Who's Afraid of Index Funds?
In this issue:
- Who's Afraid of Index Funds?—The critiques of index funds are usually directionally right, but overstate the case. Any claim about market inefficiency is implicitly a claim that nobody's going to exploit that inefficiency, which is a hard case to make
- The Market for Data—Companies are licensing their data to AI businesses, but they aren't making much just yet.
- Crowded Careers—China is experiencing a bullwhip effect in education: too much college, not enough jobs that require a degree.
- Black Boxes, Everywhere—Netflix joins larger ad companies in getting better at targeting in a way that entails sharing less information with advertisers.
- Externalities—Local news subsidies try to recreate a brief period where local news' economics were unusually good.
- Intra-Elite Conflict—The rise of asset managers over big banks.
Who's Afraid of Index Funds?
In 1976, when Jack Bogle launched an index fund First Index Investment Trust, he knew he was throwing bombs, and his actively-managed competitors threw bombs right back. Index funds looked like a tacit admission of defeat in the quest to find the right stocks to buy—but as indexers have pointed out, active management had been defeated in its effort to beat indices over long periods. Since then, indexing has gone from an obscure strategy to around 40% of the market as of a few years ago, based on the volume of trading that happens during rebalancing. And, as those index funds have competed more with other assets, they've pushed some active managers to move their portfolios closer to the index, while the ones who don't are increasingly heading in the opposite direction, by reducing their exposure to indices and other factors to zero so they only get paid based on the idiosyncratic performance of their bets ($, Diff).
This has not been without controversy. Indexers are worse than Marxism, according to a Bernstein research paper. A legal theory holds that they turn corporate America into a profit-maximizing cartel. They concentrate power in a tiny number of unaccountable people. They're increasingly concentrated on a handful of companies. And, from time to time, active managers will complain that either passive investing or "algorithms" are distorting the market and making it impossible to make money.
The short response to this is: if you are in the target market for index funds, i.e. you don't enjoy obsessing about markets and do not have special insights into some company or sector (or your insights are the special kind that it's illegal to act on), then index funds are almost certainly the right choice. They do exactly what they say, they're cheap, you'll have some exposure to every major market theme, and, given the size and adaptability of US capital markets, you will end up with exposure even to the themes that are dominated by private companies.[1]
Some of the gripes are directionally legitimate, but a big reason for that is that for a long time, the asset management business had excellent economics because it was really selling two things: general market exposure, and whatever value a portfolio manager could add in excess of transaction costs and the cost of paying them. For a fund that bought large, respectable stocks, didn't use leverage, and had a diversified portfolio, it would take a nearly-superhuman level of management skill to derive the majority of returns from manager skill rather than stock selection; in the long run, equities have done about 9% annualized, and it's very hard to produce nine points of gross unlevered alpha without doing some weird stuff. What was possible was charging a fee of, say 1-2% of assets under management for this bundle, marketing the particular bundle based on the skills of the manager in question, but mostly getting paid for giving customers market exposure.
It's a bit like how irked large airlines get at low-cost carriers: the big airlines do work extremely hard to provide a nice experience, but a substantial fraction of flyers just want to get from one place to another as cheaply as possible. Once someone is selling the stripped-down version of what you're offering, you find out what the market thinks all your effort is really worth, and the answer is usually disappointing.
But what about the "Worse than Marxism" critique? It's a snappy line for a genuine argument: at least Marxists have a five-year plan! Passive investors are treating the market price as fair value, and they're not judging whether these prices make sense, in isolation or relative to one another. If tomorrow, AICannaCoin Inc. gets added to the S&P with a $3 trillion market cap, the indexers will happily buy it.
Of course, that doesn't happen. Annoyingly, one of the best statements on this comes from minute 46 of this interview, contrasting equities and crypto:
Let's say that Apple—there's no news in Apple—but somebody sells a lot, whatever, liquidations, who knows... and Apple's market cap crashes... down to a billion dollars tomorrow. What happens next if Apple crashes to a billion dollars?
