IOSG said Hyperliquid’s HIP-4 prediction market looks less like Polymarket than a binary option sitting next to a perpetual futures position. In the report, Yes shares and leveraged positions live in the same account, use the same margin, and trade on the same matching engine, which means a user can place a bet and hedge it from one interface.
The report pointed to three other product differences. Anyone staking 500,000 HYPE can open a market. Settlement conditions are fixed in a template before the first trade. Entry is free, while exit costs roughly 0.1%, the reverse of the fee pattern seen at Kalshi and Polymarket.
HIP-4 launched on May 2, 2026. For its first four months, every market was operated directly by the protocol. There were 691 in total. External venues were not allowed in until Aug. 29. Since then, three venues have posted the required stake, and there has still not been a fourth.
Whether users can even see the product depends on where they are. Prediction markets have a dedicated page at app.hyperliquid.xyz/outcomes, but the frontend decides by region whether to place the category alongside perpetuals and spot in the trading interface. IOSG said a connection from South Korea showed the category in the market list. On the network used in its own test, the list and menu did not show the entry at all, though the page still opened and trading still worked. The report said Hyperliquid restricts frontend access in several countries, so the path visible to one user may not exist for another.
Volume is still tiny next to Kalshi and Polymarket
From May through Oct. 5, HIP-4 generated $317 million in cumulative volume. IOSG said Kalshi can do that in under four hours. Using September figures, the report estimated that for every $1,000 traded on Kalshi and Polymarket, HIP-4 handled about $0.70.
That is the report’s central divide. The design stands out. The scale does not. More than one-quarter of all historical HIP-4 volume came from a now-finished soccer World Cup market. The daily bitcoin market that carried much of the first four months fell 93% from its May peak. New venues brought September back to $51 million, in line with July, but the venue contributing the most was effectively paying traders about 0.6 cents for every dollar traded. October started more slowly, at roughly $1.25 million in average daily volume. Across the entire market, fees came to about $1,300 a day, less than half the interest one venue could earn by simply leaving its staked capital untouched.
How HIP-4 differs on fees and settlement
IOSG reconstructed fee rates from executed trades under the base fee tier and said HIP-4 charges exactly double the spot fee schedule: 0.070% for takers and 0.040% for makers on spot, with one side doubled and the other reduced to zero in HIP-4 so the round-trip cost stays the same but is charged only on exit.
The report added that standard discounts stack multiplicatively. A 4% referral discount reduces 0.140% to 0.1344%, and adding a 10% staking discount brings it to 0.12096%. IOSG said every trade it checked matched that formula exactly. Maker orders do not receive rebates, which the report described as the one clear area where market makers are worse off than in other parts of the exchange.
No venue has yet set its fee coefficient above the default, according to the report, so venue revenue is carved out of the protocol fee rather than added on top. The one place users may face a charge on entry is through frontend apps. Under spot rules, builder fees are capped at 1%. IOSG said one app charged 0.81% on entry, roughly six times the protocol’s exit fee.
Settlement is where venue risk becomes more visible. Polymarket relies on votes by UMA token holders, and the report cited an earlier dispute this year involving a $375 million Microstrategy market. Kalshi, as a regulated exchange, resolves outcomes itself under regulatory oversight. HIP-4 splits the process: crypto price markets settle automatically using the exchange mark price at a specified minute, while every other market is signed off by the market creator. There is no dispute window and no substitute signer.
At the time of the report, 223 of 228 live markets fell into that second category. IOSG argued the constraints here are financial and temporal rather than programmatic. Stakes remain locked as long as a venue has unsettled markets. Withdrawals are blocked for six months. Wording must come from validator-approved templates. Those templates are intentionally detailed. The report said the template for a company listing spends an entire page defining what counts as a listing. That, IOSG wrote, is better than arguing after trading starts, but handing the final signature to a named party without an appeal path is still weaker than a stable procedural system, assuming the procedure itself works.
World Cup volume faded, and the daily bitcoin market lost traction
June was the high point for HIP-4, largely because of the World Cup. Soccer markets alone traded $89.5 million, or 28% of cumulative HIP-4 volume. Once the tournament ended, volume fell by more than 80% over the next two months. September returned to $51 million, and that entire recovery came from external venues. Volume in the first five days of October was $5 million.
