London-based fintech startup Permutable AI today announced the successful launch of its end-to-end Large Language Model (LLM) based systematic trading platform, which has been operating live since October 2024.
The innovative system, developed by founder Wilson Chan and his team, has outperformed all existing indices across energy, metals and agricultural commodities over the last six months during a period of unprecedented global market volatility. Most notably, Permutable AI's trading network has surpassed the S&P Global Macro Commodity Index - the primary CTA benchmark for the industry.
In contrast to traditional investment vehicles, Permutable AI's system has maintained a negative correlation with the S&P 500, delivering robust returns during recent market sell-offs when investors have increasingly turned to commodities as a haven. The system has achieved a 26% annualised return with sub-10% annualised volatility while maintaining a negative 20% correlation with the S&P 500 during the Trump Tariff period.
This impressive performance comes at a key juncture for the industry, which has experienced generally disappointing results across the CTA sector in recent years, with many commodity hedge funds posting significant losses amid geopolitical uncertainty.
"While the broader CTA industry has struggled to navigate recent market complexities in 2025, our system has demonstrated impressive resilience and performance," said Wilson Chan, CEO of Permutable AI. "By harnessing the power of advanced language models to process and interpret vast datasets in real-time, we've created a trading infrastructure that not only adapts to market volatility but thrives within it."
Permutable AI is currently collaborating with several key institutions in the hedge fund industry, further validating the potential of this approach to systematic trading.