Original author: Kain.eth
Original translation: Deep Tide TechFlow
With the official launch of Synthetix V3, now is the best time to question the assumptions of the past to ensure the community stays aligned with the long-term vision. This article will focus on the SNX token and its role in the ecosystem.
Multi-chain
The vision of Synthetix is ambitious. Deploying the Synthetix protocol on multiple chains presents challenges far greater than most other DeFi protocols. This is because most protocols deploy independent instances on each chain, introducing operational complexity but is the simplest way to maintain presence across multiple networks. In many cases, the governance of these different instances is controlled through unified governance. While this adds additional complexity, governance typically operates off-chain or on a single network. Some protocols go even further, with operations on one chain affecting all other chains. This can be achieved through cross-chain bridges, liquidity management, or shared clearing, among other methods. Protocols can also support interchangeability of cross-chain assets on-chain through wrappers or bridges. As early as early 2020, the Synthetix community envisioned a grander scenario: a single protocol spanning across multiple networks, where the system behaves as if every operation on each chain occurs on a unified chain. Though we are closer to achieving this goal than we were three years ago, it remains challenging to achieve. There are many unresolved technical challenges that must be addressed. Given the developments in the cryptocurrency space over the past few years, it is worth reevaluating whether this multi-chain approach is still the best path forward.
Why Share Liquidity?
Imagine deploying Synthetix on five different EVM networks. SNX tokens can seamlessly move between these networks and be used to provide liquidity (SNX LP) on each network to support trading. Complexity arises when these SNX LP positions also support liquidity on other networks. If Alice only trusts Mainnet, she can provide SNX LP there and provide liquidity for the other four supported networks. However, there is a potential contradiction: if she only trusts Mainnet, why would she support liquidity on other networks? Bob can provide SNX LP on Optimism and provide liquidity for Mainnet, Optimism, and Base. Carol can provide LP on Avalanche and only provide liquidity for that network. The market should ideally resolve this issue, with liquidity flowing to where the trading demand is highest. This design assumes that each network has a certain proportion of unique users who only trade on that network and not elsewhere. If they cannot provide LP and trade on their chosen network, they will be completely excluded from Synthetix. We can question this assumption later; for now, let's see what it takes to implement this.
Implementing Shared Liquidity
The complex approach is to deploy different markets and pools on each network and allow liquidity providers to provide cross-network liquidity from any network. In Synthetix V3, a market represents a single asset, such as sBTC, and a pool allows liquidity providers to send liquidity to a market aggregator and then delegate to multiple markets. The main pool will be the Spartan Pool, which will contain all "officially supported" markets. In Synthetix V2 x, there is only one pool that contains all supported markets, and all liquidity providers are forced to delegate liquidity to this pool.
Alright, let's take a specific trader as an example to understand how this impacts them. Suppose this trader only trades on Base. They like to engage in ETH swing trading, and to ensure they can continually open new positions, there must be sufficient liquidity delegated to the Base Spartan Pool and/or delegated to the Base ETH pool. Remember, in this setup, it doesn't require anyone to provide liquidity on Base; all of this liquidity can come from Mainnet or other chains. This is effective for Base traders because the demand for sETH on Base will prompt liquidity providers from other networks to directly delegate liquidity to this Base market through a single market pool or a pool containing the ETH market.
In order for this implementation plan to take effect, extensive cross-chain communication is required. In fact, this is a very massive task, especially with the increasing number of supported networks. This can be achieved through CCIP and some upcoming Chainlink infrastructure, but the question we must ask ourselves here is not "Can we do it?" but "Should we do it?"
What do we want to achieve?
Ultimately, we want to have an efficient liquidity market where the demand on any chain can be met by supply on that chain. One assumption is that liquidity is limited, therefore we must be able to share liquidity between networks to avoid fragmentation, where liquidity is available on each network but poor overall due to excessive dispersal of limited liquidity. This leads to a "cold start problem": insufficient liquidity on any one network would dampen demand and reduce incentives for further liquidity provision. Currently, liquidity is limited, but some liquidity is more limited than others. Part of the multicellular complexity is that the community has realized that in the short term, SNX liquidity is unlikely to be sufficient to meet the demand across all networks. Therefore, to avoid introducing additional collateral, we must twist ourselves into various forms to achieve this cross-chain liquidity system and avoid fragmentation with limited SNX liquidity. I want to point out here that I believe in the long run, if there is demand, the SNX token will expand to support the required levels of liquidity. The issue is that in a bear market, this is a lagging process and demand does not increase SNX collateral sufficiently. Hence, we are forced to choose between increasing the complexity of the system to solidify SNX's role as the main collateral or finding a compromise.
Why not just stay on Optimism?
