The Race to Trade Everything

Introduction

Hyperliquid’s HIP-4 protocol upgrade marks the arrival of a third major prediction market model. Among incumbents, Polymarket owns consumer-native discovery and Kalshi owns regulated U.S. exchange access. Hyperliquid announced HIP-4 was coming in February, aiming to bring outcome markets to the “house of all finance,” and on May 2 activated the proposal on mainnet. Prediction markets have proven themselves as one of the most useful new financial primitives in years for hedgers and speculators. With Hyperliquid already the industry standard for perpetual futures execution and infrastructure, it is now positioning to compete for event trading volume.

Hyperliquid’s user base is different from Polymarket’s or Kalshi’s. Both incumbents have spent years building products that appeal to non-crypto people. Both have polished consumer frontends that allow users to “browse” prediction markets as if they were shopping on Amazon. Hyperliquid serves active crypto-native traders through a terminal frontend. That’s a narrower top of the sales funnel, but it’s a user base that already trades, already holds stablecoins (for the most part), and already has the wallet open. HIP-4 is built for that user first.

As of May 25, HIP-4 has expanded from recurring BTC price binaries into validator-published canonical markets covering real-world offchain events. Live markets now include:

1. “BTC above X price on Y date at Z time?” This is a recurring daily binary contract that resets at 2 a.m. EST and settles to the BTC mark price on HyperCore at expiry

2. “BTC price range on X date at Y time?” This is a multi-outcome contract launched within days of HIP-4's mainnet activation, with three buckets (above, in-between, below) settling to the same oracle.

3. June Fed rate decision (change / no change), settling ~June 17

4. May CPI YoY (exactly 4.3% / below 4.3% / above 4.3%), settling ~June 10 to official BLS data

5. NBA Finals Game 4 (Knicks / Spurs)

6. 2026 NBA Finals Champion (Knicks / Spurs)

7. 2026 World Cup Champion (France / Spain / England / Portugal / Brazil / Argentina / Germany / etc.)

These are narrow products, as they should be starting out. The interesting question is what HIP-4 can do, how it stacks up against Polymarket and Kalshi on infrastructure, fees, and distribution, and what its arrival means for a category where everyone is now converging on the “trade everything” model from different starting points.

This piece is a deep dive on HIP-4: what it enables, how it compares to Polymarket and Kalshi, and what it means for the prediction markets landscape.

Executive Summary

• HIP-4 went live on Hyperliquid mainnet May 2, 2026. The proposal was announced Feb. 2, hit testnet that week, and shipped three months later. Outcome contracts now sit alongside perps and spot under a single margin account on HyperCore, Hyperliquid’s native trading engine.

• The launch came just days after Polymarket completed its migration to a new central limit order book (CLOB v2) and stablecoin (pUSD). That same week, Bloomberg reported Polymarket was seeking U.S. regulatory approval to bring its flagship exchange onshore.

• Prediction markets crossed $150 billion in combined lifetime volume in April. Kalshi posted a record $14.81b notional volume (+13.3% MoM) and overtook Polymarket on notional, taker volume, and transaction count for the first time. Polymarket’s volume fell to $9.01b (-14.8% MoM).

• Sector volume posted its first MoM decline in seven months, but the longer trajectory is intact: monthly taker volume has grown more than 17x in two years. Analysts at Bernstein project the category scaling from $51b in 2025 to $1t by 2030.

• HIP-4 plugs directly into HyperCore, Hyperliquid's onchain CLOB, sharing the same execution layer and unified margin account as perps and spot trading. Sub-second finality, ~200,000 orders per second throughput, and cross-margin composability are structural advantages neither Polymarket (considering migration) nor Kalshi (closed centralized infrastructure) can replicate without a full product rebuild.

• By Day 25, HIP-4 had captured 20.1% of combined BTC prediction-market 24h volume (HIP-4 $2.38m vs Polymarket $9.46m).

Prediction Markets Right Now

Prediction markets have been one of the standout growth stories in crypto and broader fintech over the past 18 months. We have written about them numerous times, as well as iterations such as impact markets and decision markets. Combined monthly volume across Polymarket and Kalshi was approximately $2 billion as recently as September 2025. By April 2026 it had reached roughly $24 billion in notional volume across the two platforms, with lifetime combined trading volume crossing $150 billion during the month. Total monthly taker volume has grown more than 17x in under two years.

