Trading all things, never closing the market: RWA perpetual contracts (Part 1)

This is a deep research report from @OKX_Ventures. Due to its length, we will publish it in two parts, with the next part expected to be released on March 12. This first part contains a total of 6,700 words, with an estimated reading time of about 20 minutes. Feel free to like and bookmark it for a leisurely read!

Introduction:

Currently, the exploration of RWA in the Crypto market mainly focuses on asset tokenization—specifically, how to map the ownership of real-world assets such as government bonds, stocks, or real estate onto the blockchain to achieve more efficient settlement and holding. However, this solution, centered on efficient holding and settlement, cannot fully meet another side of demand in the financial market, which is the leverage trading and risk management of asset price fluctuations.

In fact, the true engine of liquidity in the global financial market is not the static holders of assets but the traders seeking leveraged directional exposure. From the U.S. end-of-month nominal value of about $50 trillion in the options market to the non-U.S. regions with a monthly trading volume of about $30 trillion in the CFD (Contract for Difference) market, retail investors' thirst for high leverage and short-term risk exposure has never ceased.

Despite the enormous trading scale, existing traditional financial instruments still struggle to meet this demand: 0DTE Options (Zero Days to Expiration) force traders to bear both Theta (time decay) and Vega (volatility) non-linear risks while engaging in pure directional speculation. Meanwhile, the CFD market has been criticized for its opaque black-box mechanisms and centralized counterparty risks.

From the perspective of traders purely seeking directional exposure, what many traders truly desire is not "options" or "tokenized stocks," but a pure Delta One exposure—where the fluctuations in asset prices can be simply and directly converted into proportional investment gains and losses, without any loss or deviation in between (Arthur Hayes wrote an article titled "Adapt or Die" at the end of last year, reviewing the complete background of their development of crypto perpetual contracts, which is worth reading).

It is within this structural mismatch that DeFi protocols have keenly captured this market opportunity. Some DeFi entrepreneurs are attempting to introduce the perpetual contracts that the Crypto market has spent nearly 10 years validating into the traditional asset domain. These products adopt a synthetic derivatives architecture, anchoring the underlying asset prices through oracle feeds and funding rate mechanisms, providing around-the-clock leveraged trading services for stocks, commodities, and foreign exchange without the need to actually hold or deliver the assets.

(Image: Current main trading asset types of RWA Perps Dex)

---

I. Market Background (Opportunities in the RWA Perps Market)

1.1 Market Entry 1: The U.S. 0DTE Option Market

In the past decade, the U.S. options market has undergone profound structural changes. According to data from major options exchange Cboe Global Markets, the trading volume of 0DTE options in S&P 500 index options has surged from less than 5% in 2016 to over 60% currently, with a monthly nominal trading volume of $48 trillion (about 40 times the monthly trading volume of perpetual contracts on CEX exchanges). This data not only reflects the increase in trading frequency but also reveals the existence of a massive capital force in the market seeking extremely high leveraged exposure on a daily basis.

Note: 0DTE stands for "Zero Days to Expiration," referring to options that expire on the same day. These options expire at the end of the trading day. Traders use them for ultra-short-term speculation to obtain quick returns and avoid overnight holding risks.

(Image: The above two images show the proportion of S&P 500 index options with different expiration times from 2016 to 2025. It can be seen that 0DTE Options accounted for nearly 5% of the options market in 2016, and by 2025, its market share skyrocketed to 61%, indicating that nearly half of the S&P 500 index options trading is betting on the direction of the day for ultra-short-term speculation.)

(Image: The above image shows that retail investors are the absolute main force in the 0DTE market.)

From the first principles of financial instruments, financial derivatives can be divided into Delta One products and non-linear products. Traditional Delta One tools like stocks and futures have symmetrical risk exposure: the gains from price increases are linearly proportional to the losses from price decreases. However, the original intention of options design is to manage asymmetric risks.

For example, a fund manager holding a large amount of Apple stock may be optimistic about the company's long-term fundamentals and unwilling to sell the stock, but he is concerned about short-term earnings volatility leading to a significant drop in stock price. At this point, he can purchase put options to insure his position. In this structure, his potential for profit remains intact as the stock price rises (symmetric upside potential), but his loss is strictly capped at the premium paid (asymmetric downside risk).

