Decentralized Options Trading Protocols #101

Since 2nd half of year 2020, even with whole world was suffered from Covid-19, blockchain geeks and finance pioneers worked together setting off a storm of DeFi into both blockchain and finance industry. One of the main methodologies of DeFi is staking, which has drawn loads of crypto assets from CEX into various types of DeFi vaults.

As Uniswap's trading volume surpassed Coinbase in September last year, taking the fourth place, it officially announced the complete start of the era of DEX. However, the market size of derivatives, whether in traditional finance or CEX of crypto market, is more than a few times that of spot trading volumes. In contrast, the market size for derivatives is still far less than spot trading in DeFi (see the figure below), so there is inevitable gigantic room for growth.

Among all derivatives, options should be the brightest rising star, based on two main reasons and potentials:

1, Currently, major active crypto investors or traders are more like speculators, who are looking for high leverages and high returns regardless high risks.

2, The dramatic fluctuations of cryptocurrencies, or professionally put, implied volatility are quite high. That means protections or hedging are rigid demand for any crypto players in all categories: miners, whale holders, traders in either professional or amateur level.

Options, by its nature, perfectly fits the above market demands, which giving one high leverage with small input on speculating and hedging purposes. With Wall Street mainstream stepping in since the beginning of 2021, the variants and combinations of options would be fully exploited by professionals, which will make options market more interesting and diversified.

Therefore, just like the competition between China and the United States over the 5G network standard protocol, various protocols for realization of decentralized options trading are entering the arena in light speed, targeting to become the next unicorn to dominate the market. So, if you are a blockchain enthusiast, this article will be your “first class” about the must-known information of on-going decentralized options trading protocols.

Basic Concept1: What is AMM - Automated Market Maker? Why is it playing such important role for DeFi and Decentralized Options Protocols?

Firstly, traditional market making, is a strategy in the secondary market to accelerate the volume (inject liquidity) in the market by providing a large number of two-way counterparty transactions, and to achieve benefits through spreads. Writer once worked in the world's largest market maker: Knight Capital in the risk control department, knowing that there two most important considerations for market making:

1. Asset Inventory/Capital Efficiency: in many cases, due to the absorption of liquidity, assets will stay in the positions. How to maximize asset utilization would be always the key for any MM.

2. Sharpe Ratio: relationship between the volatility of the underlying asset and the change in the rate of return. The higher the stability, the higher the return.

While Automated Market Making, paired with On-Chain Shared Liquidity Asset Pool are invented by spot trading Dex like Uniswap’s XYK model, which uses a fixed formula to automatic defining the prices for exchanging between a currency pair. The attractiveness of it was to solve the liquidity issue of peer-to-peer trading in decentralized environment without the centralized order matching machine in CEX. Of course, we then saw with liquidity mining incentives, AMM became the igniter of DeFi rocket lifting off last year. Naturally, like XYK model and other pricing
algorithms of major spot trading DEX were recognized as standard AMM protocols.

However, isn’t there any shortcomings of XYK model and its similarities of AMM formula? Definitely there is: entering the infamous Impermanent Loss (IL). We will delve into the IL in the later sections with examples.

Now, for options, any protocol inventor must build AMM mechanism accordingly for option trading, which in both traditional and centralized exchange, are quite different than spot trading, as it is a SCATTERED market.

Take Deribit for example, the first and biggest option centralized exchange in the world:

Liquidity diluted by different expirations, strike prices and ask prices with size by various options writers/sellers. Each one represents a standalone market and proper pricing relies on professional option sellers. Scattered liquidity and lack of professionalism are main two obstacles for options not even popularized in centralized exchanges.

When in DeFi, more challenges are to be cracked. Creating a liquid
market for all trading pairs on chain would be extremely expensive and capital-intensive. Each trading pairs has its own lifespan to be removed after expiration, which is also costly for the platform. The sellers’ obligation and buyers’ right are unbalanced, adding operational complication as well.

So how to generate the AMM mechanism and liquidity providing system specifically for options are the problems to be tackled for the decentralized option protocol creators.

Basic Concept2: Forerunners of decentralized options protocol separate them into three paths: Order-Book Matching, Using Standardized Spot Trading AMM, Shared Collateral Pool of Automated Market Makers. Writer myself is a believer of Collateral Pool approached. So, I will simply introduce the first two categories and dive deep on the 3rd.

Opium is well-known for its border dreams for solution not only for options but also for all kinds of derivatives. They are implementing a off-chain + on-chain order book matching system. They’ve even created their own Opium token standard called ERC-721o, for their platform token to be traded. Their framework is illustrated as following:

Replayer are external actors who match orders of users off-chain and broadcast to blockchain. Match contract would verify the result of Replayer and create positions. Core is consisted with multiple contracts to generate financial contracts based on derivative recipe and oracle recipe. Minter is to create long/short positions upon ERC-721o for trading.

Interesting highlights of Opium are including options on gas fee and some tryouts on exotic options. They also launched sub-system for the first DeFi CDS (credit default swap) and IRS (interest rate swap) tradable products.

In short, Opium is using partial off-chain matching mechanism and oracle verification for simulating the complex option trading logics.

