Author: Jae, PANews
Computing power has become the "new oil of the 21st century" fueling global AI operations. The computing power arms race driven by AI is transcending the physical boundaries of information technology and deeply penetrating the veins of modern financial infrastructure.
Larry Fink, the head of global asset management giant BlackRock, once pointed out that in the context of scarce AI ecosystem resources, a futures market linked to computing power might emerge. This prophecy found concrete evidence in May.
Within just one week, two leading players in the traditional financial market, CME Group and Intercontinental Exchange (ICE), the parent company of the New York Stock Exchange, announced plans to develop GPU computing power futures markets one after another.
Computing power is transforming from an intangible technological resource into a standardized financial asset that can be speculated on, traded, and hedged. The fierce competition among Wall Street giants for the pricing power of this new macro commodity also marks the official opening of the era of computing power asset financialization.
GPU Futures Become New Wall Street Battleground: ICE Aims for Universal Coverage, CME Seeks First-Mover Advantage
In this beachhead landing war for the financialization of computing power assets, the two Wall Street titans have chosen different entry paths.
On May 19th, ICE, in collaboration with data provider Ornn, entered the fray, planning to launch a series of GPU computing power futures contracts based on the Ornn Computing Price Index (OCPI).
The OCPI introduced by ICE is the world's first computing power index constructed from real transaction records. Ornn ensures the transparency of this pricing data by distributing the index in real-time to Bloomberg Terminals through its subsidiary Ornn Data, thereby avoiding the issue of "listed price distortion."
Ornn co-founder and CEO Kush Bavaria believes computing power has grown into a trillion-dollar market, and ICE's futures listing will provide a risk transfer layer for institutional buyers and computing power operators.
ICE's computing power futures contracts cover not only mainstream enterprise-grade high-end GPUs like H100, H200, and B200 but also include high-end consumer-grade graphics cards like the RTX 5090, offering refined hedging options for computing power needs across different scenarios. This means ICE is attempting to seize pricing power for computing power across the entire spectrum, from cloud to edge, and from training to inference.
To further solidify the index's industrial foundation, Ornn has also brought in Hyperbolic Labs, one of the world's largest GPU marketplaces, as an ally. Its co-founder and CEO, Jasper Zhang, pointed out that the current GPU market is increasingly resembling the global commodity market, and ICE's move precisely addresses the risk management pain points of new computing power service providers (Neoclouds) and AI labs.
Rather than saying ICE is actively entering the computing power futures market, it's more accurate to say it is rushing to catch up. In fact, CME had already made the first move a week earlier.
On May 12th, CME announced it would partner with GPU market intelligence and benchmark data provider Silicon Data, backed by trading giant DRW, to launch the world's first computing power futures contracts. As a benchmark in the global derivatives market, CME's entry signifies that computing power has officially been included in the sequence of "macro commodities" recognized by Wall Street.
Unlike ICE's broad approach, CME's computing power futures are anchored to Silicon Data's compiled "H100 Lease Index." By providing daily standardized tracking of real-time on-demand leasing rates across major cloud service providers and new GPU cloud platforms, it establishes a unified pricing benchmark for a highly fragmented and opaque spot market.
To avoid depreciation and transportation losses associated with physical hardware delivery, CME's GPU futures contracts will adopt a cash settlement model. The underlying asset being traded is not the physical chip but expectations for future H100 leasing prices.
For large-scale cloud service providers, this provides a much-needed hedging tool. When a cloud service provider invests billions in procuring H100s, it only needs to establish a short position in the CME computing power futures market to lock in a minimum return on investment (ROI) for its servers, thereby mitigating the risk of asset devaluation caused by plummeting computing power prices.
This approach closely mirrors the logic that once turned crude oil, natural gas, and electricity into commodities.
Computing Power Futures Ignite Pricing Power Battle, Financialization Brings Both Opportunities and Tests
Since the wave of large language models swept the globe, computing power has leapt from an "IT resource" to a "strategic material" contested by the AI triumvirate (OpenAI, Anthropic, Google) and Silicon Valley giants like Meta. Simply put, whoever hoards more GPUs holds the entry ticket to the AI era.
But problems have also emerged: the computing power market is too expensive and too unpredictable.
