MarketsBlockchain
|13 min ReadUS Stock Perpetuals on DeFi: Trading 100x Apple and Nvidia
Maya Chen
Senior Analyst
Published
Jan 16, 2026
Alpha Briefing: On-chain US stock perpetuals are taking off, with trade.xyz, Lighter and Ostium offering 10 to 100 times leverage on names like Apple and Nvidia, backed by billions of dollars in volume and deep liquidity. The tools are powerful and the mechanics are clever, but the risk profile is brutally leveraged for anyone who misjudges Wall Street’s next move.
Smart crypto money is quietly changing games. Instead of chasing the next random memecoin, a growing crowd is using crypto rails to trade the biggest market on earth: US stocks. Bitcoin’s market cap sits around 1.7 trillion dollars. The US equity market is over 50 trillion. Any one of Apple, Microsoft or Nvidia can dwarf half of the entire crypto complex by value. Traders are looking at that scale and thinking one thing. This is where the real liquidity and real stories live.
US stocks are tied to everything. Global growth, interest rates, geopolitics, AI chips, social media. Every earnings call and policy statement can move trillions. No memecoin can compete with that level of attention. That is why perpetual futures on US stocks are becoming the next frontier for aggressive traders who already understand on-chain leverage.
Perpetual DEXs like Hyperliquid, trade.xyz, Ostium and Lighter are racing to bring US stock contracts on-chain. They promise 24/7 access, high leverage and permissionless accounts. They strip away the slow, painful parts of traditional brokerage that used to block people outside the United States. No more waiting on KYC approvals and settling for 2 to 4 times margin. Now, on the right platform, you can push Apple to 100 times notional exposure. It looks beautiful when you are right. It is devastating when you are wrong.
Where the leverage is: trade.xyz, Lighter and Ostium compared
If you think like a trader, three questions come first. Which stocks can I trade. How much leverage can I use. What does it cost me each time I click.
Trade.xyz is the first perpetual DEX built on Hyperliquid’s HIP-3 protocol and today it is the largest DEX on that stack. Its main pitch is simple. True 24/7 US stock trading on-chain. The platform focuses on a custom US tech index called XYZ100, which tracks the Nasdaq, plus a basket of single-stock perpetuals. Right now it supports 11 US equity assets. Most single-name stocks offer 10 times leverage. The XYZ100 index can go to 20 times.
Trade.xyz uses a central limit order book model, so active traders see familiar depth, spreads and limit orders. The fee structure is in growth mode. Management has cut fees by over 90 percent. Actual taker fees sit around 0.009 percent, roughly 9 dollars per 100,000 dollars traded. Maker fees can go as low as 0.003 percent. The team operates under the Hyperunit brand in a low-profile, semi-anonymous style and has not announced external fundraising yet. It is still in a pre-token stage, with community chatter linking it to the Hyperliquid ecosystem. Background pieces such as “10 天 20 亿美金交易量，Hyperliquid 的又一个爆款” are already treating it as a breakout product.
Lighter takes a different path. It runs on a custom zk-rollup on Ethereum and sells itself as a zero-fee, provably fair perpetual platform. All order matching and liquidation paths are verified using zero-knowledge proofs, which means execution can be audited cryptographically. Lighter has just rolled out US stock trading, starting with 5 names, all at 10 times leverage. It also uses a CLOB engine.
The team behind Lighter is not shy. Founder Vladimir Novakovski is a Russian-born math and physics competition winner who graduated from Harvard at 18, then went on to trade at Citadel and build a long career in quantitative finance and tech. He co-founded AI social platform Lunchclub, which raised 30 million dollars. Lighter itself raised 68 million dollars in a series B round at a 1.5 billion dollar valuation, backed by Founders Fund, Ribbit Capital, a16z crypto, Lightspeed, Coatue and others. Retail users trade US stocks with zero maker and taker fees. High frequency and market making flow pays a small fee, roughly 0.002 percent maker and 0.02 percent taker. On top of that, Lighter charges and pays funding based on the premium time-weighted average price, capped at plus or minus 0.5 percent per hour.
