Skip to main content
ETFs vs mutual funds

The Flash Crash and ETFs

Pomegra Learn

The Flash Crash and ETFs

Quick definition: The Flash Crash was a 2010 market event in which major stock indices and many ETFs plunged and recovered within minutes, exposing structural vulnerabilities in ETF mechanics during extreme liquidity disruptions and prompting regulatory reforms.

On May 6, 2010, the U.S. stock market experienced a sudden, severe, and largely unexplained crash. Major indices fell 5% to 10% within minutes, then recovered almost as quickly. During this chaos, many ETFs traded at extreme premiums and discounts to NAV, and some ETFs were "circuit breaker halted" (trading suspended) while their underlying holdings continued trading. The Flash Crash exposed the structural limits of the ETF model under extreme conditions and raised important questions about ETF suitability for passive investors. Understanding this event clarifies the risks and resilience of ETF structures.

Key Takeaways

  • The Flash Crash of May 2010 saw major indices fall 5% to 10% in minutes, with ETFs trading at extreme premiums and discounts to NAV
  • ETFs trading at a discount to NAV during the crash meant passive investors could only exit at prices significantly below the value of underlying holdings
  • The creation-redemption mechanism relies on authorized participants to arbitrage premiums and discounts; under extreme stress, AP liquidity can evaporate
  • Regulatory reforms after the Flash Crash, including circuit breakers and position limits, reduced the probability and severity of flash crashes
  • Long-term passive investors are unlikely to be affected by flash-crash-type events, but the event illustrates the risks of illiquid or leveraged ETFs

The May 6, 2010 Event

On May 6, 2010, starting around 2:30 PM Eastern Time, the S&P 500 index fell precipitously. The market fell 5% in a few minutes, the largest intraday decline since the 1987 Black Monday crash. Panic selling overwhelmed buying pressure. Liquidity disappeared. Bid-ask spreads widened dramatically.

Within minutes, the market began to recover. Within an hour, the indices had largely recovered the losses. The entire event lasted less than 30 minutes, but the damage was done: approximately $1 trillion in market value had vanished and reappeared.

The cause remains somewhat debated, but a common narrative implicates a large automated trade that exacerbated existing selling pressure. As selling accelerated, liquidity evaporated, and computers programmed to avoid losses triggered further selling, creating a feedback loop.

The interesting aspect for ETF investors: many ETFs traded at the extreme discounts and premiums to NAV during the crash, while their underlying holdings (individual stocks) continued trading. Investors who tried to exit ETF positions received prices far below what the holdings were worth.

ETF Dislocations During the Flash Crash

Several ETFs experienced notable dislocations during the Flash Crash:

Some major ETFs traded at discounts of 5% to 10% to NAV. An investor holding an ETF worth $100 per share in underlying value might see it trade at $92 to $95, with no rational explanation for the gap.

Some leveraged inverse ETFs—which move opposite to the market—traded at extreme premiums as panicked investors sought protection. A 3x inverse Nasdaq ETF that should have gained 15% on a 5% Nasdaq decline was trading up 40% or more, a massive dislocation.

The dislocations were temporary, lasting minutes to hours. By market close, most ETFs had normalized. However, for investors trying to exit during the dislocation, the losses were real.

The Flash Crash revealed a critical vulnerability: the creation-redemption mechanism, which normally keeps ETF prices aligned with NAV, can break down under extreme market stress when authorized participants disappear and underlying markets become illiquid.

Why the Creation-Redemption Mechanism Failed

The creation-redemption mechanism relies on authorized participants standing ready to arbitrage premiums and discounts. When an ETF trades at a discount, an AP can buy the ETF, redeem it for the underlying basket, and sell the basket for NAV, capturing the spread.

However, during the Flash Crash, this arbitrage broke down because:

  1. Underlying market liquidity evaporated. Authorized participants could not quickly sell the baskets of securities they received from ETF redemptions because buying interest had disappeared. They faced the risk of holding illiquid positions. This risk made arbitrage unprofitable and dangerous.

  2. Market uncertainty. APs did not know if the market crash was the beginning of a catastrophic decline or a temporary panic. Holding inventory during uncertain conditions is risky.

  3. Regulatory halts. Circuit breakers halted trading in some securities and indices. When trading halts, the creation-redemption mechanism cannot function—APs cannot execute trades to complete arbitrage transactions.

  4. Cascading fear. As market dislocations worsened, selling pressure accelerated, creating a negative feedback loop. APs retreated from providing liquidity.

