Skip to main content
AI Fin Hub
Menu

Benchmarks, in AI in Markets

AI Agents and Automation in Markets Statistics

About 37% of US equity volume ran through algorithms in 2023, and regulators at the IMF and FSB treat that figure as a systemic-risk variable, not just a market-structure stat. Each figure below names its source and year; none was generated by this site. The Coalition Greenwich data and the regulatory warnings are primary sources. Market-structure shares shift with how 'algorithmic' is defined, so the relevant figure is reported with its source's own definition.

6 figures. Published May 26, 2026. By AI Fin Hub Research.

Statistics

The numbers worth quoting

1

About 37% of overall US equity volume was executed through algorithms and/or smart order routers in 2023, up from 35% the prior year

The figure reflects buy-side execution channels as measured by Coalition Greenwich in a January 2024 release. Measures of 'algorithmic' share differ across studies depending on what is counted.

Source Coalition Greenwich, US Equity Markets research
2

Among the most active institutional traders, about 59% of flow by notional value was channeled through algorithms

Algorithmic share is markedly higher among the heaviest traders than the market-wide average, indicating automation concentrates among the most sophisticated participants.

Source Coalition Greenwich, US Equity Markets research
3

Buy-side firms projected algorithmic trading to reach about 40% of volume within three years

The projection is the surveyed managers' own expectation of a continued upward trend, with crossing-network use also expected to rise.

Source Coalition Greenwich, US Equity Markets research
4

The IMF warned that AI-driven trading can amplify market volatility and increase the speed and correlation of asset-price moves

The Fund flagged that wider AI adoption could deepen flash-crash dynamics and herding while also improving liquidity provision in normal conditions, a two-sided assessment.

Source International Monetary Fund, Global Financial Stability Report (October 2024)
5

The FSB identified market correlations from common AI models as a channel that could increase systemic risk

The FSB noted that institutions running similar AI systems could behave in a synchronized way, raising the risk of correlated moves and herding in stressed markets.

Source Financial Stability Board, The Financial Stability Implications of Artificial Intelligence
6

On the FinBen financial benchmark, leading LLMs performed well on extraction and analysis but struggled with forecasting and decision-making tasks such as stock trading

FinBen evaluated 15 LLMs across 24 tasks and was the first benchmark to include stock-trading evaluation, finding agentic market tasks among the hardest, which tempers expectations for autonomous trading agents.

Source Xie et al., FinBen: A Holistic Financial Benchmark for Large Language Models

Key takeaways

Automation already executes a large share of US equity volume, with Coalition Greenwich putting algorithms at about 37% in 2023.
Algorithmic share is highest among the most active institutional traders, near 59% of their notional flow.
Regulators (IMF, FSB) warn AI-driven trading can amplify volatility, correlation, and herding.
LLM benchmarks find forecasting and trading decisions among the hardest tasks, limiting autonomous-agent claims.
Market-structure 'algorithmic share' figures depend on definitions, so compare like-for-like sources.

Methodology

Figures are drawn from named research (Coalition Greenwich), regulator reports (IMF, FSB), and a peer-reviewed benchmark (FinBen), each reported with its source and year. Market-structure shares reflect the source's own definition of algorithmic execution. No statistic on this page is derived from data collected by this site.

Run the numbers next

  • Playgrounds Execution Simulator

    Estimate execution cost in closed form — square-root permanent impact, linear temporary impact, half-spread, and a latency-drift band. See the slippage a naive backtest hides.

  • Comparators Model Selector for Finance

    Input task, latency budget, cost budget, context size, and quality sensitivity; get ranked model recommendations with rationale — grounded in published.

  • Playgrounds Forecast Scoring Sandbox

    Paste a forecast stream (probability + outcome) and see Brier score with full decomposition, log loss, reliability diagram, and bootstrap confidence.

Sources

  1. Electronic Platforms Capture Growing Share of U.S. Equity Trading Volume , Coalition Greenwich (2024)
  2. Global Financial Stability Report, October 2024 , International Monetary Fund (2024)
  3. FinBen: A Holistic Financial Benchmark for Large Language Models , Qianqian Xie et al., NeurIPS Datasets and Benchmarks (2024)

Keep the topic connected

  • Statistics AI in Finance Adoption Statistics

    Sourced statistics on AI and generative-AI adoption across financial services: deployment rates, use cases, governance gaps, and where the money is going.

  • Statistics LLM Reliability in Finance Statistics

    Sourced statistics on large language model reliability for financial tasks: hallucination rates, numerical-reasoning accuracy, citation faithfulness, and benchmark gaps.

  • Statistics AI Governance and Regulation in Finance Statistics

    AI governance statistics for finance: EU AI Act timelines and fines, regulator-flagged risks, and the gap between AI adoption and formal strategy.