All right, welcome to Benny and the Squirrel. It is amazingly already the end of January. Lucky today to have my buddy Mike Green here. Got the squirrel here. And we have solved the three-body problem of podcasts. Mike is in New York. It's 2 p.m. there. It's 11 p.m. here in Dubai and it's 6 a.m. Thank you, Rupert, for waking up early. And we are going to bring to you a nice conversation with Mike talking about inelastic markets and just extending that a little bit. So, thank you, Mike. Mike is the chief strategist of Simplify Asset Management and somebody once called him the Cassandra of passive management. Not me, but I've heard that's been said. [laughter]
Let's remind everyone though that Cassandra's curse was being right and nobody listens. So, well, my curse is that I optimize for being right, not making money often. So, I'm with you.
Yeah, sometimes I do that as well. All right, let's start here, you know, and I remember the first time I ever came across your name was this XIV trade, and it was maybe on Macro Voices, but Raoul Powell's thing, and you were talking about it, and this was an all-timer trade. So it's an interesting thing to start with because this is what I understand from you. You told me it was the genesis of how you went down this journey.
Yeah. So, you know, when I first started doing the work that I've done on the impact of passive investing, you know, the most important thing to understand is there is no such thing as a passive investor. Right? At the end of the day, the definition of a passive investor is somebody who always holds. They never buy. They never sell. And as a result they can't get into the market and they can't get out of the market. It's an interesting thought exercise that can't exist in the real world. And so what it actually turns out that it is, is it's a large systematic allocation strategy that simply says, did you give me cash? If so, then buy. Did you ask for cash? If so, then sell. And when I began to realize the implications of this for the broader equity markets, I started searching around for opportunities to express this in markets where there was even more systematic penetration. And I realized fairly quickly that one of those markets was the XIV, which was an inverse VIX product. That meant that it tracked negative 100% of the movement of the VIX index. It also had a feature to it that had a positive carry component to it because volatility tends to be elevated in forward space relative to the current level. And so it had two sources of return, right? Falling VIX and it also had the shape of the curve that created carry.
As a systematic strategy, it had to rebalance itself every single day to reflect the size of the move that occurred in the VIX. And so this created a couple of interesting features. One is you knew exactly what they had to do in terms of transactions given the size of the fund and any form of predicted price change. The second thing that was really interesting about it was that because it's an inverse product, all that has to happen is the VIX has to double. Really, the VIX futures is what it referenced. The VIX futures had to double in order to cause it to go to zero. In fact, it was even easier than that because there was a force majeure clause built into the operating document that specified on an 85% one-day decline, it would be liquidated. And so it actually became very easy to calculate what type of move was required for this to occur. In the 2015-2016 time period, shorting volatility via the XIV became very popular. You may remember there were articles...
It became the shoe shine boy trade. I lost some money doing it. Just to be fair, for full disclosure, I did lose some money doing it.
All right. Well, I actually had traded XIV profitably from the long side, effectively replicating the strategy using options on the VIX futures. But as I began to really dig into this impact of systematic strategy and the size that this had become, what I discovered was basically the same thing that Stan Druckenmiller, that Paul Tudor Jones discovered in 1987 when he looked at the portfolio insurance trades, that the size of them had become so large that they were dominating the underlying trading and the activity. In specific terms, the XIV complex, the related funds associated with it represented about 70% of the average daily volume of the VIX futures. And this was in a time period in which the VIX futures were exhibiting extraordinarily low volatility. You may remember in 2017 the VIX touched as low as 9.
And so to create a doubling in a low VIX is actually quite easy. All you have to do is have a move in the S&P of sufficient size. That also took on interesting characteristics in which we are seeing echoes in the S&P today in which the VIX itself, the VIX futures had taken on a much higher than historical beta to the S&P 500 because this had become such a popular trade. And so it actually became remarkably easy. You can do the math. We were running a 22 beta on the VIX futures to the S&P. It's usually about five, by the way. We're running a 22 beta on the VIX futures. That means in order to surpass that 85% threshold, you needed to have a 4% decline in the S&P 500. And actually, that type of outsized move is fairly common. It's roughly a 95% probability over any two-year period.
And we recognized that this was actually creating a very different distribution than was priced into the derivatives on the XIV. Actually, the derivatives were expressed against a product called SVXY. And they had made the mistake of listing two-year options on this thing. So, what I recognized was that, you know, that decline of 4% was a high probability event and it was being priced at basically zero probability in terms of the price of the XIV itself. That allowed me to take a risk-limited trade that I could also hold for an extended period of time without worrying about being short something like the XIV, which a lot of people got carried out on.
And it was really that simple. Now, the reason you heard about it and the reason it became famous is also a good lesson on Wall Street. One, I had presented the trade in May of 2017 at a derivatives conference in Las Vegas, and unbeknownst to me when I was giving the presentation, the founder of the product was in the audience, and he and I had a very nice discussion. I think the word is...
