Original Title: "Navigating DeFi Option"

Original Author: 0x Keyu

Original Translation: Shenzhen TechFlow

Main Points

Considering that the monthly trading volume of the overall cryptocurrency options is only US$40 billion (the monthly trading volume of perpetual contracts is about US$1.6 trillion), it may take another market cycle for on-chain options to achieve significant development.

Currently, the futures trading volume of decentralized exchanges (DEX) and centralized exchanges (CEX) is only about 2%. If we assume that the market share of on-chain options will reach a similar level as on-chain perpetual contracts, the expected trading volume of on-chain options should be about 800 million US dollars, which may not be enough to incentivize market makers to provide liquidity on-chain.

Utilizing derivative products with centralized liquidity provider (LP) positions may be a way to address liquidity issues, as there is already a large amount of liquidity existing in centralized liquidity pools. These products (Infinity Pool, Panoptic, Itos, Smilee, etc.) also provide significant product differentiation for CEX, such as the ability to speculate on long-tail assets. The main challenge for them is to convince centralized liquidity providers to deposit their liquidity into their pools.

In the long run, I am optimistic about on-chain options protocols based on CLOB (Central Limit Order Book), which first utilize off-chain order book matching and on-chain settlement similar to dYdX, and then gradually move the order book on-chain as the underlying blockchain infrastructure improves.

Options Market Overview

In traditional finance, options are the most commonly used financial instruments. According to FIA data, the global exchange-traded derivatives (ETD) trading volume (measured by the number of futures and options contracts traded and/or settled) increased by 34% compared to last year, reaching an astonishing $83.8 billion in 2022. The trading volume of options ($54.5 billion) and futures ($29.3 billion) accounted for 65% and 35% of the total trading volume, respectively, and increased by 63.7% and 0.1% compared to last year.

Contrary to the futures trading volume being greater than the options trading volume in TradFi, the current trading volume of crypto derivatives is primarily dominated by futures: In June 2023, the trading volume of ETH and BTC options (around $30 billion) only accounted for 2% of the futures trading volume (around $1 trillion). This may be due to Bitmex introducing perpetual futures, which has concentrated liquidity on expiry dates, resulting in higher capital efficiency compared to traditional financial futures. Higher capital efficiency translates into better futures pricing, lower slippage, and higher leverage opportunities, catering to the risk preferences of crypto traders. For emerging asset classes like cryptocurrencies, capital efficiency is crucial because their liquidity is not as high as traditional stock markets. In the case of crypto options, the dispersed liquidity between expiry dates and strike prices has led to significantly lower trading volumes compared to crypto futures.

On-chain options only account for a small portion of decentralized financial derivatives trading volume and total value locked (TVL).

Currently, the total TVL of on-chain derivatives protocols is approximately $1.5 billion, while the TVL of on-chain options protocols is only about $110 million, indicating a huge untapped market. In terms of trading volume, DEX options trading only represents a nominal trading volume of $114 million (premium of $3.7 million) out of the monthly derivatives trading volume of around $30 billion. This means that the on-chain options market is still in its early stages and has significant market potential.

Differences between Futures and Options

Due to options granting rights instead of obligations, the buyer of an option needs to pay a premium upfront to afford this flexibility when exercising their position at the strike price. This mechanism gives the option buyer a convex payoff function (fixed downside risk and unlimited upside potential), while the option seller has a concave payoff function (fixed upside potential and unlimited downside risk). In contrast to the asymmetric payoff functions of option buyers and sellers, futures buyers and sellers have symmetric payoff functions. Different payoff functions lead to different user investment profiles and use cases:

Individual User Investments:

• Options: Higher entry barrier; Buyers: Long Gamma, Short Theta; Sellers: Short Gamma, Long Theta.

• Futures: Lower entry barrier (especially in perpetual futures, which eliminate physical delivery and expiry dates), suitable for typical cryptocurrency users demanding high leverage.