Obviously, what happens next is that somebody buys Apple for $2 billion, or that Apple starts buying chunks of itself and restores the price to something like fair value. That happens in reverse, too; if a stock is too expensive, the company's managers will consider secondary offerings, or use it as currency for mergers, increasing the supply of whatever the market wants more of until things equilibrate.[2]
The cartelization question is harder to answer. Yes, it's true that every time oil prices go down, Blackrock gets hurt from declining oil stocks, but it also gets helped because profits at airlines and logistics companies go up.[3] So in theory their incentive is to direct low-multiple companies to buy as much as possible from high-margin companies. But we don't see a lot of cases where index funds are pressuring steel mills to buy more Nvidia chips, or to sell their steel through an app so Apple gets a 30% cut that increases aggregate market cap. It's hard enough to maintain a cartel when there's an explicit agreement; CEOs don't get paid based on what the rest of the market does, but what their company does, and if a CEO ever agreed to some pay scheme that rewarded them for maximizing the value of customers' or suppliers' shares instead of their own, shareholders would revolt and the FTC would have questions. It becomes a public choice kind of question: passive investors are big in the aggregate, but at any given company the people who have the largest interest in the stock, and the ones who pay the closest attention to the details of how the company operates, will be discretionary investors.
That model also partly addresses the question of whether companies will start following ESG rules mandated by the big passive funds rather than what's in their own interests. A passive fund has a sort of light regulatory role, but it's hard for an issue to be important enough that Blackrock pushing it is a marketing angle without being big enough that voters will care about it and it'll be a literal rather than metaphorical regulation. On the margin, there's probably an effect, but that effect is likely to show up more in the pace and degree—and when regulations are written by professional financiers who are maximizing the value of their investments, it's at least a bit less likely that they'll create gaping loopholes, second-order effects that more than reverse the intended first-order effect, etc.
On concentration: yes, a passive investor today is making a big bet on tech, especially large-cap high-growth tech. But that reflects the fact that the economy itself is also, increasingly, a bet that those companies will keep growing. So it's true in the same sense that an index fund doesn't protect your portfolio from nuclear war, violent revolution, or a worse pandemic: those are hard things to protect yourself from, and a broad sampling of assets is also going to double down on some fundamentally unhedgeable bets.[4] There are plenty of passive options that reduce this risk, like running an equal-weighted index instead of a market-weighted index. That would still be passive, but it's a different bet, and it doesn't always pay: if you'd bought an equal-weighted S&P 500 index fund at the beginning of 2023, you'd be up 15.6% since then, compared to 31.9% for the S&P itself.
Gripes about the impact of passive investing on the markets are part of a classic genre, which goes like this: "I used to have a job that made a lot of money considering how hard it is. But now I have competitors who do a better job and/or charge less, so my choices are to work harder or to accept less money. I'd prefer not to do either—help!" It is a lot less lucrative to manage a diversified portfolio of large-cap stocks than it used to be. Oh well! There are still plenty of good portfolio managers putting up good returns, whether that's by choosing a different kind of market or choosing a strategy that's resilient to the effects of passive.
There are two cases where there is a plausible argument that passive investing distorts the market:
- If active investors do better than chance, but just charge too much, then every shift from active to passive means selling a bit of stocks that will outperform and buying a bit of the ones that will underperform.
- When a stock gets added to an index, particularly one of the larger ones, passive investors will systematically buy it. This creates a sort of air pocket, where the short-term price target goes up when the stock goes up.
But both of these are just new special cases of market dynamics that investors have had to deal with for a long time. Some stocks move for non-fundamental reasons; if a portfolio manager is levered and overweight X and Y, then a big decline in X will produce a decline in Y because that investor is more likely to liquidate. If a company flips from trading just below $1 to just above $1 (or, more common historically above/below $5), there will be a new set of investors who, for psychological or investment-mandate reasons, now have permission to buy it. It's not incoherent to buy something at $0.80 saying that the price target is $0.90, unless it goes to $0.95 in which case the price target is $1.10.