April was the month Kalshi’s lead became structural.

Kalshi posted $14.81 billion in April notional volume, a 13.3% increase over its prior March record and an all-time monthly high. Polymarket’s volume fell 14.8% to $9.01 billion. Kalshi’s lead widened from $2.5 billion in March to $5.8 billion in April, the largest monthly gap between the two platforms on record. Kalshi also flipped Polymarket on taker volume ($5.42b vs $1.99b) and transaction count (94.4m vs 87.4m), the latter reversing a long-running pattern where Polymarket led on transactions.

Two structural drivers explain Kalshi’s pull-away. The first is product mix. Sports contracts made up 74.3% of Kalshi’s weekly volume in late April, and Exotics (the platform’s parlay-style combination contracts) pushed that share to roughly 85%. The Masters golf tournament alone generated $545 million in notional volume on Kalshi, matching the platform’s Super Bowl single-game total. Kalshi’s setup heading into the NBA Finals and 2026 FIFA World Cup is strong. That said, sports markets skew closer to gambling than to the kind of "information signaling" value geopolitics or tech markets generate, even if some commercial hedging use cases exist. The second driver is regulatory clarity. Kalshi’s status as a CFTC-regulated designated contract market (DCM) has enabled distribution partnerships and a faster path to U.S. retail than Polymarket has had access to.

Polymarket’s category mix is more diversified (sports at 46%, politics at 27%, crypto at 22% as of late April) which is structurally an advantage during periods of political volatility or major crypto cycles, but a disadvantage during sports-heavy stretches without political catalysts. April was one of those stretches. Polymarket’s active user count also softened, falling from approximately 733,000 in March to 646,000 in April.

Still, Polymarket’s economics remain strong. Polymarket collected $47.7m in April fees. The user base is also still significantly larger than Kalshi’s implied retail base. Polymarket retains the deepest non-sports markets in the category, an official MLB partnership announced in March, and a $2 billion strategic investment from NYSE parent ICE. The platform already runs a limited US-compliant product through an acquired CFTC-licensed exchange and is pushing to bring its flagship exchange onshore (more on this below). It also just completed a major infrastructure upgrade (CLOB v2 plus a migration from USDC.e to pUSD, its new internal collateral token) on April 28, four days before HIP-4 went live.

April also marked the first month-over-month decline in sector-wide volume for prediction markets after seven consecutive months of records. Whether that is a calendar artifact (April has no Super Bowl, no March Madness, no NFL playoffs) or the first real demand ceiling is the open question as we see how May plays out.

Sector open interest stood at $1.11 billion on May 1, with Kalshi at $630.7 million and Polymarket at $449.9 million (a combined ~98% share). Limitless, predict.fun, and Opinion each held under $15 million.

Every major venue is moving toward the same “trade everything from one account” model from a different starting point.

What’s interesting here is the strategic convergence underneath the volume numbers. Kalshi and Polymarket are reportedly both building perpetual futures. Hyperliquid has already built outcome markets with HIP-4. Every major venue is moving toward the same “trade everything from one account” model from a different starting point.

HIP-4 Mechanics

HIP-4 adds outcome markets to HyperCore, Hyperliquid’s onchain trading engine. These are fully collateralized binary instruments that settle to either 0 or 1 at expiry based on whether a discrete real-world event occurred. The mechanism is simple and familiar to prediction markets traders: buy YES at price P. If the event occurs, the contract settles at 1 and you collect (1 – P) profit. Buy NO and the inverse applies. Maximum loss is always your entry cost. No leverage and no liquidations. Under the Coinbase/Circle AQAv2 arrangement, USDC is becoming the aligned quote asset across Hyperliquid's markets; canonical HIP-4 outcome markets settle in USDC, with the Native Markets–issued USDH being phased out.

To deploy a market, a builder must stake 1,000,000 HYPE and then define the event title, resolution time, oracle source, and optional window for challenging resolutions. The 1m HYPE stake applies to builder-deployed markets. Canonical markets are published directly by the validator set without a builder stake, with deployment and settlement governed by onchain validator votes.

