Inside Active: Simplify's Michael Green on Passive Distortions

Well, when I focus on passive, first of all, I just want to be clear: there is no such thing as a passive strategy. That's one of the key points that I raised in my analysis of the concept of passive investing as a thought experiment from the 1960s that was projected forward and given additional depth by Bill Sharpe in 1991 with a paper called "The Arithmetic of Active Management," which is the source of all the analysis that you hear that active managers and passive managers own the same underlying securities and therefore the only difference between them is the fee structure. That's what drives the outperformance of the passive strategies, quote unquote. Unfortunately, Sharpe's definition, as was highlighted by a partner at AQR, Lars Pedersen, in 2016, relies on magic, because his definition of passive is somebody who never buys or sells. They simply hold. That means it's impossible for them to get into the market. It means it's impossible for them to get out of the market.

Once you realize, and he defines an active investor as anyone who is not passive, so by definition all investment is active. What we are really referring to when we talk about passive is the attempt to replicate market cap weighted indices that largely dominates the passive world. There's lots of discussions around active ETFs, etc. Most of those are some variant of two-times levered Nvidia or call overwriting strategies, etc. They aren't really active in the traditional sense. True active investing relies on the discretion of the allocator, the portfolio manager, to select securities on the basis of candidly whatever they want, right? They could choose all stocks starting with the letter A, and they could also decide that they wanted to focus their research on predicting the forward cash flows of a company. Both would be considered active.

When we're talking about passive and when we talk about the scale and size and the impact, what I am largely referring to is market cap weighted indices. In particular, the S&P 500, to a lesser extent, the NASDAQ 100 and extensions thereof like the total market indices and the Russell 2000.

Well, I think it all starts with that label, right? They are not passive investors. So, I think we actually have to start with that observation. The moment we recognize that they aren't passive and instead they regularly buy and sell securities, we know what they actually are, which is they are systematic algorithms that when given cash will buy things in proportion to their market capitalization, and when asked for cash, they will sell things in proportion to their market capitalization.

That has very predictable effects on price behavior, on what's referred to as market elasticity, how prices react to changes in supply and demand. And at this point, the academic work and the theoretical work has moved so far beyond where the practitioner space is that we increasingly are aware that this is turning into the driver in markets today.

Oh, we've been there for a very long time. So this is actually what I refer to as Passive 2.0. Passive 1.0 actually began in the early 1990s as Vanguard and other participants in the passive space grew to 1 to 2% market share. They began to experience the bane of passive investing's existence: tracking error versus their indices. This is because they'd grown to a size that when they went to transact in all the individual securities in the S&P 500 or in smaller indices, they were having a noticeable impact on those securities, creating conditions under which they would experience tracking error. They were diverging from the indices. They went to Wall Street for helpful advice: how can we solve this? The street kind of looked at them weird and said, you know, why are you buying every security? Why don't you just buy futures? And the answer to that was, well, we can't, because the 40 Act restricts the ability of mutual funds to access leverage. Margin accounts are required to trade futures, and so we are restricted from using futures for this entry process. The street recommendation was to lobby the SEC. Vanguard did so, received what's called a no-action letter in 1994 that allowed them, somewhat uniquely along with other index investors, to ignore the rules of the 40 Act and to begin to transact in futures. That ran headlong into a market structure issue, which is at the time the indices were market cap weighted. Most people think they still are. They're not. They're float adjusted today. That change happened in 2003.

But in the early 1990s, what it meant was that passive investors were trying to buy more shares than were actually freely available in low-float stocks, companies like Microsoft, Cisco, Dell, etc., stocks with high levels of insider ownership. And we began to see the effects of this almost immediately in market structure and market behavior. Passive 1.0 was really what facilitated the dot-com bubble. In 2003, we changed the structure of the indices to float adjusted. That bought us additional capacity, and my work suggests that somewhere around 2015-2016 we exceeded the thresholds that that increased capacity allowed us and began to meaningfully impact markets again. Today, I think it is the single most dominant force in markets.

Sure. So the premise of elasticity within markets is something that is assumed within the efficient market hypothesis, which is really the primary underpinnings of the passive investment revolution. The idea that market prices represent all the best information that is available at any one point in time, and people are able to invest almost unlimited sums without disrupting that market. The actual specification for the efficient market hypothesis is that a dollar into the stock market, because there has to be a seller on the other side, really only changes market value by roughly the bid-ask spread. So a dollar in creates about a penny of additional market cap, as the EMH framework.

In 2020, two academics, Efim Gabay at Harvard and Ralph Koijen at Chicago, did a very disciplined analysis of this and found that the average over the time period from 1992 to 2019 was actually between about five and eight cents of market capitalization being created for each dollar that was going in. Their paper was called "The Inelastic Market Hypothesis" and was really one of the primary academic revolutions that is underway in finance right now, understanding this concept of elasticity and how it is changing in markets. But really importantly is that observation that it was an average from 1992 to 2019 that was cited in that paper. Subsequent work by Valentin Haddad at UCLA has identified that this is actually a trending series that is a function of passive share, and my work at this point, which Valentin would largely echo, is that we are looking at somewhere for the mega cap stocks like Apple, Nvidia, Microsoft about $70 of market capitalization being created for every dollar that is going into the market.