The way to deal with systematic unfairness is to find out what the system is and exploit it instead of complaining about it. Total alpha before transaction costs is $0 by definition; every instance of outperformance comes at the expense of someone else's underperformance. So anyone who claims that passive investing is systematically costing them money is also saying "There are some market scenarios in which I am, predictably, the patsy." Not a point I'd want to make publicly! There are plenty of games where my counterparties have more resources, better information, and more talent than me; those are games I try to avoid playing. And plenty of excellent firms have prospered by mastering some initial strategy and then moving on to another one when the original got too competitive or otherwise infeasible.
All this is not to say that index funds don't affect the market. They do, but in a surprising way. An equity investor is buying a stream of future returns, and there's a price (i.e. an expected return) at which they're willing to do so. Index funds lower this price: if you were paying 1% management fees for a 9% pre-fee return, and your fees drop to roughly zero, you can now accept a valuation implying an 8% pre-fee return and get the same after-fee return. So index investing does raise equity valuations, because it raises the after-fee value of owning equities! This is hard for active investors, who don't get the same benefit. On the other hand, active investors have benefited from lower transaction costs, something that matters to index funds but that matters much more to active traders. Index funds create a different world with less easy money, but that's not a surprise. Ask any smart fund manager what happens to an industry when a new entrant shows up with a lower price than anybody else, and with comparable quality to the median participant. That manager would expect profits to go down. It would be weird to expect this phenomenon to apply to everyone else's business, but not to your own.
Not only that, but you end up dodging a bullet: whenever there is some exciting idea in markets, like cannabis, crypto, EVs, or AI, there will be a rush of small companies that either pivot their entire business to the new hot sector or go public with that as their pitch. (My favorite of these is a company that went public as "Jet.AI," the first air charter company that allowed bookings with a chatbot; its Form 4 from when it de-SPACed still has the old corporate name, Jet Token, from back when they were the first air charter company to accept payment in crypto.) But these tend to be small companies, so they end up not being in S&P 500 index funds and being only a tiny piece of broader passive strategies. So one attraction of passive is that it's a way to avoid the drag on returns from hearing about hot sectors at their peak, which is statistically when most people will hear about them. ↩︎
In fact, this happens not just at the level of the stock price but of volatility; a convertible bond, with its embedded option, is basically a way for a company to raise money from its implied volatility rather than just its share price. When a company issues a convert, it's a message from the CFO: this one won't be as fun to trade forever. ↩︎
With slightly different mechanisms: fuel surcharges mean that the cost of fuel gets passed through for someone like FedEx, but the pricing of a product will be impacted by changes in its marginal cost, whether this shows up as a line item on the bill or just in the evolution of future prices. ↩︎
In almost every case, any time the index is overweight some category today, it means that you would have been glad to own exactly that category in the time leading up to today, because the way a stock increases its index weight is by going up. Technically, there is an exception: an index fund could change its industry weighting if a small, fragmented industry were consolidated by one big company, or if a large private company in a given sector went public. But the scale of the biggest public and private companies means that this is only a theoretical concern. At least unless OpenAI achieves AGI internally and sets it to work finding ways to maximize OpenAI's valuation. ↩︎
Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:
- A seed-stage startup is using blockchains to enforce commitments and is in need of a fullstack developer with Solidity experience. (Remote)
- A company building ML-powered tools to accelerate developer productivity is looking for frontend engineers. (Washington DC area)
- A company reinventing the way Americans build wealth for the long-run by enabling them to access "Universal Basic Capital" is looking for fullstack engineers with prior experience in fintech. (NYC)
- A diversified prop trading firm with a uniquely collaborative team structure is looking for experienced traders and PMs. (Singapore or Austin, TX preferred)
- Several companies in the Diff network are looking for smart generalists interested in their next challenge, especially for people who like solving interesting problems at the intersection of math and humanities.
Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.
If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.
Elsewhere
The Market for Data
Alex Izydorczyk has a good rundown of publicly known AI licensing deals. What stands out is that the numbers, so far, are small. Shutterstock is literally in the business of aggregating a lot of data and making it available, but it's only getting about 16% of its revenue from the "Data, Distribution, and Services" line item. That implies one of three scenarios:
- The value of the core stock photo business has been permanently impaired, and this is the least damaging way to realize its value.
