The journey of DeFi’s poster boy from a stablecoin payment network to an API for liquidity
Chapter 1: The Pivot
Synthetix started in February 2018 with a $30M ICO. Only, it wasn’t called Synthetix then. It was Havven, a decentralized cryptocurrency payment network similar to MakerDAO(Havven’s stablecoins were called nomins). But as other stablecoins captured the market, this direction was rendered less valuable and the project rebranded to Synthetix in November 2018, nomins becoming synths.
Synthetix, the exchange launched in February 2019. At the time of release, it was a place to trade spot synths, with derivatives teased in the roadmap. Having realized from Havven the importance of a strong incentive to build a debt system based on a single token collateral, the rebrand introduced inflationary SNX as a reward to stakers. This was their way of saying: Stake, or get your share diluted.
Below is the proposed inflationary tokenomics in 2019. It has had several revisions since then, including the introduction of a target staking ratio that adjusted inflation dynamically.
Chapter 2: Atomic Swaps
Synthetix exchange did not have the best trading experience at the beginning. Trouble started at oracle latency. Whenever a trade occurred, there used to be a waiting period (10 mins) during which users could exchange, transfer, or burn the synth they traded into. This waiting period gave the oracles enough time to check the difference between the initial price and the new price. Fee reclamation was introduced to remove this delay, by calling settle function to the carryover from the previous trade.
This later gave way to Atomic Swaps. The idea here was to use a combination of Chainlink and DEX oracle (Uniswap V3) prices to deliver atomic transactions between synths. Right after 1inch integrated with Atomic Swaps, it proved to be a massive driver of volume. A few weeks after integrating, an average of $100-200 million in volume was being routed through the protocol daily, with about $300-600k in trading fees being returned directly to SNX stakers. It provided the best fills for 1inch users at the time and 1inch routed over 2x the volume of the entire first quarter of 2022 through Synthetix in Q2 2022, coming in at just over $2 billion.
To retain a competitive advantage and keep up with improving AMM DeXs, SIP-258 was introduced to allow the protocol to further lower fees in atomic swaps. But soon after, in mid-Q3 2022, it was discovered that traders had been successfully frontrunning the very low fees offered by atomic swaps, and volume from 1inch for the quarter dropped to $1.1 billion. Though through Q4 2022, 1inch routed $1.9 billion through Atomic Swaps, they were sunsetted towards the end of the quarter, sparking a wave of debates in the community and also between members of the Spartan Council.
The decision to sunset Atomic Swaps was motivated by toxic flow. It was discovered that the majority of volume coming through 1inch was toxic, which was detrimental to SNX stakers and increased the amount of debt in the protocol (similar to how impermanent loss works in a traditional liquidity pool). But there was a quite strong community sentiment that this decision was taken too soon when ETH prices were trending up, and if the team had held on, the value of debt would have come down and also driven more fees to SNX stakers.
Even though this was a major change to the protocol’s positioning, there wasn’t a typical forum discussion or a public statement as you usually see in DAOs. Right now, the core team members note that there was too much competing liquidity at lower prices for atomic swap liquidity to get used except in cases of extreme volatility, basically saying that Atomic Swaps took an L to the market.
Chapter 3: Perps
Having encountered problems around oracle latency on mainnet, Synthetix had decided to focus its innovative efforts to Optimism early on. Perps beta (V1) was launched in March 2022 on Optimism, allowing both leveraged longs and shorts on a large selection of assets. Kwenta, which spun out of Synthetix, was the first platform to integrate and is the leading integrator for Synthetix Perps by trading volume.
In 2 months, it collected $5.7M in fees and processed $2B volume.
https://twitter.com/synthetix_io/status/1527787081430224896?s=20
With a successful beta release, Perps V2 was launched in December 2022, with trading accessible through Kwenta and Decentrex. Its goal was to rectify the most notable limitations of V1, i.e., capital inefficiency (restrictive open interest limits), funding rate volatility, and high fees / wide spreads.
Perps V2 also introduced three new features:
New Off-chain Oracle System: an off-chain oracle system provided by Pyth Network, reducing perps trading fees to 5-10 basis points. There was some initial reluctance in the community, given the longstanding relationship with Chainlink.
Adjusted Funding Rate Mechanism: The funding rates in Perps V2 could continuously drift higher or lower based on the presence of long or short skew. This created a natural price discovery mechanism for funding rates.
Premium/Discount Function: A new premium or discount was be applied to market prices depending on the current skew of the market. This encouraged balanced trading and adjusted prices based on the imbalance of long and short positions.
