I’m again having conversations with Google Gemini as part of my research into the new story I’m working on. Here’s the first question:
How much of the financial market activity is carried out automatically by AI?
Gemini’s response:
A very significant portion of stock and financial market activity is now carried out automatically, often driven by algorithms and increasingly by Artificial Intelligence (AI).
Here’s a breakdown of the current situation:
- Algorithmic Trading Dominance: The vast majority of trading volume in major markets, particularly in the U.S. stock market, is initiated through algorithmic trading. Estimates often range from 70% to 80% or even higher (some sources indicate around 92% in the Forex market). This dominance of Algorithmic Trading involves computer programs that follow predefined rules and instructions to place trades based on factors like time, price, and volume. Strategies Include:
- High-Frequency Trading (HFT): A subset of algorithmic trading that executes trades in milliseconds, capitalizing on tiny price discrepancies. HFT accounts for a substantial portion of overall trading volume.
- Market Making: These algorithms continuously place buy and sell orders to capture the bid-ask spread, providing liquidity to the market.
- Arbitrage: Arbitrage Algorithms identify and exploit small price differences for the same asset across different exchanges.
- Trend Following: Programmed algorithmsto buy or sell based on identified market trends.
- Execution Algorithms: Execute large orders in the most efficient way possible, minimizing market impact.
- AI’s Evolving Role within Algorithmic Trading: “Algorithmic trading” broadly refers to automated rule-based trading. Integrated AI and Machine Learning (ML)makes this trading more sophisticated and adaptive.
- Self-Learning and Adaptation: Traditional algorithms follow fixed rules. AI-driven algorithms, however, can learn from new data, identify complex patterns, and adapt their strategies in real-time as market conditions change. This is a key differentiator.
- Sentiment Analysis: AI can analyze vast amounts of unstructured data, like news articles, social media, and earnings call transcripts, to gauge market sentiment and incorporate it into trading decisions.
- Predictive Analytics: AI models produce more accurate forecasting of market trends, potential risks, and even customer behaviour.
- Risk Management: AI enhances risk management by dynamically adjusting positions, setting stop-loss levels based on volatility, and identifying fraudulent activities.
- Efficiency and Speed: AI further improves the speed and accuracy of algorithmic trading, processing more information faster than humans can.
In summary:
While it’s difficult to put an exact percentage on “AI-only” trading (as AI is often embedded within broader algorithmic strategies), it’s safe to say that:
- Most financial market activity (70-80%+) is automated through algorithms.
- AI and Machine Learning are increasingly powering and enhancing these algorithms, making them more intelligent, adaptive, and capable of handling complex, unstructured data, moving beyond simple rule-based systems.
The trend is towards even greater integration of advanced AI, including generative AI, in various aspects of finance, from trading to research, risk management, and customer service.
This begs the question. If humans were locked out of the 80% AI run financial markets, would they continue to function?