Use Cases:

• Options: Income generation (sellers), hedging, speculation, volatility risk exposure;

• Commodities Futures: Hedging, speculation, high leverage (for perpetual futures).

Competitive landscape

According to its mechanism, options agreements can be roughly divided into five categories: structured products, options infrastructure, automated market makers (AMM), central limit order books (CLOB), and protocols utilizing centralized liquidity pools.

Structured products aim to generate returns for LP (liquidity providers) through various strategies and typically rely on options infrastructure for on-chain options minting/settlement. Currently, structured product offerings include options treasuries that offer covered call/put positions (e.g., Ribbon, Psyoption) and more complex yield products (e.g., Cega), which include exotic options as well as opportunities for collateralization, borrowing and liquidity provision. Options infrastructure is essentially smart financial custodian protocols that allow users to create, mint, and settle various financial derivatives. Options AMMs take on the opposite side of traders by utilizing liquidity pools. They programmatically price options based on a modified Black-Scholes formula and the supply-demand situation of options contracts. Options CLOBs are markets that actively match buying and selling orders of options buyers and sellers. Protocols utilizing centralized liquidity pools are protocols that create derivative primitives by taking the opposite side with centralized liquidity pool providers.

Currently, AMM-driven options exchanges dominate the majority of the options trading market share, similar to early on-chain perpetual models. This is because AMM-driven options exchanges are generally easier to establish and launch their liquidity compared to CLOB-based models. A fully on-chain CLOB options exchange requires a matching engine that can scan the order book and match orders quickly, which is challenging to achieve on the blockchain and requires substantial business development resources to find centralized market makers to bootstrap initial liquidity.

Therefore, most order books only support option minting and settlement, without supporting trading. Trading is usually conducted through over-the-counter transactions with market makers. Zeta solves this problem by combining its order book infrastructure with Serum, checking its order book up to twice per second. However, due to the FTX incident causing a loss of liquidity for Serum, all users and trading volume migrated to the community-forked Openbook, resulting in a temporary suspension of option trading for Zeta. Another possible solution could be a hybrid model of off-chain order books and on-chain settlement, which is exactly what Aevo is currently building. However, due to the market being in its early stages, the current market share is not indicative. With the emergence of more CLOB-based models and further innovation in AMM pools in terms of delta neutrality and capital efficiency, the market landscape will undergo significant changes. Next, I will outline the history of option protocols, identify some key challenges they face, and the improvements made.

The first wave, obstacles faced by AMMs

Option mispricing: The problem with AMM spot pools is that they require accurate pricing of options. In traditional markets, the price of options is determined by the supply and demand of options. However, in spot pool models, options have no supply and demand, as the supply is fixed (acting as automatic counterparty to option buyers). Therefore, the main problem faced by AMM liquidity pool models is how to derive an effective pricing model to price options.

The most common method of pricing is using the Black-Scholes formula, which takes into account the asset price, strike price, risk-free interest rate, time to expiration, and implied volatility. Among these five factors, the only unobservable factor is implied volatility (IV), which is an indicator reflecting the market's expectation of future volatility for a specific security. Traditionally, IV can be calculated based on the supply and demand of option contracts: high demand and low supply result in high IV, and vice versa. However, IV is challenging for on-chain option protocols due to the inconsistency of demand and supply flows. For example, in Hegic, its IV is calculated off-chain and manually updated on-chain on a weekly basis. This means that whether you open a $100 or $10,000 at-the-money option, the pricing will be the same. This is important because unlike market makers in CLOB-based systems who can dynamically reprice quotes based on new information about the true price, AMM LPs rely solely on the pricing function embedded in the smart contract. Therefore, during market volatility, LPs may face significant impermanent loss, where IV could be significantly lower than realized volatility. As shown in the diagram below, most first-generation option AMMs have static volatility inputs and do not adjust dynamically based on real-time trading volume.

