Key Takeaways:
Two protocol upgrades turned Hyperliquid from a crypto perpetuals exchange into a platform where builders deploy markets on top of its infrastructure, the way developers deploy apps on cloud computing.
Hyperliquid spent its first two years as the fastest decentralized perpetuals exchange in crypto. Since late 2025, the network has been executing a more ambitious plan. HIP-3, live on mainnet since Oct. 13, 2025, lets any builder who stakes 500,000 HYPE deploy an independent perpetual futures exchange on HyperCore, choosing the assets, oracle, collateral, and fee structure while inheriting Hyperliquid's matching engine and margining systems. HIP-4, live since May 2, 2026, adds fully collateralized outcome markets that settle to exactly zero or one at expiry, with merged Yes and No order books and no liquidation risk.
"Hyperliquid now looks less like a stock exchange and more like Amazon Web Services," Grayscale Research wrote in a June 2026 note cited by Stocktwits. The comparison captures the architectural shift: HyperCore processes around 200,000 orders per second, and builders deploy markets on top of that engine without needing core team approval for each new listing.
Open interest across HIP-3 markets grew from roughly $790 million in January 2026 to a peak of $3.2 billion by June, according to Grayscale. Cumulative trading volume since launch has exceeded $200 billion. Seven of Hyperliquid's top ten markets by volume are now tokenized equities or commodities rather than traditional cryptocurrency pairs, including Nvidia, Tesla, gold, and a licensed S&P 500 perpetual contract. At peak HIP-3 activity, the platform generated $2.3 million in daily fees, funding $11 million in HYPE token buybacks.
The economic design of HIP-3 uses the 500,000 HYPE stake as a slashable bond. Deployers earn half of trading fees, which are set at roughly double the native rate — 0.09% for takers versus 0.045% on validator-operated pairs. A Growth Mode feature introduced in November 2025 lets deployers reduce fees by 90% to accelerate adoption, according to OAK Research. All HIP-3 markets are margined in USDC, priced against off-chain oracles, and trade 24 hours a day, seven days a week.
TradeXYZ dominates, but concentration is the caveat
TradeXYZ, built by the Hyperunit team, accounts for more than 90% of all HIP-3 open interest. The platform offers exposure to U.S. equities such as NVDA, TSLA, GOOGL, and AMZN, a synthetic Nasdaq-style index called XYZ100, and commodities including gold and silver benchmarked to COMEX front-month futures. Non-crypto assets achieved 60% trader retention in late March 2026, indicating sustained engagement rather than speculative activity.
The concentration is the caveat. Blockworks Research has flagged the deployer economics as a structural risk: with a roughly $30 million lockup, auction costs, and stiff competition, a smaller deployer's break-even period can stretch to four years. Hyperliquid's documentation says the 500,000 HYPE threshold is expected to fall as the infrastructure matures. Until it does, HIP-3 is permissionless in principle and an oligopoly in practice.
HIP-4 removes the token vote from settlement
HIP-4 addresses a structural weakness in existing on-chain prediction markets. Polymarket outsources contested resolutions to UMA's optimistic oracle, where token holders vote on disputed outcomes — a system that produced repeated controversies in 2026, including a $60 million market on a Strategy Bitcoin sale that resolved against the documented facts, as covered in a companion explainer.
HIP-4 replaces the token vote with deterministic settlement through Hyperliquid's validator set, executing automated resolution against pre-specified objective data sources. There is no dispute window, no escalation, and no path for a market participant to also vote on the outcome. The trade-off is scope: deterministic settlement only fits questions with clean data sources, which is why the first HIP-4 contracts are Bitcoin price thresholds that reset daily, run by the prediction platform Outcomexyz.
Initial markets are curated and validator-deployed. A later phase will open permissionless deployment, with builders staking 1,000,000 HYPE per market slot, slashable and burned if validators find oracle manipulation or invalid state transitions. Opening or minting an outcome position costs nothing; fees apply only on closing, burning, or settling, and makers pay zero. That pricing targets Polymarket and Kalshi, which processed a combined $44.8 billion in June on the back of the World Cup.
What to watch from here
Three markers will tell the story over the next year. First, whether the HIP-3 stake requirement drops and the deployer set widens beyond one dominant builder. Second, whether HIP-4 volume becomes measurable against Polymarket and Kalshi once permissionless deployment opens and categories expand past crypto prices. Third, whether regulators treat builder-deployed stock perpetuals as an innovation to license or a loophole to close. The upgrades themselves are shipped and working. The open question is what survives contact with scale.
This article is for informational purposes only and does not constitute investment advice.