To achieve this "separation of rights and obligations" insurance function, the cost structure of options must not only reflect the intrinsic value of direction (Delta) but also include the time value (Theta) reflecting the possibility of volatility (Gamma) and the passage of time.

The significant growth in the share of the 0DTE market in recent years reveals a paradox: many traders are not seeking to manage asymmetric risks or engage in complex volatility trading, but rather view it as the only way to obtain intraday directional leverage. In this case, traders are forced to pay a high cost of time value (Theta Decay) for an "insurance function" they do not need. As long as the speed of the underlying asset's increase is insufficient to cover the decay of time value, even if the directional judgment is correct, the trade will still result in a loss.

Time value is the main part of options that shrinks over time and is the core of 0DTE options traders' struggles.

Therefore, perpetual contracts, as a Delta One product, provide value by stripping away excess time and volatility costs, offering pure linear leveraged exposure that can mathematically and logically match the speculative needs of this capital more accurately than 0DTE Options.

1.2 Market Entry 2: The CFD Market Outside the U.S.

In markets outside the U.S., retail leverage demand is mainly met by CFDs (Contracts for Difference), with the monthly average trading volume of the CFD market reaching $30 trillion by 2025.

Although CFDs provide a linear return Delta One structure, their market operation model is based on the brokerage model, which has significant transparency issues.

The vast majority of CFD brokers adopt a B-Book (internal market-making) model, where brokers act directly as counterparties to clients (there are serious risk control brokers in the industry that hedge against profitable clients to avoid risks, but since the top few companies in the CFD market only account for 20% of the market share, the remaining 80% is filled with many small and medium brokers, making the overall market rife with brokers relying on client losses to make profits), in this zero-sum game structure and opaque black box, brokers have the technical authority and economic motivation to modify quotes, slippage, and execution speed.

Compared to CFD products, RWA Perps can also be understood as a "transparent CFD based on smart contracts." By putting the clearing logic, funding rate calculations, and oracle pricing on-chain, DeFi protocols eliminate the possibility of centralized brokers interfering with trading outcomes. At the same time, the atomic settlement mechanism based on stablecoins enhances the efficiency of capital flow to the second level, achieving true self-custody of funds and real-time clearing.

---

II. Challenges in Building RWA Perps Products

RWA Perps are not simply a replication of the perpetual contracts focused on crypto assets that we have seen before. Crypto assets have the characteristics of 24/7 trading, real-time pricing, and T+0 on-chain settlement, while traditional assets are constrained by the legal frameworks of the physical world, holiday systems, and outdated banking clearing protocols.

This asynchronous nature of the underlying attributes creates an "impossible triangle" in the product design of RWA Perps:

• High Leverage: Meeting retail users' speculative demand for high leverage.

• 24/7 Availability: Maintaining the core value of DeFi transactions anytime, anywhere.

• Risk Externalization: Ensuring that the protocol and market makers do not bear directional betting risks, achieving systemic long-term survival.

2.1 How to Anchor On-Chain Prices of RWA Perps When the U.S. Stock Market is Closed?

The essence of Perps products is "the mirror of price discovery," requiring continuous external spot price feeds. However, when the Nasdaq or CME is closed on weekends and at night, it creates a break in the oracle data sources.

This pricing vacuum and misalignment during the U.S. stock market's closure give rise to two core risks:

• Risk 1: Market Makers Lack Sufficient Risk Hedging Channels During Weekend Closures

  Professional market makers can provide extremely narrow spreads and deep liquidity because they do not bet on direction but pursue neutral positions and only collect spreads. This means that for every $1 million worth of Tesla stock contracts sold to traders on-chain, market makers must immediately buy an equivalent amount of assets in the traditional spot or futures market to hedge this risk exposure.

  When traditional markets are closed, the hedging channels are shut down, and market makers cannot adjust their hedging positions. To avoid this risk, market makers can only choose to withdraw orders or add a significant risk premium to their quotes during closed periods. This explains why traditional order book models see spreads expand non-linearly to dozens of times normal levels on weekends, making liquidity dry up easily.

• Risk 2: "Gap Risk" of Extremely High or Low Openings on Monday

  Due to the 24/7 trading nature of crypto-native assets, the price curves of assets are usually continuous, and the clearing engine has enough time to liquidate users' positions when prices drop. However, in the RWA Perps domain, the pressure of price increases and decreases accumulated during the closure of traditional assets will be released instantaneously when the market opens on Monday. If there is a significant gap at the opening on Monday, the clearing engine may find itself in a vacuum during the "price gap," unable to find counterparties to execute liquidation before the position is liquidated.