A one-time, ~15-minute single-price clearing auction opens the market and clears the price that maximized matched volume. Continuous limit/market order trading follows, with prices bounded from 0.001 to 0.999 until expiry. At settlement, the oracle posts 0 or 1 and USDC pays out automatically. The PnL cards look similar to those of a perps trade, just with the market’s question as the title instead of an asset ticker, and your position size expressed as a number between 0 and 1 rather than a leveraged notional. A dispute window can delay finality if the resolution is challenged.

One technical detail worth noting: each outcome market has two tokens (YES and NO), but their orderbooks are merged to share liquidity. An order to buy YES at price P is equivalent to an order to sell NO at price 1-P. Under the merged book, price-time priority generalizes to price-side-time priority. This means that resting sell orders sort before resting buy dual orders at the same merged price level. For advanced users, balances can be manually split and merged between primary and dual sides. Most of this is abstracted away at the API level, but it's the mechanism that lets HIP-4 markets quote tight spreads with relatively thin nominal liquidity.

HIP-4 charges zero fees to open positions. Fees apply only when closing, burning, or settling. (In practice, all outcome market fees are zero across the board while the system is in initial testing.) For traders, it is much more economical to predict the future on Hyperliquid than on Polymarket and Kalshi, which charge meaningfully higher fees on winning positions. The fee design also plugs directly into Hyperliquid’s tiered fee structure (outcome market volume counts toward protocol-wide tier calculations), meaning active prediction-market traders qualify for lower rates on their perp trading through the same unified account.

The first live contract covers a recurring daily BTC price threshold event that resets at 2 a.m. EST. The market is essentially “BTC above X price on Y date at Z time? YES or NO.” Within days of debuting on mainnet, Hyperliquid added multi-outcome markets, starting with a recurring BTC price range contract that settles daily into three buckets (upside, downside, in-between). Planned categories for expansion include politics, sports, macro data releases, crypto, and entertainment.

On May 25, the canonical market set expanded to include offchain events. Hyperliquid validators now publish markets directly using automated newsfeed software run as part of chain operations, with onchain validator voting governing both deployment and settlement. The first offchain markets were the June Fed rate decision (change or no change, settling ~June 17), May CPI YoY (three-bucket multi-outcome around 4.3%, settling ~June 10 to BLS data), and Champions League soccer winner (PSG or Arsenal). Settlement remains closed-loop within Hyperliquid L1: no external oracles, no UMA-style optimistic dispute layer (arguably Polymarket's Achilles' heel), just validator votes against predefined rules and automated feeds.

The natural reaction to HIP-4 is to lump it together with HIP-3, Hyperliquid’s framework for builder-deployed perpetual futures. Both involve builders staking HYPE to deploy markets on HyperCore. Both share the same matching engine and order types. But the two primitives are structurally different in ways that matter for what each can trade.

HIP-3 perpetuals use continuous oracles with roughly 1% price deviation limits per update; a design suited for ongoing leveraged trading on assets with continuous price discovery. That works for stocks, FX, commodities, and crypto, the categories trade.xyz and other HIP-3 deployers have built out. It does not work for discrete events.

A perpetual cannot cleanly express “did the Fed cut?” or “did Trump win in Oregon?” The payoff isn’t continuous, neither is the oracle, and there is no funding rate that makes the contract converge to the right answer.

HIP-4 uses fixed-range settlement, no funding, and no liquidation engine. The contracts are isolated to their entry premium and rely on a single oracle post rather than a streaming price feed. This is why HIP-4 had to exist as a separate proposal rather than as a HIP-3 sub-feature.

Probably the strongest part of HIP-4 is what it lets a trader do across the rest of their Hyperliquid account. The unified margin unlocks a few specific use cases that no standalone prediction market platform can match without a full product rebuild.

First is discrete event hedging for perp books. A trader long ETH perp going into a Fed meeting has historically had two choices: reduce the position or accept the binary exposure. HIP-4 introduces a third option which is to buy a YES or NO contract on the event itself in the same margin account. The outcome position partially offsets the perp’s directional risk without unwinding it. For example, a $10,000 ETH-PERP long going into a Fed decision could be paired with $1,000 in “Fed cuts rates” NO contracts. If the Fed holds rates steady and ETH sells off, the outcome position pays out (1 – P) and offsets some of the perp drawdown. If the Fed cuts and ETH rallies, the trader loses the outcome premium but keeps the upside of the perp. The hedge sits in the same margin account and settles in the same collateral.