You had an Eisman moment.
Yeah, I had a Steve Eisman moment in which, you know, he predicted that one, he claimed I didn't understand the product, and two, he claimed that it would survive a crash of '87, and I said no, I think it's going to go to zero on a 4% decline in the S&P. On February 5th, 2018, the S&P declined 3.9% and the product went to zero, and we had about a quarter-billion-dollar position on against it that ended up paying out very nicely. And that's the story of the trade.
That's fantastic. You know, I mean, and the best part about it is that, you know, as you know, Mike, I have a short V or a long V fund. The carry can be a disaster, right? We both own high yield CDS, you and I, and it's not been very fun. It was fun in Q1 last year. It has not been fun for a while here.
No. And it's been very, very painful.
Yeah. Exactly. And so when you can find carry neutral or maybe even carry positive, which is kind of a unicorn of long vol, that is amazing. So talk about then how it proceeded into you going down this path of becoming the Cassandra of passive.
Well, what was interesting about it, and you know, this is kind of another warning for people, right? The work on passive actually preceded the trade, and so the theory behind it was what informed the structure of the trade. The recognition that these types of events could occur and the mechanics here were very simple. My analysis on passive investing in 2016 was tied to a paper that was released by a partner at AQR, Lars Pedersen, absolutely brilliant individual, who highlighted what I would describe as an error in the assumptions on Bill Sharpe's, who won the Nobel Prize for his work on the Sharpe ratio, you know, Bill Sharpe's 1991 paper, 'The Arithmetic of Active Management,' which really is kind of the tool and thought process that was used to really promulgate the growth of passive investing. It's the source of all the words that you hear, you know, all the phrases that you hear that ultimately passive and active have to be the same in aggregate because they both have to own the market in aggregate. And therefore the only difference between the aggregate performance of active managers and passive managers is going to be the fees. Since passive charges lower fees, they will outperform over time.
Lars pointed out something really important in that paper, which is that Sharpe had articulated the definition of a passive investor is somebody who never trades. Right? You can never get in and you can never get out. Well, that's obviously stupid. But Lars focused on a particular aspect of that. He highlighted that during index reconstitution, when there was an index rebalancing action, that would force trading on the part of passive investors, and that then opened them up to being active investors. They would become active during that time period because their portfolio had to change. His observation was that created an opportunity for active managers to compete directly with passive, and it actually is very interesting that has become the single largest hedge fund strategy: index arbitrage. So he really provided the thought process behind that. My observation was very different, but I think actually much more powerful after reading Lars's paper, which is that there's two ways that portfolios change. One is the index reconstitution, and the second is actions by clients that increase or decrease the cash balance in the fund. That also requires trading. And since those inflows and outflows occur on a continuous basis, it means these strategies can never be passive. They're always active. And once you recognize that, then you start devolving what the algorithm actually is. It really is as simple as: did you give me cash? If so, then buy. Did you ask for cash? If so, then sell.
And once you have that algorithm, you can start to model the implications of an investor with that algorithm growing in size in a market. That's where the insight ultimately came that at a certain size you effectively almost guarantee a terminal event with these types of strategies. We know that systematic strategies can become crowded. And in the case of the XIV, it was the market that effectively had exceeded the threshold that I'd identified for passive to really matter. So it was our first implementation of it. Now, there's also the gambler's curse of don't win first, right? You want to lose money on your first trades on this so that you actually think more deeply about it or decide not to do it. Having had that very successful trade, it turned into, you know, an element of notoriety around these insights that, you know, in some ways locked me into becoming the Cassandra of passive investing. [laughter] And in many ways, I gotta be honest with you, if you had told me a decade ago that I would be still talking about passive and still be considered kind of the world's leading, you know, Cassandra on the topic, I probably would have run screaming from the room. It's, you know, 10 years is a long time to spend of your life on anything, to say the least.
Particularly for guys like us that tend to, you know, like to hunt things and then run to the next prey.
So that was the core insight, right, that this was effectively a systematic algorithm, and then I began the process of basically saying how would this impact the market. And what I did was I built agent-based models, I surveyed active managers to derive their reaction functions, and programmed agents with the active manager functions, and then introduced this passive manager strategy of if you give me cash then buy, if you ask for cash then sell, and effectively simulated the market on a run-forward basis.
Let me, Mike, let me stop you for a second. When you say I'm going to take cash out though, how do you know when people are going to take cash out?
That's actually the really interesting part about it, right? Is that when it becomes large enough, like something like the XIV, and it is a derivative event, you can actually understand what is the impact of an underlying market. In the case of the VIX, the VIX itself is a derivative of the S&P implied volatility. And so the focus actually became on the relationship between the S&P and the VIX. One of my other insights tied to this analysis was that around 2017, we crossed about a 25 to 30% threshold in terms of passive share in which my model suggested that volatility should actually become an increasing function in markets. Something that has largely been borne out. We really haven't seen a return to those levels of realized or implied volatility since. But you know, that was one of my forecasts that we were effectively shifting into a rising volatility regime. It ended up being right, and that obviously facilitated the trade as well.
