Since the launch of the first retail index mutual fund by Vanguard in 1976, the passive investment paradigm has expanded from a philosophical curiosity to a structural feature of global capital markets. By the mid-2020s, passive funds—including index mutual funds and exchange-traded funds (ETFs)—commanded assets under management in excess of $15 trillion in U.S. equity markets alone, eclipsing their active counterparts in aggregate flow dominance.
In the canonical Grossman-Stiglitz framework, informed active traders perform the socially valuable function of incorporating dispersed private information into asset prices. Passive investors, by construction, abstain from this activity—they hold market-cap-weighted portfolios and trade only at the margin to track an index. If the share of passive capital becomes sufficiently large, the marginal incentive to gather and trade on information is diminished, raising the possibility of systematic deterioration in market quality.
The foundational tension between passive investing and informational efficiency derives from Grossman and Stiglitz. Define the fraction of passive investors in the market as \(\lambda \in [0,1]\). Let \(p^*\) denote the fundamental value of a risky asset and \(p\) its observed market price. The price efficiency of the market can be characterized by the variance of the pricing error:
\[\sigma_{\varepsilon}^2 = \text{Var}(p - p^*)\]
In equilibrium, active informed traders invest in costly information acquisition up to the point where the marginal benefit—the expected profit from superior pricing—equals the marginal cost \(c\) of information gathering. As \(\lambda\) rises, the pool of active traders shrinks, reducing the aggregate information production:
\[\frac{d\sigma_{\varepsilon}^2}{d\lambda} > 0\]
This result formalizes the intuition that passive investors are free-riders on the price discovery externality produced by active traders. At some critical threshold \(\lambda^*\), the active trading ecosystem may be insufficient to sustain informationally efficient prices.
A more dynamic version of this narrative considers how rising passive ownership affects the incentive to become informed. Let \(\Pi(\lambda)\) denote the expected profit to an active trader as a function of the passive share:
\[\Pi(\lambda) = \underbrace{\alpha \cdot (1 - \lambda)}_{\text{Trading opportunity}} \cdot \underbrace{\Delta(\lambda)}_{\text{Mispricing}} - c\]
where \(\Delta(\lambda)\) is the average magnitude of mispricing. The key insight is that \(\Delta(\lambda)\) and \(\Pi(\lambda)\) may be non-monotonic. Initially, as \(\lambda\) rises, mispricings widen (increasing \(\Delta\)), but the trading opportunity \(\alpha(1-\lambda)\) shrinks. If the latter effect dominates, the equilibrium number of informed agents collapses, producing a discrete deterioration in price efficiency—a tipping point phenomenon.
Empirically stocks with higher passive institutional ownership exhibit lower earnings response coefficients and greater post-announcement drift, consistent with degraded price discovery.
A second and perhaps more intuitive channel concerns the potential for passive fund flows to amplify asset price volatility. The mechanism operates through inelastic demand schedules induced by index inclusion and mechanical rebalancing.
Let \(F_t\) denote the aggregate net flow into passive funds at time \(t\). By construction, passive funds must translate flows into proportional purchases (or sales) across all index constituents. Define the flow-induced price pressure on stock \(i\) as:
\[\Delta p_{i,t}^{\text{flow}} = \frac{\omega_i \cdot F_t}{\text{Float}_i}\]
where \(\omega_i\) is the index weight of stock \(i\) and \(\text{Float}_i\) is its tradeable float. Critically, this pressure is uninformed—it bears no relationship to the fundamental value \(p_i^*\). Vayanos and Woolley (2013) construct a rational equilibrium model in which fund flows generate price momentum and subsequent reversals, both of which represent excess volatility relative to fundamentals.
