AI-assisted trading is a growing area in quantitative finance. However, concerns have emerged that it may destabilize markets. We recently discussed how trading strategies generated by large language models could introduce new systemic risks to financial markets.
Continuing this line of research, Reference [1] examines how AI trading affects market volatility, liquidity, and systemic risk. The authors used daily data from the S&P 500 index, applying an OLS regression and a Poisson model to estimate the frequency of extreme price jumps, and a GARCH(1,1) model to analyze volatility clustering. They pointed out,
One of the key takeaways is the lack of a strong direct relationship between AI trading and market fluctuations. While AI trading does not appear to significantly drive volatility under normal conditions, its effects may depend on broader market structures, liquidity availability, and macroeconomic shocks. This suggests that AI-based trading systems do not inherently destabilize markets but may interact with other variables in ways that influence financial stability. In periods of normal trading activity, AI may enhance price efficiency and liquidity provision. However, during financial distress or economic uncertainty, algorithmic decision-making could amplify volatility through feedback loops and self-reinforcing mechanisms. The persistence of volatility observed in the GARCH model supports this argument, indicating that once volatility spikes occur, they tend to last longer.
A significant finding of this study is the strong correlation between energy consumption and market volatility. Unlike AI Presence, energy consumption emerges as a key driver of financial fluctuations. The high computational demands of AI trading suggest that energy-intensive models may contribute to extended periods of heightened volatility, raising concerns about both financial stability and sustainability.
In short, the results show that AI-driven trading is positively associated with more frequent market jumps and higher volatility. Interestingly, AI does not impact markets directly, but rather through the energy consumption used to train the models.
Let us know what you think in the comments below or in the discussion forum.
References
[1] Zorina ALLIATA, Andreea-Mădălina BOZAGIU, The Impact of AI on Market Volatility: A Multi-Method Analysis Using OLS, Poisson, and GARCH Models, Proceedings of the 19th International Conference on Business Excellence 2025
Article Source Here: Jumps and Volatility Clustering in AI-Driven Markets
source https://harbourfronts.com/jumps-volatility-clustering-ai-driven-markets/