And so, you know, the interesting thing that I've realized since then, and this literally is a breakthrough that you're aware of, Ben, that has come through in just the last couple of months. I candidly give some credit to being in St. Barts with you and forcing me to think about some of this stuff, is that the unique feature of the XIV was that it was a single security. And so a single security behaves in an extraordinarily predictable way. But the S&P 500 is composed of 500 securities, 503 to be precise. As a result, most of the interesting action actually happens initially within the index. It's effectively a concentration formula because what you are saying is every dollar that comes in rewards the stocks that have already gone up. You're buying more of them the next day with the inflow because they have appreciated. That drives concentration in the index, and it effectively means that the largest stocks in the index, not quite the largest stocks but basically the largest stocks in the index, become the centerpiece of where this activity is actually happening.
I had predicted the consolidation part and had, you know, used that from an investment framework, but the actual really understanding the mechanics of exactly how every individual stock was being affected is something that eluded me until literally about four months ago. [snorts] So we've had some incredible breakthroughs. We're now at a point where we can effectively derive the actual impact of passive on markets, how much it's inflating valuations, how much it's causing individual stocks to rise, etc. This is the paper that you've read, the price-weight feedback loop. The math is really straightforward, right? Stock price increases more than the market average. That means weight increases in the index. The next dollar that comes in buys even more of that stock, forcing it to move higher than the index, driving the concentration. And the equation for this is very straightforward. It's actually been long known. It's used in any sort of block trading analysis of market impact. It's the equation on the right that's unfortunately...
Yeah, sorry about the slide. I don't know what happened there.
It's not that big of a deal. What it's showing you is the impact of an order size Q is approximated by the standard deviation of the stock. We typically will use a 200-day trailing moving average of that standard deviation times the square root of the size of that order in dollar terms versus the dollar value of trading activity. And so you can immediately see why the calculus became very clear that you would have a very large impact in something like the XIV on a large trade because it had become effectively the market itself. It was 70% of the daily trading volume. That math tells you that you're going to have a large impact if a sizable trade tries to go through, which is exactly what occurs when you have a 4% decline in the S&P and the VIX futures move violently. A $2.5 billion fund has to rebalance and try to push a ton of money through a very small hole that breaks markets. And it's really that simple.
W times 1 minus W is basically, you know, at 0.5 you kind of max out at that 0.25. And like it's just kind of math, honestly. I mean, I remember you sent me your paper in the spring or the fall, and I know you're doing a new paper with Harry. And like once you kind of run the math and then you do your own, you're like, 'Hey, it's just the math, right?' And then...
Yeah. Right. It's just the math, and then there's amplifiers, right? There's nonlinear amplifiers like...
I wanted to talk about one of those amplifiers, which is the dispersion trade, right? Where you... I mean, I think this is the sort of, if you like, the nitro that's being poured on top of this effect because you have got just such enormous growth in the positions being taken. And just to clarify for everyone, this is traders that are long single-stock vol against short positions in index-level vol, yeah, and concentrated in the S&P 500. And if you like, they're pushing in exactly the same direction as this passive effect.
Yep, that's exactly right. So, if you think about what the dispersion trade is, the dispersion trade is basically replicating the volatility of the index in the most cost-efficient manner possible. Effectively, you short the index volatility and then you buy single-stock volatility. When you have a highly concentrated or an increasingly concentrated index like the S&P 500, and again, think about that market impact formula, the orders that are coming through for a stock like Microsoft or Nvidia. Microsoft is a great example today for all the excitement around gold. How would a four-trillion-dollar market cap company lose 11% today, like just very quickly? I need people...
Simply because Meta was going in the opposite direction.
Yeah. Right. Basically the power had to go in one way or the other.
Oh, absolutely. The trade I put on yesterday of long gold and short Microsoft paid off brilliantly this morning. No, I'm joking about that. That's not true. [laughter]
Hindsight Capital, Hindsight Capital, until it didn't, right? Until gold put that little dirt nap in.
Yeah. Well, you know that I had already traded against myself before that occurred. Hindsight Capital is wonderful. It's really a really high-return business.
Best returns out there. Better than the Renaissance fund. It can actually be somewhat closely matched by data-mining backtests that also benefit from the hindsight, right? Anyway, the long and short of it is that when you see a company like Microsoft fall by $400 billion in market cap, you have to ask yourself seriously how much trading volume actually happened, right? I mean, the decline happened before the market even opened. We know that we're talking millions of dollars, not billions of dollars of flow in order to generate that type of a change. And so, you know, people don't understand how illiquid and inelastic markets have become. And one of the reasons for that is because of the phenomenon of passive investors. They always hold, right? Microsoft announces bad earnings. Was Vanguard on the call and anticipating good earnings for their holdings of Microsoft? No, they don't care. And so passive... analysts, like, given the updates...
