Options are back in focus on-chain, but Castle Labs says one of the biggest problems in the market is that people still use a single term to describe several products that are structurally different.
In the article published by TechFlowPost, Castle Labs splits the space into three active product families: exchange-style options platforms, binary options and prediction-market venues, and structured products or yield strategies. The piece argues that users can easily pick the wrong tool if they do not separate those categories first.
Three product families, three different user needs
Castle Labs groups Derive and Paradex into the exchange-style options category. Polymarket, Kalshi and Hyperliquid HIP-4 sit in the binary options or prediction-market category. Rysk and Hegic are treated as structured-product and yield protocols.
The report draws a sharp line between them. A trader buying a capped-risk call option on Derive is not doing the same thing as a HYPE holder selling covered calls through Rysk. A user betting on an election result or BTC price on Polymarket is also not using the same product as a trader pricing skew on Paradex.
That distinction matters, the article says, because each platform has a different value proposition and a different ideal user.
Exchange-style options platforms are moving from passive pools to CLOB and RFQ
Castle Labs describes this segment as the closest thing to a traditional options exchange in crypto. Users can go to these platforms and build strategies that match a specific exposure or view.
The article says the earliest versions of these platforms relied on passive liquidity pools, which sharply limited liquidity and scalability. That model has been giving way to hybrids that combine central limit order books, or CLOBs, with request-for-quote, or RFQ, systems. Professional market makers are then used to deepen liquidity around the book. Derive and Paradex are presented as the clearest examples.
According to Castle Labs, this category is best suited to professional traders, market makers, funds and active retail options traders. It is much less suitable for passive holders, users who want one-click yield, or prediction-market users looking for a simple yes-or-no outcome.
Derive is keeping the complexity and aiming at mature volatility traders
Derive grew out of Lyra, one of the earlier options AMMs. It now runs its own layer 2 on the Optimism stack and offers cross-margined options through an order-book interface.
The article says Derive’s volume has kept rising through the year and now accounts for more than 90% of total daily options volume. Castle Labs argues that the product’s interface makes its target user obvious. Rather than hiding options complexity, Derive is built for users who already understand volatility trading and want a complete set of tools.
Users can select among assets, strikes and expiries, then combine them into custom payoff structures. The report highlights a recent ETH call spread shared by KoolAid. In that trade, the user bought an ETH call option with a $5,000 strike expiring on March 26, 2027, while selling an ETH call at a $7,000 strike with the same expiry.
Castle Labs explains that the structure amounts to a bet that ETH will end up between $5,000 and $7,000 by March 2027. The post drew attention not only because of the target range, but because of the payoff. The user put in $300,000, and if ETH reaches $7,000 the maximum gain would be $20 million. The position only makes money if ETH settles between $5,000 and $7,000. Otherwise, the options expire worthless.
For Castle Labs, that example shows what Derive is good at: giving experienced users a plug-and-play interface for building their own options strategies. The article also says Derive recently posted its largest single trading day on record, with more than $500 million in notional volume.
On the backend, Derive uses an off-chain matching engine for fast execution and settles trades on its layer 2. The report says a V3 upgrade is coming soon and will turn the protocol into a zkVM on Ethereum mainnet. Castle Labs lists the expected additions as higher throughput, cross-margining across portfolios, faster listings for RWA and long-tail assets, native vaults and the ability to borrow additional non-cash assets.
The article says those changes should let Derive offer options on a wider range of assets, including stocks and commodities. Native vaults would also allow third parties to create strategies and act as curators of structured products. Castle Labs concludes that Derive is best suited to users who want full options infrastructure rather than simplified yield packaging.
Paradex is using Paradigm RFQ to address execution and liquidity constraints
Paradex announced in mid-September 2026 that its options platform had integrated Paradigm RFQ. Castle Labs says users can build anything from simple trades to multi-leg structures, send those requests into RFQ, and receive the best quotes from market makers and institutional liquidity providers.
The article argues that this matters because RFQ removes the need to bootstrap deep order-book liquidity for every individual market, which was one of the biggest hurdles for early options protocols. It describes Paradigm RFQ as one of the deepest and most liquid implementations in the industry and says it has processed more than $1 trillion in volume since its launch in 2019.
Castle Labs adds a comparison point: Deribit uses the same RFQ system, and it accounts for an average of 27% of Deribit’s volume.
