Why Trade[XYZ] Came to Dominate Hyperliquid HIP-3 Trading Volume

WuBlockchain republished a long-form article arguing that Trade[XYZ] has become the clear leader within Hyperliquid’s HIP-3 ecosystem. Using data through June 2026, author Mohit Pandit says Trade[XYZ] should not be viewed as an existential threat to Hyperliquid. His case is the opposite: Trade[XYZ] adds value to both Hyperliquid and HYPE.

The article’s central claim is that Trade[XYZ] spent eight months building what it calls the hardest thing in this category: a genuinely liquid perpetual market for equities, indices, commodities, and foreign exchange. In the author’s framing, that outcome shows HIP-3 can support professionally operated, institution-grade non-crypto perpetual verticals. Hyperliquid, meanwhile, still keeps the users, matching-engine activity, fee sharing, auction demand, and broader ecosystem narrative, without directly taking on listing or regulatory responsibility.

HIP-3 as an access layer, not just open deployment for its own sake

The article splits derivatives exchange strategy into two routes.

• A vertical route, where a venue builds every market itself, sources the assets, runs oracles, recruits market makers, takes the risk, and keeps the economics. The piece cites Lighter and Ostium, described there as a pure RWA venue, as examples of vertical integration.
• A horizontal route, where the base layer is provided by the platform and permissionless deployers build markets on top, sharing fees with the exchange. In that framework, HIP-3 is Hyperliquid’s model, and @tradexyz is one deployer operating on it.

But the author says it is a mistake to understand HIP-3 as horizontal just for the sake of being horizontal. His preferred description is that HIP-3 is an application for access. Hyperliquid, in this view, is betting that durable onchain financial advantage sits at the level of core infrastructure: the L1, the clearing layer, and the matching engine. The job of the protocol is to keep improving that stack in performance and neutrality so the strongest operators choose to build there.

The article compares that logic to traditional market structure. There is one CME, one NYSE, one HKEX. Liquidity attracts liquidity. A category that never produces a single, deep liquidity winner has effectively already lost. On that reading, Hyperliquid’s ambition is not to handpick category winners but to open the rails and let the best operators compete to build the deepest markets on a neutral base. If one deployer ends up heavily concentrated, that does not mean the model failed. The author says it may simply mean the model is working the way finance usually works.

The main criticisms the article addresses

The piece lays out two common objections.

The first is that Hyperliquid is giving away future value. A deployer keeps roughly half the fees and holds a franchise over the market category, while Hyperliquid could have captured more if it had built equity perpetuals itself.

The second is sharper. One deployer accounts for around 98% of HIP-3 volume, which has triggered accusations of favoritism. The article says those claims often point toward links between Trade[XYZ] and the Unit ecosystem, while Hyperliquid still takes 50% of exchange fees.

The author’s answer is that these criticisms understate how difficult it is to build an institution-grade real-world-asset market. The article sets out to test, with data and first-principles reasoning, whether the current model has actually succeeded at all.

What it takes to build an equity perpetual venue

One of the article’s recurring points is that “just list the asset” misses the real problem. Listing is the easy part. The moat is making a newly listed market tradable in meaningful size.

Using Trade[XYZ] as the case study, the piece says there are three main challenges:

1. Listing fast enough to capture demand.
2. Securing market makers that can create real depth.
3. Keeping liquidity economically real while operating these markets day to day.

Launch speed

Measured from onchain registration of an asset to its first trade, the median launch time for Trade[XYZ] markets was 3.3 days, according to the article. It says 65% of markets launched within one week and 47% launched within three days.

Tradable markets are the real moat

The article argues that Trade[XYZ] has not just deep books but rationally distributed liquidity. Flagship index and commodity markets show the strongest numbers. It lists $2.6 million of resting depth within 10 basis points of mid on XYZ100, $964,000 on the S&P 500 market, and $759,000 on gold. Single-stock markets such as NVIDIA and Tesla are also described as deep enough for comfortable trading size.

That compares with a median market carrying only about $20,000 within 10 basis points. The article presents this as a sign of how rational market makers allocate capital rather than a flaw in the design.

Across 73 markets with enough data, the piece reports a -0.72 correlation between daily distinct market-maker wallets and spread, a -0.82 correlation between volume and spread, and a +0.96 correlation between volume and open interest. It also gives a book-wide volume-weighted average spread of 2.33 basis points and says daily turnover runs at roughly 2.9x open interest.