No, not that I'm aware of at least, right? I'm sure that there is an analyst somewhere in the active strategies at Vanguard who was like, 'Oh yeah, I nailed this Microsoft call.' But you know, the reality is that 95% of the assets at Vanguard and other institutions that are passively managed literally don't care, right? So they're not going to change their position.
That effectively, when you have passive strategies running roughly 50% of the market capitalization of Microsoft, it means that you effectively have a very low-float stock, so you end up with incredible responsiveness to relatively small changes in order size, which is what inelasticity ultimately means, to Benny's point that he was highlighting in terms of that 50% weight. Like, this is crazily enough where these theories ultimately lead: that you will eventually see the US market, if we continue down this path, with a single stock somewhere north of 50%. And that's going to happen far faster than you can imagine if these theories continue to play out. Like, it feels insane to highlight this, but that is really the implications that come from it, and it is what's driving that dispersion trade. So basically, you can replicate the S&P volatility with roughly 10 stocks at this point. The vast majority of the volatility is going to be captured by those.
And that's a very unhealthy situation, right? Like, I mean, we're in a situation where you have such concentration, and there's the gamification from what we're talking about with the dispersion trade. I mean, like, let's be honest. I remember being short Tesla in 2020 and all these [__] were sitting there long just on the S&P inclusion trade, right? Yeah. I mean, Lululemon topped out, S&P inclusion, right? There's, I mean, I can go down the list. There's a lot of these things. S&P inclusion, Carvana, which, you know, I don't want to get into Carvana, but let's just say I love the down move that it took yesterday. But it doesn't matter, right? Because the fundamental stuff that we were taught, which is what should matter, and probably over the long enough time scale will matter, in the short run when you have enough rope to run money for somebody, it doesn't matter more or less.
Yeah. I think there's an element of that, and I also think people also remember, you know, the famous Keynesian adage that markets can remain irrational far longer than you can remain solvent, right? The index inclusion part, you know, we referred to this earlier when we were talking about Lars Pedersen's work and the impact that that had. You know, many academics or many practitioners will point, particularly if you work at S&P or Vanguard, they'll point to a diminishing S&P inclusion effect, right? That is absolutely untrue. It's just that the hedge funds are arbiting it out in advance. They've decided to aggressively use statistical techniques to basically identify what stocks are likely to be included in the S&P, and they recognize that they can bid them up to absurd levels. And if, like Tesla, you can bid it up to the point that upon inclusion, Vanguard and BlackRock and other index replication strategies have to buy an ungodly number of shares, you basically get a get-out-of-jail-free card for any form of speculation that you want to engage in, right? You are rewarded for being as irrational as you can possibly be.
Yeah. Yeah. And you know, I, as I told you and you guys know, I started my career at Fidelity, which at the time was the biggest asset manager in the world. And I remember the closet indexers. There was guys who would just closet index and they'd keep their job, right? Because if you closet index often, you beat the index. You know, like you would just by not taking bets, you could keep your job. And you know, at Fidelity, five, ten million bucks a year, go to all the conferences, you know, there was an... I'm not going to say any names, but I remember a number of guys who did that. So there's a whole active part of the world that's doing that to amplify as well, I think.
Yeah, I can actually show you a chart very quickly. Let me share my screen here. This is actually from Paul Hubner, who along with Valentin Haddad wrote a paper called 'How Competitive Is the Stock Market,' which is part of a growing body of research that is really, you know, now coming out afterwards and validating a lot of my stuff. But this actually captures exactly the phenomenon that you're referring to. This is looking at the active manager database. You can see our communication here on the right. This is looking at the active manager database. These are the managers that are still classified as active. And what we're looking at here is their elasticity. In other words, their sensitivity to changes in supply and demand in terms of their impact on price, excuse me. And in the active manager community, traditionally we would run somewhere between 75 and 80% elastic. You would expect that because active managers tend to do the analysis on what a stock is worth and then evaluate the price relative to it. That contributes to a mean-reversion strategy, which is a high-elasticity strategy. Ben, what you're hitting on is basically post-2008, and really it was post-2007's adoption of the Pension Protection Act and the designation of passive strategies as the qualified default investment alternative in 401ks.
That meant that firms like Fidelity and others were forced to go increasingly passive in order to compete for the retirement assets. And so you can see that their elasticity has fallen as of 2020. This has continued to decline, and declining actually at an accelerating rate now, had literally fallen by 50%.
So Mike, let's go to just define, as we talk about elasticity a lot. You know, not everybody likes to read academic papers. But I think there's been, you know, you had this idea a long time ago, which is why I wanted to start with the XIV, and you've done the work, and now you're getting, like, you know, supported by the academic community, you know, Gabay and Cojocaru and the Frenchman, JP Bouchaud. Like, let's just talk about what, when you talk about what an inelastic market is, just define that for us real quick.
