Institutional

Wall Street banks post record trading revenues on the AI-stock frenzy

Goldman Sachs and JPMorgan Chase led a quarter of blowout trading numbers, with AI-related equities providing both directional volume and the volatility that market-makers monetise.

· Source: Financial Times


The FT and CNBC framing - AI is the trade - is doing two different jobs at once, and it is worth separating them. The first is a market-directional story: Nvidia, hyperscaler capex names, US mega-caps with AI exposure, and a broadening cohort of software and infrastructure companies have delivered a fat tape for both prop books and client flow. The second is a market-microstructure story: elevated cross-sectional volatility, high-quality liquidity in the AI complex, and unusually consistent retail interest have produced ideal conditions for systematic and market-making desks.

For an AI-driven trading system, the second story matters more than the first. High-turnover names with tight bid-ask spreads and heavy options activity are exactly the environment in which multi-agent LLM systems can add measurable signal on the sentiment/news side without being murdered by transaction costs. But the same tape is what makes reproduced Sharpe ratios in the recent multi-agent papers (TradingAgents on AAPL, FinAgent on GOOG) hard to trust - the underlying market makes almost any long-biased AI-momentum strategy look competent.

Two operator takeaways. Backtest your AI stack on a genuinely different regime - Q4 2018, March-April 2020, autumn 2022 - before deploying capital that depends on the current AI-driven tape continuing. Treat the current environment as sample-of-one: strategies that need this specific volatility surface and this specific concentration of retail attention will not survive a mean-reversion or a rotation into other sectors.

The banks are earning fees on the frenzy. The retail builder should be pricing the frenzy into how much of their alpha is real.


Read the original: Financial Times - Wall Street banks post record trading revenues on the AI-stock frenzy. Commentary is the independent editorial view of Share Trading; the original article is credited to its publisher.