The interpretation is straightforward: Trade[XYZ]’s edge lies in market-maker business development and capital coordination. That work, the author says, is what produced tighter and deeper markets.

Why hedging and risk controls matter so much for stock perpetuals

The article then turns to first principles. Market makers earn the spread, but they survive only if they can manage the inventory left on their books after each fill. For stock perpetuals, the key variable is hedging.

Crypto perpetuals can be hedged around the clock on another crypto exchange. Stock perpetuals are different. The natural hedge is the underlying stock, an ETF, or a listed future, and those instruments trade only when their cash or futures markets are open.

During regular hours, a desk can hedge TSLA perpetual inventory with TSLA stock and quote tight, deep markets while collecting spread. After the underlying market closes, that same desk is sitting on unhedged inventory. The rational response is to widen spreads, cut depth, or stop quoting. In pre-IPO names, the article says there is effectively no true hedge at all before listing, which is why those books stay thin.

The report also lists several related problems: adverse selection, because a larger share of after-hours flow is informed; funding and carry constraints, because funding rates need to keep the perp anchored without making hedging uneconomic; and oracle or gap risk, because stale, manipulable, or jumpy marks create liquidation risk that makes size provision hard.

Discovery Bounds and liquidation protection

To address that, the article describes a Discovery Bounds mechanism that keeps mark prices within plus or minus one times the maximum leverage distance from a reference price. At 20x leverage, that corresponds to about 5%. The mark is re-anchored in discrete, market-capped steps and acts as a hard ceiling until outside pricing returns. The article says liquidation protection is layered on top, preventing positions from being liquidated when their liquidation price lies outside the active bounds.

In plain terms, the author argues, a market maker now faces a known upper bound for how far price can move in one step, and the exchange will not liquidate the desk inside that range. Unhedgeable overnight inventory therefore becomes bounded and measurable rather than open-ended.

Funding-rate multipliers

The article also says each market can apply a funding-rate multiplier. Standard funding is scaled to 0.5, which it says is roughly a 5.5% annualized baseline, while pre-IPO names are scaled down to 0.005. The stated purpose is to keep perpetuals tied to fair value without making the position itself uneconomic for market makers. In pre-IPO products where there is no stock to arbitrage against, that setting brings carrying costs close to negligible.

Taken together, the author describes these features as a toolbox for markets that, on first-principles reasoning, should become unquotable once hedging disappears.

Overnight depth held up, even if the article notes clear limits

Using the top 10 stock order books by depth, the article says overnight depth remained at roughly 116% of regular-session levels. It adds that single names such as NVIDIA and Tesla actually deepened after cash markets closed because the perpetual became the only active venue for price discovery and quoting concentrated there.

On weekends, when even index futures are shut and hedging disappears for two full days, depth drops to about 37%.

The author is careful not to oversell that result. He says it does not mean Trade[XYZ] books are somehow magically better after hours. The durable differentiator is still daytime depth, order flow, and breadth across genuinely hard-to-make markets. What the data does show, in his reading, is that Trade[XYZ] built risk systems strong enough to keep market makers quoting overnight where first-principles logic would predict collapse.

This is not a one-time listing business

Another point in the article is that Trade[XYZ] did not list markets and walk away. Over the most recent window of roughly 300 onchain operational actions, it executed 294 distinct risk-management actions. Those included 54 position-limit changes, 35 growth-mode switches, 34 funding-rate multiplier changes, 28 trading pauses, 11 margin-mode changes, plus asset-level labeling actions.

These actions spanned 92 underlyings. The article presents that as evidence of ongoing, market-by-market risk management tied to real trading sessions, halts, and funding conditions. In other words, this is described as a full-time market-operations business.

Comparison with other onchain efforts

The article uses several comparables to frame the scale of the task.

On Solana, tokenized spot equities under xStocks have generated more than $25 billion in total volume, but actual DEX volume is only about $517 million, according to the piece. Ostium, described there as a dedicated and funded RWA perpetual DEX, has roughly $59 billion in cumulative volume but only about $115 million in open interest, which the author says is one twenty-fourth of Trade[XYZ].

Newer entrants also appear in the analysis. Variational, the article says, does not even try to build native depth, instead aggregating liquidity through RFQ from Hyperliquid, Lighter, and centralized exchanges, with routing ultimately flowing back toward Hyperliquid for part of the liquidity under discussion.

