

I don't know the rules exactly. I suspect, my suspicion is that there is a similar set of rules, but then the structure of the economy itself has changed such that

you very often see companies that stay private for a long time and then it becomes public and then almost instantaneously gets upgraded to the large cap. Airbnb is a recent good example, Uber and so on and so forth. I think there is not structural change in the economy. And then with that, there's a sectoral change. So you increasingly have a lot of smaller, unprofitable biotech companies that are found often in

in the small cap index. And their exit strategy typically is they have a drug and they don't have any revenue. They just keep working on the drug and eventually get bought up by one of the large pharma companies. And they don't really graduate to become from like a kid company to an adult company. Instead, they just sort of get acquired. So I think, again, go back to what we were saying before. I think

what we are seeing in the financial markets is actually mirroring, you know, the structural changes in the real economy. And I don't know what was the explanation. We'll have to go back to look at the sectoral composition and the type of companies that were held in the Russell back in the dot-com era. But I suspect, you know, like,

to the extent that you had this pet or companies or whatever, they probably were not massively included. Steve, I know you spend your days focusing on indices, not necessarily macro, but what is your macro view right now? How are you thinking about the performance of the US economy and just asset allocation as well?

Yeah, I mean, it's interesting. I don't explicitly think about macro, although macro does, I think, guide my even thinking about design indices and the type of things that actually matter for investors.

to the extent that one thinks about the last 20 years and the persistent deficits the US government has run, that may in part have had some big contributing effect on the profitability of US companies. Maybe that's actually a driving factor behind the quality factor, why it has done so well. Now, 20 years later,

Starting from Clinton's surpluses, people worrying about the Fed run out of Treasury securities with which to conduct open market securities. So now we have a world of debt and you have Doge trying to deal with government deficits. Clearly, we are in an environment where

you know, sort of stock bond correlation fluctuation notwithstanding, bonds are just becoming much less desirable, you know, assets, right? From an allocation perspective, you're thinking of values whereas previously you'd be comfortable with a simple 60-40 portfolio where stock and bonds mutually sort of

Now you think about maybe adding in things that will actually help you form a better set of correlations. But Steve, the 60/40 portfolio, like the 10-year bond, it has to be more attractive relative to stocks now than when the 10-year bond was at 70 basis points in 2020. It just has to be. I understand the correlation point, but you're just earning like the 400 basis points higher.

I mean, I don't know. People were very eager to buy bonds when yields were at zero. I know, I know. I mean, I was recently, I mean, you had Citrini on and I spoke with him a little bit and he was making this interesting point where he said, oh, you know, like the, the, the,

the 2022 experience and the Fed hikes kind of broke the bond market, broke bonds as an asset class because people suddenly have had experience of being burned by bonds, having this massive drawdown that I haven't had in decades. Now, I mean, look, I

I think bonds are certainly still going to be there as part of the allocation and you clip a coupon and the bond is going to be high enough to justify whatever risks the market is going to be in equilibrium. And you're just going to have whatever correlation it's going to have with stocks. People are still going to want to hold bonds because outside of the loan-only beta context, there's nothing else you can hold that can protect you in a deep deflationary recession. If we go for a bit like halve,

a bad recession now, inflation is going to plunge and suddenly there's not going to seem like there's not enough bonds to go around. So people are still going to want to hold bonds. But I think if you look at that chart that I've shared with you previously of the stock bond correlation, how it has actually co-varied over the last half a century or more, I mean, until 1990 or 2000,

stock bonds were generally positively correlated. The next 20 years, it's been negatively correlated. Now we spike upward to positive and we are fluctuating a bit. But my suspicion is that we're going to find ourselves, meander ourselves eventually towards a definitively positive territory.

Between stock bond correlation, meaning that as stocks sell off, bonds sell off, as stocks rise, bonds rise. Correct. And the reason that is, is because fundamentally what undergoes this regime is the salience of inflation. Because inflation, if you think about it from asset pricing point of view, golden formula,

Like you are discounting cash flows, future cash flows by some discount rate to the present, right? You know, inflation enters as the discount rate, right? And then, you know, denominator, right?

