Mike Green: Updated from Passivania

The case for passive-impact is compelling. So what?

Mike Green, Chief Strategist at Simplify, has been kicking around the active-passive debate with me for years now, and I find him to be one of the best-read, and honestly most open-minded people on the topic. We’ve had an absolute raft of new academic research on the impact of passive on markets and market structure, so rather than hop on a phone call on a Sunday morning, we recorded it so you can watch along, over at Excess Returns podcast:

This is a topic that makes people deeply uncomfortable for one big reason: it forces us to embrace a seeming paradox. It is, to my mind, almost irrefutable that the mathematically correct allocation for most investors will include at least some, if not a predominance, of low cost, passive products. It’s how most of my money is invested. It’s likely how most of your and your clients money is invested. And that is good, smart, and logical.

At the same time, the impact of passive flows (not ownership) have demonstrable, now multiply tried and explained impacts on how capital is actually distributed, which axiomatically impacts prices and how markets function.

This is uncomfortable because it feels like you should do something about it. If passive is breaking price discovery, driving real capitalism into the less-regulated dark, changing volatility and how news is processed and favoring certain kinds of activity over others, it really makes my brain itch to not do anything with that belief.

So why learn about it? Because it will help you understand reality, and as my friend Barry Ritholtz says/steals all the time “Your first job as an investor is to accurately perceive reality.”

So here are three additional resources for ya:

• Key Quotes

• Academic Research

• Full Transcript

Key Quotes

[00:04:58] "There is no such thing as a passive investor. There can't be a passive investor."

This is the brain breaker for a lot of people. One can only be passive if one never takes action, and the simple act of buying and selling the supposedly passive vehicles, like an index fund, actually have “active” market impacts in unintuitive ways. Quite a bit of the recent research shows mathematically how this sharp departure from Bill Sharpe's original theoretical framework has manifested. For all the big “it can’t be!” hand wringing about this observation, I’ll reiterate one of the the most important footnotes ever, from Sharpe’s original work:

“We assume here that passive managers purchase their securities before the beginning of the period in question and do not sell them until after the period ends. When passive managers do buy or sell, they may have to trade with active managers; at such times, the active managers may gain from the passive managers, because of the active managers' willingness to provide desired liquidity (at a price).”

Obviously, this is a ludicrous assumption about how actual trading and price setting happens. Sharpe assumes that NO AMOUNT of passive buying — a Trillion dollars at the market — has NO impact on pricing. That’s obviously silly. Just consider that last week, 0.4% of Bitcoin being sold (the ultimate passive, inelastic security), resulted in about an initial 10X reduction in market cap (about 4%).

[00:25:45] "For [the mag seven stocks], the multipliers are now crossing into triple digits."

I think this is something folks really, really miss about the passive-perterbs argument. Think about how likely a non-index owner is to sell you stock if you show up with an order. If you want to buy out all their NVDA, well, that’s a big market risk for them. NVDA is *important*. It’s performance is wildly idiosyncratic. It’s a meme. If you want to buy out all their Mohawk, a midcap industrial with .01% in the S&P 500, the really don’t mind just taking the spread. It doesn’t impact their portfolio that much.

That “reluctance to sell” is also known as “inelasticity.” You can express it as “how much something moves based on a unit of flow, ” or “k” in the literature often. In Oil, for example, k can be 20. $1mm in new flow — new demand for oil — translates into $20mm in “value” (like market cap) to the market. A small movement in price is a big movement in “capitalization.” A lot of oil trades. But it’s hard to turn on new supply overnight.

Not every individual stock shows aan identical elasticity or “k”-multiplier effect. There are multiple reasons for this (well documented in the literature now). The most obvious is substitution. Active managers don’t mind not-owning a little Mohawk (MHK) (0.01% of the S&P 500) if a big buy order comes in. They really care about being light Nvidia (NVDA). There are lots of substitutes for mid-cap industrials. There’s really no substitute for NVDA. It’s more important, so active-management sellers are reluctant.

Another reason for elasticity differences is that liquidity doesn’t scale with size. NVDA isn’t 740 times more liquid than MHK at the top of the book, where prices move. If NVDA is even SLIGHTLY less liquid than that capitalization spread suggests, then there are in fact real “run the book” issues with the larger stocks, more than there are with the smaller ones, for a given amount of index-allocated buying.