The conclusion is that the onchain stock-perpetual category leader, by a wide margin, is Trade[XYZ] on Hyperliquid.

User acquisition appears to accrue mainly to Hyperliquid

A common assumption is that the deployer owns the user because it owns the frontend. The article says the data points the other way.

By tagging each trade with the frontend or builder code that generated it and measuring on the taker side, the report says about 97% of Trade[XYZ] market volume was traded through Hyperliquid’s own app and API. All third-party frontends combined account for around 3%, and Trade[XYZ]’s own frontend is only a small slice of that.

That means nearly every trade associated with Trade[XYZ] products, in the article’s telling, is still happening inside the Hyperliquid interface.

The user-acquisition effect is described as large and ongoing. Trade[XYZ] has brought more than 300,000 distinct wallets to Hyperliquid cumulatively, the article says, and is still adding 36,000 to 48,000 each month. In March, during the burst of listings and the SpaceX surge, the monthly figure nearly reached 79,000.

The author’s framing is that stocks and RWA perpetuals function as a top-of-funnel acquisition channel. The assets are the hook, but Hyperliquid is where the user lands, trades, and stays.

How the fee split works

The article says HIP-3 traders have paid about $37.9 million in total fees. That pool is divided into three parts.

• About $9.2 million went to builder-code fees paid to third-party frontends, not the deployer.
• The remaining exchange fees are split 50/50 between Hyperliquid and the deployer.
• That leaves around $14.3 million for the protocol side, which the article says flows toward HYPE buybacks, and about $14.3 million accrued to the deployer.

The report adds that HIP-3 caps the deployer side of the split. Hyperliquid’s protocol fee matches any deployer share above 100%, so the deployer can never collect more than half.

The broader argument is that cheap, deep markets attract the kind of scale that generates fee revenue in the first place.

The growth-mode debate

Each HIP-3 deployer can select a fee mode per market, according to the article.

• Standard mode charges takers 9 basis points and makers 3 basis points.
• Growth mode charges 0.9 basis points and 0.3 basis points, roughly a 90% reduction.

Growth mode is limited to non-crypto real-world assets. The article says that excludes crypto wrappers such as MSTR and also excludes GOLD because it overlaps with the existing PAXG-USDC market.

The author treats that exclusion as a natural experiment. Books that qualify for growth mode currently carry fees near 0.86 basis points, while excluded markets are closer to 7 basis points. On the same matching engine, the gap is around 8x.

His argument is that RWA perpetuals compete with the all-in cost of traditional finance. A 9-basis-point fee cannot compete with CME index futures or stock commissions, while a 0.9-basis-point fee can, especially when paired with leverage and 24/7 access.

Why the article says low fees are not the root cause of volume

The author gives three reasons.

First, the onchain control group. Of seven other HIP-3 deployers, six have the same fee toolset but almost no volume. The second-ranked deployer, dreamcash, even quotes tighter spreads, yet remains about 30 times smaller. If low fees alone produced activity, the article says, dreamcash should be much closer.

Second, the GOLD example. GOLD pays around eight times the fee of growth-mode books, yet it is still the single largest fee market and ranks in the top three by both volume and open interest. The interpretation is that traders will pay full freight when the liquidity is there.

Third, the article argues that turning off growth mode would not kill volume. It would shift more value to HYPE. Because exchange fees are split 50/50 in both fee modes, moving from roughly 0.9 basis points under growth mode to roughly 9 to 12 basis points under standard mode would raise HYPE-linked value by about 9x to 15x. The article says the protocol’s buyback share would still rise even if volume fell sharply, unless volume collapsed by more than about 85%.

At the roughly 7-basis-point level observed in GOLD, Trade[XYZ] would need only about 11% of today’s volume to match current buyback levels, the article estimates. At 5 basis points, it would need about 15%. At 3 basis points, about 25%.

The monetization path proposed in the article is to move mature markets toward 5 to 7 basis points while keeping one-half to three-quarters of the volume because of the moat. Under that scenario, annual buyback flow could reach roughly $90 million to $185 million, or 3x to 5x current levels. The author says this is not purely theoretical because GOLD already operates at standard fees and converts 4.3% of volume into 23% of all buybacks.

That leads to the two-stage strategy described in the article: use low fees now to build the moat through users, volume, open interest, and price-discovery status, then monetize later. In the author’s framing, both stages direct value toward HYPE.