Impact on Different Types of ETFs

The Flash Crash affected different ETFs differently:

Large, liquid broad-market ETFs (SPY, QQQ): These experienced discounts and dislocations, but because they hold highly liquid underlying securities and because the creation-redemption mechanism still functioned partially, discounts were limited to 1% to 2% in most cases. Recovery was quick.

Leveraged and inverse ETFs (SSO, SDS, SPXU): These experienced dramatic dislocations. Some leveraged inverse ETFs that were supposed to track 3x movements in opposite directions experienced extreme premiums as panicked investors sought protection. Some leveraged long ETFs experienced extreme discounts as investors fled. These dislocations were severe and lasted longer.

Less liquid sector and specialty ETFs: These experienced wider dislocations and slower recovery because their underlying securities were less liquid.

Bond ETFs: Credit spreads widened dramatically, but bond ETFs generally experienced smaller dislocations than equity ETFs because the underlying Treasury and corporate bond markets remained more liquid than stock markets.

Regulatory Response: Circuit Breakers and Safeguards

The Flash Crash prompted regulators to implement safeguards:

Circuit breakers: The SEC implemented a market-wide circuit breaker that halts trading if the S&P 500 falls 7% in a single day (level 1), 13% (level 2), or 20% (level 3). Individual stocks have single-stock circuit breakers that halt trading in a stock if it moves 10% in 5 minutes.

Revised quotation rules: The SEC implemented a "limit up-limit down" rule that prevents trades from executing at extreme price levels, preventing some of the most severe dislocations.

Position limits: The SEC implemented position limits for certain index derivatives to reduce the possibility of large directional bets destabilizing the market.

ETF-specific oversight: The SEC increased scrutiny of leveraged and inverse ETFs and required risk disclosures. Some leveraged inverse products were delisted.

These safeguards have reduced the frequency and severity of flash-crash-type events. A similar event has not occurred since 2010, though smaller dislocations happen occasionally during periods of high volatility.

The March 2020 COVID Crash

The next major stress test for ETFs came in March 2020 when the COVID-19 pandemic triggered a sharp market decline. Despite severe selling pressure, ETF dislocations were more limited than in the Flash Crash, partly due to regulatory safeguards and partly because the market crash, while severe, followed a more rational pattern (widening credit spreads, forced liquidations) rather than the unclear chaos of the Flash Crash.

However, some liquidity disruptions still occurred. Bid-ask spreads widened, and some bond ETFs, particularly high-yield bond ETFs and emerging market ETFs, experienced notable discounts to NAV. The underlying bond markets also experienced dislocations as liquidity dried up.

By March 2020, market-making had also evolved, with more sophisticated technology and deeper pockets of capital available to provide liquidity. This helped limit dislocations compared to 2010.

Implications for Passive Investors

What should passive investors conclude from the Flash Crash and similar events?

Broad-market ETFs are resilient: Large, liquid ETFs tracking broad indexes (S&P 500, total market) experienced temporary dislocations but recovered quickly. The underlying creation-redemption mechanism, while stressed, still functioned sufficiently. Dislocations were 1% to 2%, not 5% to 10%.

Leveraged and inverse ETFs are risky: Leveraged ETFs (2x or 3x) and inverse ETFs (which profit from declines) experienced extreme dislocations. Passive investors should avoid these. They are designed for active traders, not long-term investors.

Buy-and-hold investors are largely unaffected: If you buy and hold a broad-market ETF for decades, flash-crash-type events are minor bumps. You do not realize losses unless you sell during the dislocation. Most investors should avoid selling during panics.

Illiquid ETFs can be risky: Specialty ETFs, particularly those holding illiquid underlying assets, are vulnerable to wider dislocations. Passive investors should stick to liquid, widely-held ETFs.

Market safeguards matter: Regulatory reforms have reduced the frequency and severity of flash crashes. The probability of another extreme event similar to May 6, 2010 is lower today than it was in 2010.

When Not to Panic

The Flash Crash is a useful historical reminder:

  • Do not sell in a panic. If an ETF trades at a discount during a market dislocation, you are locking in the loss. Patience usually pays as market stabilizes and dislocations correct.
  • Stick with liquid ETFs. Large, widely-held ETFs are much safer than specialty or leveraged products.
  • Understand what you own. Passive investors in broad-market ETFs were largely unaffected by the Flash Crash. Investors in leveraged or inverse products were devastated. Know the structure of your holdings.
  • Long-term horizons matter. Flash-crash dislocations last minutes to hours to days, not weeks. Long-term investors are unaffected.

A Mermaid Diagram: Flash Crash Cascade

Next

The passive investing series on ETF structures concludes with this article; additional related information on advanced ETF topics and risk management can be found in Chapter 6: Costs, TER, and Tracking Error.