Perps V2 processed 3 times more volume in the first 6 months compared to the V1. Kwenta dominated cumulative fee generation, contributing to more than 90% of the value.
With Atomic Swaps out of the picture, Synthetix needed to get its core offering, Perps right. While Q1 2023 showed promise that Perps could be the next breadwinner for Synthetix, Q2 cemented this thesis, as can be seen from the volumes below. Total quarterly volume crossed $15 billion, for the first time in the protocol’s history. Some of this volume was driven by OP trading rewards that launched in April but Perps V2 also came at a critical inflection point in the wider crypto sphere where users began to seriously search for alternatives to centralized offerings. Currently, there are over 42 assets with a Perps market enabled via Synthetix.
Decentralized derivatives have one of the fiercest competitions in web3 right now. Order book-based DYDX still occupies nearly half of the decentralized perp market trading volume. However, the liquidity pool DeX derivatives landscape has shifted quite dynamically in the last months.
It is impossible to analyze the volume breakup across these exchanges without taking into consideration the incentivization mechanism each employs. One can note how volume in protocols like Vela was driven by the initial airdrop hype, after which activity has almost died. Similarly, the surges of volume in Kwenta can be attributed to the trading incentive campaign upon launch and presently, OP token rewards. Though Synthetix isn’t a direct player here, it is helpful to note that there are levels to this game, and as it stands, dYdX > GMX >=Kwenta.
https://twitter.com/0x_____________/status/1680964672055328768?s=20
(Above is a tweet about Synthetix flipping GMX in open Interest briefly earlier this week)
With Synthetix gradually evolving from being a user-facing protocol to focusing on liquidity provisioning, it need not be thought of as directly involved in the DeX derivatives competition (even though at present, it is). For example, MUX, a derivatives exchange similar to Kwenta, decided to integrate with Synthetix rather than bootstrapping their own liquidity when they wanted to deploy on Optimism.
The value offered with Perps is a liquidity layer that can power an array of on-chain derivatives and financial instruments. SNX stakers in their unified collateral pool, serve as temporary counterparties as needed, and synthetic derivatives are minted against it. In the current positioning, Synthetix, the liquidity as a service provider, is a friend of all DEX derivative platforms trying to solve their liquidity bootstrapping problem as they expand to more chains.
The protocol recently announced that they have integrated Chainlink Cross-Chain Interoperability Protocol (CCIP) into Synthetix V3. This will help Synthetix offer shared collateral to all EVM chains.
Chapter 4: The (lack of a) DAO
The project began its operation as a not-for-profit foundation domiciled in Australia. In early 2020, the foundation was decommissioned, giving way to three distinct DAOs: SynthetixDAO (for controlling the treasury), the protocolDAO (for protocol upgrades), and the grantsDAO. SynthetixDAO was later split into two bodies: Treasury Council & Core Contributor Committee and a delegated governance council called “The Spartan Council” was also introduced.
This is its current governance structure:
Treasury Council(4 members): Manages the Synthetix Treasury and provides funding to other governance bodies, stipend payments to Core Contributors, and discretionary incentives.Spartan Council(8 members): Acts as the central governing body of the Synthetix protocol, voting on overall improvement proposals and parameter changes.Grants Council(5 members): Funds beneficial and high-quality public goods projects through grants, initiative bounties, or competition prizes.Ambassador Council(5 members): Actively promotes and advocates for the interests of Synthetix in the Ethereum and DeFi ecosystem. Also involved in governance at other DAOs, as a delegate.
Each council sits for an epoch, which is currently 4 months and is voted on again by SNX stakers through its governance app. Anyone interested in becoming a councilor can self-nominate themselves. Council members receive a monthly stipend of 2,000 SNX from the Synthetix Treasury.
In the current epoch, only 4 out of the 22 councilors are on their first term, others having been a councilor at least once previously. In the last 6 epochs (January 2022 to present), two contributors have been councilors for every epoch. Also, every council other than the Spartan Council has had four contributors who were a councilor for more than 4 epochs (out of the previous 6).
The mode of operation of Grants Council, in particular, is quite different from other DAOs. They also act as project managers, deeply involved in the implementation details of initiatives. The number of initiatives funded to date by the Grants Council is quite low. One could argue that a lot more infrastructure for and on top of Synthetix may get built if their focus shifts more towards identifying the right initiatives for Synthetix. Because the council is also heavily involved in implementation, it creates an unnecessary carryover that hampers its ability to rotate freely after each epoch.
The two bodies that are out of the structure defined above but still as important as the Spartan Council are the protocolDAO (or pDAO) and the Core Contributors Council (or CCC). Core Contributors of over 6 months get added as signers to pDAO. The CCC is the most powerful, yet faceless part of Synthetix.
