IWhat a market is for
Before asking what passive investing does to markets, it is worth stating plainly what markets are supposed to do. A stock market has two jobs. The first is allocation: directing society's savings toward the enterprises most likely to use them productively. The second is information: aggregating the dispersed judgments of millions of participants into prices, so that everyone — managers, workers, policymakers, savers — can read from a single number what the collective assessment of an enterprise's prospects is.
This second function is the profound one. It is Hayek's argument from 1945: no central planner can gather the knowledge scattered across society — the factory manager's sense of demand, the engineer's knowledge of a technical bottleneck, the customer's shifting preference. But prices can. Every trade is a small act of information disclosure. The market price is the running total of everything everyone claims to know, weighted by how much money they are willing to stake on it.
The whole moral case for free markets — the reason we entrust capital allocation to them rather than to committees — rests on this machinery working. Prices must mean something. And prices mean something only if the people setting them are trying to figure out what things are worth.
This is the standard against which passive investing must be examined. Not whether it is good for the individual investor — it demonstrably is — but whether a market can remain a market when a growing share of its participants have, by design, stopped doing the thing that makes prices informative.
IIThe scale of the shift
The numbers first, because the scale is the argument.
Two of these numbers deserve elaboration. The first is the 8%: over the decade to mid-2025, only about one in twelve US large-cap active funds survived and outperformed its passive rival, after fees. The rational response of investors to this fact — switching to index funds — is individually correct and collectively transformative. Nobody did anything wrong. That is what makes the passive question interesting rather than scandalous: it is a story of rational individual choices aggregating into a structural change nobody chose.
The second is the Chinco–Sammon finding. Official statistics count index mutual funds and ETFs. But a large volume of "active" institutional money is closet-indexed — benchmark-hugging funds, pension mandates that track indices internally, quant strategies keyed to index composition. By measuring the volume of trading that occurs mechanically at index-reconstitution moments, Chinco and Sammon estimate that true passive ownership of the US market was already around a third by 2021 — roughly double the headline figure. The real number today is plausibly north of 40%.
IIIPrice discovery — the Grossman–Stiglitz machine and its fuel
In 1980, Sanford Grossman and Joseph Stiglitz published a paper whose title states its thesis: On the Impossibility of Informationally Efficient Markets. The argument is elegant. If market prices perfectly reflected all information, there would be no profit in gathering information — so nobody would gather it — so prices would stop reflecting it. Perfect efficiency is self-defeating. Markets must be imperfectly efficient: inefficient enough that research pays, efficient enough that prices are usable.
This means price discovery is not a natural feature of markets. It is a manufactured product, produced by active investors who spend real resources — analysts, models, channel checks, time — hunting for mispricing. Their profits are the wage society pays for informative prices. Passive investors consume this product without paying for it. That is not a criticism; it is the design. Jack Bogle's genius was recognising that the marginal retail investor gains nothing from paying for price discovery they can free-ride on.
Price discovery is not a natural feature of markets. It is a manufactured product — and the workforce that manufactures it is shrinking.
The question is what happens as the free-riders multiply and the producers exit. Here the recent academic literature has produced a genuinely important finding.
Classical theory assumed equity demand is highly elastic: if a price drifts above fair value, arbitrageurs sell until it corrects, so flows barely move prices. Gabaix and Koijen measured what actually happens and found the opposite: investing $1 into the aggregate stock market raises aggregate market value by roughly $5. Demand is extraordinarily inelastic — because the marginal holders of equity (index funds, pension mandates, insurance allocations) operate under rules that fix their equity exposure. When money flows in, almost nobody is willing to sell against it, so prices must move a lot to clear the market.
Haddad, Huebner and Loualiche connect this directly to the passive transition: as the passive share grew from roughly 19% to 41% (2000–2020), the price-elasticity of demand for individual stocks measurably declined. Fewer participants react to prices; mispricings can grow larger and persist longer; liquidity provision thins.