So it's going to move both stock and bonds at the same time up and down because nominal bond, the coupons are fixed as a fixed number. So when the discount rate goes up, bonds go down. Whereas stocks, it has numerator and denominator. And the numerator,

If the inflation is very low or generally muted, the cash flow component can dominate such that bonds and stocks move in the opposite direction. But if inflation becomes first order and is always on the back of people's minds, then it's going to dominate and move stock and bonds in the same direction. And I think that's what we're going to see over the next however long that we are, in my opinion, modestly

generally inflationary environment where I don't expect us to go back to the 70s by any means. But I also don't think that in the absence of the types of structural positive supply shocks that we saw coming to 2000 with China entering the WTO and automation, and then we also got hit by a GFC,

that you're going to have inflation really come down that much. I mean, the Fed can only do so much. Ultimately, you need something else to take over. And whatever is taking over, I think, is going to be a generally more inflationary set of factors. But of course, that's more secularly the case. Cyclically, we could very well go through

a few months or two a year of further disinflation. Who knows? We'll link in the description your first paper for your dissertation, When is the Supply Effect Large in the Government Bond Market? As well as hopefully your second piece, A Realistic Model of US Treasury Debt Management. Steve, people can find you on Twitter at Steve Ho F. Where can people find you? Where can people find the work of Bloomberg Indices?

They can go on to BloombergIndices.com and we should put the link of all of our research pieces at the bottom of our chat today. And yeah, I look forward to people reaching out and chatting about things we've discussed today. And so what's that link? Bloomberg.com. What's the link? BloombergIndices.com. BloombergIndices.com. Yes.

Got it. Cool. All right, Steve Ho, thanks for coming on Monetary Matters. Thank you everyone for watching. A reminder, people can find Monetary Matters not just on YouTube, but on Apple Podcasts, Spotify, wherever else they get their podcasts. Until next time. Thanks again. Cheers. Three, two, one.

I'm here with Steve Ho. We just finished wrapping our over 90 minute interview. It was very good. Talks about term premia, talks about his work he does for Bloomberg Indices, index construction, all very good. After we finished recording, I asked Steve a simple question. Does he think the 10 year is a buy if the 10 year hits 5%? It's now at 4.7%.

Does he think the 10 years are by? He starts to go and talk about something very interesting, how he thinks bonds are by, and we're going to have a repeat of a dot-com bubble. I couldn't help myself. I started yelling at Steve. I said, Steve, this is such great podcast content. How dare you hold this in for 90 minutes and only start saying this after we start recording? So I had to turn the cameras back on. Steve, the floor is yours. Dot-com bubble, go.

Yeah, so I have a nuanced view about 10 years and I mean, I already gave you why I think secularly, I think bonds as an asset class is bearish because of macro forces and with all the uncertainty supply. But I think cyclically, what's going on right now,

I increasingly see us in an environment similar to what we witnessed just during the dot-com era. If you recall, in 1998, in 2000, during that period, the Fed similarly cut interest rates three times, paused, and eventually had to resume hiking. And during that period when the Fed paused, 10-year yield surged.

And stock market surged even more, right? Market went berserk, right? And of course, what eventually followed was a stock market crash and a recession. During the recession, the 10-year obviously yield fell a lot with inflation as it does during every recession.

I mean, the reason I didn't mention it before is because this will feel a bit too, I think, I don't know, what's the right word? In any event, it's highly speculative, right? But I increasingly find myself thinking that we are going to have a repeat of the dot-com playbook where

the Fed almost becomes a little bit of a secondary player in this whole show, whereas the previous 20 years that it was all about the Fed, increasingly, I think it's going to be increasingly not about the Fed and other sources of other macro factors will come to dominate market sentiments. In this case, clearly what's been happening over the last few months since the election has been Trump, right?

the new administration, everybody is just projecting their wildest dreams, the most bullish expectations on the Trump, the blank canvas, ignoring the bearish factors. And all the Trump-related trades, which I think Satrini has discussed with you plenty on your previous episodes, have just gone absolutely parabolic and, in fact, stayed parabolic.

If you look at the overall index as a whole, we actually retraced a lot of the gains post-election. But if you look at the names that are intimately related to the various Trump trades, including, of course, Elon's, Tesla, and BFF stock, it's actually almost doubled before the election. And it's retraced a bit. I don't

what the market is doing now but it's actually held most of the gains since the election i think that's true for a lot of the sort of names and then of course it's the ai stuff and i just

You can have an environment where the stock market mania gets so big that it trumps the discount rate channel. Now, if you think about the last few bubbles we've had, the dot-com bubble was about internet and proliferation of connectivity.

'08 was about the US real estate market and the subprime lending thereof. And then you had fast forward to 2021, that was COVID initially, and then liquidity and then crypto junk rally.

Now, what is the bubble now? I mean, apart from the previous run-up in the quality evaluation expansion, we're really talking about an AI bubble, AI and automation bubble. Now, if I ask you out of the blue, does AI as a theme deserve to be the biggest, greatest as a bubble in human history?