Sure. So elasticity in the economic sense means changes in supply and demand to changes in price. This is simply inverting that. It's basically saying how much does price change in response to changes in supply and demand. A highly elastic market is what had always been presumed to be the case. And so this is the efficient market hypothesis actually specifies that the market has nearly unitary elasticity, meaning it can accommodate an almost unbelievable amount of change in volume or supply with very minimal impact on price. This is at the core of the idea behind the efficient market hypothesis, that markets are ultimately somewhat random walk and only lightly steered in the direction of individual expressions by people putting money to work, excuse me, by people putting money to work or taking money out of the market. The actual specification, as you kind of capture here in a quote of, I think it was on another slide, you actually show the practitioners' perception of it, which roughly matches with the EMH.
I had it here and then I put the extra paper, so I gave the Frenchman a little love and you don't have that anymore. I bought...
Yeah. All right. Well, so you used to have a quote basically saying that most practitioners think there's very little change for a buy and a sell. The market impact analysis that was done by JP Bouchaud, which we've had since roughly 2012, contradicts that strongly. It was presumed that markets would revert to that very quickly. That's actually the core of Valentin Haddad's paper, 'How Competitive Is the Stock Market?' The answer is not very. But in standard EMH models, if I put a dollar into the market, because by definition it requires somebody to sell a dollar to me, you know, to buy that dollar from me in exchange for shares, it should have very little impact on the market. Theoretically, it's only supposed to have about a penny increase in market capitalization, roughly the bid-ask spread, right? You execute the trade at the ask, as a result, the price goes up by a penny. Therefore, you had very little impact on aggregate market capitalization for your contribution of $1.
It actually turns out that markets are surprisingly inelastic, as it says on the right-hand side. So one of the benefits of growing computational capability through computers is the ability to do higher-level mathematics around this. Gabay and Cojocaru use something called instrumental variables to effectively isolate the impact of that dollar in and dollar out. And they came to the conclusion that over the period of 1999 to 2019, using an average over that time period, it was between five and eight dollars of market impact on an individual security.
In the equity market, right?
We're talking about the equity markets.
So, just to very quickly parameterize that, right? That means that the EMH has misspecified 800 to 1, using an average from 1999 to 2019. Now, my work also suggests that that inelasticity is growing. That has been backed up by the work of Valentin Haddad, the professor we've mentioned repeatedly who's at UCLA. And what we are actually seeing is trending declining elasticity in the market that is only partially offset by more aggressive trading from active managers for the very simple reason that Ben already articulated. It's really dangerous to do, right? You're basically stepping in front of a steamroller. And so the right thing to do is actually hitch a ride on the steamroller and start to mimic the passive strategies, which then further reduces the elasticity of the market by lowering the elasticity of the active participants, as I just showed. And so that actually is kind of sitting at the core of this. What we have is a market in which we had assumed under a theoretical framework that markets could absorb almost unlimited amounts of passive investing. And what we unfortunately have discovered, and has now been backed up by the academic research, is just that that's just not true, right? This wouldn't be a huge deal if it weren't for the fact that the US government, the Australian government, to the squirrel's credit, and... [laughter]
What's that? I'm not taking any credit for their decisions in the super complex.
No, but this unfortunately has become what has played out in market after market around the world. I'd highlight Korea is a really good example of one that actually has a high degree of passive concentration because when people buy Korea, they don't buy individual securities, they simply buy the index. And so you actually are seeing these types of impacts in markets around the world. It's exacerbated by the size that we're trying to trade. If we try to reallocate from the S&P to the Korean market, you're talking about going from a $70 trillion market to a $2 trillion market.
Yeah. And look what happened this year as a result of that. I mean...
And that is exactly what has happened.
And you know, if you think about it, I remember being in grad school over two decades ago, and they taught us, because that was in the early days of passive, they're like, 'Hey, this is a great idea, but here's the thing: the larger it gets, the more inefficient, ironically, the market will become because, you know, you take the limit Lopold rule math, that if it's 100% passive, that there's going to be no one who's, you know, we laughed about the Vanguard guy, but there's going to be no one who's actually paying attention to anything. So you'll get more and more inefficiency in essence.'