The inelastic markets finding reframes the passive question. The concern is not that prices become subtly less accurate — it is that the market's shock-absorbing capacity is draining away. In an elastic market, a wave of buying meets a wall of valuation-sensitive sellers and dissipates. In an inelastic market, the same wave moves prices five times its own size. Flows, not fundamentals, increasingly set the level of the market. This cuts both ways: the same multiplier that amplified the 2023–24 melt-up would amplify a sustained outflow — for instance, when demographics turn 401(k) contributions into retirement withdrawals.
IVThe concentration engine
Cap-weighted passive investing contains a feedback loop. Inflows buy companies in proportion to their existing weight — so the largest companies receive the most price-insensitive buying, which raises their prices, which raises their weight, which directs even more of the next inflow toward them. In an elastic market, arbitrage would lean against this. In an inelastic one, it compounds.
This NBER work models and measures exactly the loop above: flows into passive funds disproportionately raise the prices of the largest firms in the index — and especially the largest overvalued firms, because passive buying is indifferent to valuation. The effect they estimate within the index is large (their empirical estimate of the within-index effect on the biggest stocks is on the order of 30%). Related work by Gutiérrez and Philippon finds that firms with high passive ownership invest less in their own businesses — a governance effect layered on the valuation effect.
The observable outcome is the concentration data from earlier in this series: the top 10 companies at roughly 37% of the S&P 500, semiconductors at 18% of index weight, NVIDIA alone at 7% — larger than entire sectors. Concentration has many causes — network effects, AI economics, genuine earnings dominance — and it would be wrong to attribute it to passive flows alone. But passive mechanics are an accelerant: they ensure that whatever the market's current bet is, the next dollar doubles down on it automatically.
Note also what this does to the meaning of "diversification." The investor who buys the index believes they hold 500 companies. In weight terms they hold a concentrated position in a handful of AI-exposed mega-caps, plus a long tail. And because index weight determines passive flow, the top of the index is precisely where passive ownership is heaviest and price discovery weakest — while the bottom 300 stocks, which barely move the index, still trade in something resembling a textbook market.
VGovernance — owners who cannot leave, and cannot really stay
Ownership of a company traditionally came bundled with two disciplining mechanisms: exit (sell the shares, punishing management through the price) and voice (vote, engage, agitate). The passive vehicle disables exit by construction — an index fund cannot sell Boeing because it disapproves of Boeing; it holds whatever the index holds, forever. That leaves voice. And voice has been concentrated into remarkably few hands.
Vanguard, BlackRock and State Street collectively manage over $23 trillion and sit among the top three shareholders of roughly 90% of S&P 500 companies — around 17–20% of nearly every large American firm, rising toward 25% at the bottom of the index. This is ownership without precedent: the same three institutions are simultaneously the largest owners of every airline, every major bank, every chipmaker, every retailer. They did not choose these positions and cannot exit them; stewardship teams of a few dozen people oversee governance for thousands of portfolio companies, necessarily by generic checklist rather than company-specific judgment.
Three structural problems follow. First, the common ownership concern: an owner of all competitors in an industry has attenuated incentives to push any one of them to compete hard, since one holding's gain is another's loss (the Azar–Schmalz–Tecu airline evidence — contested, but structurally serious). Second, the governance vacuum: as active managers who did company-specific engagement shrink, monitoring of management quality thins — and Gutiérrez and Philippon's finding that passively-owned firms invest less suggests the vacuum has real effects. Third, the accountability inversion: several of the index's largest companies (Alphabet, Meta, Palantir) run dual-class structures in which founders control the votes regardless of what shareholders think. The Big Three formally oppose dual-class shares — and are compelled to buy them anyway, because the companies are in the index. The world's largest shareholders have no voice precisely where their holdings are largest.
The index fund is an owner that cannot sell, cannot meaningfully engage, and in several of its largest holdings, cannot even vote to effect.
VIThe honest counterarguments
A fair treatment requires the other side of the ledger, and it is substantial.
