Impact of arbitrage between leveraged ETF and futures on market liquidity during market crash

A team of Japanese researchers has used sophisticated AI-powered market simulations to challenge conventional wisdom on high-frequency trading. The study, led by Ryuki Hayase, Takanobu Mizuta, and Isao Yagi, specifically modeled the relationship between leveraged exchange-traded funds (L-ETFs) and their underlying index futures. By creating an artificial market capable of simulating crashes, they tested how the presence of algorithmic arbitrage trading—which exploits tiny price differences between these linked assets—affects market stability when prices plummet.

The key finding is counterintuitive: this arbitrage activity, often blamed for market fragility, actually acts as a shock absorber. When a crash originates in the L-ETF market, the simulation showed arbitrageurs supply liquidity from the futures market, improving the L-ETF's SellDepth (available sell orders) and Tightness (the bid-ask spread). Conversely, when a futures crash occurs, liquidity flows from the ETF market to support futures, while also boosting trading Volume in the ETF market. The AI model allowed the researchers to isolate and analyze these internal mechanisms, providing empirical evidence that this specific form of automated trading can mitigate liquidity dry-ups during extreme stress.

Published in The IEICE Transactions on Information and Systems, this research offers a nuanced, data-driven perspective on a complex market structure. It suggests that the interconnectedness created by products like leveraged ETFs and the algorithms that trade them is not purely a source of risk. Instead, under the conditions modeled, this electronic linkage functions as a private-sector liquidity backstop, transferring stability from one venue to another during a panic, which could inform future financial regulation and product design.