- AI-generated and human-generated stock photos will coexist for a long time; in some contexts it's worth paying up not to have AI artifacts, but the same high-quality images can be used as raw material.
- The most interesting possibility is that, like with early streaming rights deals, early lead generation from online travel booking, food delivery platform fees, etc., the bid/ask spread is wide because buyers are underpaying. The more data becomes a limiting factor, the more being able to aggregate particular kinds of it amounts to a royalty on the growth of AI.
There's a wide range of outcomes for AI companies themselves, but arguably a wider one for the data companies that interact with them. A business like Shutterstock, or Reuters, or Yelp, has revenue at risk if AI doesn't pan out, and even more if it does, but if those businesses end up being the main supplier of a critical raw material to the fastest-growing economic sector, they end up being closer to real estate or oil, as a lucrative option on growth.
Crowded Careers
China has stopped reporting some data on youth unemployment, but extrapolating from recent trends implies that roughly a third of college graduates under 24 are unemployed there ($, Economist). It's a weird cycle, but still fits the usual model of an economic cycle: it takes four years to finish a degree, and longer than that to start a new school, so in a country with growing demand for educated workers, the planned capacity to produce them will typically exceed current demand. That's a good thing when demand grows, but when it doesn't, things get messy fast:
Some students are trying to dodge the tough private-sector job market. The number of people sitting for China’s civil-service exam hit a record high of 2.3m in 2024, a 48% increase year on year. Others are pursuing postgraduate studies. The number of master’s and doctoral students has increased by so much that some campuses have run out of housing.
Black Boxes, Everywhere
For a while now, The Diff has been covering the trend of ad platforms getting less transparent, both so they can use more customer data without sharing it and so they have more pricing power. Netflix is following the same path: they're using a few of the typical measurement providers for viewership and the like, but they're also tracking things like brand awareness lift and, through data on some chains, actual purchase behavior. This is more useful to track than viewing—an ad that's more memorable than what it's selling or which brand it's associated with is a waste even if it does get a large audience. But they also involve linking personal viewing data to personal spending data, and the closer the link—the more Netflix can say that a particular variant of an ad sells better than a different one—the less granular the information Netflix will be willing to share with advertisers.
Externalities
New York has passed a payroll tax credit for local news. The local news business has attracted plenty of capital in an effort to remake the economics of the pre-Internet newspaper. There is a case for this: local news is a bundle that provides a nice way to target people with narrowly-targeted content (city council meetings, local crime, high school sports, obituaries) that offers similarly targeted ad opportunities for geographically-constrained businesses and commerce. On the other hand, that bundle's economics existed as an incidental side effect of some other changes—rising costs and changing readership patterns killed off the afternoon papers in two-paper towns, turning local newspapers into a monopoly ($, Diff), but that was already unraveling in the early 90s due to cable TV and talk radio. Meanwhile, there's a feedback loop where local political issues matter less than national ones, in part because those local issues get less news coverage. So recreating that local news product does require subsidies to offset its worse economics, but the economics are worse in part because customers want different things than they used to, and there are more cost-effective ways to get them.
Intra-Elite Conflict
The largest asset managers now control twice the assets of the largest US banks ($, WSJ), after being close to parity in 2008. This reflects the growing institutionalization of asset management in two ways:
- More fund managers have switched from a superstar approach to a team approach. (Sometimes the superstar is kept around, as a mascot/front man rather than as the main decisionmaker on trades.)
- More importantly, many funds have realized that they'll maximize their fee base by earning a small upside on a large amount of assets rather than maximizing returns by taking the highest risk their customers will tolerate. For a given asset-based fee, the lower the volatility a fund runs at, the higher its fees will be as a share of alpha generated.
So the economic incentive for any firm that can produce above-average returns is to pursue mediocrity in terms of absolute return, since that gives them the biggest fee base (and means that they don't have to spend quite as much time thinking about liquidity—a much bigger factor when leverage is higher).