Yeah. Well, what you actually... so ironically, the emergence of the theories of inelasticity helps to explain other features that sometimes confuse people and may lead people to believe that markets are becoming more efficient. One of those examples are things like post-earnings announcement drift. Right? So a company reports its earnings, its price rises in response to that, and then it continues to rise for a period of time. The rationale behind that is basically an information diffusion, right? People aren't entirely certain what to do with the earnings, but they see the price moving higher. They interpret it that everybody else thinks those earnings are great, therefore they should own the stock, causing the price to drift. As a market becomes more passively dominated, that actually ceases to matter nearly as much because what you actually end up with is the largest market participant, once the price has changed, they're like, 'Oh yeah, that's the right price, right? We accept that price.' And so [laughter] they will buy in proportion to that market capitalization, roughly causing the stock to begin behaving just like every other security. There's no attempt to say, 'Oh, that's too big of a move, let's hold off on buying it' or anything else. They simply follow through in it, and it basically causes an extraordinary move to occur followed by the stock behaving very much like the stock market. And that's interpreted as a form of efficiency, right? The information...
And it's an efficiency that is being exploited by the hyper-financialized management teams. You know, ultimately, presenting the optics of a quarter is more important than the quarter itself. Once you can trigger the flywheel in a...
certain direction, you know, good news for those RSUs.
You have discovered the dirty little secret, which is I am certainly not the first person to have figured this out, or even the first practitioner to figure this out. I would suggest Elon Musk figured this out long ago.
And it has gotten bold enough at this point that in 2023, I think it was 2024, it might have been. Time flies as you get old. A board member at Palantir was dumb enough to go on to Twitter and announce that they were switching their listing to the NASDAQ so they could capture the sweet, sweet flows of the NASDAQ, of the QQQs, right? And you're just sitting there like going, 'Oh my god.' Like, 'Dude, you just admitted to stock price manipulation.' Like, what's wrong with you?
It's kind of like when Trump, you know, I was down in Medellin with our friend George Gammon over the new year, and the squirrel and I have been long the Schlumberger on the Venezuela thesis, right?
And they didn't even beat around the bush, you know? They were just like, 'We got to get the oil flowing.' You know, we're at this point now where that's what's happening. It's no different, right?
Yep. Yeah. I know it's, I mean, it really is unfortunately. And so that is one of the phenomena that Benny, that I'm sorry, that Rupert correctly highlighted, is that we increasingly see management teams who are like, you know what, it really doesn't matter if we tell the truth, right? We can massage this seven different ways from Sunday. After we've announced it, somebody's going to come out with a report that says, oh, we disagree, we think there's a low-quality earnings, etc. The act community is too small for it to really matter.
Well, or too late, right?
Yeah.
The price impact is created by the optical beat on an adjusted EBITDAR basis or some kind of BS around RPOs or whatever, right? And that's all you need. You just need that nudge over the cliff in a certain direction.
Yeah. And it's typically released into a market that is closed. That market then trades in the after hours, right? The international fundamental guys will trade, right?
Well, actually, what's interesting about it is it'll typically be released and then you'll actually see passive investors from around the world say, okay, that's the new right price, right? So if you can get the price to move exactly as Rupert is highlighting, you're going to end up in a situation where you cause permanent shifts. And this is again part of Valentine Haddad's work, and the reality is that it is death to short sellers, right?
I can attest to that. I can attest, I can attest to that. Trying to trade after earnings and be like, this is [__] and it just, you're just done. And man, those years you were talking about, 15, 16, 17, were my least favorite years of my career by far.
It's interesting, you know, there's a sort of muscle memory about, oh, this initial move has to be faded, right? Which has just got to be completely erased from your mind. Do not fade initial moves because that's probably got 90% of the future predictability, right? Because of this effect.
Yeah. No, I honestly think that is absolutely correct. And you know, to Benny's point in our connecting at an event, like in all seriousness, one of the side roles that I've taken on in the last decade is being a therapist to short sellers. You know the scene from Good Will Hunting where Robin Williams says to Matt Damon, 'It's not your fault, son. Son, it's not your fault.' [laughter] It's like I've had to use that line with so many friends who are short sellers over the past decade. Just like, guys, first look at who owns this thing, like the most important. And they're like, 'Well, no, it's all Vanguard. They don't know what they, you know, they have no impact. It's passive.' I'm like, 'You have no idea what you're in for.'
That goes back to why, you know, I came to you and I was like, let's talk about the inelastic market hypothesis because, you know, I'm on a chat with a bunch of my old grad school buddies and we were all indoctrinated because we were still before these times of, hey, what's the narrative? What's the story? What are they going to earn? And you know, I know you have a slide somewhere that shows mean reversion. That's how we kind of think, right? And we've moved from this mean reversion world to positive autocorrelation.
And you need to know, look, you know, we'll talk about silver in a little bit, but like it could be silver, it could be Palantir, it could be Bitcoin, it doesn't matter. You know, they rotate what the market is, but it just keeps going. And then if your natural inclination is this is [__], which eventually the Christmas tree of death usually comes through, but you might be out of your seat or out of your client's money or out of your own money well before you get to the other side of the Christmas tree.
Yeah, the market remained irrational longer than you could remain solvent. Actually, let me share that slide because that is actually a really good slide since we're doing some of this in detail.