Running underneath those event markets was the same protocol-run product every day starting in May: whether bitcoin would be above a given price at 6 a.m. the next morning. IOSG described it as the most Hyperliquid-native market on the platform, and also one that kept sliding for four months before bottoming. It traded $16.6 million in the week of May 11, then only $1.2 million in the week of Sept. 28, with the two prior weeks at roughly the same level. Repeating the same question each day eventually exhausted attention.
Five weeks of open deployment brought volume, but mostly with subsidies
Outcome was the first external venue to go live on Aug. 29. Skew registered markets the next day and has seen almost no trading. Events, under Trade.xyz, registered on Sept. 5 and opened to the public on Sept. 10.
Over the latest day cited in the report, Outcome accounted for 68% of total HIP-4 volume, Trade.xyz for 21%, protocol-run markets for 10%, and Skew for the remainder. Looking only at external venues since Aug. 29, Outcome held 92%, while Trade.xyz rose from zero to nearly one-quarter of the latest day’s external flow. IOSG said Trade.xyz was more opinionated in market selection, listing binary contracts tied to prices of its own stock and commodities perpetuals, NFL markets, and the exchange’s only questions on whether Anthropic and OpenAI would go public before a deadline.
Still, the report said those volumes should be discounted. Outcome is running a $1 million incentive program that pays makers and traders by market and by day, but only for orders carrying its own app code. As of Oct. 5, Outcome had paid $273,409 to 2,487 wallets, an average of about $7,400 per day. IOSG said that was around 0.6% of Outcome’s traded volume and roughly four times a full round-trip trading cost. One-quarter of the budget had already been spent, and the daily payout pace had not slowed. External venues traded between $1 million and $2 million a day through September, with weekends the busiest, then eased to around $1 million a day in October while rewards kept flowing.
Open listing also produced the first duplicated markets. Outcome and Trade.xyz both listed a market on the Federal Reserve’s October rate decision. In the past seven days, Outcome’s version traded $354,000, while Trade.xyz’s version traded just $90. Both venues also listed a set of markets on Anthropic’s first-day public valuation, and both saw little activity. IOSG said this is the direct consequence of permissionless listing without shared books: two venues can list the same question, but their order books do not merge, so liquidity follows whichever venue is paying for it.
The mix of trading is tilting toward sports. Over the past 30 days, soccer, American football, and esports accounted for 45% of volume. Crypto price questions made up 42%. Stocks and commodities were 7%, and Fed markets were 6%. Open interest leaned even more heavily toward sports, with two-thirds of at-risk capital tied to sports markets, much of it in season-long books such as the English Premier League, UEFA Champions League, and NFL champions. Crypto price questions were only 17%.
Template supply is growing faster than venue count. Validators had approved 28 templates by the time of the report, up from 18 in mid-September. New templates covered spreads and totals, the Ballon d’Or, head-to-head outcomes between two AI models on a ranking board, and first-day market capitalization for a public listing. A coming network upgrade will double the cap per venue, allowing 200 concurrent markets and as many as 1,000 new markets in a day. On testnet, 95 deployers had already registered. On mainnet, there were still only three.
Order flow appears to run mainly through Outcome’s frontend
IOSG also examined how orders reached the system. A frontend app that submits orders on a user’s behalf needs a separate signing key, and that key is named after the app. That means a wallet registered through outcome.xyz keeps a public tag such as Outcome-9d3c74de. Hyperliquid’s own mobile login works similarly and shows up as Mobile QR. Desktop users sign every action with their own key and leave no tag.
The report only analyzed Outcome because it accounts for roughly 90% of external venue volume. It found that about 70% of volume came from wallets previously registered on outcome.xyz, roughly one-fifth came from wallets with no tag at all, and around one-tenth came from scripts and third-party tools with self-named keys. The larger the wallet, the higher Outcome’s share. Among the 20 biggest wallets, Outcome-tagged wallets represented about three-quarters of traded value. In the rest of the sample, they represented about half. IOSG said that is what the reward design should produce: only orders routed through Outcome’s code qualify for incentives, and the wallets collecting those incentives are also the largest market-making wallets.
The method has limits. A tag only shows that a wallet once registered through an app, not that any given trade was routed through it. The untagged fifth of volume also should not be read as exchange-native retail flow, because most of those wallets were quoting both sides across many books using their own signing keys. The sample covered 121 wallets drawn from recent trades in the 60 most active books and was weighted by 48-hour traded value, which naturally biases the result toward market makers. Fees do not resolve the issue either. Less than 1% of sampled volume paid builder fees, totaling just $9 over two days, and Outcome’s own code does not charge a builder fee, making free app routing indistinguishable from no app routing at all.