Second is market-maker hedging on discontinuous risk. A delta-neutral perp market maker faces tail risk on one-off events (regulatory announcements, protocol upgrades, sudden macro decisions) that streaming perp hedges cannot cleanly express. A market maker quoting size on ETH or BTC perps going into a CPI print can use HIP-4 contracts to hedge the discontinuous portion of the risk without moving capital to a separate platform or running a parallel options book. For professional liquidity providers on Hyperliquid, this is a real improvement in risk management.

For a trader who already runs perps and spot from the same account on Hyperliquid, the cost of entry to prediction markets is essentially zero.

How Does HIP-4 Compare?

HIP-4, Polymarket, and Kalshi are now competing for the same end-state: a venue where users can express views on any event from a single account. But they are starting from different products, different infrastructure, and different regulatory positions. The table below lays out the structural differences. The subsections after expand on the dimensions where the gaps matter most.

UX and Discovery

Polymarket and Kalshi have spent years building consumer-grade front ends, and both work. They just work for different audiences, and they generate market depth in different ways. Kalshi’s depth comes from programmatic, vertical-specific market generation. A single NBA basketball game does not get one market on Kalshi. Instead, it gets markets spanning player propositional bets, point spreads, and combination contracts. Multiply that across every major sports league, daily weather and economic data releases, and macro print cycles, and Kalshi’s active contract count runs into the hundreds of thousands. The result is massive depth inside a narrow set of CFTC-permitted verticals. The app is sports-first (sports including exotics now drive roughly 85% of weekly volume) and the onboarding flow is the cleanest in the sector. U.S. Dollar deposit, no crypto wallet, no bridge, fully regulated (provided users have no problem KYC-ing). Kalshi is live inside the Robinhood and Coinbase apps, which puts prediction markets in front of tens of millions of retail traders who would never download a dedicated app. That combination of programmatic depth, regulated onboarding, and embedded distribution is what flipped Polymarket on April volume.

Polymarket has the opposite scope. Its depth is in breadth. Operating outside U.S. commodities rules, Polymarket can list long-tail markets that Kalshi legally cannot touch: hyper-specific international elections, foreign policy events, court rulings, Twitter drama, crypto protocol upgrades, and much more. The category mix is the broadest in the space: politics at roughly 27% of weekly volume, sports at 46%, crypto at 22%, with culture and macro filling in the rest. The front end is built on browsing with curated categories, “trending” markets, a social layer (comment sections) where users debate active markets, and an onboarding flow that walks users through their first trade. The result is 646k MAUs in April and the deepest non-sports liquidity in the category.

These are different products solving for different users. If you want to day-trade micro-movements in NBA player props or Fed funds futures, Kalshi has the programmatic depth. If you want to take a position on a U.K. election or a celebrity court ruling, Polymarket is currently the only game in town. Both work. Neither is going to lose its respective audience to the other any time soon. It will be interesting to see which one is able to capitalize on the institutional flow. Prediction markets as hedging instruments for institutions is perhaps the most lucrative market for a platform to win, and Kalshi appears best positioned to do so.

HIP-4 has none of this. Outcome markets are tucked inside Hyperliquid’s trading terminal, alongside perps and spot. There is no “browse markets” view, no trending feed, no social layer, no distribution partnership with a retail brokerage, no programmatic market generation across sports or macro. To be fair, HIP-4 just launched. Discovery, to the extent it exists, runs through builder front ends like Outcomexyz and Stratium, early projects whose footprint is still small.

This is a real near-term gap.

The question is whether the builder ecosystem can fill the gap fast enough to matter. Hyperliquid’s framework explicitly invites third parties to deploy markets and build front ends on top of HIP-4, which is how perpetuals scaled on the platform under HIP-3 (trade.xyz alone now accounts for the majority of HIP-3 open interest). If a comparable builder ecosystem materialized for outcome markets on Hyperliquid, the discovery gap could close faster than expected. If it does not, HIP-4 will likely remain a niche product inside an active trader’s account.