So, every stock in the S&P has it’s own supply and demand influences, and thus, it’s own elasticity. It’s own “k.”

For these reasons (and more), the actual elasticity of big “liquid” stocks certainly seems to be lower than that of less-headline stocks. This isn’t theoretical, it’s observed (links below, but here’s a banger). So if NVDA’s elasticity is lower than MHKs, each incremental dollar into the index increases the next index-strike’s NVDA weight. I’m doing an ELI5 on this this week, but it’s an important nuance.

I have a lot of concerns about the runaway nature of American capitalism right now: it seems that more and more power is accreting to fewer and fewer actual individual decision makers — billionaires we all know by name. Being at the top of the index-cap-table actually accelerates this problem, driving the top firms to higher and higher multiples. While I’m skeptical of some of the wildest claims not in the research, Mike highlights that these inelastic market pricing effects can become extreme — with Bitcoin being a petri dish.

[01:20:09] "What we have actually created is a narrative that, oh my gosh, we're losing control of the long end. My work says there's absolutely no truth to that whatsoever."

Bond indexes are nearly a self-writing joke: they’re almost universally based on giving more money to the biggest borrowers, and within a given borrower (like the Treasury), giving the most money to whatever that borrower issues. This has the effect of, essentially, forcing the bond indexes to “chase” the issuance pattern of the Fed. So if the Fed never issues anything but 2-years ever again, over the next 20 years, the indexes would all just be come 2-year Treasury indexes. It’s a sort of dumb point but one that obviously has market impact. Mike identifies a specific distortion: older bonds (issued with low coupons) now trade well below par, and thus will get less pressure from passive flows. This technical factor, (more than or in addition to) fundamental concerns about inflation or fiscal policy may be driving the yield curve. A new angle I hadn’t thought through.

[00:33:47] "Eventually you get to the point where the withdrawals exceed the contributions and then it becomes a question of does it correct quickly or does it correct slowly? … What passive suggests is that it would be a very quick, very sharp and nearly continuous correction."

This is the kind of “doom quote” that gets headlines and which I have mixed feelings about. Yes, mathematically, there is real, measurable inelasticity, and that should be bidirectional. So yes, if a LOT of passive all sells at the sametime, you would expect a dramatic and violent reaction — like, say, perhaps we saw in April in both directions.

But, the trigger for this kind of coordinated unwinding isn’t obvious. Yes, a narrative common-knowledge panic where everyone from Boomers to Zoomers smashes the Red Button would be bad, and passive dominance would make it worse. But structurally, the simple “rolling off” of Boomer portfolios into bequests doesn’t get it done, because one would expect the risk tolerance of those inheriting boomer portfolios to be higher — more market oriented — then the 90 year old dying with their bond ladder. You have to imagine all the boomers selling their last slugs of equity perfectly to spend on consumption before they die, which seems unlikely. The devil, as always, will be in the minute details of who sells, why, and how quickly.

[00:45:00] "Is it a great thing to provide [people] with an incentive to encourage people to save for retirement? Absolutely... [But] now the government is very firmly putting its thumb on the scale and providing differential capital costs for this selected group of companies."

And here’s the uncomfortable nugget laid bare: “Trump accounts” are both doing a good thing and a problematic thing. $1000 accounts for every child? Honestly, as a left-leaning person, I love it. It’s one of the most progressive redistribution policies I can think of, given that birth rate is inverse to wealth (while the very wealthy have more kids than the middle class, by numbers, lots more poor kids are going to get these accounts than rich kids).

But even as someone who’s “pro indexing” for most use cases, I found the mandate of sub-10bps indexing in a piece of legislation to be pretty wild. I’m not saying its a bad idea for Junior’s account to be in VOO, I just think it’s a pretty clear giveaway to the largest asset managers — no upstart asset manager can run a fund for 10bps. These accounts, if successful, will simply funnel more “never-selling” money into the S&P 500, which will keep the inelasticity train rolling.

Something can be both great and problematic at the same time.

Academic Research

Here are all of the key academic citations followed by the full transcript. Note these are just the ones I have read, and generally in preprint. If you want a quick take on the latest research, read Larry Swedroe’s piece at Alpha Architect. For anyone else wading into this debate, I’d suggest spending a weekend with a notepad seeing what the PhDs have been up to (AI used to help format, not read):

Dozens more, and transcript, follows: