The February 2026 Selloff › The Liquidity Illusion and Why Calm Markets Lie to You

Modern markets create an illusion of liquidity that holds up when things are calm and breaks down when they aren't. Before decimalization and the rise of high-frequency trading, the New York Stock Exchange relied on human specialists with formal obligations to maintain orderly markets, providing liquidity even during stress. Today, algorithmic market makers account for over 60% of equity trading volume and provide abundant liquidity in normal conditions, quoting tight spreads and deep order books. But these firms have minimal presence obligations and activate kill switches when volatility spikes, pulling out when liquidity is needed most. Researchers have called this "the liquidity illusion": markets that appear deep and liquid in calm periods but become thin and unreliable under stress, with bid-ask spreads widening and order book depth collapsing. ETFs amplify this illusion. With over \$6.2 trillion in U.S. assets, ETFs account for 31% of equity trading volume and create the appearance of liquid access to every market and sector and asset class. But research has shown liquidity shocks at the ETF level transmit to underlying securities through the arbitrage mechanism, with spillovers increasing during crises. During the August 2015 China selloff, major ETFs traded well below their net asset values. In extreme stress, the arbitrage mechanism breaks down entirely, and ETFs amplify volatility rather than dampen it. A liquidity shock in a single ETF forces price movements in every one of its hundreds of underlying holdings, regardless of those holdings' fundamentals. Leveraged ETFs add another layer. These products, which reached a record \$117 billion in assets in 2024, must rebalance daily to maintain their target leverage ratio. On September 3, 2024, leveraged ETFs sold an estimated \$15 billion in equities in a single session, contributing to a 3% plunge in the Nasdaq 100. The rebalancing is concentrated in the final 30 to 60 minutes of trading, creating predictable end-of-day volatility surges that other algorithms front-run. A 2x leveraged ETF tracking an underlying that drops 10% and then recovers 11.1% ends up losing 1.82% even though the underlying is flat. This is volatility decay, and it quietly erodes leveraged positions during choppy markets. Good fundamentals alone did not explain the calm, liquid, low-volatility market that prevailed for much of 2025. It was partly an artifact of the same leverage, gamma positioning, and algorithmic market-making that would amplify the eventual reversal. Low volatility attracted more capital into volatility-targeting strategies, which increased their equity exposure, which suppressed volatility further, which attracted even more capital. This reflexive loop is the inverse of the crash dynamic. Equally mechanical, equally divorced from fundamentals, and bound to reverse.