RWA Perps: Where They're Growing & What You're Really Trading

Transcript

[applause] >> Hey everyone. My name is Meredith and I'm the CEO of Stork. Um I'm excited to talk to you guys about RWA perps. Um we build Oracle infrastructure for DeFi's fastest markets. On any given day, Stork price feeds power more than half the volume in on-chain perpetuals trading. Thanks for having me at ETHDenver. Stork offers three products: price feeds, perps, prediction markets data, and research. All of them matter to what I'm going to talk about today.

Here's what we're going to cover. First, a quick look at how the RWA perps market has grown and where what's been driving it. Then we'll get into the mechanics, what you're actually trading when you trade a commodity perp, an index, or equity perp on the weekends, or pre-IPO perps. And finally, why pricing is the variable that determines whether this market will reach its potential. To start, we'll dive into RWA perps growth story, how we got here, oil, gold, and indexes. Let's start with the numbers. Monthly volume in RWA perps has grown by an order of magnitude over the past 4 months. Perpetual futures on traditional

assets had existed before this, but with little meaningful volume or attention. Then war broke out in the Middle East on a Saturday at the end of February. Major events would continue to hit on weekends throughout the month of March, closing the Strait of Hormuz, on-again, off-again peace talks. Crypto-native traders were ready. RWA perps on WTI oil had already gone from 76 million in January volume to 488 in February. In March, they hit 21.5 billion monthly volume. The following month, they reached 60.2 billion. Gold, silver, and oil perps would go on to dominate RWA perps trading through the early growth

period in 2026. Around the same time, Trade XYZ, an exchange operator on the Hyperliquid platform, was building out its index product. XYZ listed its own Nasdaq 100 proxy, the XYZ 100. Then, it secured a licensing deal with S&P Dow Jones Indices to list an officially sanctioned S&P 500 perp. That mattered. Most index perps are unlicensed proxies, and here was a brand name index with a clear chain of custody.

The third wave has been equities. The AI trade, which has been running hot in traditional markets, moved into perps. Equity perps volume more than doubled in May alone, from 25.6 billion to 62.8 billion. The leaders are AI-adjacent names, Micron, SanDisk, Nvidia, but there's a growing long tail. The exchanges we track nearly doubled doubled their number of equity listings from 196 in April to to 391 at the end of May.

There are now 569 different RWA perp markets, and that's just on the dozen or so leading exchanges we track. Every crypto exchange is now doing this, and it happened pretty fast. And then last week, the CFTC stepped in. On May 29th, the commission approved Bitcoin perpetuals contracts for Kalshi and authorized Coinbase to serve foreign derivatives markets. They also issued an advisory specifically on 24/7 trading. Commodities and indexes.

Here's where it gets interesting. The volume is real, price discovery is happening even when traditional markets are dark. But what exactly are you trading? For commodities and indexes, the short answer is you're usually trading a proxy. The canonical price data for WTI crude comes from CME. S&P 500 data comes from S&P Dow Jones indices. Both are expensive and neither has licensed its data broadly for use in perpetual DEXes. So exchanges pull WTI futures data from secondary sources, typically trading desks or crypto exchange order books.

For indexes, they construct their own facsimile index design to track the brand name index closely. Right now, Stork Research is tracking four separate Nasdaq 100 proxy perps and six separate S&P 500 proxy perps. Each printing a slightly different price. The Trade XYZ S&P 500 contract, officially officially licensed from Dow Jones, makes a seventh. These are not the same product.

Layer on top of that, these index perps are always quoted in stablecoins, not dollars. That introduces basis risk between USDC and USDT and USD. And in the case of foreign equity markets, Hong Kong, Seoul, Tokyo, you're often trading in dollar-denominated stablecoins, not yuan, won, or wen. So, you're also taking implicit currency exposure you may have not signed up for. All that said, it's working.

Weekend index volume averaged 164 million per day in May. WTI on weekends average 635 million per day. These are not rounding errors. So, how does a WTI perp print a price on a Saturday afternoon when CME is closed? Here's the mechanism. At Friday 4:00 p.m. Central close, the perp the perp exchange switches to an internal oracle.