To address the above dilemmas, there are currently two main handling solutions for RWA Perps:

1. Internal Simulated Pricing (e.g., TradeXYZ / Hyperliquid): Introduces an Exponential Moving Average (EMA) algorithm to allow prices to slowly "drift" based on on-chain buy and sell pressure when the oracle disconnects, maintaining a 24/7 shell, but theoretically still a manipulable "shadow market."

2. Mandatory Risk Downgrade (e.g., Ostium): This is a more pragmatic risk control solution. Ostium introduces 0DTE attributes: requiring all high-leverage positions to be automatically liquidated or significantly reduced in leverage before the market closes. Only low-leverage positions (with sufficient margin buffers to cover 5%-10% gaps) are allowed to be held overnight. This approach sacrifices some "perpetuality" in exchange for absolute safety for the system when facing Monday's opening gaps, preventing the LP pool from being penetrated by systemic bad debts.

2.2 How to Provide TradFi-Level Trading Depth On-Chain at Low Cost?

In the development of DEXs, the choice of liquidity provision and order execution mechanisms is a core variable determining the system's capital efficiency, risk allocation logic, and user experience. The two mainstream solutions currently are: CLOB (Central Limit Order Book) and Oracle-based Pool (Oracle-driven liquidity pools).

Hyperliquid has validated the success of the order book model for crypto-native assets, with its core being zero-friction hedging execution: market makers can transfer risk across platforms in milliseconds using stablecoins. After receiving orders on the on-chain order book, market makers can use stablecoins for millisecond-level risk hedging in CEXs that operate 24/7. Since crypto funds and assets operate within a highly interconnected crypto network, the hedging costs are extremely low, allowing market makers to compress quote spreads to very narrow ranges, thereby attracting trading volume and forming a positive feedback loop.

In the RWA domain, market makers face significant cross-border hedging friction: on one hand, the time mismatch between on-chain USDC (T+0) and traditional fiat currency settlements forces market makers to keep a large amount of dollars idle in traditional accounts as hedging reserves; on the other hand, the traditional banking weekend and holiday closure mechanisms mean that market makers cannot hedge in a timely manner when sudden market movements occur on non-working days.

This is also why founders like Kaledora of Ostium have insisted on adopting a pool-based model rather than an order book, as she believes that the zero-friction hedging seen in crypto-native asset exchanges is difficult to achieve in the RWA perps domain. When market makers receive an NVDA order in RWA Perps, they cannot use stablecoins to hedge in milliseconds on Nasdaq because they must navigate numerous traditional banking obstacles.

2.3 How Does the System Ensure It Does Not Go Bankrupt When Traders Are Continuously Profiting from One-Sided Markets?

The third dilemma involves how the protocol ensures long-term solvency through external hedging. The pool model of GMX has been able to survive in the crypto market for a long time because it plays the role of a "passive dealer," leveraging statistical advantages under large samples to stabilize the absorption of position wear and liquidation profits generated by high-leverage positions amid frequent fluctuations. In the crypto market, which exhibits significant oscillation characteristics, the mathematical expectation of this model is favorable for pool LPs.

However, the risk distribution of RWA assets is entirely different. Mainstream indices like the S&P 500 often exhibit prolonged one-sided bullish trends lasting for years. In the absence of risk externalization (hedging) mechanisms, users' continuous profits will directly translate into net losses for the LP funding pool, leading to a situation where the system not only fails to capture volatility dividends but is also completely drained by one-sided positions, ultimately facing solvency exhaustion.

---

III. Representative Projects and Structural Game: Oracle Pricing + Pool-Based vs. Order Book

(Image: Daily trading volume of RWA Perps Dex, showing a significant drop in trading volume over the weekend.)

The core contradiction of RWA Perps revolves around the "break in physical time": although various RWA Perps Dex platforms have generated over $20 billion in trading volume within 30 days, trading volume shrinks dramatically by 70-90% during weekends. This data reveals the current true state of the industry: despite DeFi's attempts to break free from the gravity of traditional finance, liquidity still heavily relies on the opening hours of TradFi.

In the face of this break, the market has evolved two distinctly different architectural paradigms: the active hedge pool model represented by Ostium and the internal pricing order book model represented by Trade.xyz in the Hyperliquid ecosystem.