There are also other projects are trying out on relatively “traditional” order-book approaches, while Opium probably is the shining jewel on the crown, so I will leave this section with overview of only Opium.

Concerns about decentralized protocols of order-book matching for option trading:

Off-chain order book solution is not invented by Opium, there are other DEX unlimited to trade options adopting such mechanism: market makers broadcast an order off-chain to be picked up by a counterparty who then passes the full order to a smart contract for fulfillment. Nevertheless, with fast evolution of DeFi, off-chain market maker direction seems out-of-dated due to success of AMM by Uniswap. Especially for scattered option market, order book revision does not in the right direction of solving the real pain point.

Opyn represents the type of protocol of mapping option contracts into different ERC20 tokens like "ETH-USDC-100 Put" which can be traded in Uniswap, naturally taking advantage of AMM mechanism of spot trading. 

OpynV1 provided physical settled American options on Convexity Protocol while V2 added cash settled European auto-exercise options on Gamma protocol. V2 promote the capital efficiency mentioned in the market maker section.

Opyn as well have mint smart contract for generating tokenized options. Especially, they provide 8 parameters for sellers as each option instance: (1) time of expiry, (2) underlying asset, (3) strike price, (4) strike asset, (5) call or put, (6) type of collateral, (7) margin requirement of collateral, and finally (8) whether the option is either American or European.

What Opyn have moved forward to DeFi-ish from Opium would be utilizing standardized AMM, instead of focusing on order-book matching infrastructure. Writer thinks it would be a smart move, since Uniswap have become the largest hub of DeFi ecosystem. And AMM is more well-known and straightforward for entry-level users.

Concerns on Opyn and similar protocols using Uniswap AMM and its similarities mechanism:

Firstly, we should admit the simplification that Opyn have made, and using AMM is a huge step forward to DeFi approach. But still, the model still heavily relies on the liquidity provider and their professionalism, as the oToken are generated by them, and also, in scattered way. And what we have learned from even CEX, the lack of option writer/liquidity provider would be the major roadblock for option market on cryptos.

Secondly, it is the time we talk about the IL: impermanent loss on XYK-like AMM model.

IL happens for any LP(liquidity provider) when prices of deposited asset changed and the loss ratio is the not neglectable most of time. The chart below is a simulation of joining in the ETH/USDC pool on Uniswap on Nov. 27, 2019 and staying for a year.

As we can see more than 50% of time it is more profitable for one holding the asset than staking in the pool.

Given such situation, and oToken heavy reliance on Uniswap, and with the time decay nature of any option contract, oToken writers are essentially forced to sell to harvest premiums, rather than provide long-term liquidity, as the speed of decay accelerates as time moves closer to expiration. Therefore, LP providers are highly likely to suffer great IL in providing individual oToken liquidity. In high chance Opyn are taking the IL themselves for boosting liquidity of oTokens that are not deep enough in swap, which would limit the circulation of tokens and harm the growth of the entire system.

What we have seen from first two approaches reflects the process of gradual option protocol exploration, from solutions on order-book, to “borrow” imperfect but widely accepted AMM mechanism. However, in writer’s opinion, Shared Collateral Pool model is the real innovation and predictably the next generation of DeFi’s liquidity pool concept explicitly fitting into derivatives trading field.

The key progress is gathering scattered market maker structures into peer to pool trading, with the asset pool contributed from anyone, not limited to option professionals, act as collaterals for option sellers, also the option automated market makers.

There are three promising projects to be introduced to use this model: Hegic, FinNexus and Asteria.

As the most successful option protocol for now, from the TVL metric point of view, Hegic was rather the fastest growing option trading platform:

Within 6 months, Hegic successfully pulled off 2 version migration on peer-to-pool model, with a creative IBCO for public sale of the platform token.

Hegic’s edge would be also the advantages of the shared collateral pool, providing much convenience of both sellers and buyers.

Now options sellers could be anyone without even knowledge of options but looking for the fixed income profits of market maker/options seller whose role usually can only be executed by professional teams from
institutions. In this model, liquidity is shared collectively with zero slippage.

Options buyers now can choose any prices as strike prices instead of limited choices pre-defined by sellers.

Hegic V1 support DAI and ETH and single direction of physical settlement options, while its v888 changed to bidirectional cash settlement with ETH and wBTC pool. Their staking rewards are quite generous.

Beauty and Concerns of Hegic:

1, Their IBCO, even not the first project to apply it, had quite satisfied result: with 31,000 ETHs raised.

2, Previously single direction pool facing one side risk while physical settlement with100% staking would be guaranteed delivery. Bidirectional
pool in v888 reduces, to some extent, the directional risk/delta risk, but cash settlement mechanism based on non-stable currencies (for now is ETH and wBTC) could suffer huge risks of rapid value changing of collaterals. Check the put-call balance in each pool on March 10th 2021:

From above we can see the put/call are so unbalanced indicating same market expectation, so that huge loss could happen when market go for buyers favorable direction.