The four cloud giants—Amazon AWS, Microsoft Azure, Oracle, and Google GCP—control about 78% of global IT power capacity and 69% of H100 supply. Spot leasing prices sometimes surge severalfold, only to plummet during chip generation updates. If an AI lab wants to lock in computing power a year in advance, it may have to pay a significant premium; if not, it faces the risk of supply disruption.
What's more troublesome is that the computing power market lacks hedging instruments.
DRW founder Don Wilson frankly stated: The explosive growth of capital-intensive investments like data centers has historically been constrained by the lack of effective risk management tools. The launch of a computing power futures market is a solution to this pain point.
One could argue that whoever masters the pricing power of computing power masters the Bretton Woods system of the AI era.
The battle for computing power pricing rights between the two Wall Street giants reveals that this emerging factor of production is at a historical intersection of "financialization" and "commoditization." This evolution is underpinned by industry cycles but also accompanied by non-negligible potential risks.
From a supply-demand cycle perspective, the global computing power market is entering a new phase of rebalancing. Although early-stage explosive growth in AI applications led to extreme supply-demand mismatch for high-end GPUs and leasing prices soared multiple times, with the large-scale completion of data center construction and chip generation iterations, spot prices will exhibit high volatility. The market urgently needs forward pricing tools to smooth out risks.
However, the "intangible nature" of computing power means it cannot replicate the delivery logic of traditional commodities. Physical chips have a short lifecycle, typically facing technological obsolescence or depreciation within 18 to 24 months, rendering physical delivery forward contracts invalid. Therefore, using a "standard computing unit," such as converting to a benchmark unit like 1 hour of H100 runtime, supplemented by cash settlement, has become the industry's recognized optimal solution. However, this also increases the complexity of pricing models.
Furthermore, the supply side of computing power is highly concentrated, and the spot market is essentially an oligopoly. Building a derivatives market on such a structure creates an inherent fragility in the price discovery mechanism, making futures prices susceptible to indirect manipulation via spot prices.
More importantly, once the computing power derivatives market fully opens, its leverage attributes might amplify price volatility in the spot market. Inflows of leveraged capital and speculative fervor could drive up procurement costs for computing power, turning small and medium-sized AI enterprises into the "harvested," potentially evolving into a "financial hunt" that further exacerbates the uneven distribution of computing resources.
Wall Street Awaits Approval, Crypto Players Already Charging Ahead
While Wall Street's two major exchanges still await regulatory approval, players in the crypto market have already taken the lead.
As early as January this year, Architect Financial Technologies, founded by the former president of FTX US, partnered with Ornn to launch perpetual contracts linked to the OCPI-H100 index through its AX platform.
As more platforms follow suit, it's not impossible that centralized exchanges (CEX) will successively list related computing power futures markets. Additionally, they might launch structured wealth management products for ordinary users or fixed investment products tied to GPU leasing rates, further achieving seamless integration between the crypto market and traditional financial macro assets.
Compared to the heavily regulated CME and ICE facing lengthy approval processes, perpetual decentralized exchanges (Perp DEX) operating on smart contracts possess higher agility and the institutional advantage of permissionless innovation.
Perp DEXs also don't need to go through the lengthy listing procedures of CEXs. For instance, developers only need to stake 500,000 HYPE tokens (or potentially lower thresholds in the future) to list computing power perpetual contracts linked to GPU indices on Hyperliquid's HIP-3 market. This product development capability will allow DeFi to establish a global computing power speculation market unrestricted by geography or barriers, operating even outside Wall Street's regular trading hours.
However, computing power futures is, after all, a nascent asset class with relatively high risk coefficients in its early stages. The computing power market is predominantly over-the-counter (OTC), making data sources susceptible to manipulation. In more extreme cases, such as facing black swan events like technological breakthroughs or chip embargoes, computing power indices might experience non-continuous, sharp price fluctuations. Both scenarios can lead to price distortion, triggering large-scale liquidations of highly leveraged contracts.
Regardless, the Wall Street giants' scramble for computing power futures marks an inflection point in the convergence of AI infrastructure and modern finance.
GPU computing power, once viewed more as an IT resource, is being experimented with as a quantifiable, tradable, hedgeable standardized asset, embedding the logic of technological resource allocation into the global financial system.
With the commoditization of computing power assets, its resource allocation logic may also shift from relying solely on spot procurement to being increasingly influenced by financial market price signals. In the future, computing power might gradually form more mature price discovery mechanisms and capital allocation systems, much like foundational factors of production such as energy and electricity.
