Wall Street Wants To Trade AI Compute Like Oil. Is The AI Boom Creating Finance’s Next Big Market?
Wall Street’s growing interest in artificial intelligence infrastructure is extending beyond chipmakers and cloud providers, with financial firms now working to create markets where computing power can be bought, sold and hedged in much the same way as traditional commodities.
The push comes as governments and technology companies race to secure computing capacity amid intensifying global competition over AI development, semiconductor supply chains and data center expansion. That competition has taken on added strategic significance following continued export restrictions on advanced AI chips to China, growing investment in domestic computing infrastructure across the United States and Europe, and increasing concerns over securing sufficient power and hardware for next-generation AI models.
Startup Ornn has raised $33 million in seed funding to build a marketplace for trading GPU computing power, according to Axios, which reported that the Andreessen Horowitz-backed company aims to provide financial infrastructure similar to commodity markets used for oil, metals and agricultural products. The company is developing pricing benchmarks that would allow buyers, sellers and lenders to value and hedge computing resources more efficiently.
Unlike commodity markets where companies routinely use futures contracts to lock in prices for oil or jet fuel, AI companies have largely relied on long-term purchasing agreements to secure access to computing resources.The news outlet pointed out that investors increasingly view compute as a commodity that requires standardized pricing and risk management tools as spending on AI infrastructure continues to expand.
The scale of that investment is substantial. Goldman Sachs estimates that roughly $7.6 trillion will be invested globally between 2026 and 2031 across compute infrastructure, electricity generation and data centers, the report said, citing the investment bank’s analysis. The report said the financial systems needed to support spending at that scale remain underdeveloped.
The idea of treating computing power as a commodity has already attracted major financial exchanges. In May, MarketWatch reported that CME Group plans to launch compute futures tied to benchmarks developed by Silicon Data, pending regulatory approval. The contracts are intended to allow AI companies, cloud providers and investors to hedge against fluctuating GPU prices in much the same way businesses manage exposure to energy or metals markets.
CME Group Chairman and Chief Executive Terry Duffy described compute as “the new oil of the 21st century,” MarketWatch reported, while DRW founder Don Wilson said he believes computing could eventually become one of the world’s largest commodity markets. Carmen Li, founder of Silicon Data, told the publication that volatility in GPU pricing and changing supply-demand conditions have created a need for standardized financial products.
Intercontinental Exchange is pursuing a similar initiative. Benzinga reported in June that ICE plans to launch GPU compute futures based on Ornn’s Compute Price Index, while CME’s products will rely on Silicon Data’s pricing benchmarks. Both exchanges are awaiting regulatory approval before introducing the contracts.
According to Benzinga, Ornn Chief Executive Kush Bavaria said compute has grown into a trillion-dollar market despite lacking the pricing transparency and risk-transfer mechanisms that exist for other major commodities. The publication reported that the proposed futures contracts are designed to provide AI developers, cloud providers and financial institutions with tools for valuation, long-term planning and price hedging.
The effort faces significant technical challenges because computing power differs from traditional commodities. GPU hardware depreciates quickly as newer generations of processors deliver better performance, reducing the value of older equipment. Unlike oil or metals, unused computing capacity cannot be stored for future use, making standardized pricing more difficult.
Those issues have also drawn attention from researchers examining the financial implications of AI markets. Bloomberg reported earlier this month that growing adoption of AI across Wall Street may shorten the lifespan of profitable trading strategies as more firms rely on similar models and datasets. Researchers cited by the publication also found that increased use of AI systems could contribute to more crowded trades and expose investment models to manipulation through altered financial news.
Broader financial markets are also beginning to examine whether computing infrastructure can become a financial asset beyond futures trading. The Economist said companies are exploring products ranging from GPU-backed financing to structured financial instruments linked to computing resources. The report noted that computing assets remain difficult to value because hardware becomes obsolete quickly and data center capacity varies significantly by location.
Ornn has already integrated its pricing data with Bloomberg Terminal and other financial information providers, allowing traders to monitor GPU prices using platforms already familiar to institutional investors. The company is operating under a de minimis exemption while larger exchanges continue working through regulatory approval processes.
Pending approvals, CME plans to launch compute futures tied to Silicon Data’s benchmark, while ICE intends to introduce GPU compute futures based on Ornn’s pricing index.
