Z. Zhang
6/15/2026
Research Articles
Volume 2, Issue 2
https://doi.org/10.67419/jyi.v2i2.14
Abstract
This paper reviews empirical research on algorithmic trading in the U.S. stock market and argues that its effects on price efficiency, transaction costs, and market volatility all stem from a single underlying behavior: the voluntary supply of liquidity by algorithmic and high-frequency trading firms. These firms are not required to keep trading when conditions turn unfavorable, so they supply liquidity when doing so is profitable and withdraw it when it is not. Under ordinary market conditions, this behavior narrows bid-ask spreads, speeds up the incorporation of new information into prices, and lowers trading costs for participants who can move as quickly as the market does. The same behavior also explains why these benefits break down during rare, correlated market shocks: firms that reliably supply liquidity on calm days can withdraw from the market in unison once conditions become risky, an event best illustrated by the Flash Crash of May 6, 2010. Bringing these three strands of research together shows that algorithmic trading has not simply made U.S. equity markets more efficient or more fragile. Instead, it has tied everyday market quality to a form of liquidity supply that is dependable under normal conditions and prone to sudden, collective withdrawal exactly when markets need it most.