This survey's key assumption is that there must be a market on every chain. I think this is a reasonable assumption. But it is an assumption that we should question. A similar situation is if Coinbase had a separate exchange for each operating system, so no matter which operating system you prefer, you can still trade. On Debian, how much liquidity can we expect to have on Coinbase? This analogy deviates slightly from the point; a better analogy is if Coinbase had different markets for each database engine to cater to user preferences. Do users have strong enough database preferences to influence their use of the trading platform? However, this is exactly the situation we currently face in the cryptocurrency space. Some people firmly support specific chains and even refuse to recognize or transact on other networks. Why does this happen in smart contract platforms but not in databases? In this analogy, the database is the state storage layer of CeFi, while the L1/L2 networks are the execution and/or state storage layers of DeFi. The reason is simple: it's called tokens. While supporters of database engines may be annoying, there is no mechanism to transmit this technological tribalism to end-users. In the cryptocurrency field, we have built a very powerful incentive mechanism to promote this user preference. This phase may eventually pass, but for now, I think it is reasonable to assume that there are different user groups on most networks.
Perpetual Fragmentation?
Well, user fragmentation at the execution level is not surprising, and we seem to be temporarily stuck in this situation. Despite ultimately wanting practicality, users are highly motivated to have this practicality only exist on the networks they support. Therefore, even if we build the most optimized exchange in history, if it only exists on one chain, it will be isolated and unable to receive the attention it deserves. This becomes particularly evident when competitors - centralized exchanges - can choose any tech stack and put all users into a single database, and no one cares. This means all their liquidity is on one island, a huge advantage that decentralized cross-chain exchanges cannot match. By the way, one way to address this problem is to abstract the network and make it more like a centralized exchange; this is exactly the experiment that Infinex intends to try. From the perspective of Synthetix, this is a valuable experiment. But Synthetix must optimize for the reality of Infinex's failure, meaning as a protocol, it must meet users on every chain where they want to trade or provide liquidity.
What are our options?
If we need to have the Synthetix protocol appear where users are, how can it get there? In my view, there are three methods.
Deploying a forked version of Synthetix on each chain;
Implementing a unified cross-chain protocol where state changes must be propagated to all chains;
Deploying a separate instance on each new chain and minimizing cross-chain messaging and liquidity fragmentation by utilizing non-SNX collateral.
Option One: Forking All the Way
This is not a viable method, although it might be interesting to do so. The main problem that prevents us from doing this is the reliance on SNX tokens as collateral. If you have ten forked SNX tokens, they are likely to devalue and significantly reduce liquidity on each fork, thereby defeating the purpose of avoiding liquidity fragmentation. Most protocols simply deploy new instances of their contracts and let liquidity flow to the new network on its own; Aave and Uniswap are good examples. For Synthetix, we need a substantial amount of SNX as collateral on the network, and we also have a lot of governance overhead. Therefore, we may need to fork SNX tokens on each network. You can test this process with Base. We can deploy an entirely new instance of Synthetix including SNX, and prefix all tokens with 'b'. Thus, we would have bSNX, bsUSD, and bsBTC. Although this might be interesting, I don't think it is our best hope. But again, this can be very interesting, and I'm not saying we shouldn't consider this method. One advantage of forking today compared to a few years ago is that deployment has become much easier. In the past, deploying Synthetix on one network alone could take several days. At that time, deploying multiple instances across networks, even if each deployment was independent, was not feasible.
Option Two: Unified Liquidity Theory
This method only needs to solve all technical problems and build a network protocol that can maintain all the necessary cross-chain message communication. Then we can continue to rely only on SNX collateral. This is theoretically good, and we are moving in this direction, but what if our assumptions above are wrong? What if there are not enough independent users on all these chains to generate incremental trading volume? What if we spend too much time building this supporting infrastructure, and another protocol builds a much better single network solution than all other decentralized exchanges, and users start migrating to that protocol? There is a precedent here: dYdX, but they have built a powerful enough trading engine that traders are willing to migrate there. If their new engine on Cosmos is much better than all other DEXs, will Ethereum Virtual Machine users migrate on a large scale? I think we should focus most of our efforts on core product functionality rather than supporting infrastructure. But at the same time, we should at least conduct some experiments to determine if there is enough trading volume on other chains to justify the need for a cross-chain solution.
Option 3: Explore Forbidden Waters
Some OGs, including myself, have a religious belief in pure SNX collateral. Anyone who challenges the sacredness of pure SNX collateral must be burnt to death or thrown into a lake (metaphorically speaking). Even Fifa and ha-oN, this noble pair, would not be spared if they were tempted by the dirtiness of Ethereum collateral. However, we have an excellent opportunity to test the potential network demand hypothesis without the need for cross-chain SNX collateral. We can deploy perpetual contracts to Base and use Ethereum as the only collateral. This will reduce the risk of transferring SNX liquidity from Optimism. It requires almost no cross-chain communication. The main problem will be migrating and burning fees to Optimism. But there is another potential option: we can use the fees generated on that network to buy back and burn SNX. This allows us to test two novel approaches simultaneously with relatively low risk. If one of the methods is ineffective, we have already established a stronghold on the new network to later upgrade and replace Ethereum collateral with SNX. We can also switch to burning debt on that network instead of conducting buybacks. This brings up one last issue: the best incentive mechanism for liquidity providers is to burn debt only on the network where they provide liquidity, rather than burning it globally. Regardless of which network they provide LP on, we will create a "free-rider" problem, and it may be beneficial to camp on the "safest" network, even if it is not the most demanding network, even if that network does not have the highest demand, to potentially profitably stay on the "safest" network.
