Brokers Prepare AI Compute Derivatives as Funds Eye Futures

ADM, BGC and Nanhua are preparing compute derivatives in the US while exchanges await CFTC approval for futures linked to AI infrastructure costs.

Major commodity brokers are preparing derivatives tied to the cost of computing power used in artificial intelligence, creating potential products for hedge funds and institutional traders.

ADM Investor Services, BGC Group and Nanhua USA are positioning themselves for the possible launch of compute futures in the US. The contracts could help companies manage exposure to data-center capacity and allow traders to compare prices for electricity, natural gas and computing power.

Potential users include hyperscalers, cloud providers and AI developers seeking to manage infrastructure costs. Hedge funds could use the contracts to trade price differences between energy and computing markets. The size and liquidity of the market remain uncertain.

BGC created a division in June to support secondary trading in compute and memory capacity. The division also provides real-time risk management. Nanhua USA has reported growing interest from technology-focused clients, while ADM Investor Services is considering expanding its research on the market.

“There’s definitely going to be significant interest for it from those looking to hedge,” according to Zach Espinosa, BGC’s co-head of compute infrastructure markets. He added that standardized contracts could improve liquidity by giving traders common terms for risk management.

CME Group, Intercontinental Exchange, Nodal Exchange and Architect Financial Technologies are awaiting approval from the Commodity Futures Trading Commission to launch contracts linked to computing power. In August, the CFTC prevented exchanges from listing compute derivatives while it completed a public comment period. Prediction-market operator Kalshi later withdrew related event contracts.

The proposed products come as investors assess the capital required to build data centers and expand AI processing capacity. Peter Keavey, CME’s global head of energy and environmental products, described the leverage and financial exposure involved in AI infrastructure development as a reason for risk-management instruments.

If approved, compute futures would let traders take positions on changes in computing costs and compare those costs with energy prices. Exchanges would need to establish benchmarks for computing power before the contracts could support trading at scale.

Articles by this author