Infrastructure

HyperCore is the strongest piece of HIP-4's pitch and the dimension where the gap to Polymarket and Kalshi is widest. Hyperliquid’s execution layer is an onchain CLOB with sub-second finality, ~200,000 orders per second throughput, and a unified margin engine that handles billions of dollars in daily perp volume. HIP-4 plugs into that infrastructure without modification. Outcome markets share the same matching engine, the same order types, the same wallet, and the same collateral as perps and spot.

For now, Polymarket is running a freshly rebuilt central limit order book on infrastructure that wasn’t designed for it.

Polymarket is on Polygon (an Ethereum layer-2 blockchain). The April 28 CLOB v2 launch was a full rebuild of the matching engine and a migration from USDC.e to pUSD as the internal collateral token, and it’s a meaningful upgrade. But Polymarket’s chain dependency remains structural. The platform’s own VP of engineering, Josh Stevens, has publicly acknowledged that hitting the performance levels active trading requires would mean moving to a dedicated chain. Whatever Polymarket does next on the infrastructure side will require multi-quarter engineering work and likely a chain migration of some kind. For now, Polymarket is running a freshly rebuilt CLOB on infrastructure that wasn’t designed for it.

Kalshi runs on centralized exchange infrastructure. It is fast, stable, regulated, and closed. Kalshi cannot natively compose with onchain products, cannot offer cross-margin with crypto positions, and cannot let users self-custody collateral. The tradeoff is frictionless USD onboarding and regulatory clarity. For institutions that need a CFTC-regulated counterparty, Kalshi’s infrastructure is the feature.

The ranking across the three: HyperCore is the most flexible execution layer in the category, Polymarket is mid-transition, and Kalshi is closed by design. But infrastructure does not win volume on its own. It has to be paired with markets, users, and discovery. That is the part HIP-4 still has to prove.

Oracle Scope

At mainnet launch on May 2, HIP-4 was restricted to HyperCore’s price feeds. The only markets it could resolve were variations of “what is the price of BTC at time T?” Both initial markets, the daily binary and the multi-outcome range contract, were BTC price events. The May 25 expansion changed this. Hyperliquid validators now publish canonical offchain markets directly, running automated newsfeed software as part of regular chain operations and onchain validator votes for deployment and settlement. Resolution remains a closed loop within Hyperliquid L1. There is no UMA-style optimistic dispute layer and no Chainlink or Pyth integration.

This is a third path that the category has not seen before, and it is worth being explicit about. Polymarket has UMA’s optimistic oracle, which resolves arbitrary real-world events through a stake-and-challenge mechanism. It is permissionless and general-purpose, which is why Polymarket can list markets on any tweet, foreign election, court ruling, or protocol upgrade on demand. Kalshi has CFTC-certified resolution sources, where every contract goes through a filing process that approves the market and the resolution method. The oracle layer is slower and more rigid than UMA’s, but it is highly defensible in a regulatory sense.

Hyperliquid’s validator-as-oracle model is curated rather than permissionless, which means it cannot match UMA’s long-tail coverage. The validator set has to actively choose, which filters for quality but constrains throughput. But it is materially faster and more flexible than Kalshi’s filing process, and it ships without dependence on a third-party resolution layer. The model is well suited for high-signal events with unambiguous outcomes, but poorly suited for long-tail markets where the resolution criteria are contested or the event itself is obscure.

The watch-item for the next 6-12 months is what Phase 2 (permissionless builder deployment, think HIP-3) will look like. The canonical markets question has been answered (validators handle them directly), but the permissionless layer is still open.

Regulatory Posture

The three platforms occupy three different regulatory positions, and the asymmetry matters more than my table can capture.

Kalshi is a CFTC-regulated designated contract market (DCM) with frictionless U.S. retail access. That status is what unlocked the Robinhood and Coinbase distribution. But Kalshi’s sports contracts are being legally challenged by states claiming jurisdiction over such betting. Massachusetts has issued an injunction. Wisconsin has filed suit. Multiple state attorneys general are pushing back on the argument that CFTC supervision preempts state gambling law for sports event contracts. VC firm (and Kalshi backer) a16z's letter to the CFTC in April backing federal preemption was a notable industry-side intervention, but the issue is not settled. If federal preemption holds, Kalshi will keep its biggest growth category. If it doesn’t, the sports business that drove 85% of Kalshi’s April volume will get carved up state by state.