Ostium is built on Arbitrum and lives in the real-world asset niche. It lists perps on US stocks, stock indices, commodities and FX. Its calling card is aggressive leverage, with a maximum of 200 times on some products. The founding team includes former Bridgewater staff and Harvard classmates. The project raised 3.5 million dollars in a seed round led by General Catalyst and LocalGlobe, with well-known names like Balaji Srinivasan, SIG and GSR on the cap table. It is still pre-token and runs a points campaign for active users.
For US stocks, Ostium currently has the widest shelf. It supports 13 names. Mainstream tech giants like Apple, Amazon, Meta, Microsoft, Nvidia and Tesla can be levered up to 100 times. Crypto-linked equities such as Coinbase, Robinhood, MicroStrategy, SBET and Circle sit between 30 and 50 times. Unlike trade.xyz and Lighter, Ostium does not use a book. It uses a pool-to-pool AMM design on Arbitrum.
Fees are simple but strict. The opening fee is a flat 0.05 percent. There is no closing fee. Each trade also pays an oracle fee of 0.10 dollars. Depending on execution, that fee can be refunded after close. Details are laid out in the official fee breakdown. On top of that sits a rollover fee, which is similar to a funding rate but calculated per block, with compounding and potentially different rates for longs and shorts.
How these perps trade when Wall Street sleeps
There is a structural problem when you put US stocks on 24/7 crypto rails. The stocks themselves do not trade around the clock. Traditional US sessions have regular cash hours, extended pre-market and after-hours windows, and then long dead periods. Perpetuals never sleep.
So the question is obvious. When the underlying price is asleep, what do you feed the oracle. How do you handle orders in the dark.
Trade.xyz splits the problem by asset and time. For its XYZ100 index, it ignores the stock market and looks at CME NQ futures, which trade 23 hours a day. Using a cost-of-carry model, the platform maps the futures price back into a synthetic spot index. That keeps the index perp priced almost all the time.
Single stocks are trickier. They do not have nearly 24-hour futures. For those, trade.xyz leans heavily on Pyth price feeds, which cover pre-market, after-hours and overnight sessions as long as there is quoting activity. Only when all external feeds are dark does the system fall back to an internal engine. It then reads its own on-chain order book and computes an exponentially weighted moving average with an eight-hour time constant to smooth prices. It also adjusts spreads based on order book depth, so that thin books show more slippage. As soon as external data returns, the oracle snaps back to that reference.
Ostium went for a custom, pull-based RWA oracle. The team itself handles data sourcing, exchange hours, futures contract rolls and opening gaps. Nodes such as the Stork network run the feeds in a decentralized way. That level of control costs money, which is why every Ostium trade includes an oracle fee. Even if a trade fails due to slippage being set too tight, the oracle fee is still charged. When a trade fills and settles, the fee can be refunded, as described in the same fee breakdown docs.
For the user, this means prices do not drift during full market closures. You can enter limit and stop orders during the break, but they will only trigger after the next open if the price condition is met. Market orders are disabled while cash markets are closed. The behavior is intentionally close to traditional equity trading. The system even respects holiday closures. With leverage, Ostium is strict. High intraday leverage is allowed only during the regular trading window. Before the close, leverage requirements tighten. Positions with leverage above allowed overnight levels are force-closed roughly 15 minutes before the session ends to cut off gap risk.
Lighter’s approach is simpler and more conservative. During trading hours, US stock perps behave like any other market on the platform. Prices update, orders match, positions can be opened, adjusted and closed. Once the stock market closes, Lighter flips the switch into reduce-only mode for those products. You can trim or close positions. You cannot open new ones or crank up leverage. Index prices are allowed to move only inside a narrow band of plus or minus 0.5 percent from the last reference level, to prevent runaway moves while the underlying is frozen. Funding still runs in the background, but fresh directional risk cannot be added.
Dividends, funding and who eats the juicy yield
US stocks do something that Bitcoin never does. They pay cash dividends. That is free money for cash equity holders. It is a nightmare for careless perp design.
Imagine a stock trading at 100 dollars that pays a 2 dollar dividend. On the ex-dividend date, the stock price drops to roughly 98. In spot markets, everyone understands this. You lost 2 dollars in price, but you received 2 dollars in cash. You are flat. In a perpetual, if you do nothing, the short side gets a free 2 dollar win from the price drop, while the long side takes the hit without the dividend. Traders will immediately farm that edge by shorting before dividend and covering after. That is pure arbitrage if you let it stand.
