Arthur Hayes Proposes Futures Market for AI Compute

Arthur Hayes Takes Aim at AI Compute Trading

Crypto pioneer Arthur Hayes is stepping up with another ambitious financial product idea. The BitMEX co-founder has published a theoretical blueprint for trading artificial intelligence compute power using quarterly futures and perpetual swap contracts.

Hayes views this market design as his attempt at a third major innovation in crypto finance. His earlier contributions include launching the perpetual swap at BitMEX in 2016 and developing the synthetic dollar concept that powers EthenaENA-1.70%'s USDe stablecoin.

The core goal of his new proposal is simple: create a standardized way for market participants to price and hedge the cost of raw AI computing capacity.

Can One Network Set the Global Price for Compute?

According to Hayes, the AI sector currently lacks a unified global price index for computing power. Right now, costs are fragmented across physical chip rental rates and varying per-token model access fees.

To solve this fragmented pricing model, Hayes proposes the Flop Reference Compute Rate (FRCR). This metric would function as an hourly price index calculated directly from computing jobs processed on the Flop Network.

On the Flop Network, automated AI agents pay network miners in native FLOP tokens to execute AI model queries and tasks. In simple terms, FLOP acts as the digital currency that fuels compute activities on the blockchain.

This summer, I unveiled my idea for the Flop Network. FLOP is food for your AI agent. Agents will use FLOP in commerce because FLOP is... — Arthur Hayes, October 26, 2026

How the Derivatives Contracts Would Work

The proposed market setup relies on specific financial mechanics to allow traders, compute buyers, and sellers to lock in prices or hedge their operational costs.

The design includes several key components:

• Quarterly Quanto Futures: These contracts are denominated in FLOP tokens for index calculations, but settled in US Dollars. This quanto setup allows traders to gain exposure to compute prices while posting collateral in cash.
• Perpetual Swaps: These contracts have no expiration date and rely on an eight-hour premium index to balance long and short positions via funding payments.
• Market Maker Hedging: Liquidity providers can replicate risk positions by rolling forward strips of 30-day compute units.
• FLOP Performance Bonds: Compute sellers must post FLOP tokens as collateral above their fees. If a seller fails to deliver promised compute within 24 hours, their bond is forfeited.

Evaluating Track Record and Project Timelines

While Hayes has a history of introducing popular financial instruments, market participants remain cautious about building a complex derivatives system on top of a single unlaunched token ecosystem.

Looking at previous innovations, Ethena's USDe synthetic dollar grew significantly to a peak supply near $14.8 billion in October 2025. However, data from DefiLlama shows its supply later declined below $5 billion, illustrating how market conditions can shift over time.

Please note that past performance of digital assets or financial products does not guarantee future results or performance stability.

Additionally, the underlying network infrastructure is still in development. Flop Labs has not finalized a public testnet release date as of mid-October, and the team is targeting Q1 2027 for the network's genesis block.

Hedging Against Potential AI Industry Risk

Hayes has separately suggested that massive capital expenditure in AI data centers could eventually require financial risk management tools or lead to sector adjustments. A functional compute derivative market could help institutions manage those risks.

However, whether broad market adoption occurs will depend on whether institutional buyers and market makers accept a compute price benchmark tied exclusively to one network token.