Polymarket is pursuing a full U.S. re-entry. It already has a domestic foothold: after acquiring a CFTC-licensed exchange (QCEX) last year, Polymarket launched a limited, U.S.-compliant product. But that version is a popcorn stand next to the offshore flagship, with a fraction of the markets, liquidity, and category breadth. The real prize is bringing the flagship itself onshore. The week of HIP-4's launch, Polymarket was reportedly seeking CFTC approval to do exactly that. If the approval lands, Polymarket would gain the full U.S. retail access that has eluded it for years, putting its deepest markets and broadest category coverage in front of American traders for the first time, rather than the stripped-down version it runs onshore today.

HIP-4 inherits Hyperliquid’s broader regulatory exposure, which is the same legal gray area that the rest of crypto-native DeFi occupies. CME Group and ICE, the parent company of the NYSE, have reportedly urged U.S. regulators to scrutinize Hyperliquid. According to news reports, the two exchanges told the CFTC and lawmakers that Hyperliquid’s decentralized environment poses manipulation and sanctions-evasion risks, with particular focus on the 24/7 commodities and crude oil contracts running under HIP-3, categories CME and ICE have long dominated.

Onchain analyst ZachXBT pointed out that the NYSE reportedly took issue with Hyperliquid but not with Polymarket, despite ICE’s $2 billion stake in the latter. ICE’s stake might give it a direct interest in slowing Hyperliquid’s expansion into prediction markets via HIP-4, not just its commodities push under HIP-3. Critics read the lobbying as an attempt to neutralize a fast-growing competitor, particularly as both incumbents face their own CFTC and DOJ scrutiny over well-timed oil trades. Hyperliquid called the manipulation claims "unfounded concerns," arguing that public blockchains expose manipulation rather than enable it. HYPE sold off the day of the Bloomberg report.

The broader regulatory picture is also shifting. The Senate Banking Committee advanced the CLARITY Act on May 14 in a 15-9 bipartisan vote, allowing the bill to be considered on the Senate floor. The legislation would create a federal framework that classifies most digital assets as commodities under CFTC jurisdiction, formally extending the CFTC's authority over derivatives to include crypto spot markets. The bill also includes protection from money transmission laws (see Sec. 604 of the CLARITY Act, the “Blockchain Regulatory Certainty Act” or “BRCA”) for developers of non-custodial software. Though that safe harbor provision was designed for a Tornado Cash and Samourai pattern of developer prosecution (a coder who writes and publishes open source software without ongoing operational control), and Hyperliquid is a different fact pattern (an actively operated trading venue that happens to settle onchain), nonetheless it could be supportive for Hyperliquid if its developers are determined to be “non-controlling” in the context of Sec. 604. Crucially, though, Hyperliquid would need to contend with Title III of the CLARITY Act, specifically the Sec. 301 framework, which deals with securities law and Bank Secrecy Act obligations. That includes exemptions for decentralized finance trading ledgers applications from the Title III, but Sec. 301(a)(2)(A) removes those exemptions if the protocol is determined to be a non-decentralized trading protocol. If Hyperliquid failed to qualify for the exemptions and regulators implementing the Act determined it to be a non-decentralized trading protocol, Hyperliquid would need to accept Bank Secrecy Act obligations under Sec. 301(b)(d)(D).

Some amendments were made to the latest text of CLARITY during the May 14 Senate Banking markup and the latest publicly available version does not include those changes, some of which are reported to involve the BRCA. Furthermore, negotiations over CLARITY Act language are ongoing and issues related to illicit finance activity and the BRCA are among the top unresolved items. So, all of this could change between the time of writing and the ultimate passage of the bill into law.

Beyond the BRCA and BSA-related issues, Hyperliquid also wants the emerging acceptance of perpetual futures by the CFTC to include onchain perpetuals. That hasn’t yet occurred and it isn’t clear that it will, but Hyperliquid founder Jeff Yan spent a week in Washington in mid-May with the Hyperliquid Policy Center meeting with lawmakers.

All of this to say: a lot of regulatory questions remain for whether Hyperliquid can replicate its offshore dominance in the U.S.

The summary across the four dimensions of this section: Polymarket leads on oracle scope and consumer UX, Kalshi leads on regulated U.S. distribution and sports, and HIP-4 leads on infrastructure and unified margin, with a regulatory path that is potentially starting to open. None of the three has a dominant position across all four. What happens when each platform extends into the others’ territory is the story of this category over the next 12-18 months.