That oracle takes the CME closing price as a starting point and moves it conservatively based on order book activity on the exchange itself. The keyword is conservatively. The algorithm uses impact prices, exponential moving averages, and hard caps on how much the price can move in any given window. These are intended to be manipulation-resistant methods. The results have actually been good. Over 19 of 21 weekends in 2026, Trade XYZ's WTI market on Hyperliquid moved with directional accuracy towards the CME Sunday open 90.5% of the time, according to Stork

research. Stork has now built 24/7 oracle methodology covering four commodities and four of the largest equity perp markets. Our inputs come from five exchanges: Binance, Bitget, Hyperliquid, Lighter, and OKX. When you trade on any of these markets over a weekend, your price is muted interpretation of order book activity anchored to the last known TradFi closing price. Many smaller exchanges also incorporate price updates from Trade XYZ, so in a very real sense, you may be trading a price that was formed somewhere else with your own exchange activity adding some marginal influence.

Pre-IPO perps are the story everyone is watching. SpaceX, Cerebras, Quantinuum. These have captured serious attention both inside and outside crypto. SpaceX is the largest pre-IPO perp for obvious reasons, but AI chip may chip maker Cerebras was the first and the one that got everyone's attention. In the hours leading up to Cerebras' May IPO, the trade XYZ contract on the hyperliquid traded much higher than the IPO price set the night before. At the opening of trading, the perp had predicted the opening minute within 0.5%.

And bank investment bankers were literally following the perp price on their screens. The share price fell back to $311 by the day's close. The markets the perp markets were frothy, but at the end of the day, they were closer to the stock stock market value than the IPO price of $185 set by the underwriters the night before. Pre-IPO perps did not look quite so good the second time out. Traders had bid up the pre-IPO perp for Quantinuum, a quantum computing company, to around $100.

Around 6:00 a.m. on Quantinuum's IPO day, the price started to falter. By the time the stock started trading at 1:34, it had fallen to $68 to match the Nasdaq trading open. The S-1 price of $54 boosted on the Monday before the IPO was closer to the actual start of trading than the frothy pre-IPO perp markets by a factor of two. Instead of an IPO pop predi- predictor, it looked like easy money for sophisticated participants who knew where the price was likely to open.

What you're actually trading in pre-IPO perps? Shares in private companies do trade on established secondary markets and those prices feed into the Oracle. But tick by tick, you're trading order book activity on the perp itself. With all the liquidity and manipulation risk that entails. These markets are thin. SpaceX is the deepest pre-IPO market to date with 1.45 billion in total volume through June 3rd.

The Cerebrus perps price discovery is remarkable with only 213.6 million in trading volume up to its IPO. Quantinuum was about 40% of that with $82 million. At this point, pre-IPO price discovery, like we saw with Cerebrus, is remarkable, not guaranteed. All this points to the same underlying constraint pricing. Oracle infrastructure in crypto perps was built to solve a specific problem.

Spot markets are fragmented, price differs across venues, and perp exchanges cannot afford data mishaps. A single errant print can cause a liquidation that shouldn't have happened. The Oracle filters noise and delivers the true state of the market. Stork was the first Oracle to do this at sub-millisecond latency for perp DEXes. We still are. The RWA version of this problem is different in structure, but harder in practice.

During market hours, price discovery for equities and most commodities happen on one canonical venue. CME for oil, New York Stock Exchange or Nasdaq for equities, for US equities. The question of what is the price has a clear answer. But, outside those hours, on weekends, in pre-IPO markets, there is no canonical value. That's where DeFi has stepped up, and it has done so with thin liquidity, proxied indexes, and nascent order books that are nonetheless producing directionally accurate prices.

Even overnight US equity trading now has a reference. The Blue Ocean ATS runs sessions from 8:00 p.m. to 4:00 a.m. Eastern. There's a thin but real spot market, and Stork delivers those prices to subscribers. But, 24/7 is a poor value proposition, and weekends remain a liquidity gap. Here's the real barrier to the next volume unlock. Professional traders. The 24/7 price discovery story has their attention. The issue is hedging. Professional trading desks wind down their RWA perp positions on Fridays.

Without liquid spot markets to hedge against, the risk reward on weekend RWA perps doesn't hold up. So, the market remains retail-dominated and thin. The path forward runs through a few parallel tracks. Tokenized stocks trading 24/7, expanded overnight and weekend commodity sessions, and deeper pre-IPO secondary markets. Each of these requires the same thing at the core. Reliable reference prices running continuously across a wide range of global assets, including where the fragmentation of crypto price discovery is now entering traditional markets for the first time.

This is where Stork is focused. We built the latency and uptime profile that perp decks is required for crypto assets. We are now applying those same capabilities to the hardest parts of RWA problem. The weekends, the off hours, the assets with no canonical pricing venue. The oracle is not the passive is not passive infrastructure here. Thank you.