So this was the proprietary research that I'd mentioned around the behavior of active managers. I went out and I surveyed investors and I asked the question: you're a portfolio manager with 5% cash in your portfolio. That is not going to be a constraint either too much or too little in an active portfolio. Typically, it's fallen significantly as more and more active managers have to mimic passive strategies. But the question was, you receive a new inflow or outflow. What is the likelihood that you're going to deploy funds or sell securities to meet the redemption given some type of valuation?
I surveyed roughly 450 asset managers. The responses that came back were totally unsurprising, right? So, your marginal propensity to sell rises as valuation rises.
Your marginal propensity to buy falls as valuations rise. That's totally unsurprising. Those are standard downward-sloping demand curves. The intersection of the two is actually what's really interesting because it hits at 50/50 at exactly the market's historical average.
And if you build an agent-based model based on these inputs and then you feed it cash or you take cash away simulating economic cycles, what you get is a mean-reverting market.
And that's why we're all [__] because that's the natural way of things essentially. It is absolutely the natural way if you analyze the fundamentals first, right? If you are...
That's what I mean. Yeah, that's what I mean.
And so the question then becomes what happens when you introduce passive? Well, again, the right way to think about passive is not as an actual passive investor, but as somebody with 100% propensity to buy or sell on flows. And so as they grow in share, what's actually happening is you're shifting that market away from that historical mean reversion towards a level at which you actually experience much greater propensity to buy or sell. Flows become much more important and the market effectively starts rising in valuation as passive share grows. It's really that simple. All right, we've removed the filter of valuation.
So, I'm gonna put this slide up because I know you're writing a book, buddy. And I know we all like AI. So, I fed this into the NotebookLM. And this is a shout out. It gave a summary of your whole spiel here in like one minute. By the way, all of us who grew up making slides, that job's over.
Yeah. Pretty incredible. This is, I was really impressed with this actually. But um...
You can have it. I'll send it to you.
Well, it's also funny though because you can see that AI doesn't fully understand what it's doing. So, you know, money is coming in and then automatic selling somehow shows up here. But it's, you know, the order is really what you should be thinking of in terms of that component. It is just a super simple algorithm and if you play this through it has some really significant impacts. One is the dangerous consequences on the right. You know, as passive gains share, the market becomes more inelastic, that dollar flow rises from five to eight, I currently estimate that it's around 20 as you correctly capture here, that price-weight spiral drives concentration as the largest stocks attract more demand, their volume doesn't necessarily increase in a proportional manner. In fact...
One thing that requires you to be a deep practitioner to understand is that the process of market making, the profitability of putting capital up to making markets in stocks, is a byproduct of the width of the bid-ask spread and the volume that is traded. The largest stocks have very low bid-ask spreads and their volume roughly scales with about the square root of their market capitalizations.
So it doesn't grow proportionally. That means that the vast majority of market making operations in the largest stocks are actually kind of just loss leaders. They really don't want to put that much capital up against them. And more importantly, they're interested in market making of derivatives on those products which are trading in a totally different manner as well. So when you end up with those contributions, you end up with exactly what we see: a market that becomes increasingly concentrated, increasingly volatile at an aggregate level. The individual securities that are the largest components of it become a larger and larger fraction of the actual price movement in the index and the rest of the components.
This is actually very important that you're hitting here on the correlations. This is showing the correlation between the S&P 100, which is the largest stocks, the S&P 500, and an equal-weighted index of the same things. And you can see here the last time the dispersion trade actually was really popular was in the dot-com cycle when you could do very similar things. You could effectively replicate index volatility by buying the most volatile stocks like Microsoft, Cisco, etc., the technology stocks that gave rise to the dispersion trade that disappeared in 2003. The reason it disappeared was because we changed the structure of indices from market cap weighted to float adjusted. That meant that low-float stocks no longer had a disproportionate impact from passive flows. I actually call passive...
Small caps? Not all the small caps, but like IPOs maybe, or...
Actually, yeah. So it's worth explaining why I call the dot-com cycle passive 1.0. In the early 1990s, Vanguard and other index providers, as they grew to roughly 2% of the market, began to experience an individual problem called tracking error. As they would try to put money to work buying the smaller stocks in the index, they would cause their prices to diverge from what the index printed. That meant that their performance was diverging from the index.
They went to Wall Street and they said, 'Hey, what can we do about this?' Wall Street said, 'Well, why are you buying every individual stock? Why don't you just buy futures?'
And like, well, we can't do that because futures require a margin account and a margin account requires leverage and leverage is not permitted under our offering documents in the 40 Act. And the street said, well, that's kind of silly. Why don't you just lobby the SEC to get an exception because you're really not trying to speculate. You're just trying to deliver the lowest cost, best access to the markets. And the SEC looked at it and blessed it and said, 'This is great. Of course, you're allowed to do this, right? You uniquely, by the way...' That's how it works, right?
We don't let people speculate, but we're gonna let you do this, right?