This is why you can see a $340 billion market cap swing on a company like Microsoft. Nowhere close to $340 billion, or more accurately, on the EMH, you know, $340 billion would be $34 trillion in order to get that impact. It really was more on several hundred million to possibly a billion dollars in relatively thin markets that are closed that cause that type of impact. So we're discovering that markets are radically less elastic than we had assumed.

Well, the primary component that's being distorted is what's called price discovery, right? So we actually effectively have a market that no longer has a form of memory. We as individual humans experienced a variant of this in the COVID pandemic. We'd go to the grocery store, we thought that eggs were supposed to be $3 a dozen, they're $5 a dozen. That creates chaos, right? What's going on? This doesn't make any sense. As we continued through that pandemic, people increasingly became unhinged from the historical price level and they basically accepted whatever price was available for eggs. That's the same phenomenon that happens with passive. You effectively lose the memory of what is a meaningful price level and you effectively adopt whatever the last price is. Passive is like an investor who has zero memory.

Well, it is unfortunately a true systemic risk, right? What we have actually done is we introduced an agent who we effectively, in a paperclip world variant, right, when you think about automation, we just said, well, go buy at what price? Whatever price you can get right now. What that has actually done is inflated valuations. If you think about the impact of a growing pool of agents whose response to "at what price?" is "whatever price, I don't care," the people selling to them will be able to extract higher and higher prices. Over time, those prices will become increasingly disjointed from the fundamentals. We'll construct a narrative around it, right? They're better companies. They have better growth opportunities. Productivity is going to surge right around the corner, etc. Because that's what we do as human beings, right? Remember, we as a species for thousands of years believed that the sun transited the sky being dragged behind a golden god. We will literally believe anything, including stuff that Elon Musk says.

You know, the simple reality is that when you create those conditions, you set the stage for what was very well articulated in a 2004 academic paper by Michael Jensen called "The Agency Cost of Overvalued Equities." And he highlights that effectively, if an equity becomes overvalued, there really is no mechanism to discipline the management team, and they have every incentive in the world to justify the existing valuation in their behaviors. If they believe that they are being rewarded for incredible growth opportunities, they will invest like they believe there are incredible growth opportunities. And unfortunately, I think that's what we're largely seeing. On the flip side of it, many industries that suffer from neglect under this framework are told that capital is not available to them, and as a result, they underinvest.

And that broadly is a pattern that's playing across our economy where privileged companies that happen to be large enough to be in the public equity indices are receiving differential cost of capital and a much lower cost of capital that allows them to invest, while businesses that are not privileged in that way, your local mom and pop grocery store, hardware store, are suffering from neglect and an inability to obtain investable funds.

I think it depends on which regulators you refer to. I like to point people to a conversation I had in 2018, soon after the events of Volmageddon, which I think you both know I was involved with, in which I met with the Bank of International Settlements and their financial crisis group. And their reaction to my work was, "Yeah, we think you're right," and they've subsequently written a few papers on it that tread very lightly on this topic. My reaction to that was, "Oh, that's fantastic. How can I help? What can we do to ameliorate this condition?" I did not go public. My goal was not to incite panic in the crowds. My goal was to get regulators to change it. And their response, unfortunately, was there's nothing we can do because what you don't understand is the regulatory apparatus is controlled by BlackRock and Vanguard. And if we raise the alarm and the event didn't happen immediately, we would simply be fired.

Well, the answer is both, but it's just a function of time scales, right? So, this is a variant of what John Maynard Keynes observed: the markets can remain irrational far longer than you can remain solvent. As long as people are piling money into passive strategies and passive continues to gain share, there's really not much prospect for active managers on a continuous basis to underperform. That's actually a direct byproduct of how we calculate things. So if you think about a market that has a function that causes valuations to increase over time, in other words, the returns at any unit of risk are actually rising, that creates a positively convex surface.

The measures that we use, as you talk about alpha, for calculating active manager performance is actually a linear equation, y = mx + b. The return on my portfolio is the beta times the market plus some idiosyncratic factor we call alpha. That alpha is simply the intercept. And if you just draw a curved surface and you run linear equations on it over time, you will actually find that those intercepts are forced negative, that perfectly matches the experience of the active manager community. So very quickly, this is actually proprietary research that I did back in 2017. I asked the question of portfolio managers: You're a portfolio manager. You are given new capital or capital is taken away from you. What's the likelihood that you'll deploy funds or sell securities to meet a redemption given some type of valuation?

Valuation is on the x-axis. Your marginal propensity to either buy or sell is on the y-axis. These are the cumulative results across roughly 450 responses. And unsurprisingly, you find a marginal propensity to sell as valuation rises and a marginal propensity to buy as valuations fall. That's not surprising. Those are standard downward sloping demand curves in economics. It's this intersection between the two that's actually really interesting. The two cross, and remember this is just a random survey by an idiot, at the market's historical valuation average. That's actually really important because if you build an agent-based model, feed these responses in them, then randomly give them cash and take cash away, what you discover is a mean reverting valuation in markets. This is why things like the Shiller PE have historically worked.