Risks

HIP-4 launched 25 days ago into a hyper-competitive category, and the case against it becoming a meaningful product is concrete. The risks below are the ones most likely to determine whether HIP-4 scales beyond a niche feature inside a trader terminal.

Canonical Market Breadth

The remaining constraint is breadth rather than category coverage. Both HIP-4 and Polymarket curate which markets go live, so this is not a curated-vs-open distinction. The difference is in what each resolution layer can support. UMA's optimistic oracle resolves arbitrary real-world events, so Polymarket's curation is a product choice, not a technical ceiling, and it can list a near-unlimited long tail whenever it wants. HIP-4's validator-as-oracle resolves only what the validator set can cleanly adjudicate, so its curation is partly forced by the resolution model. The validator set must actively choose to publish each market, an arrangement that, as noted above, filters for quality but caps throughput. Long-tail markets (specific foreign elections, court rulings, anything socially contested) would require Phase 2 permissionless deployment with broader oracle integration, which has not yet shipped. Until Phase 2 lands, HIP-4's catalog stays constrained by what validators can resolve, which limits how fast the market count can scale even as category coverage expands.

Discovery Gap

Hyperliquid is a trading terminal, not necessarily a consumer product. Outcome markets have no native browsing layer or social component. The bet is that builder front ends fill the gap the way trade.xyz scaled HIP-3. If they don’t, HIP-4 will remain a feature for active Hyperliquid traders and never reach the audience Polymarket and Kalshi have built. Discovery is also the one risk Hyperliquid cannot solve itself. It depends on third parties choosing to build. Luckily for Hyperliquid, it is one of the most attractive places for builders to build in crypto.

Validator Centralization

Hyperliquid runs on 24 (soon 27) validators, and the chain’s history includes the March 2025 JELLY delist incident in which the team froze a market mid-flight. That precedent is well-known to crypto-native users and is the version of “trust” Hyperliquid operates under. For prediction markets where settlement is binary and contested resolutions matter, this is a real source of execution risk. The May 25 expansion makes this risk more concrete in principle: canonical market resolution now relies on an explicit validator vote rather than an external oracle post. In practice, the validator-curation model is doing the work of avoiding this risk upfront. The initial canonical markets (Fed change/no change, CPI bucketed around a fixed number, a binary Champions League final) are all engineered to have unambiguous resolutions tied to a single authoritative source. The harder cases (macro data revisions after settlement, contested sports outcomes, ambiguous Fed statement language) are the kinds of markets the validator set has incentive not to publish. The closed-loop design removes third-party oracle dispute risk but concentrates resolution authority in the validator set, and the early market selection suggests Hyperliquid understands this trade-off.

Regulatory Exposure

Hyperliquid operates in the same DeFi gray zone as the rest of the category. The Clarity Act offers a plausible path to operating legally in the US, but it is not law. Until it is, HIP-4 inherits Hyperliquid's current status as an offshore, unregulated venue, and the CME/ICE lobbying push shows that risk is active, not theoretical. A regulatory action against Hyperliquid, or a final version of the Clarity Act that excludes onchain derivatives or simply fails to address their legality, would meaningfully constrain HIP-4's addressable market. The 12-24 month outlook is brighter than six months ago, but the near-term reality is unchanged.

Convergence Pressure

This risk stopped being hypothetical on May 29, when the CFTC approved Kalshi's Bitcoin perpetual contract (BTCPERP) and Kalshi moved to launch. Kalshi had telegraphed the move since April, but approval is what turned the plan into a live competitive threat. It is the first CFTC-regulated perp in US history, and the framing was explicit: a regulated, onshore alternative to the offshore platforms that have dominated perps, Hyperliquid chief among them. The "trade everything" pressure now clearly cuts both ways. HIP-4 was Hyperliquid moving into prediction markets; Kalshi launching perps moves directly against Hyperliquid's strongest and most defensible product.

Two things temper the threat in the near term. First, Kalshi's approval is BTC-only so far, with the CFTC reviewing additional contracts case-by-case, so this is a single regulated perp rather than the full multi-asset book Hyperliquid runs. Second, the CFTC currently has just one sitting member, chair Michael Selig, a Trump appointee who has championed bringing perps onshore, which makes the regulatory posture supportive but also thin and potentially fragile. Polymarket is reportedly pursuing its own perp product as well.