Gabaix and Koijen develop the Inelastic Markets Hypothesis (IMH), which provides perhaps the most provocative formal treatment of this channel. Their central result, derived from a model with limited arbitrage and slow-moving capital, is that the aggregate market-level demand elasticity is far below unity:
\[\varepsilon \equiv -\frac{d \ln Q}{d \ln P} \approx 0.1 - 0.3\]
This implies that a \(\$1\) trillion inflow into equity markets raises aggregate market capitalization by approximately \(\$5\) trillion—a multiplier of five. The mechanism: with inelastic demand, price-taking behavior by passive funds means that flows must clear via large price movements. Formally, their price-impact equation is:
\[\Delta \ln P_t = \frac{1}{\varepsilon} \cdot \frac{\Delta Q_t}{Q_t}\]
where \(Q_t\) is the quantity of shares demanded. This implies that the secular growth of passive AUM mechanically amplifies the sensitivity of market prices to macroeconomic sentiment shocks, redemption cycles, and risk-off episodes.
A distinct sub-narrative concerns exchange-traded funds specifically. Unlike mutual funds, ETFs trade continuously on secondary markets, creating a two-layer price system: the ETF’s market price \(p^{\text{ETF}}\) and the intraday indicative NAV of its underlying basket. Arbitrage by authorized participants (APs) is supposed to keep these aligned. However, Ben-David, Franzoni, and Moussawi (2018) document that stocks with higher ETF ownership exhibit elevated daily and intraday return volatility, finding:
\[\text{Volatility}_i \propto \text{ETF Ownership}_i\]
The mechanism proposed is that ETF trading creates a volatility conduit: sentiment shocks, liquidity events, or margin calls in the ETF market are transmitted mechanically to all constituent securities, irrespective of their individual fundamentals. This was vividly illustrated during the Flash Crash of May 6, 2010, and the ETF dislocation events of August 24, 2015, when dozens of ETFs temporarily traded at discounts exceeding 30% to their underlying NAV.
A third narrative concerns the liquidity illusion engendered by the apparent ease of trading ETF shares. Superficially, ETFs offer high liquidity—they trade on exchanges with tight bid-ask spreads. However, critics argue that this surface liquidity is contingent on an underlying liquidity in the constituent assets that may be unavailable during stress conditions.
Define two layers of liquidity:
Under normal conditions, AP arbitrage enforces:
\[|p^{\text{ETF}} - \text{NAV}| \leq \frac{s^{\text{basket}}}{2} + \text{Transaction Costs}\]
However, during liquidity crises, if the cost of transacting in the underlying basket rises sharply (i.e., \(s^{\text{basket}} \rightarrow \infty\) for illiquid constituents), the arbitrage mechanism fails, and the ETF premium or discount can widen dramatically. This is not merely a pricing anomaly—it represents a withdrawal of effective liquidity precisely when investors most need it.
The liquidity illusion hypothesis gained empirical urgency during the COVID-19 market stress of March 2020, when the largest investment-grade corporate bond ETF (LQD) traded at discounts of 3–5% to its NAV for multiple consecutive days—a historically extraordinary divergence. Falato, Hortaçsu, Li, and Shin (2021) document that bond mutual fund outflows during this period were amplified by ETF-induced selling, with a first-order fire-sale externality:
\[\Delta \text{Yield}_{i,t} = \beta_1 \cdot \text{Fund Outflow}_{i,t} + \beta_2 \cdot \text{ETF Ownership}_{i} + \epsilon_{i,t}\]
Their estimates of \(\beta_2 > 0\) and statistically significant imply that securities with higher ETF ownership experienced greater yield spikes during the stress episode, consistent with liquidity withdrawal rather than liquidity provision.
A related formal mechanism is the mutual fund fire sale amplification model extended to the passive context. When passive funds face outflows \(R_t > 0\), they are constrained to sell a proportional slice of their portfolio:
\[\text{Sell}_{i,t} = \omega_i \cdot R_t \quad \forall i \in \mathcal{I}\]
where \(\mathcal{I}\) is the index constituent set. If \(R_t\) is correlated across passive funds (as it tends to be during risk-off episodes), aggregate selling pressure concentrates mechanically across the same securities. The price impact per dollar sold is amplified by the shallow residual demand from a now-smaller active trading community:
\[\Delta p_i = -\frac{\omega_i \cdot R_t}{\eta_i \cdot \text{Active AUM}_t}\]
where \(\eta_i\) is the price elasticity of active demand. As active AUM shrinks with rising \(\lambda\), this price impact grows without bound in the limit, implying that the same outflow shock produces increasingly severe price dislocations.