Unhedged LP position: In the first wave of AMM, protocols like Hegic, Dopex, and Premia provided a hostile experience for LPs because their collateral was not hedged. This mechanism was driven by concerns about capital efficiency, as delta hedging typically requires taking liquidity from the pool to engage in long/short trades based on the net delta of their short position. However, this resulted in LPs accumulating a high exposure to underlying assets. As a result, it increased the liquidity cost for LPs and hindered the growth of the liquidity pool.

From the trader's perspective, their choices in asset selection, execution price, and expiration date are limited compared to CEX, and pricing efficiency is low. The limited choices in asset selection, execution price, and expiration date stem from limited liquidity, as broader choices would further scatter the already scarce liquidity in the pool.

Improvements made by the second wave of AMM

Emphasizing delta neutrality to attract liquidity: Lyra's Valon update introduced delta hedging in AMM for the first time. Lyra reduces risk by taking hedging positions through GMX or Synthetix. For example, when a trader holds a long position in ETH call options, Lyra’s Market Making Vaults (MMV) takes an equivalent long position in ETH based on the net delta when the position is established. By doing so, MMV protects itself from the potential unlimited profits that traders holding long call options may make when the ETH price rises.

Similarly, Siren Flow also introduced a delta hedging system by partnering with Prennial to hedge delta exposure and provide liquidity. Meanwhile, other AMM liquidity pool protocols also attempt to address LP issues in different ways: Hegic, Premia, and Dopex divide the liquidity pool into bullish and bearish vaults, giving LPs more control over the options they underwrite. However, this approach is not as effective as the methods used by Lyra and Siren because it still transfers the burden of hedging to LPs and disperses liquidity.

Improving inefficient pricing models and introducing partial collateral for a better user experience: Protocols like Lyra, Dopex, and Siren Flow implemented new strategies to improve pricing efficiency compared to Hegic's static IV problem. The core mechanism of Lyra AMM adjusts the implied volatility (IV) and option cost based on market conditions. When there is high demand for options, the AMM increases the implied volatility, and when there is excess supply, it decreases the implied volatility. This approach allows the AMM to converge to the market clearing value of IV for each execution price and expiration date.

When listing on Lyra, the initial baseline volatility value IV and the ratio of listed strike volatility to IV are initialized. These initial values are derived from the IV at the 50 Delta (ATM) strike price based on current market data. After initialization, IV and strike volatility ratio (skew) are determined by the supply and demand of options for specific strike prices and their respective expiration dates. On the other hand, Siren Flow implements a hybrid on-chain/off-chain request-for-quote (RFQ) system to provide competitive options trading pricing. This innovation allows Siren Flow to offer pricing comparable to centralized exchange derivatives while retaining the benefits of self-custody and decentralized trading. However, Siren's approach determines that it cannot serve as a price discovery platform as it relies on data from centralized exchanges to derive its IV.

In addition, Lyra also employs an innovative approach to provide partial collateralization for option sellers, which increases capital efficiency by 4-5 times. Avalon allows traders to partially collateralize short positions, enabling them to sell 4-5 times the quantity of options with the same capital. Partially collateralized short trades are crucial for two reasons: it provides option traders with a more comprehensive experience comparable to CeFi platforms, and it allows AMMs to offer more efficient pricing. The main challenge in implementing partial collateralization lies in calculating the initial collateral amount based on various factors and establishing a robust risk management system to mitigate counterparty default risks.

CLOB: Limited current appeal, but promising products are being launched

Due to the FTX incident leading to the cessation of operations for Zeta's options exchange, Aevo is a promising CLOB-based options exchange incubated by Ribbon, adopting a hybrid model of off-chain matching and on-chain settlement. Aevo is built on a custom EVM-rollup, reportedly providing access to hundreds of instruments that can be traded across various strike prices and expiration dates with deep liquidity. To kickstart its liquidity, aside from partnering with professional market makers, Aevo also plans to integrate with Ribbon's DOV as the venue for settling sold options contracts. Currently, Ribbon has a monthly trading volume of approximately $30 million, laying the liquidity foundation for Aevo. Additionally, this also addresses the current issue of conflicting interests faced by DOV, which may bring more trading volume to Ribbon. DOV depositors can also profit or reduce losses before expiration, greatly increasing the flexibility of Ribbon DOV. Lastly, Aevo will create liquidity for market makers who currently purchase DOV options off-chain, allowing them to hedge positions directly on the exchange. Aevo also has the potential to generate synergies with other DOV protocols as an infrastructure layer.

