Do ETFs Distort Markets?

Concern 1: price discovery

The argument. Prices are informative because someone does the work of analysing companies and trades on the conclusion. Index funds do no such analysis; they buy in proportion to size. If indexing grows large enough, the marginal price-setter disappears and prices stop reflecting information.

The strongest form. Grossman and Stiglitz's classic result establishes that perfectly efficient markets are impossible, because if prices already reflect everything, nobody is paid to gather information, and prices stop reflecting everything. Indexing pushes toward that boundary: the more free-riders on price discovery, the thinner the layer doing the work.

The rebuttal. Three parts, each substantive:

1. Price setting is done by trading, not by holding. Index funds hold enormous assets and trade relatively little (Chapter 5's self-maintenance). The share of trading volume driven by fundamental active investors is far larger than the passive share of ownership suggests, and it is trading that moves prices.
2. The correction mechanism is self-enforcing. If prices genuinely became uninformative, active management would become highly profitable, capital would flow back to it, and prices would be corrected. That the returns to active management remain poor (Chapter 6) is evidence prices are still hard to beat, which is evidence they remain reasonably informative.
3. Fundamental investors are not the only informed traders. Quantitative funds, corporate insiders, arbitrageurs and market makers all incorporate information without being traditional stock-pickers.

Where the evidence sits. Research on whether higher index ownership degrades price efficiency has produced mixed results, with some studies finding reduced information incorporation at the individual-stock level and others finding no meaningful effect or improvements in liquidity. There is no consensus, and the honest summary is: measurable, small, and disputed.

Concern 2: co-movement

The argument. When a stock joins an index, it becomes something every index fund must own and every index seller must sell. Its price then partly moves with the index rather than with its own fundamentals.

The evidence. This is the best-supported concern in the chapter. Studies of index-inclusion events find increased correlation with the index and reduced idiosyncratic movement afterwards, and research on high-ETF-ownership stocks finds similar patterns. The mechanism is straightforward: basket trading transmits flow to constituents indiscriminately.

What it means practically. For an index investor, mostly nothing: you own the whole basket. For active managers, it is a real complaint: if stocks move together for non-fundamental reasons, the payoff to distinguishing between them falls, which is a genuine (and somewhat self-serving) grievance. For market quality, it suggests some information is being crowded out, which loops back to Concern 1 without settling it.

Concern 3: corporate governance concentration

The argument. The three largest index managers together hold significant stakes in nearly every large listed company. That concentration of voting power in a handful of firms is a governance question regardless of how those votes are cast.

The substantive worries:

• Under-engagement: index managers cannot exit a holding, and their fee revenue does not scale with stewardship effort. The incentive to invest in serious oversight is weak.
• Over-influence: conversely, if they do engage, an enormous amount of corporate influence rests with a small number of unelected asset managers, on behalf of investors who never expressed a view.
• The pass-through response: partly in reaction, several large managers now offer voting choice programmes, letting underlying investors direct how their shares are voted. Adoption has been modest so far, and whether it meaningfully redistributes influence is an open empirical question.

The rebuttal. Index managers have built substantial stewardship teams, publish voting records and engagement statistics, and have demonstrably influenced board composition and disclosure practices. And the counterfactual matters: the alternative owners are not engaged citizens but dispersed retail holders and short-horizon traders, whose governance participation is historically minimal.

Our position on this one: the concern is legitimate and structural rather than a matter of anyone's bad conduct, and the most useful investor-level response is the Chapter 46 one: if stewardship matters to you, evaluate providers on their actual voting and engagement records, which are published.

Concern 4: common ownership

The argument. If the same institutions own significant stakes in competing firms, those firms may compete less aggressively, because vigorous competition harms the common owner's aggregate returns.

The evidence. A prominent line of research argued for anticompetitive effects in concentrated industries; the empirical claims were subsequently challenged on methodology, and the literature has become contested rather than conclusive. Regulators and competition authorities have examined the theory without, so far, restructuring policy around it.

The honest status: theoretically interesting, empirically unsettled, and the mechanism by which passive owners would transmit anticompetitive preferences to management remains poorly specified.

What is genuinely agreed

Away from the contested ground, several points command broad assent:

• Some threshold exists beyond which indexing would impair price discovery. Nobody credibly claims otherwise.
• Nobody knows where it is, and estimates vary enormously.
• The self-correction mechanism is real, and its continued operation (active management remaining difficult) is the best available evidence that we have not crossed the threshold.
• The concerns are about indexing, not about ETFs. An index mutual fund raises identical questions. The ETF wrapper is incidental to every argument in this chapter, which is worth remembering whenever the debate is framed as an ETF problem.
• Growth is unlikely to continue to 100%. As indexing rises, the incentive to be an informed trader rises with it.

What it means for you

Very little operationally, and that is not a dodge:

• If the concerns are valid, the correction runs through active management becoming profitable, at which point the evidence in Chapter 6 changes and this book's recommendations would change with it. We would tell you.
• In the meantime, the arithmetic of Chapter 6 holds, and choosing active management as a hedge against the theoretical possibility that active management might one day work better is a strange trade to make with your retirement.
• The one actionable item is stewardship: if governance concentration concerns you, provider voting records are public and are a legitimate selection criterion (Chapter 46).

A worked example

A commonly-cited concern in practice: a mid-cap company is added to a major index. In the weeks around inclusion its price rises, its correlation with the index increases measurably, and analysts note that its stock now moves on days when nothing about the company changed.

• The distortion reading: its price now reflects index membership rather than business fundamentals.
• The counter-reading: its liquidity improved, its cost of capital fell, its shareholder base broadened, and the fundamental investors who follow it still trade on earnings, which still move the stock.
• What is measurably true: the co-movement effect is real (Concern 2). Whether it degrades price accuracy rather than merely price behaviour is the disputed part.

Both readings are defensible from the same data, which is precisely why this chapter refuses to hand you a verdict.