That actually really kicked off the growth of the futures business in the 1990s. You know, Ben, you're a little bit too young. Rupert, you're probably roughly my age. And so, you may remember like the futures markets were largely a backwater after '87. There was no real use of them. People were not aggressively trading them. And then over the course of the 1990s, they exploded in volume. Of course, nobody really asked why, but the answer was ultimately that the indices were using them for replication of the index on the initial inflow and then slowly moving into the individual stocks once they had position, effectively exchanging the future for the individual positions.
If you're going to replicate the index, you've now turned that problem into a streetwide arbitrage issue. And so maintaining the relationship became the responsibility of basically everybody on Wall Street.
And because the indices at the time were market cap weighted, that meant that they were trying to buy in proportion to all of the shares that existed, whether they were held by Bill Gates or whether they were held by a mutual fund.
Bill Gates was unwilling to sell his shares to you. As a result, that meant you were trying to buy twice as many shares of Microsoft as were actually trading. And of course, that has a very large market impact. We exceeded that market impact metric relative to the dollar trading volume because we're trying to buy twice as much as actually existed in free float.
In 2003, they changed the rules. They changed it to float-adjusted weighting. And when they did that, they destroyed that mismatch between the liquidity of insider ownership stocks and broadly high free-float stocks that killed the dispersion trade for almost two decades.
Would you argue then, here, let me put this up.
Would you argue that was why small caps had their nice run in the post-dot-com? Because you look, this is the Russell divided by the SPY or SPX.
I think that was more, there's a factor element to that as well, right? It was more of a...
There wasn't that many factor flows back then.
Well, but the one factor that did exist was actually the size factor and the value factor that has been identified by Dimensional...
Chicago boys. I remember grad school. Yeah, for sure. That's right.
Right. Now, that is in and of itself another systematic algorithmic strategy that attempts to replicate what they think active managers are doing, right? And biases you towards a particular outcome. What very few people realize about a systematic value strategy is it's actually a short-vol trade. The reason it's a short-vol trade is you have agreed in advance how you are going to behave if your stocks behave in a certain way.
Right. You are long stocks that by definition you are willing to sell if they rise above the threshold at which they switch from a small cap index to a large cap index. You've agreed in advance you're going to buy stocks that fall below that threshold. That means you are short call options on all the stocks that you own and you are short put options on all the stocks you don't own.
I just, absolutely terrible strategy. So it actually makes sense why value guys only know pain, to be honest.
There is an element of it, but I also want to emphasize like this led any number of value guys, somewhat myself included, you know, down the wrong thread, which is this idea that well, small cap value outperforms, right? And so portfolios used to allocate on a factor, there would be a factor allocation that say, okay, we want to have a 10% allocation to small cap value because small cap value outperforms over time. Ken French and Eugene Fama told us so, right?
What I actually think happened to cause this reversal is people stopped. We used to allocate basically to managers on a three-year time horizon. And if their performance was good, we would let their portfolios accumulate relative to the rest of our portfolio. And if their performance was bad, we would basically give them time to see their strategy prove out because we didn't want to make short-term decisions. And so what happened over that '97 to 2000 time period was basically the technology fund managers' allocations grew exponentially and the small value managers really didn't grow very much at all because they weren't receiving this flow tied to the indices.
That's again why I call dot-com passive 1.0. Somewhere around, you know, if you think about the math of why it broke in March of 2000, the answer is pretty straightforward. Investment committees get together in January and discuss how they're going to reallocate their portfolios. And I literally felt this and lived this. I was a small cap value manager sitting at my Bloomberg when the NASDAQ cracked. The thing that people don't realize is the NASDAQ cracked and the Russell 2000 value soared.
What was actually happening was people were reducing their allocation to technology stocks and buying small cap. No different than what we just saw in this most recent blip.
Right. So when we go back to that, you know, I'm again back to my grad school chat and the guy, you know, 13 days of Russell outperformance and you know, I guess it outperformed today. It's just beginning, right? Like so there's this narrative that it's going to come back. Would you, it looks like it's still in a downtrend to me, you know, as I see it. It's like, you know, there's volatility around these things, but what is this the year of small caps or you think that this is just a, we reallocated some money?
So, there's another really interesting feature to this. If you notice, that trend peaked and began to reverse right around the 2012 to 2014 time period. That is another government regulation at work. So in 2012...
Under lobbying from Vanguard and other participants, the SEC changed the... it was not the SEC, it was the Department of Labor changed the specified qualified default investment alternative from balanced funds, which in many cases were managed actively. So they would have an allocation to equities, they would have an allocation to bonds. What type of allocation to equities, kind of up to the manager, right? They try to pick individual securities. They changed the qualified default investment alternative, what you are defaulted into when you participate in a 401k. You can actively choose to change that, but something like 95% of participants never change their allocations once they've been opted into these programs. And we switched it to target date funds.