A fourth narrative concerns the distortion of the cross-sectional correlation structure of returns. Barberis, Shleifer, and Wurgler (2005) document a striking empirical regularity: upon addition to the S&P 500, a stock’s return correlation with existing index members increases discontinuously, while its correlation with non-index stocks decreases. This index comovement effect implies that passive investing induces a non-fundamental correlation structure across assets.
The formal argument is as follows. Decompose the return \(r_{i,t}\) of asset \(i\) into:
\[r_{i,t} = \underbrace{\beta_i^F \cdot F_t}_{\text{Fundamental factor}} + \underbrace{\beta_i^{\text{Index}} \cdot \tilde{F}_t^{\text{Index}}}_{\text{Non-fundamental index factor}} + \varepsilon_{i,t}\]
where \(F_t\) is a fundamental macroeconomic factor and \(\tilde{F}_t^{\text{Index}}\) is a non-fundamental index flow factor driven by passive fund activity. As passive AUM grows, the loading \(\beta_i^{\text{Index}}\) on the non-fundamental factor increases for all index constituents. The pairwise correlation between two index stocks \(i\) and \(j\) thus has a spurious component:
\[\text{Corr}(r_i, r_j) = \frac{\beta_i^F \beta_j^F \sigma_F^2 + \beta_i^{\text{Index}} \beta_j^{\text{Index}} \sigma_{\tilde{F}}^2}{\sigma_i \sigma_j}\]
The second term in the numerator grows with passive market share, inflating measured correlations beyond their fundamental level. This has direct implications for portfolio diversification—if correlations are artificially elevated, the risk-reduction benefits of diversification are overstated.
The comovement narrative connects to systemic risk through the concept of endogenous correlation amplification. In a stylized model, suppose the market consists of \(N\) stocks, each with idiosyncratic variance \(\sigma_\varepsilon^2\) and a common factor with variance \(\sigma_F^2\). In the absence of passive distortions, the portfolio variance of an equally weighted portfolio converges to \(\sigma_F^2\) as \(N \rightarrow \infty\):
\[\sigma_{\text{portfolio}}^2 = \sigma_F^2 + \frac{\sigma_\varepsilon^2}{N} \xrightarrow{N \to \infty} \sigma_F^2\]
Now introduce passive-driven non-fundamental covariance \(\Sigma^{\text{passive}}\). The portfolio variance becomes:
\[\sigma_{\text{portfolio}}^2 = \sigma_F^2 + \frac{\sigma_\varepsilon^2}{N} + \mathbf{w}^\top \Sigma^{\text{passive}} \mathbf{w}\]
where \(\mathbf{w}\) is the portfolio weight vector. Because \(\Sigma^{\text{passive}}\) does not vanish with diversification (it is systematic, not idiosyncratic), it represents an irreducible floor on portfolio risk. This “phantom” systematic risk cannot be hedged through standard diversification and is invisible to models that rely on historical correlations without correcting for passive-induced comovement.
The passive investment revolution, by removing informed discretionary traders from the price formation process and concentrating ownership in diversified hands, may compromise the self-correcting mechanisms that capital markets depend upon for stability.
Each of the five channels—price discovery degradation, volatility amplification, liquidity illusion, non-fundamental comovement, and governance externalities (ownership of large institutions, not discussed here)—rests on well-specified theoretical foundations and is supported by at least partial empirical corroboration. At the same time, the countervailing arguments are far from negligible, and the macro-level verdict remains contested.
The stakes are high. If the leading narratives are correct, the passive investment revolution—celebrated for delivering efficient, low-cost market exposure to ordinary investors—may undermine the very market qualities upon which its value proposition depends.
\(\textcopyright\) All rights reserved, Orange County Science Institute