But the direction of travel is now unambiguous. A CFTC-regulated competitor offering perps onshore attacks exactly the institutional hedging flow that is the most lucrative prize in this category, and it does so with the one thing Hyperliquid cannot yet offer: US regulatory status. If Kalshi expands beyond BTC before HIP-4 ships permissionless markets and broad category coverage, the convergence race tilts toward the incumbent that ran the full stack first under US regulation. The next 12 months will be shaped less by who builds outcome markets best and more by who can credibly run the full stack, and Kalshi has now fired the first shot.

So What?

The prediction markets category is in the middle of a structural shift that most coverage is missing. The story is not Kalshi vs. Polymarket, and it is not HIP-4 vs either incumbent. The story is convergence. Every major venue is moving toward the same end-state: a single account, a single margin engine, every market in one place. Hyperliquid built outcome markets on a perp platform. Kalshi and Polymarket are building perps on prediction market platforms. The race to trade everything is now the category’s defining competitive dynamic, and the prize is the most lucrative customer base in crypto-adjacent finance: institutions that need a single venue to hedge and speculate across asset classes, events, and time horizons.

Each platform is starting from a different moat. Polymarket has consumer distribution and the most flexible oracle layer. Kalshi has U.S. regulatory access and the cleanest sports product. Hyperliquid has, in our view, the best execution infrastructure in crypto and a unified margin engine that competitors cannot match without rebuilding their stack from scratch.

The question is which moat is hardest to copy. Polymarket and Kalshi are both, in some sense, trying to build what Hyperliquid already has. CLOB v2 was Polymarket trying to build HyperCore on Polygon. Polymarket reportedly pursuing its own chain is Polymarket trying to build the HyperCore-equivalent execution layer.

Kalshi is a different case. It already runs a high-performance regulated exchange, and its May 29 perp approval proves it can ship the product. What Kalshi cannot easily replicate is not the perp itself but the composability around it: cross-margin between perps, spot, and outcome contracts in a single self-custodied account. Kalshi's closed, centralized design is the source of its regulatory clarity, but it is also what prevents that kind of unified, composable account. Kalshi can match Hyperliquid product by product; what it cannot easily match is the integrated stack.

Hyperliquid, by contrast, needs to expand category coverage and attract a consumer discovery layer. HIP-3 has shown the platform can attract builders to its primitives. Building markets and discovery on top of an execution layer is significantly easier than rebuilding the execution layer itself, and the unified, composable account is something Hyperliquid has already built. The asymmetry favors Hyperliquid, though it is narrower against Kalshi than against Polymarket.

The story is not Kalshi vs. Polymarket, and it is not HIP-4 vs. either incumbent. The story is convergence.

We are bullish on HIP-4 long-term for two reasons:

1. The regulatory frame for HIP-4 has shifted faster than most coverage suggests. The Clarity Act cleared the Senate Banking Committee on May 14. Jeff Yan and the Hyperliquid Policy Center are in DC actively engaging the process. Six months ago, the offshore-DeFi exposure looked permanent. Today, there is a credible path to U.S. regulatory clarity, and Hyperliquid is positioning itself to take it.

2. HIP-4 has moved faster than expected. By day 25, the protocol had taken a 20% share of combined BTC prediction-market 24h volume, and the validator-published canonical market set already covers the Fed, CPI, and sports. The validator-as-oracle model is a credible third path between UMA's general-purpose resolution flexibility and Kalshi's CFTC-certified rigidity. Phase 2 permissionless deployment is still the long-term unlock for arbitrary builder markets, but the near-term category gap has closed faster than the original launch suggested.

Hyperliquid spent two years becoming the standard for onchain perpetuals. It did that by building infrastructure first, attracting builders second, and letting the consumer layer assemble itself on top. HIP-4 is the same playbook applied to a new product surface. The platform that wins this race is the one whose execution layer is hardest to replicate, whose builder ecosystem ships fastest, and whose regulatory path opens widest.

On all three, the case for Hyperliquid is stronger than the current market positioning suggests. As they say in the community: Thank you Jeff.