If you introduce passive investors into the mix, remember that they have a 100% marginal propensity to buy and sell. Did you give me cash? If so, then buy. Did you ask for cash? If so, then sell. That pushes the marginal propensity to both buy and sell off of that mean reversion towards what's called mean expansion, that takes the form of rising valuations over time, which unfortunately perfectly matches the underlying data that we see.

This is the valuations rising. This is the median stock. This is the market cap weighted stock, which again plays into theoretical frameworks of why the largest stocks are more impacted on this. So we see rising valuations over time. If we then think about that in the context of how we measure manager performance, historically we're used to thinking about a mean reverting market. The reality is now we have a mean expansionary market that causes those valuations to rise. If I take that same linear equation and I run it at various points in time against different passive share, my alphas are pushed negative. And that is unfortunately exactly what we're seeing.

So contrary to the argument that most people focus on, which is this idea under a Grossman-Stiglitz type framework that the opportunity set for active managers is growing, the path to that is actually very destructive to the process of price discovery and market functioning and ends up killing the active managers along the way. That creates a feature that's been very well documented, what's called effectively defection. Active managers increasingly become closet indexers, basically trying to preserve their career by minimizing their tracking error against that benchmark.

It's largely a byproduct of regulation and the repeated experience. So in 2006 we passed the Pension Reform Act, the Pension Protection Act, I'm sorry, that changed 401ks from an opt-in framework where you would get a job and then choose to participate in a 401k to an opt-out framework where you would get a job and you would automatically default it into a 401k.

When you created that, you had to establish what you were going to automatically default them into. That established what's called the qualified default investment alternative. Initially that was balanced funds. There was some competition from active managers, but very quickly over time we established legal precedent that said if you choose anything other than the lowest cost passive exposure as the sponsor of a 401k, you run an inordinate risk of being sued by your own employees. My favorite example is a class action lawsuit by Fidelity employees against Fidelity for offering them actively managed Fidelity products.

So unfortunately what you're describing, James, is exactly what you would expect because of two features. One, fewer and fewer active managers have access to that pool of capital because of the QDIA. About 85% of 401k participants never change their designation away from that qualified default investment alternative. And as a result, there's really no new entrant customers into the active manager space, or a very small fraction of it. The second feature is that then means because there are no new customers, the worst thing that can happen is losing an existing customer. And the easiest way to lose an existing customer is by underperforming versus your benchmark. Outperformance will retain it, but...

Underperformance guarantees that you're going to get fired and your business moves rapidly into a terminal stage. This has driven a very clear defection that's actually well documented by the work of Valentine Hadad and Paul Hubner in the behavior of active managers themselves. They have shifted from being historically highly elastic and sensitive to valuation to today they're effectively very close to passive investors in many ways.

Well, first, we just described actually that they are, right? As they move to become more benchmark-like and less elastic, they are mimicking the behaviors of passive investors. Secondly, the largest and most successful active managers have done exactly that. By far the largest business in the hedge fund space as a single distinct strategy is index arbitrage, effectively trying to front-run the passive investors on events like index inclusion. Going beyond that, if you look at the production of research and where value is created in research today, the single most valuable piece of research is not 10-Ks or analyst reports. It's what's called payment for order flow, which effectively gives details of what the noise traders, the remaining retail traders who now outnumber active managers roughly 2:1, what they are doing. Citadel pays on the order of $4.5 billion a year for access to transparency in the order books of Robinhood and other participants precisely so they can position and trade ahead of that and maximize the value that they're creating in their option models and option market making. So James, the quick answer is they are doing it, right. There's also a lot of people out there that are still unaware of this phenomenon and really don't understand how to exploit it. And then finally, there's a lot of people, myself somewhat included, who are basically saying, look, you know, this is a systemic risk that we are creating. Your question to me, James, is some variant of, you know, well, if everybody is getting high, why don't I get high with them, right? Somebody has to keep an eye out for the cops.

Yeah, I mean, effectively you can isolate, like, and this is a lot of my more recent work, and it highlights, you know, just how dumb I actually am, right? It took me almost a decade to get to the point that I fully understood some of the mechanics at the single stock level, and I'm still learning. I want to be very, very clear on that. But you know, the simple reality is that the only thing you can do is basically build predictive models to understand the impact and how that is changing. And again, that leans right into the index arbitrage, which I would highlight is happening not only in equity markets but is also actually the story behind the emergence of the basis trade, which is simply index arbitrage happening in fixed income markets where in many ways I would argue the current impact is equally if not more distorted.

I mean, the quick answer is at scale it's incredibly hard, and you're asking people to develop a skill set with a lack of understanding of the mechanisms, right? And so, you know, we're ostensibly having a debate about whether this is actually happening. If your assertion is it is definitely happening, and you're willing to go out and make that claim, and identify that Vanguard and BlackRock and the growth of passive investing is having a clearly caustic effect on the historical functioning of markets, and Bloomberg would stand behind that, I would bet that many more people would actually do it. But that's not where we are today. And so I think it's unsurprising that most managers try to continue doing what they were doing before.