The gap with Kalshi and Polymarket is about 1,400x and still widening
IOSG put the volume gap at roughly 1,400x, and said the dollar gap is still increasing even though HIP-4 is growing. Kalshi traded $59.3 billion in September, up $20.6 billion from August. Polymarket, including its U.S. business, traded $13 billion, with most activity now in the U.S. app. HIP-4 traded $51 million.
In IOSG’s view, incumbent platforms are not winning because they designed a cleaner market structure. They are winning because they have the two things that actually move volume. First, they have thousands of events people care about and can list new ones within hours of a major news break. Second, they have distribution: apps, brand, sports audiences, and in Kalshi’s case a compliant shell that U.S. brokerages can connect to directly. HIP-4 currently has 28 templates, three venues, and a user base that was already trading perpetuals.
At the same time, incumbents are moving toward Hyperliquid’s product shape. Polymarket launched perpetual futures for users outside the United States in early September, with 67 markets across crypto, equities, indexes, and commodities and leverage up to 20x. Keeping a bet and a leveraged trading position under one roof used to be a clear HIP-4 distinction. Now it is turning into a features race, and Polymarket entered that race with event inventory already in hand.
The economics do not yet work as a standalone business
At the pace of the past 30 days, HIP-4 traders would pay about $480,000 in fees over a year. Venues can take at most half of that, and there are three of them. By contrast, the mandatory stake required for one venue would earn $1 million a year if left untouched, according to the report. The largest venue is also distributing incentives at a $2.7 million annualized pace to support its current flow.
On those numbers, IOSG said no venue is running a real business today. They are buying an option. That option has value. Hyperliquid has said deployers may eventually be able to set higher fees. Volume may one day grow enough to justify those costs. Private-market comps in the sector were cited at $22 billion and $21 billion, while Kalshi was said to be discussing a $40 billion valuation. Even so, it is still an option, and holding it costs a $47 million stake. If volume does not arrive, the option expires worthless.
Conclusion: the design is ahead of the market
IOSG’s closing argument was simple: the building blocks are there, but the market is not. By the metrics product engineers tend to care about, Hyperliquid has made the stronger system. One account. One margin pool. Bets and hedges on one screen. Settlement logic fixed before the first trade. A shared order book for Yes and No so liquidity is not split in half. Over the past five months, the report said, there has been no evidence that design is the limiting factor.
The strongest case for HIP-4 is that users never need to leave Hyperliquid. Margin, hedging, and prediction positions stay in one account. For a desk already trading perpetuals there, adding a prediction position is just another click. No second platform. No new balance. No transfers back and forth. Whatever happens in its competition with Kalshi, that convenience has value, and IOSG said it likely explains why HIP-4 can persist at some scale even if it never becomes a major standalone business.
But the report was just as clear on what is missing. The product is no longer as unique as it looked in August, because Polymarket now offers leverage next to betting as well, and it started from the side that already had users. Depth is still thin. Across the 25 most active books, the median bid-ask spread was 0.29 cents, and visible size within 1 cent of the best ask was about $4,700. A $1,000 market buy moved only a few basis points beyond the best ask. A $10,000 buy cost 2% to 3% more, and in four or five of those 25 books could not be fully absorbed by visible liquidity at all. Retail-sized flow can trade. Larger capital still struggles.
There is also not much to trade. Prediction market volume is driven by topic supply, and topics come either from editorial teams that list hundreds of events each week or from sports calendars. HIP-4 has 28 approved templates. Three venues each locked $47 million in stake to use them, and two of those venues are already listing some of the same questions. Kalshi, the report noted, can open a new market off a news event in a single afternoon.
The user base is also narrow in a very specific way. IOSG said each address trades eight times a day on average. That looks like a trading desk, not a mass-market betting audience. If the goal is to build a capped-upside tool for crypto traders, that is fine, and arguably exactly what this design does best. If the goal is to reach the users Kalshi and Polymarket are chasing, it is not enough. The fact that two-thirds of at-risk capital sits in season-long soccer and NFL books suggests venues are still trying to push in that broader direction.
The most likely outcome, according to IOSG, is the less dramatic one. HIP-4 remains a feature inside Hyperliquid, earns a modest amount, and matters because perpetuals traders do not need to open accounts somewhere else. That may be worth building. The report said it is not enough, at least for now, to justify valuing HIP-4 as a standalone prediction market business.