Comparison between CLOB-based model and AMM liquidity pool model:

In general, when evaluating on-chain options exchanges, there are three most important indicators: liquidity, capital efficiency, and asset selection.

Liquidity: The AMM liquidity pool model has a clear advantage in attracting initial liquidity as it can easily attract retail liquidity to automatically provide liquidity. However, in the long run, the CLOB-based model has a higher upper limit as it allows professional market makers to provide liquidity on the platform. Additionally, protocols like Elixir also offer market-making capabilities for retail users on CLOB-based systems. Therefore, the CLOB-based model has an advantage in attracting liquidity compared to the AMM liquidity pool model.

Capital efficiency: CLOB and AMM both have different advantages. The CLOB-based model has the potential to act as a price discovery platform when the options trading volume is large enough, resulting in more efficient options pricing. It also does not face any impermanent loss issues like the AMM model. On the other hand, the AMM-based options protocol can improve capital efficiency by combining its LP positions.

Asset selection: I believe the CLOB-based model has a relative advantage as it is similar to centralized exchanges and potentially allows more assets to be listed on the platform. On the other hand, the AMM liquidity model would struggle to achieve a broader range of assets as pricing these out-of-the-money options would be challenging. Additionally, traders can deplete the liquidity pool funds by purchasing long straddle options and earning significant wealth.

For structured products, there are two issues: misaligned incentives and risk mismatch.

Incentive misalignment between market makers and structured product vaults: Currently, due to the limited adoption of on-chain options exchanges, most options vaults construct covered call strategies using LP deposits and then sell them to market makers through auctions. Market makers typically buy call options from Ribbon's depositors at a negotiated premium price and hedge themselves by selling an equal quantity of call options with the same specifications (expiry date, strike price). By doing so, they make a profit from the price difference between what they purchased from Ribbon and what they sell on Deribit. The problem is that market makers and options vaults have conflicting interests as market makers want to buy call options as cheaply as possible, while depositors want to outperform the market within their risk range. However, since DOV can only sell options to a limited number of market makers, most options sold are undervalued, resulting in minimal profits for LPs.

Risk mismatch with current cryptocurrency user's portfolio: In reality, LPs in DOV are long theta and short gamma as most vaults cover call/put options. However, cryptocurrencies are inherently highly volatile, which makes LPs typically perform poorly in bull markets and suffer similar losses in bear markets. This return profile does not match the interests of the majority of cryptocurrency users who participate for the asymmetric upside potential rather than negligible gains.

Future: The maturity of CLOB-based options exchanges can solve the problem of conflicting interests between market makers and structured products. With the emergence of CLOB-based options exchanges like Aevo, this can provide a price discovery venue for DOV to find its sellers and resolve the imbalanced power dynamics in over-the-counter transactions. Furthermore, as mentioned above, DOV combined with the on-chain CLOB options exchange can allow depositors to profit or minimize losses before the expiration of DOV, providing depositors with greater flexibility.

In summary, current on-chain options protocols, whether based on CLOB or AMM models, have made little progress in terms of trading volume and liquidity. This raises a typical chicken and egg problem – without liquidity, there is no trading volume, and vice versa. From a liquidity perspective, on-chain liquidity providers face the issue of inconsistency with options pricing, resulting in realized volatility typically being higher than the calculated implied volatility, making them reluctant to participate in liquidity pools. For traditional market makers, they lack sufficient incentive to provide liquidity in these protocols due to negligible trading volume. From a long-term perspective, I am optimistic about the CLOB-based model occupying a market share similar to DyDx in terms of options trading volume. However, the current on-chain options protocols are insufficient in product differentiation compared to CEX, which leads us to the next wave of protocols that utilize centralized liquidity pools to leverage the substantial liquidity present in these pools.