3.1 Early RWA Perps Projects: Synthetix, Gains Network

Before Ostium and Hyperliquid attempted to reconstruct RWA trading through complex hedging mechanisms or order books, the DeFi market had already conducted the first round of "synthetic asset" experiments. Early protocols represented by Synthetix and Gains Network completed the concept validation of RWA Perps, demonstrating a strong demand for on-chain capital exposure to traditional assets, but also fully exposing the limitations of the first-generation mechanisms in capital efficiency and risk control.

Synthetix: Global Debt Pool Model

Synthetix was one of the first protocols to attempt to bring real asset prices onto the blockchain. Between 2020 and 2021, Synthetix aggressively tried to launch synthetic stocks like sAAPL and sTSLA, attempting to bring U.S. stocks onto the blockchain.

As a pioneer of the "pool counterparty" model (where all SNX stakers are counterparties), Synthetix's design philosophy was to establish a model of exchange without an order book and infinite liquidity: all synthetic assets could be freely exchanged at prices provided by oracles, allowing users to trade without matching counterparties, which greatly solved the liquidity cold start problem in the early days (especially under the liquidity mining incentives at that time).

After 2021, Synthetix delisted most RWA assets, primarily due to the lack of active hedging mechanisms at the protocol level, making it vulnerable to attacks when prices of assets like sTSLA could not be updated during market closures.

Overall, Synthetix pioneered a model of providing on-chain RWA synthetic asset liquidity through derivative collateral pools. The design of no order book + oracle pricing remains influential today, but the product side effectively began to withdraw from the RWA Perps market around 2022.

Gains Network (gTrade): Oracle Pricing-Driven Market Making Pool Model

Gains is another early representative project exploring on-chain RWA synthetic leveraged trading, supporting various trading pairs including cryptocurrencies, foreign exchange, and U.S. stocks. Its design concept is to use independent asset pools as counterparties: users can open synthetic leveraged positions by collateralizing USDC, DAI, or ETH, with trading gains and losses borne by the funding pool (gToken Vault).

Liquidity Model and Market Making Game Mechanism:

• Unilateral Treasury: Gains' market-making funding pool mainly consists of stablecoins like USDC/DAI.

• GNS Token as Risk Buffer and Incentive: To prevent the market-making funding pool from being penetrated during extreme market conditions, the protocol introduces the GNS token as a last line of defense. When the market-making funding pool has a surplus, the protocol will use excess profits from the pool to buy back and burn GNS tokens to reduce inflation. When the market-making funding pool incurs losses, the system will mint GNS and sell it off-market to replenish the market-making funds.

In terms of pricing, Gains obtains real-time prices based on Chainlink and adds a fixed spread, with spread income distributed as fees to LPs and GNS stakers. Risk control measures include price impact fees (charging additional fees for large orders to simulate slippage and compensate for funding pool risks), limit protections (setting upper and lower limits on individual profits and losses to enforce take-profit or forced liquidation), among other designs.

Overall, Gains provides a highly leveraged, multi-market coverage synthetic trading experience and is regarded as an important example of decentralized exchanges benchmarking against centralized platforms, demonstrating that the "oracle + liquidity pool" model can support large-scale trading under reasonable risk control. However, it also exposes challenges such as the funding pool needing to bear concentrated profit risks and the lack of hedging mechanisms, providing experience for subsequent projects' mechanism innovations.

3.2 Ostium: Breaking Through the Limitations of the Pool-Based Model to Create an On-Chain CFD Broker

Ostium is a rising RWA Perp DEX that officially launched on the Arbitrum mainnet in August 2025.

In terms of liquidity provision and order execution mechanisms, Ostium still chooses the pool-based model as its core architecture. However, based on reflections on the early mechanisms of GMX, Gains Network, and others, they deeply realize that the opposing game relationship of "traders profit means LP losses" in traditional pool models is long-term unfavorable for LPs and, more importantly, limits the upper limit of trading volume, making it impossible to expand market scale (as analyzed in the previous Perp Dex study). To address this, they have made some special designs to integrate traditional brokers' A-Book (hedging) and B-Book (internal digestion) on-chain to alleviate this zero-sum game conflict.