Because if AI works out, if AGI really is a thing, kind of none of what we do matters. So you have this fundamental uncertainty that's to be resolved, potentially with the upside. And while we are still resolving this uncertainty, that's what makes me think that this has the potential, thematic potential, of being the greatest asset bubble of all time.

Right. And against this backdrop, you can have, you know, sort of sort of other macro forces unfavorable, you know, inflation, let's say it takes up a bit, right? Ten-year yield takes up a bit. Unless we get something where inflation is literally going to the roof like what we did in the 70s, you know, rates were really high in the Dream.com bubble too, you know, and you

you can look up the 10-year yield, and inflation was a bit lower, but inflation was turning around. So anyway, that's what really gives me the intuition. And the more we see market melting up like this, unrelentingly, we took a little breather pause during Christmas, but then we are very much resuming that.

Right. I'm not a technical analyst. I don't put much stock in technical analysts, to be honest. But just looking at the technicals of the market, like breadth is weak. It is getting narrower and narrower. And what people said a year ago about how breadth is weak was BS, but it actually is true now. Like most indices actually were down in December. Excuse me, most sectors were down in December.

So you're saying that there's going to be a, we're in the middle of a.com bubble 2.0. That sounds like a bearish thing because all bubbles pop and then the disaster, of course, but it's not on a short-term basis. It's not necessarily bearish because if you're halfway through a.com bubble, the craziest part is still to end. And that honestly does, I could see that happening because I've been saying on this program for close to a year now that you're maybe longer than a year that we haven't had all these IPOs of these junkie companies and

So all the money has been flowing into high quality companies. We said there's kind of a bubble in quality stocks. And, you know, the thing about a quality stock is if you buy it when it's super overvalued on a long term basis, you're probably total return is probably still pretty good. On a short term basis, you can get killed, of course. But like, I think that, you know, I mean, there's like SPACs that we're going to merge with some AI company. You know, I own the warrants. I sold the warrants. I'm not going to mention the company, but like they haven't done a deal yet.

I feel like if this was a true AI bubble, that company would be public. I feel like there's so many crappy companies that are still privately traded. Venture capital has so much more to dump on the public markets. Do you know what I mean?

And they're going to start going on all the podcasting about how it's the best company ever. I feel like that hasn't happened yet. So far, it feels like as far as a bubble is concerned, there is still a lot of discussion about price to earnings ratios and cautionary tales and so on. I don't know what it was like during the dot-com bubble mania, but from what I have read, it feels like to me,

bubble of that kind gets to the point where everybody throws in the towel and just think it declares a new age. And I don't know if we're really quite there yet. So anyway, I mean, that's an interesting question. Is that a bearish prognosis to predict a full repeat of a dot-com bubble?

I guess it depends on where in the calendar we are. Between 2001 or 2000, I guess the crash and prior, are we still in 1998? Are we in 1995? Or are we in the last quarter of 1999?

So it feels like to me, based on what we just discussed, and you might agree as well, that we are possibly still in 1998 or something, where things can still get substantially crazier before everything comes crashing down. But going back to what we were saying before, in that context, the 10-year bonds, I mean,

in the event of a market crash, the 10-year yield will go down a lot. Yes. But so you said before, I think you were referencing to 1998, the Federal Reserve pause. Maybe I don't know exact timing. Maybe it's because of the blow up of long-term capital management. But anyway, you had this period of the Fed pause, the 10-year yield shot up alongside stocks. So that is a

negative correlation, but bonds selling off. So yields going up, stocks going up. So bonds down, stock up, negative correlation the other way. So I guess stocks are a hedge for bonds and that's what you see. But so if we have that bubble, I mean, the correlator would play it with yields continuing to go up as stocks go up and then stocks crashing as yields collapse.

That's right. And if you see the plot of the stock one correlation, which I can share with you, at least my version, I plot the two year weekly return trailing a two year correlation. You do indeed see that sort of decline towards the negative, right? And it spikes just during the dot com, right? You know, and then it comes back down, you know, towards the definitely negative territory. I mean, sort of as fate will have it, I'm not

ruling out the possibility of seeing something similar where we see the reverse, right? Where the stop-warn correlation plunges towards negative before it jumps back towards positive again. So anyway, that's my little bit of pure speculation about where we are in the macro landscape.

- Yeah, macro most of the time, maybe all the time is speculation. And yeah, needless to say, nothing that you said of Steve Ho, the, you know, sharing his own opinion on a podcast, not representative of your employer Bloomberg, you know, not representative of your PhD work. Steve, we will leave it there. Thank you for giving us the extra 10 minutes and, you know, sharing your macro views. I appreciate it. - Likewise. - Thank you. Just close this door.