Next Wave: Protocols Utilizing Centralized Liquidity Pools

The core idea behind these new wave protocols is that Uniswap v3 liquidity providers (LPs) positions can be viewed as tokenized short put options. This is because the economic return function of LPs is mathematically equivalent to selling put options. For Uniswap V3 LPs, they are essentially short gamma and long time value, as they suffer losses when the underlying asset price fluctuates rapidly and gain swap fees over time. Therefore, various protocols including Panoptic, Infinity Pool, Smilee, and Itos attempt to build their derivative primitives by utilizing a large number of short option positions in centralized liquidity pools. Though conceptually similar, these products have significant differences in terms of design and offerings.

Mechanism Overview

Panoptic

Overall, Panoptic consists of liquidity providers, traders (option buyers/sellers), and clearers. Liquidity providers need to deposit interchangeable tokens in the Panoptic pool in any ratio of token 1/token 0. Option sellers can borrow this liquidity to create short options by depositing liquidity into the corresponding Uniswap v3 pool. Similarly, traders can create long options by withdrawing liquidity from the Uniswap v3 pool. For example, suppose a trader wants to buy a put option with a strike price of 1000 USDC and a width of 10%. When the trader purchases this option, the portion of liquidity in the Uniswap V3 pool between 909 and 1100 USDC per ETH range will be taken out and returned to the Panoptic pool. The cost of this option is the amount of fees that would have been generated if the liquidity were kept in the Uniswap pool. Now, let's consider different scenarios:

If the ETH price is higher than 1100 USDC when the option is purchased, the option is out of the money (OTM) - not yet profitable, and therefore, no premium will accumulate.

If the ETH price remains between 909 and 1100 USDC during the tenure of the option, the option remains out of the money, and its cost remains zero. Users can decide to close their option positions without paying any premium.

If the ETH price drops below 1100 USDC, the option starts to appreciate - the premium begins to accumulate. If the ETH price further drops below 909 USDC, the option is now in the money (ITM), meaning it is profitable. At this point, the option stops accruing fees, and users can decide to exercise the option.

When exercising the option, users must return the borrowed liquidity, which now exists in the form of ETH. Therefore, they send ETH back to the Panoptic pool and retain the USDC they initially received when purchasing the put option. This means that they effectively sold their ETH at a price of 1000 USDC, even if the market price may be below 909 USDC.

On any available token in the Panoptic liquidity pool, option sellers can obtain 5x leverage, and option buyers can obtain 20x leverage.

Itos

Unlike Panoptics, which relies on Uniswap V3 AMM to create long/short options, Itos creates similar products by establishing its own centralized liquidity AMM (CLMM) on top of the forward liquidity provision, utilizing negative liquidity provision positions (referred to as takers). The AMM structure has three market participants: maker, taker, and trader. In Uniswap V3, maker and trader are the same concept, where maker refers to the liquidity provider who deposits interchangeable token pairs into the liquidity pool, and trader refers to the participant who exchanges one token for another in the liquidity pool. Unlike makers who provide liquidity, takers retain liquidity for trading. These takers always trade with the exchange, paying fees to ensure the transactions enter more valuable tokens. This enables the creation of TakerPuts and TakerCalls, similar to put options and call options. For example, when trading occurs on a decentralized exchange (DEX), takers profit from participating in the trade while paying funding rates to ensure sufficient liquidity for executing trades. By always utilizing each trade, the value of their positions increases with price fluctuations, while the value of market-making positions tends to decrease. Therefore, this allows takers to experience a return function similar to a long option (shown as the green line in the figure below) and effectively hedge against any maker positions.