Market-level growth patterns

The top 30 markets hold about 95% of total open interest, led by the S&P 500, the XYZ100 index, Brent crude, and WTI, according to the article.

The author is more interested in how quickly markets reached those levels. Measured by the number of days from listing to 25%, 50%, and 75% of current open interest, the median market reached one-quarter of its eventual size within 9 days, one-half within 15 days, and three-quarters within 30 days.

The dispersion is wide. SpaceX reached half of its current open interest in 14 days, while the S&P 500 and silver took about 15 days. Early single-stock markets launched when the venue’s liquidity infrastructure was still immature took far longer: Microsoft needed 192 days and Meta 159 days.

The article reads this gap as a visible learning curve. Recent markets scale much faster because the market-maker relationships and tooling are now in place from day one.

Market-maker concentration and the core quoting wallets

How concentration changed over time

The article uses a heat map to track the share of passive volume supplied by the top five market makers each week in each market. Early markets are dark blue, meaning a handful of firms provided nearly all passive liquidity during the first months, often with top-five concentration above 90%.

Over time, the largest and most liquid markets become lighter as more market makers compete at the top of book. Many single-stock books remain concentrated. The article says concentration itself is not necessarily unhealthy because that is how markets are bootstrapped, but the flagship markets becoming competitive is a sign that liquidity provision on Trade[XYZ] has turned into a genuine business rather than a favor from one or two firms.

The “workhorse” wallets

The report also looks at the top makers over the prior 30 days and checks which wallets repeatedly appear at the top across markets. It finds a clear core group. The single largest workhorse wallet ranks in the top five makers in 47 of 73 markets and is number one in 22 of them. The top three workhorse wallets combined rank in the top three makers in 57 of 73 markets.

Some of those wallets quote across all four asset groups: stocks, commodities, foreign exchange, and indices. The article says they display textbook market-making traits, with directional exposure inside 1% and realized PnL effectively rounding to zero.

Where the fee base comes from

Commodities and indices account for most of the fees. Commodities alone make up 54% of all earned fees, indices add 24%, and the long tail of single stocks and FX contributes the remaining 22%, even though stocks make up most of the listed markets.

At the single-market level, GOLD is the largest fee contributor, accounting for 23% of fees, or $8.7 million. It is followed by the XYZ100 index at 18%, WTI crude at 13%, and silver at 10%. The top 10 markets together generate 84% of all fees.

The article notes an important nuance forced into view by GOLD. Fee rankings are not the same as volume rankings because fee modes differ across markets. GOLD is the only large market excluded from growth mode, so it pays around 7 basis points while much of the rest of the book pays around 1 basis point. As a result, GOLD is only 4.3% of volume but 23% of all fees. It is a secondary market by activity, the article says, but a major market by buyback fuel.

Could Hyperliquid’s core team have done this itself?

The author’s answer is no, and more importantly, it should not.

The strongest reason given is regulatory. Listing perpetuals on NVIDIA, TSLA, and pre-IPO SpaceX sits squarely in securities-derivatives territory, the article says. HIP-3 is intentionally designed to externalize that responsibility to deployers. If the core team directly listed these markets, the protocol, the foundation, and HYPE would move more directly into regulators’ line of sight.

In that sense, keeping listing activity at arm’s length is not a missed opportunity. The article calls it design.

The rest of the argument follows from that. Hyperliquid’s value is in being credible neutral infrastructure. If the core team handpicked assets, it would undermine the permissionless case and the deployer-auction fee market that HIP-3 is supposed to monetize. Running 92 stock, FX, and commodity markets, sourcing oracles, handling market hours and halts, cultivating market makers, and carrying out hundreds of visible onchain risk actions is an operating business of its own, separate from building a high-performance exchange.

The empirical record, the author says, supports that point. If the task were easy or suitable for in-house execution, one would expect the core team to have already done it, or for many strong deployers to exist. Instead, the second-largest deployer is 46 times smaller, dedicated RWA venues are 24 to 33 times shallower, and newer entrants route liquidity back toward Hyperliquid. Scarcity, in the article’s view, is evidence of difficulty.

The piece closes with an analogy: what Tether did for global access to dollars, Trade[XYZ] is doing for global access to global stocks. It also states that all data in the article was provided by @hydromancerxyz.