Trade.xyz fixes this by baking dividends into the funding rate. Suppose the oracle shows 100 dollars today and needs to jump to 98 at a future moment due to the dividend. The mark price is pushed down gradually into a discount curve as that moment approaches. In each hour before the cut, the mark trades slightly below the oracle. The funding rate is set by the mark-oracle spread divided by the oracle price plus a truncation term. A mark below oracle produces a negative funding rate. Shorts pay longs. Over many small steps, the sum of funding charged to shorts equals the 2 dollar price move. On paper, the short makes 2 dollars when price drops from 100 to 98. In reality, those 2 dollars are paid out in funding. The long loses 2 dollars on price but earns 2 dollars in funding. Economically, the long has captured the dividend.
Ostium takes a more purist view. In its design, a perpetual tracks price only. It does not pretend that you own the underlying stock, so it does not try to simulate dividend entitlement directly. If a stock goes ex-dividend and the oracle drops from 100 to 98, the perp will simply follow. Longs lose on price. Shorts win on price. There is no special one-off funding adjustment to “equate” the dividend.
The way Ostium balances things is through the rollover fee. This fee is a function of time and direction. It behaves like a continuous holding cost baked into the contract. It is displayed as a net long or short rate, often labeled L or S on the screen. If you hover over that field, you see how the rollover fee is accruing. For one side it can be positive, meaning you are effectively paid to hold. For the other side it can be negative, meaning you pay to carry the position. The rates compound on a per-block basis and only settle when you close the trade. In practice, the rollover fee reflects funding costs, dividend expectations and other carry components. The result is still a fair split of economics between longs and shorts over time, even if the platform does not label it explicitly as “dividend treatment.”
Lighter does not call out dividends as a special case in its docs. Mechanically, the effect is similar to trade.xyz. When a stock goes ex-dividend and the cash equity price drops, the oracle follows. If the perp mark drifts away from that oracle, funding turns negative or positive to pull it back. With the 0.5 percent per hour cap, the system also protects traders from insane funding spikes. In practice, any dividend adjustment flows through a mix of price and capped funding.
Volumes show demand, but the risk is savage
For all three platforms, early product-market fit is visible in the numbers. Trade.xyz has already cleared more than 2 billion dollars in cumulative volume. Around major events, such as Nvidia earnings, it has printed single-day volumes of roughly 200 million dollars. Over a recent 24-hour window, total volume was about 734 million dollars.
Lighter’s global metrics are even larger. Its total perp volume and open interest are strong. One recent day showed 7.16 billion dollars in volume and 1.634 billion dollars in open interest. Because US stock trading on Lighter is brand new, there is not yet a clean public breakdown of how much of that is US equities versus crypto pairs, but the direction is clear. The pipes are already built for size.
Ostium’s total lifetime volume stands above 27.2 billion dollars, with roughly 138 million dollars turned over in a recent 24-hour window.
On Dune dashboards that track Ostium’s US stock contracts specifically, daily notional can exceed 50 million dollars. US stock perp open interest on Ostium hit a peak of around 64 million dollars before easing back toward 45 million dollars. US stocks now represent about 20 percent of total open interest on the platform.
This is what early adoption looks like. On-chain traders are learning they can trade Nvidia earnings, Apple product cycles and Coinbase sentiment from the same wallet they use for memecoins, but with institutional-style leverage. It feels new and efficient. It is also unforgiving. Market efficiency in leverage products is brutal. A 100 times long is not just “more exposure.” It is 100 times the risk of liquidation.
If you are drawn to US stock perpetuals, the smartest move is the least flashy. Start with small size and small leverage. Treat 100 times leverage as a rare, tactical tool, not a default setting. In this game, the market will always be there tomorrow. Your account will not, if you let the beautiful leverage run your life instead of your brain.
Disclaimer: This document is intended for informational and entertainment purposes only. The views expressed in this document are not, and should not be taken as, investment advice or recommendations. Recipients should do their own due diligence, taking into account their specific financial circumstances, investment objectives and risk tolerance, which are not considered here, before investing. This document is not an offer, or the solicitation of an offer, to buy or sell any of the assets mentioned.