New Hope
If we release ETH collateral on a new network, I believe Base is the best choice. This will allow us to increase trading volume without threatening trading revenue on Optimism. It also has lower risk compared to Arbitrum. This should be a win for SNX liquidity providers. The counter argument is that if we incentivize people to migrate SNX to Base and provide liquidity there, SNX will capture 100% of the fees across both networks instead of sharing them. This is true, but for SNX liquidity providers, the risk is small because we control governance. We can conduct this controlled experiment and then decide on the best option for SNX holders based on the data. This test is not one-way; if we have too much SNX collateral on Optimism, we can allow SNX to migrate to Base and use it as collateral. This will dilute the yield on ETH and achieve a new balance. We can even impose a limit on ETH liquidity or remove it as collateral altogether. That's why governance must be fully in the hands of SNX liquidity providers. It is crucial that this experiment proves the point that there is incremental trading volume on a new network without harming existing volume on Optimism. As long as the fee split for SNX is high enough, we should be able to understand the incremental revenue we can generate. I think we should start from a lower level and adjust as we progress. Capturing 40% of the fees from SNX liquidity providers is a good starting point from which we can make adjustments. We need to determine if ETH liquidity providers are willing to join the network; if the fee split is too high for SNX, we will not be able to properly test people's inclination to provide ETH liquidity.
What if this strategy works?
If we conduct this experiment and see a strong demand for LP ETH and trading on the new chain, would we burn SNX tokens? I don't think so. If this strategy works, we can expand the experiment to other chains; Arbitrum and Polygon could be the most obvious choices. By then, we may have enough data to recognize that Optimism trading is also limited by SNX liquidity. If that's the case, we can also allow ETH as collateral on Optimism. Furthermore, if it appears that liquidity demand supported by ETH is much higher than that supported by SNX, we can even decide to disable SNX liquidity on Optimism. However, if we do that, I think we need to take additional measures.
Our Own Chain
If demand for ETH-supported transactions is much higher than for SNX-supported transactions after conducting these experiments, we must accept this and leverage it. The next step will be to create an AppChain on the Optimism superchain. This will allow us to migrate governance to a chain we control, and it will also be where you can obtain leveraged sUSD loans on SNX. This remains a key functionality of the Synthetix network and a major benefit of staking SNX. I don't think we even need to enable trading on this network as there is very little ETH liquidity on it. This will be the network where all the backend functionality of SNX resides. If in the future we believe we can efficiently develop a cross-chain shared liquidity system, this will be where liquidity exists and is delegated to other chains.
Open Network
Providing SNX liquidity has always been challenging and high-risk due to the need for hedging. This problem, combined with high inflation, means that few integrators are willing to create staking solutions for SNX. By moving to an AppChain, we can lower the risk of staking SNX and completely eliminate inflation. This would potentially enable staking SNX to be integrated into more platforms and open up the network to a wider user base. Now, due to the hedging requirements, liquidity provider risk-adjusted returns are much lower. SNX may still serve as an insurance fund for the network, but it doesn't need to be actively maintained and the risk is much lower than continuous hedging.
Cross-chain Swaps?
They can still be supported in this new design; they would require CCIP, but in principle, even with liquidity on each chain, we can still support the interchangeability of Synths between chains. This does require cross-chain messaging, but it's not as cumbersome as supporting cross-chain liquidity. Cross-chain is still a very lucrative and emerging market, and although focusing on perpetual contracts right now is the optimal strategy, we should definitely continue to explore Synth Teleporters and other novel mechanisms.
Summary
Summary, we are facing several technical challenges: we can avoid these challenges and test one of our core assumptions. Is there potential demand for other network transactions? We can achieve this by using ETH collateral with low risk. If this approach works, we can expand this experiment to other networks. If this route seems promising, we can migrate to our own AppChain and charge fees from multiple networks, utilizing ETH collateral and coordinating and managing the protocol using only SNX. We give up a portion of the total fees, but we can quickly scale and avoid another protocol dominating. We can test exclusive SNX collateral on any network at any time and limit ETH collateral or other collateral as needed. This allows us to quickly expand to multiple chains with minimal technical overhead. There is a lot of data to collect in this process, and we should test each step gradually, as we have always done in Synthetix.
The key question I want to raise is: are we willing to obtain a certain proportion of the total addressable fees now, or wait longer to capture 100% of future fees when preparing to deploy a real cross-chain implementation? The good news is we may have the best of both worlds, and if it proves more profitable, we will ultimately have a more accessible network with lower entry barriers for anyone to easily stake and join.