Careful empirical work by Coles, Heath and Ringgenberg (2022) and Koijen, Richmond and Yogo (2024) finds that the growth of passive investing has not had a measurable negative impact on overall market efficiency to date. Price informativeness measures have not collapsed. One reason is the Grossman–Stiglitz equilibrium itself: as passive grows, the remaining active capital faces less competition for mispricings, so each remaining analyst becomes more effective. The system may self-correct toward a smaller but sharper active sector — price discovery produced by fewer, better-paid producers.
Beyond the empirics, three affirmative points deserve full weight. First, the welfare gain is enormous and real: fee compression from ~1% to ~0.05% compounds to life-changing differences in retirement wealth for ordinary savers; the average expense ratio across all funds has fallen by more than half since 2000 under passive competitive pressure. Second, most active management was never producing price discovery anyway — it was producing closet indexing at 1% fees; its displacement destroyed less information than the industry's marketing claimed. Third, the market has absorbed structural transformations before — the shift from individual to institutional ownership in the 1960s–80s prompted similar anxieties about the death of the prudent shareholder, and the market adapted.
The honest statement of the debate is this: the direction of the structural concerns — inelasticity, concentration amplification, governance thinning — is well-supported empirically. The severity is not yet established. We are running the experiment in real time, and the market's current concentration and valuation levels mean we may learn the answer under stress rather than in a seminar.
VIIWhere the market is going — the multiple as a concentration measurement
All of this converges on the question every investor eventually asks: what does it mean that the market trades at 28 times earnings, against a century-long average of 16–18×? The trailing multiple stood at 14.87× at the end of 2011; it has nearly doubled since. The textbook reading is that multiples mean-revert, that a contraction from 28× toward 20× would subtract roughly three percentage points a year from returns over a decade, and that investors anchored to the last decade's 13% annual returns are in for a repricing of expectations.
But the aggregate number conceals the structure — and the structure changes the diagnosis. Decompose the index into the Magnificent 7 and the remaining 493 companies, and a striking picture emerges: the S&P 493 has traded at a flat 17–20× for a decade, sitting almost exactly on its hundred-year mean. The Mag 7 trade at roughly 40×, and have for most of the period. What changed is not the price of the average business — it is the weight of the expensive ones, which grew from roughly 9% of the index in 2015 to over 30% today. The market did not re-rate. Its composition changed. Almost the entire aggregate multiple expansion of the past decade is a composition effect: high-multiple companies growing their share of the index, with passive flows — as the mega-firms literature predicts — acting as the accelerant.
The index's elevated multiple is not a market-wide valuation judgment. It is a concentration measurement.
This decomposition redraws the map of forward scenarios. A broad, grinding de-rating of the whole market — the "lost decade" scenario — loses its raw material: 493 of the 500 companies have nothing to revert to; they are already at the mean. The regime-change hypothesis, under which passive flows have permanently inflated multiples across the market, is also weakened: the 493's stable multiple suggests passive has not repriced the average business at all. What remains is a sharply localised risk. The entire valuation question of the American market now lives in seven companies — the same seven where passive ownership is heaviest, the concentration feedback loop strongest, and the inelastic-flow multiplier most powerful. The question "is the market expensive?" collapses into "will the Magnificent 7 earn enough to justify 40 times earnings?" — which is an AI-monetisation question, not a market question. If the answer is yes, the index grows into its multiple. If the answer is no, the correction is not a market event but a seven-stock event that happens to be a third of everyone's index fund.
History offers one close rhyme, worth naming carefully. A two-tier market — a small cohort of glamour stocks at 40×+ floating above a normal market at 18× — is the structure of the Nifty Fifty in 1972. Those too were genuinely superb businesses; the error was never the quality but the entry multiple, and when the 1973–74 bear market came, several of them spent fifteen years reclaiming their highs while their earnings kept growing. The differences deserve equal weight: the Mag 7 grow earnings far faster than the Nifty Fifty did, their margins are structurally higher, and 40× for a company compounding earnings at 40–50% is arithmetically a different object from 45× for a razor-and-cosmetics company in 1972. The bull case is that these are the first firms in history whose moats justify a permanent premium. Perhaps. It is also exactly the sentence that was written in 1972, and again in 2000.
