High-frequency traders exploiting momentary discrepancies in stock prices can potentially profit by as much as 0.77% per day, according to a new analysis of Nasdaq ITCH data focusing on Dow Jones Industrial Average stocks. The study, published in the Journal of Investment Strategies, highlights the critical importance of speed – measured in microseconds – in capitalizing on these “microtrend” anomalies.
Researchers found that an idealized trader with zero latency could achieve returns exceeding 3% on specific stocks. Though, the profitability of this strategy is acutely sensitive to even slight delays. The maximum tolerable latency, the study calculates, is an average of 14.6 microseconds for an equally weighted portfolio, with individual stocks exhibiting a range of 0 to 40 microseconds. Beyond these thresholds, the potential for profit diminishes rapidly.
The research characterizes the exploitable trend-length anomalies through a high-frequency, microtrend-following strategy. A key factor impacting profitability, the study notes, is the crossing of bid-ask spreads. Rapid market access is identified as crucial for successfully implementing this type of trading strategy.
The findings underscore the ongoing evolution of electronic markets and the increasing sophistication of algorithmic trading. High-frequency trading (HFT), driven by automated systems, has become a dominant force in stock exchanges, utilizing high-speed computer programs to generate and execute orders. Investment banks, hedge funds, and institutional investors are all actively involved in designing and deploying these strategies, as noted in a recent paper published on arXiv.org.
Further research, including work detailed in Financ Innov, continues to explore novel modelling strategies for analyzing high-frequency stock trading data. A study published on semanticscholar.org indicates that graph neural networks are showing promise in anomaly detection within HFT, achieving a 15% improvement in accuracy compared to traditional methods.
The impact of HFT extends beyond individual markets, influencing liquidity in related asset classes like options. Recent analysis from ScienceDirect highlights the need to understand how HFT in the stock market affects options market liquidity, a gap in existing research the study aims to address.