There is a significant amount of information asymmetry within the organization. It started from the lack of a forum. For example, when Atomic Swaps, Synthetix’s breadwinner in 2022 disappeared completely in 2023, there wasn’t a public discussion or a data-rich rationale about why they were deprioritized. To investigate this rather significant positioning change, one has to dig through Discord chats of the last 12 months, which isn’t a great experience. The same goes for details about partnership/sponsorship deals, core contributor salaries, and consensus around proposals that get rejected/passed. The team will tell you that “most sip discussion is done verbally”, or in “governance or sip-forum channels in Discord”. The Discord forum started quite recently. I’m not sure what verbally means in a DAO.
In spite of this, it needs to be noted that with the sheer number of proposals that go to vote, most of them being technical, it would be impossible for a regular SNX staker to keep up or steer the protocol in the optimal direction. Coming up with a mechanism to improve the transparency of decision-making while maintaining the CCC for their expertise will greatly benefit Synthetix on its road to decentralization. As a part of V3, V3GM (Synthetix’s governance module) is being worked on to move more components on-chain. While this will reduce capture points of the protocol, the more necessary changes seem to be cultural.
Even though transparency could be improved, there is no lack of degeneracy within Synthetix. I did an experiment to see how active and self-aware each decentralized derivative community is by looking at how much they talked about their competitors. The results were quite interesting:
Chapter 5: Looking Forward
Synthetix v3 is a ground-up rebuild of the protocol. It was deployed on mainnet in February 2023, with no markets attached to it, and the V3 spot market was released at the end of May 2023. But as we’ve seen so far, Perps drives the volume for Synthetix. So until Perps V3 is live, Synthetix V3 wouldn’t have much volume.
With the liquidity-as-a-service narrative in V3 comes permissionless asset creation, better control of credit, features like cross-chain pool synthesis, reward distributors, and also some other major changes to Synthetix :
The possibility of non-SNX Collaterals: Currently, only SNX tokens can be used as collateral within the Synthetix system. However, Synthetix V3 might open the doors for using non-SNX collaterals. This would attract a wider user base and remove the limit imposed by SNX on scaling Synthetix.
The possibility for more flexibility around staking. There are discussions to migrate the current staking system to "SNX Liquidity Provisioning". A pool for SNX staking without market exposure (zero debt issuance), providing yield from an external source is being considered.
The idea to introduce alternative forms of collateral has been going around in the Synthetix community for a long time. Now that cross-chain scaling is a priority, more people see the current ceiling on scalability, since the economic bandwidth is ultimately constrained by the market capitalization of SNX. This topic is being actively debated and it would be interesting to see how it pans out. SNX token holders have had a historical reluctance to share fees with anyone else, even the front-ends that were driving the revenue. Some of these things may also need to become more flexible as the protocol tries to become the liquidity partner for more derivative exchanges.
In summary, Synthetix has come a long way from its humble beginnings as Havven. Its evolution into a liquidity provider signals its adaptation to the changing DeFi landscape. Here is a tweet from Messari in 2021 about Synthetix:
https://twitter.com/MessariCrypto/status/1389590036337905668?s=20
Note that staking APY at the time was 31% with almost all of it coming through inflation. Currently, the staking APY for Optimism stakers is around 21% with only ~8% of it coming from inflation!
Without doubt, the past 5 years were spent optimizing the supply side of the Synthetix equation. It is now more or less working. A staking ratio that has remained over 70% for years is one of the highest in DeFi. Once SNX stops being the only collateral the protocol accepts, it shouldn’t be too difficult to get more collateral into the system. But on the demand side, almost the entirety of the protocol’s customer base is Kwenta, which again is a project that branched off from the protocol. The million-dollar question is how to scale this set of customers, as acknowledged by Kain in a recent interview.
Examining Synthetix’s blog which hosts entries since its inception in 2018, it’s not hard to realize that the team has continuously shipped, solving most challenges that were thrown their way. If this continues and the pie of decentralized perps keeps growing, which seems to be the more likely outcome, Synthetix’s future as an API for liquidity is quite hopeful.
The end:)
References:
SIPs/SCCPs: https://sips.synthetix.io/
Discord: https://discord.com/invite/AEdUHzt
$sBTC on Ethereum: https://dune.com/subinium/btc-on-ethereum
Toxic trades: https://insights.deribit.com/market-research/toxic-flow-its-sources-and-counter-strategies
PMF: https://twitter.com/MessariCrypto/status/1389590036337905668?s=20
LD Capital: https://foresightnews.pro/article/detail/34484