Interpretation of Liquidity Model and Market Making Game Mechanism

• Basic Liquidity Model (Two-Tier Pool Structure)

• First-Level Buffer: Liquidity Buffer: This is the "moat" owned by the protocol, accumulated from protocol revenues. Traders' profits are first paid from here, and losses also first enter here. Although the details of the mechanism differ, its role is similar to the market-making pool protection cushion in Gains Network.

• Second-Level Buffer: Market Making Vault (OLP Vault): This is the pool funded by LPs. Only when the Liquidity Buffer funds are exhausted will the OLP intervene as a direct counterparty.

• Core Evolution to Break Through the Limitations of the Original Pool-Based Model: Complete Separation of "Settlement" and "Market Making": Ostium understands that the simple two-tier buffer cannot cope with long-term directional imbalances (the data shown in the figure below indeed proves this, as the funds in the liquidity buffer layer can be easily exhausted. When the v1 version of the product only had these two basic layers, LPs still faced long-term one-sided risks). Therefore, Ostium has introduced a more important design—completely separating the functions of settlement and market making from the original passive market-making pool of LPs.

Currently, the OMM market-making hedging vault has not officially launched. It can be expected that when carrying high trading volumes, the product's operation will require a professional market-making team with strong execution capabilities, which remains a significant challenge: the team not only needs to have compliance qualifications to connect with traditional finance but must also achieve millisecond-level cross-market hedging to avoid oracle and external real market basis risks; at the same time, they need to have strong capital scheduling capabilities to overcome the time mismatch in fund flows on-chain and off-chain, and be able to monitor delta net position imbalances in real-time, flexibly using dynamic spreads or impact fees for precise risk control and flow limitation.

Risk Control During Market Closures

Ostium deeply aligns with U.S. stock trading hours, ensuring that market orders are only executed during market open by embedding timestamps in the oracle, effectively eliminating price vacuum risks during market closures. To address the common gap risks in U.S. stocks, the platform has established strict "forced liquidation checkpoints": 15 minutes before daily closing, the system automatically forces liquidation of positions with leverage exceeding a threshold (e.g., 10x), bringing intraday maximum leverage of 100x back to a safe range.

Why Haven't Existing Pool-Based Projects Like GMX Implemented Similar Designs?

We believe the main reason GMX has long adhered to the pool model without separating directional risks is that the trade-off is too great and the market starting points are different: the current design has achieved relative balance through internal mechanisms (such as adaptive funding fees, price impact, long-short pool separation), and introducing external/independent hedge vaults would sacrifice profits, increase complexity, and centralization risks.

Additionally, GMX's pool actually bears the comprehensive exposure of all traders. In the highly volatile crypto market, individual random bets statistically tend to a negative expected value according to the law of large numbers, with the pool capturing positive expected values as a comprehensive counterparty. In contrast, Ostium focuses on markets like stocks and other RWAs, which are relatively less volatile, and they aim to penetrate the traditional CFD broker market.

Furthermore, in August 2025, there was a proposal in the GMX governance forum for a Global Hedge Vault (GHV), hoping to introduce external market maker mechanisms to achieve something similar to Delta Neutral, indicating that other pool-based projects are also paying attention to this new trend.

Why Choose a Pool Model Instead of an Order Book?

Ostium founder Kaledora has a clear theoretical logic regarding why she insists on choosing a Pool Based model and not allowing weekend trading. She has previously criticized order book projects like Trade for their outrageous high funding rates during weekends, which led to attacks from the Hyperliquid community.

Her theory is that the limitations of the traditional Pool based model (LPs bear one-sided directional risks, and the system's capital volume limits trading volume) can be solved by her new design. By introducing a mixed risk control of A-Book and B-Book, one-sided risks can be transferred in real-time to the liquidity-unlimited global market. Once one-sided risks are technically resolved, the OI upper limit will no longer be constrained by the size of the pool, and the protocol's trading volume limit will entirely depend on its distribution capabilities (similar to the business model of top CFD brokers).

In contrast, she believes that the core function of an order book is price discovery, which is very meaningful for crypto-native assets but is a significant waste of resources in the RWA domain. Because the prices of stocks and foreign exchange are already perfectly discovered in real-time at top global exchanges like Nasdaq and CME, recreating an on-chain order book means competing with these trillion-dollar giants in a "bloodless" environment. This depth from traditional exchanges' dimensionality reduction makes any large trader prefer brokers that can reference global prices rather than order books with shocking slippage.

(To be continued)

Original link