The report says options venues want to support as many assets and payoff structures as possible, but execution quality and liquidity are often the real bottlenecks. As more platforms lean on RFQ systems fed by professional market makers instead of relying only on order-book liquidity, Castle Labs sees Paradex as moving into the ranks of mature options exchanges.
Binary options and prediction markets are simpler, but less expressive
Castle Labs says prediction markets are structurally identical to binary options. Each contract pays a fixed amount at expiry if a condition is met. If it is not met, the contract goes to zero.
The article goes a step further and says that, in payoff terms, prediction markets can be understood through regular call-option pricing and look similar to a call spread, meaning a combination of two call options on the same underlying. In that setup, a user buys a lower-strike call and sells a higher-strike call, creating a three-zone payoff profile:
- below the lower strike, both calls expire worthless and the payoff is $0;
- above the higher strike, the payoff equals the difference between the two strikes;
- between the two strikes, the payoff rises linearly from $0 up to that strike difference.
Polymarket has strong short-dated adoption, but market resolution remains a weak point
The article says Polymarket uses the Gnosis Conditional Tokens Framework to create outcome contracts and aggregates liquidity through a hybrid CLOB model with market-maker incentives.
Castle Labs identifies market resolution as one of the platform’s most criticized features. In crypto markets, the report says, Polymarket relies on UMA’s Optimistic Oracle and Chainlink, and it notes that UMA has faced issues before, including a market over whether Volodymyr Zelensky had worn a suit.
Castle Labs says these markets are simpler than standard options and, because the expiries are extremely short, they can attract more retail users while complementing sports and political markets already active on the platform. Polymarket currently offers 5-minute, 15-minute and 1-hour markets where users can go long or short. Because they have fixed expiry and binary payouts, the article classifies them as binary options.
The limitation is that the final payout depends on the asset price at the market’s final timestamp, above or below the strike, which means traders cannot express a view on a price range the way they can in a standard options market. Even so, Castle Labs says adoption has been rising quickly, and it argues that options protocols can learn from prediction markets by collapsing complex outcomes into a simple yes-or-no decision instead of a multi-outcome event.
Kalshi’s main differentiator is regulation for U.S. users
Castle Labs says Kalshi’s key distinction is that it offers a regulated product to U.S. customers.
Kalshi also offers 5-minute and 15-minute markets. Its liquidity model is described as similar to Polymarket’s, sitting between a CLOB and institutional market making. The article says both Polymarket and Kalshi currently use permissioned market creation.
That could change over time. Castle Labs says a permissionless user-generated market flow could eventually become another form of binary option, letting users define their own strike and expiry and use those contracts as hedging tools. The report says this could be reinforced as prediction-market platforms expand the range of asset classes they support.
HIP-4 brings outcome markets into the Hyperliquid trading environment
Castle Labs says HIP-4 takes a somewhat different route from Polymarket and Kalshi. In its words, HIP-4 is better understood as a framework for decentralized outcome markets.
The article says HIP-4 launched in May 2026 as Hyperliquid’s implementation of outcome markets, supporting binary market outcomes, and that permissionless deployment opened at the end of August. The launch process follows a path similar to HIP-3: users must stake more than 500,000 HYPE to create a market and can earn 50% of fees.
Castle Labs says HIP-4 is not limited to prediction markets and also includes scalable options primitives. Its main unlock, in the article’s view, is that it places everything in the same environment, making strategies possible on a single venue that previously required two or three different platforms.
The piece gives one example. A user might want to vote on a rate-cut event while hedging that view with a BTC short using the same collateral. Before HIP-4, that could require moving between several venues. With HIP-4, the article says, the whole trade can be done in one place.
The rollout has been staged. So far, cumulative volume has exceeded $150 million, according to Castle Labs.
The report also looks back at the early phase. HIP-4 first launched with 15-minute BTC markets, then cooled off for a time. Activity picked up again only after permissionless deployment opened, when Outcome and Tradexyz launched their own markets. Castle Labs says that matters for builders too and calls it another example of Hyperliquid’s infrastructure-first approach. Shared accounts, shared collateral and full composability expand what can be built on top of the platform.
The article says HIP-4 is best suited to active Hyperliquid traders, retail traders and more experienced users who want to combine outcome markets with spot or perpetual positions for hedging or more complex strategies. It is less suitable for passive holders looking for one-click yield, or for users who only want a prediction-market product.