There is one further layer, and it is where the multiple and the passive transition fuse into a single object. Whatever portion of the Mag 7's 40× is underwritten by inelastic passive flows rather than by earnings judgment is a valuation resting on the continuation of everyone else's automatic behaviour — the fortnightly 401(k) contribution, the buyback programme, the target-date fund's rebalancing rule. Flow regimes are stable for decades, but they have one known, slow-moving vulnerability: demographics. The same retirement system whose accumulation phase supplied the structural bid will, in its decumulation phase, supply a structural offer. The multiplier that turned inflows into a 5× price impact is symmetric. The market's greatest source of stability and its greatest latent fragility are, in the end, the same mechanism — and the elevated multiple is where that mechanism's work is stored.
VIIIThe imperative of a free market — and what passive is actually challenging
Return to first principles. The case for free markets is not that they make investors rich. It is that decentralised, price-guided allocation outperforms centralised allocation — because prices carry information no planner can gather. The market's legitimacy rests on prices being earned: set by participants who bear the consequences of being wrong.
What passive investing challenges is not capitalism, and not markets as venues of exchange. ETFs are traded in fiercely competitive markets; the fee war is capitalism working. What passive investing challenges is the informational premise — the assumption that the participants setting prices are trying to be right about value.
Consider what the S&P 500 actually is: a list, maintained by a committee at S&P Dow Jones Indices, with inclusion criteria and discretionary judgment. When 40%+ of the market's capital allocates itself mechanically to that list in proportion to market cap, the committee's decisions become capital allocation decisions for the world's savings. Index inclusion — a bureaucratic event — moves a company's cost of capital. This is not central planning; nobody intends it. But it is drift toward a strange hybrid: allocation by list-membership and momentum rather than by judgment, wearing the institutional clothes of a free market. Hayek's machine still runs — but a growing fraction of its horsepower comes from participants who, by design, contribute no knowledge to it.
The deepest irony: passive investing works because markets are efficient, and markets are efficient because of the active investors passive is displacing. The parasite is benign — until it outgrows the host.
Where is the limit? Nobody knows the threshold at which the loss of active capital becomes binding — estimates in the literature range from "we passed it already" to "markets remain robust at 70%+ passive." What can be said with confidence is directional: every increment of passive share makes prices more flow-driven, the index more self-referential, and the correction mechanism weaker. A system can drift a long way from its design premise while appearing to function — the appearance lasting precisely until it is tested.
IXClosing — the tuning instrument detuned?
The stock market, at its best, is civilisation's instrument for hearing itself think about the future — millions of independent judgments resolving into prices that tell us what we collectively believe our enterprises are worth. Passive investing has democratised access to that instrument's output while quietly reducing the number of players producing it. The individual investor's rational choice — take the market return, pay nothing for research — is unambiguously correct, and I would not advise anyone otherwise. That is precisely what makes the question hard: the system-level concern arises from behaviour that is individually impeccable.
The passive question, finally, is a commons problem. Price discovery is a public good; nobody owns it, everybody uses it, and the incentive to contribute to it is eroding — a category of problem Elinor Ostrom taught us markets do not solve by themselves. Perhaps the Grossman–Stiglitz equilibrium holds and a leaner active sector keeps prices honest. Perhaps the multiplier of inelastic markets delivers the lesson abruptly, in some future season of sustained outflows. Either way, the market that emerges from this transition will not be the market the textbooks describe. It will be something newer and stranger: enormously efficient at intermediating savings, increasingly uncertain about what anything is worth. Its P/E multiple — once the market's opinion about future earnings — has become, in part, a measurement of the market's own structure. Reading it now requires knowing not just what investors believe, but how many of them are still doing the believing.