Infinity Pool

Similar to Itos, Infinity Pool is a leveraged decentralized exchange (DEX) built on its own CLMM (called float pool). LPs can directly deposit interchangeable token pairs or deposit their Uniswap V3 LP tokens into the float pool (the protocol will represent LPs by converting LP tokens into interchangeable tokens). The float pool has two functions: 1) spot trading and 2) lending to leveraged traders. In the second case, the borrowed LP tokens are extracted from the float pool to a private pool (swappers), where traders can perform unlimited free swaps at predetermined execution prices. In return, traders pay funding fees to LPs through fixed-term loans (1 to 40x leverage) or perpetual loans (above 40x leverage). Furthermore, traders can convert the borrowed LP tokens at the desired price range in the float pool into interchangeable tokens in the private pool, and any subsequent external swap if needed, to repay the loan in any token combination within the float pool, thus achieving a leveraged trading experience on any available asset in the float pool. For example, in the example outlined in their whitepaper, traders can borrow ETH/USDC LP tokens worth 1000 USDC, take a long position on ETH with 10x leverage within a tight liquidity range centered around 900 USDC, assuming the market price of ETH is 1000 USDC. Since the price of ETH is currently above the liquidity range, traders can exchange the LP tokens for 1 ETH at 1000 USDC and exchange the 1000 USDC for 1 ETH on any spot DEX (assuming no fees or slippage). If ETH drops below 900 USDC, the LP position will contain 1.11 (1000/900) ETH, and if the decision is made to close the position, LPs will need to buy an additional 0.11 ETH. Therefore, in the worst case scenario for the trader, when ETH is at 900 USDC, the trader would need to purchase 0.11 ETH with 99 USDC. This means that in this example, the trader only needs to provide 100 USDC as collateral to achieve 10x leverage.

Infinity Pool AMMs provide theoretically infinite leverage perpetual options for any listed token, with a user experience similar to perpetual contracts.

Key Takeaways

Fundamentally, the main challenge faced by the above protocol is convincing LPs from centralized liquidity pools to redeploy to their protocol.

In these three protocols, compared to Infinity Pool and Itos, Panoptic does not require to bootstrap as much liquidity because their option products are created through interaction with Uniswap V3 pools. However, Panoptic still needs sufficient liquidity for option buyers and sellers to borrow for leverage trading. Therefore, these three protocols implement different fee models to compensate LPs. In Panoptic, LPs earn commission rates of 20-60 basis points based on pool utilization. In Infinity Pool, LPs earn swap fees when not lent out and receive interest payments on loans based on traders' leverage levels. In Itos, LPs (makers) earn borrowing rates (holding costs) based on liquidity utilization within the active range retained by takers. In theory, all of these protocols should offer quite competitive rates compared to the fixed swap rates provided by the Uniswap V3 AMM, taking into account the expected volatility of their quoted ranges.

The most exciting aspect of these derivatives is that they provide speculative exposure to small-cap tokens, which is impossible in any other DEX or CEX. On the other hand, due to relatively high fees, these protocols may struggle to compete with CEXs for major asset pairs. This is particularly evident for Panoptic.

For any options created in Panoptic, users need to pay at least 30 basis points (20 basis points commission rate and 10 basis points Uniswap V3 swap fee), which is even higher than many other derivative DEXs, not to mention CEXs.

For Infinity Pool, they still require interaction with external AMMs to successfully close positions, resulting in relatively high costs. On the other hand, Itos' pricing may be more competitive compared to the other two protocols, as they do not require interaction with any external AMMs and adjust swap fees based on volatility. That being said, the differing offerings of these protocols mean they can attract users with different risk profiles: Infinity Pool's theoretically unlimited leverage may be more suitable for perpetual futures traders, while Panoptic and Itos may be better suited for more sophisticated retail traders and DAOs needing direct on-chain hedging.

Conclusion and Reflections

Overall, this wave of new derivatives unlocks liquidity in concentrated liquidity pools by finding similar payment functions between Uni V3 LP positions and option sellers. In short, the goal of all these protocols is to be the counterparty to impermanent loss and address the liquidity problem hindering options market growth. Additionally, if they can achieve a certain scale of liquidity, I am also looking forward to seeing various structured products built on top of these protocols.