Structured products and yield strategies are narrower tools with a specific job
Castle Labs draws a line between these protocols and general-purpose options venues. Structured products and yield strategies are designed around a narrower slice of the options stack and a more specific use case. For that reason, the article says, they are less general than exchange-style platforms.
Rysk turns options into a yield tool for HYPE holders
Castle Labs says Rysk makes options look less like a professional volatility screen and more like a yield product for HYPE holders. The protocol first launched on Arbitrum and later found product-market fit on Hyperliquid.
The pitch is simple. Rysk lets users earn yield in advance on their spot holdings. The article says many HYPE holders are sitting on large unrealized gains, and Rysk gives them a way to hedge that spot exposure and generate income through covered calls.
That is a direct difference from the other platforms in the report. Rysk is not a general venue for retail users searching across many options strategies. It serves a narrower need: users who already have spot exposure and want to hedge and earn yield with covered calls.
Castle Labs compares the product to Pendle in one respect. It says Rysk turns a non-yielding spot asset into an income-generating product. In the article’s framing, it takes what Pendle calls YT and distributes that yield to users until the covered call expires.
Users can choose the asset and expiry and see the expected yield clearly before entering. They receive upfront income shown as APR. At expiry, if the asset price stays below the strike, they get their spot asset back plus the yield. If HYPE rises above $106, they receive the equivalent amount in USDC plus the yield generated in advance.
Because of that design, Castle Labs says Rysk should not be judged primarily by options premium trading volume. TVL is the better metric, since it reflects the amount of assets locked in covered-call strategies to generate yield. The article says Rysk recently reached $50 million in TVL, which it describes as particularly notable because monthly expiries reset TVL back to zero.
Castle Labs says the protocol is best suited to institutions, large treasuries, directional funds, market makers, retail whales and passive holders who want one-click yield. It is less suited to retail options traders, neutral funds and prediction-market users looking for a simple yes-or-no payoff.
Hegic sits between Rysk and Derive with a simplified one-click experience
Hegic is described as an on-chain options protocol that reduces options trading to a one-click experience while letting users stay anonymous. The article says Hegic currently limits listed assets to ETH and BTC, but unlike Rysk, it still lets users express a wider range of views through strategies such as spreads.
Castle Labs notes that Hegic offers American-style options, which means users can exercise at any time before expiry. The protocol also has a native token that users can deposit into its proprietary Stake&Cover, or S&C, liquidity model.
On the liquidity side, Hegic uses a peer-to-pool structure. In that model, Hegic tokens in the pool collateralize options and strategies. The article compares it to early GLP, where a vault acts as the trader’s counterparty on the platform. As a result, S&C depositors receive 100% of the protocol’s accumulated P&L.
Castle Labs says Hegic is broader than Rysk, even though it focuses on a smaller asset set. It offers a wider strategy menu than covered calls. In the report’s framing, Hegic sits between Rysk and Derive: fewer assets than a full exchange, but a broad set of strategies that have been abstracted and simplified for easier use.
The article says it is best suited to large BTC and ETH holders, large treasuries, market makers and DeFi whales. It is less suited to retail options traders, market-neutral funds and users who want a simple yes-or-no prediction-market payoff.
Castle Labs says “options” is not one market but several markets evolving in parallel
The article closes by arguing that options is a broad label and that builders across crypto are working on different layers of the stack rather than building the same thing.
Castle Labs says Derive is posting strong growth as the main venue for quantitative options trading. Paradex, after adding Paradigm RFQ, should see quote quality and volume improve, in the firm’s view. Platforms such as Polymarket and Kalshi are still working out how to expand their addressable market by bringing in more options traders and tying that activity into their existing event markets.
The article also says the HIP-4 event launched by Tradexyz will be a decisive moment in showing whether Hyperliquid is the right place for event markets and whether it can eventually compete with larger binary-options platforms such as Polymarket and Kalshi.
Castle Labs is especially interested in products like Rysk, which it says have a clear niche and a clear utility. Expanding to more assets and chains is only the first step in enlarging the user base. Broadening the range of complex strategies is another priority, because the product still mainly serves users who already hold spot positions and are experienced enough to hedge or generate yield from them.
The piece ends on a broader sector view. Castle Labs says the options segment still looks early, that market conditions are lining up, and that user profiles are evolving with them. Institutional investors will become more sophisticated, the article says, and the recent popularity of posts showing outsized payouts could eventually push more retail users to try these products.
