Systemic Risk in the U.S. ETF Ecosystem: How Contagion Spreads Across 4,000+ Funds

Opening: Why this matters now

ETFs have transformed the way of investing in markets. These include intraday liquidity benefits, reduced pricing, and extensive diversification. Those two elements have driven rapid growth. By the middle of 2025, the U.S. ETF sector possessed over $11.54 trillion in assets and over 4,000 listed investments. Share ETFs just by themselves create a potential avenue of mass market stress in case conditions shrink.

But size is a bit more than that. ETF constructions, trading, and arbitrage allow them to be connected to the underlying securities, the people that are allowed to appear in it, as well as the market makers, and the market-clearing systems. The existence of such linkages generates numerous contagion channels.

The state of stress can propagate when little strain is exerted on a node. The operation of these channels is an important issue that fund managers, institutional investors, and regulators wish to avoid to avoid the localized problems from becoming systemwide. Recent scholarly literature and business reports now chart the course of such channels much more finely.

The anatomy of contagion in the ETF ecosystem

Four key contagion channels are whether or not they are practiced.

1. Market liquidity and price impact: ETFs are traded in the same manner as stocks, but the net asset values denote baskets of securities. ETFs may be traded below or above their NAVs during accidental lies. Market makers and liquidity providers intervene, but when the underlying securities are illiquid, the cost of tight ETF spreads increases, and market makers withdraw. When it occurs with a large number of ETFs, the effects of price impact multiply and may result in additional equity price movements. Recent studies, however, reveal that spillovers of returns between ETFs and their constituent stocks are substantial during political moments of stress.
2. Concentration of assets and single-firm dependency: The assets are concentrated in relatively few large ETFs and very few issuers. Such concentration would imply that a large issuer of securities would have disproportionately large impacts on trading volumes and liquidity in the market had an operational problem, or had to be forced to liquidate a mega-ETF. According to analyses of industries, the largest promoters continue to dominate the ETF assets, and the impact of their systemic footprint is greater.
3. Arbitrage and authorized participant linkages: The creation and redemption of ETFs are the arbitrage mechanisms that facilitate ETFs to serve the ETF price so as to correspond with the NAV. The shares are redeemed or created by means of a delivery of baskets of securities and equities and/or cash, by the participants who are authorized participants (APs). Arbitrage frictions increase with both a non-circulation of APs or a shortage of liquidity in the underlying basket. Once such frictions take place, then such a normal stabilizing mechanism can stick, and then mispricing will be experienced to diffuse and link the ETF with the market in general. The channel is stronger in the case of the niche or sophisticated ETF since the basket of which is not traced so much.
4. Cross-product and leverage interactions: Some of the products that can be found in the ETF universe include leveraged products, inverse products, synthetic exposures, and futures and option-based ETFs. Rebalancing leveraged ETFs in volatile markets can result in concentrated trading of the underlying instruments, including short-term feedback loops. Likewise, ETFs based on derivatives have the fund attached to margin, counter-parties, futures markets – all other nodes to contagion. Non-linear shock amplification can be accomplished by the links between certain products on an academic network, as studies have demonstrated.

How big is the exposure? Numbers that matter

Numbers help frame risk. By the end of 2025, it is projected that the count of ETFs in the U.S., both in operation and listed, will top 4,000 and 4,300, respectively, depending on the source of data and the date. The U.S. ETF industry assets, exceeding $11 trillion over the first half of 2025, exceeded the estimates made previously using $8 trillion as a benchmark. That expansion augurs the possibility of any systemic occurrence. Notably, the degree of concentration of assets is not low: a relatively small number of funds equate the bulk of assets.

The contaminating modeling is also important regarding these facts. An adjustment to a top-of-book ETF will not only impact those investments in the fund but also those in other similar funds that hold the same security or build on the same AP network. The regulation requires the regulators to measure systemic risk through both breadth and concentration.

Evidence from past episodes

Some ETFs reported large NAV disparities and decreased liquidity in their underlying baskets during the March 2020 market dislocation and other volatile bouts. Spreads were increased by market makers. Activities Creation/redemption activity increased in some of the funds. The specified episodes demonstrated how the liquidity pressure in underlying securities may extend to ETFs and vice versa. Recent scholarly works also quantify directional spillover of ETFs on constituent stocks and detect asymmetric impacts during high-stressed periods.

Modeling contagion: networks and stress tests

Contemporary system risk analysis networks ETFs. Nodes indicate finances, issuers, APs, and securities. The information held by Edges includes holding overlaps, counterparty relationships, and trading correlations. Computations on these networks indicate that even in the absence of a single large fund, shocks can propagate through overlapping holdings, in particular when liquidity evaporates and the price influence is nonlinear. The primary role of the use of granular holdings and transaction data in the recent literature is to identify the network hubs in the chosen funds and the points of regulatory interest.

Stress tests are increasingly being applied by regulators and large asset managers in historically and hypothetically driven modes. Good practice also involves testing the failure of market-making lines, counterparty margin calls, and correlated redemption, among others. The multi-dimensional stress tests are actually superior in realistic contagion as compared to balance sheet-only models.

Practical implications for managers and investors

Others are operational and strategic implications to institutional managers and to complex investors.

• Review fund liquidity profiles: Not only the liquidity of ETFs of focal interest, but also the liquidity of the combining basket. There are thinned-out names represented in some ETFs that appear liquid.
• Monitor concentration and overlap: It might exist because a fund may be a conduit (monitor) of contagion with other funds where there is significant overlap. Due diligence should include tools that assess measures of holding similarity and centrality of networks.
• Stress test funding and leverage: Margin calls and collateral squeezes can cause trades at poor timing. Strong liquidity reserves and contingency lines mitigate forced selling.
• Approach complex/leveraged ETFs with caution: ETFs have to be cautiously treated as they have been encountered by the sights and find location of the products which can act unexpectedly under the influence of stress, as well as heighten the market influence in the short-term period. The investors can have them to get to, on a tactical short-duration trade, that is known to them, how they are framed.

What should regulators focus on?

Improve transparency and frequency of holdings disclosure

Baskets of large funds would be detected in advance to assist market participants and authorities with insight into large funds based on their daily or intraday movements.

Strengthen the resilience of AP and market-making networks

Maintaining strong capital and operational health of APs and dealers can minimize the probability of a failure by an arbitrage provider stopping regular ETF operations.

Focus on secondary products

Measure rises in synthetic, leveraged, and derivative-dependent ETFs that are not fully transparent markets.

Collectively, these enablers can minimize the likelihood that the local volatility is systemic. Designing policies must be equitable in the accessibility of investors and a stable market.

Bottom line: risk is real but manageable

ETF life cycle is durable and multifaceted. Its advantages are quite obvious: cheapness, diversification, and mobility of access. That has attracted trillions of dollars and thousands of funds into the structure through those benefits. Meanwhile, plausible channels of contagion exist through the network effects, concentration, and liquidity frictions.

This is a positive aspect because they can be recognized and addressed by improved data, more resilience of participants, and future stress testing. The next stress event will be better met when investors and managers implement network awareness risk measures.