3, Hegic’s pricing model is the simplified version different from the classic, and the implied volatility (IV) is manually updated with reference to skew.com. Moreover, Hegic v888 charges an additional 1% fee on  options purchasing. Even though simplification is what make Hegic stand out, but not all elements should be shortened like pricing mechanism, the fundamental for option markets. From writer’s point of view, this could be the flaw instead of concern for the whole system.

4, Hegic options are non-transferrable and non-tradeable. Compared to tokenized option protocols, Hegic’s option are all in smart contract level only supported in their own system. Their resell mechanism to be applied could partially provide some “secondary” market liquidity, but meanwhile add more complexity for buyers.

5, Premiums are shared equally across the entire group of liquidity providers, so are the risks. Hegic, for now, does not implement any hedging mechanism which would be the major defect for any market making system of options. Without hedging, all liquidity providers actually are gambling on the opposite side to all buyers and exposed under asset price fluctuations.

FinNexus applied the same peer-to-pool shared collateral asset pool mode as Hegic, with bigger target. They have MASP stands for Multi-Asset-Single-Pool model, targeting for capability of generating options with any assets, not only limited to crypto currencies. The proposal is bold, even though for now they have only supported BTC, ETH, MKR, LINK, SNX options.

Looks to me, FNX have more complicated framework with professional team, compared to Hegic:

1, FNX implemented Black-Scholes model on options pricing, building
the solid foundation.

2, FNX have made some efforts on risk management:

2.1, Minimum Collateral Ratio (MCR), FNX assigned an arbitrary weight for each collateral asset for preventing undercollateralized situation.

2.2, Pricing Adjustment Coefficient, FNX play some mathematics tricks on option pricing based on the weight of demand in the pool, higher weight yields higher prices.

2.3, One-hour Chill Time, protection against flash loan attacks.

2.4, Moving Average IV and IV Surface Mapping, a theoretical calculation of IV (Implied Volatility)

Despite more comprehensive solution than Hegic, curiously, the current diluted market cap of FNX is floating at 67 mils with 8 mils TVL are far less than numbers of Hegic at 730 mils and 86 mils respectively.

Asteria is a brand-new decentralized peer-to-pool options trading
protocol, built with shared collateral pooled liquidity. Like Hegic and FNX, Asteria by nature inherited the beauty of pooled liquidity of accumulating liquidity from all market participants simultaneously and automatically. Asteria have made further improvements to solving the issues of Hegic and FNX on the infrastructure components level.

The main flaw of Hegic would be no protections on collateral pool, the most crucial component of the system. Asset loss or failure on the shared pool would cause irreparable liquidity exhaustion. FNX have provided some workarounds on this issue but they are only adding extra buffer before reaching risky situations.

Asteria introduces Delta Hedging mechanism which is widely used in professional traditional financial institutions. Delta hedging are actual trading volatility on either spot, perpetual swap or options. Asteria would implement the concept of DeFi aggregator to pursue the highest return from multiple DEX by algorithms.

Asteria would also stake collaterals into lending platforms like Compound for stable currency on hedging purpose. With the professional hedging mechanism, liquidity provides/options sellers/market makers could receive both premium from buyers also yield farming profits from lending protocols. 

As pointed in the Market Maker section, Capital Efficiency would play quite important role on profit rate. Even admitted that Hegic have drawn huge TVL, but besides the mining incentives, the actual income from option sellers would not support the theoretical return rate. The LP would not stick around after the most profitable mining period resulting in liquidity escape.

FNX choose to use coefficient to adjust pricing mechanism upon BS model calculation, in fact it is somehow a subjective approach for manipulating the market.

Asteria would then dynamically calculate the price and quantity with BS model based on the available stable currency exchanged from lending protocol. This approach assures both the fairness of the price and capital efficiency of the shared pool.

Asteria support customization of option contracts. With the variety of options, more buyers with multiple speculating and hedging purposes would be attracted into the platform, also the net position would dramatically lessen because of the type variety, which relieves pressure on hedging simultaneously.

Instead of tokenizing options with ERC20 token as Opyn, Asteria use NFT ERC1155 to map unbalanced option rights/obligation between buyer and seller. ERC 1155 protocol supports operations on unlimited numbers and types of non-fungible tokens within the same smart contract, such as token merge and conversion. With such functionality, structured options are possible to be synthetized by buyers. OTC of exotic options are quite large in traditional finance. There is no reason it would be different scenario if the infrastructure is established.

Another real problem happening on all decentralized options trading platform on Ethereum main net would be the rocketing gas fee. Practical experience on both Hegic and FNX reveals the awkward situation of gas fee equalizes the premium if the option size is small.

From Asteria’s whitepaper, the platform would embrace Layer2 in V2, which is quite worth looking forward.  

To sum up, Asteria to build thorough option protocol with state-of-the-art decentralized risk management. Following table listed the common risks of option market making and how would Asteria handle them.

Let’s wrap up the class 101. The potential of options market is massive. Talents and willingness to find out new solutions on decentralized option protocols are continuously increasing, and so is the money to do so. From this introductory article, we can see the multiple projects with difference on framework and development stages. It is very interesting to keep track of all those protocols and writer is very excited to be involved in such frontier exploration.

Class Dismissed!