The shifting sands of DeFi
This post will serve as a minor follow up to the recent State of Synthetix post as a number of things have changed in that time which have required a review of our project priorities.
One of my favourite startup books is “The Hard Thing About Hard Things” by Ben Horowitz. I think it captures almost perfectly the insanity required to attempt to create a new thing. The odds are hugely stacked against you, any number of things can and will go wrong, most of which are out of your control. The best you can do, and maybe the core lesson I took away from the book, is to be adaptable; you must respond to reality not attempt to control it. I think about this a lot in the current DeFi environment, because it is without a doubt the most hectic and intense environment you could choose to build in. The proportion of things we don’t know so far outweighs those we do that the very act of trying to know things feels futile at times. When your epistemology devolves into nihilism due to the rapidity of change and the sheer absurdity of attempting to distill things into a coherent narrative, then something very interesting is happening.
I believe this is because we are in the middle of a huge transition, not of finance, but of human organisation. The fact we picked finance makes sense since it is easily amenable to the current tools we have available. But this transition is purring because we are truly seeing an exponentiating technology spread across the world. Every single component built in DeFi provides another tool for someone else to build upon. Many people believe this was how the internet works, but in reality it was only one component of the internet that was truly open, which was the information itself. Excluding low level protocols, most of the interesting stuff on the internet was closed source. The difference between the early internet and DeFI is that essentially everything is open source, so you have possibly the lowest barriers to entry ever experienced. It turns out low barriers to entry are pretty powerful. When you combine permissionless money with low barriers to entry, crazy stuff starts happening. And we are only in the very beginning of this explosion of open systems.
So what does this mean practically for the current DeFi teams? Basically, it is a total clusterfuck. So many things are happening simultaneously that it is impossible to track the potential consequences of all of them let alone the combined consequences of stitching them all together. Thus you get the challenges for the DAI peg, and weird volume spikes across the DEX landscape due to shifting protocol incentives, and high yields on specific DeFi tokens on lending protocols. Communities like those of Maker, Compound, Aave, Uniswap and Synthetix no longer have the luxury of sitting in their walled gardens, where mainly their own choices had impacted them. The MakerDAO communities talk more about external protocols and mitigation strategies than they talk about Maker itself these days…
Specifically for Synthetix this has meant that right as we felt we were ready to launch several significant changes to the protocol, unintended consequences of various initiatives within DeFi have forced us to reconsider and reprioritize. Because we must also contend with the impact other projects are having on Ethereum itself, every time you assume the Gwei high watermark has been reached something else happens to push it further. The UNI distribution, as positive as it was for the ecosystem, was also a shitshow for gas prices. So now we measure gas in twei I guess.
So as much as we want to prioritise product facing features, we have been forced to review priorities for Fomalhaut and Deneb, our next two releases and ensure that they are having the maximal impact on the network. We intended to minimise the protocol optimizations we focussed on through the rest of 2020, but DeFi had other plans. Right now, two issues are clearly gas prices and demand for decentralised stablecoins. While demand is critical for sUSD, and we quietly breached the 100m supply mark last month about 18 months after launching the sUSD multicurrency system, excess demand can push the price above the target of $1 which creates significant friction and uncertainty for sUSD users.
Based on the continued impact of these two issues we have shifted some of the priorities of the outstanding SIPs. The scope for the next two releases is below:
Fomalhaut - 24th September
SIP-85: Ether collateral v0.3
SIP-86: Exchange rates aggregator V3
No SIP: L2 migration Phase A
Deneb - 29th September
SIP-83: Debt pool snapshots
SIP-TBC: Re-enable all paused Synths
Here is a brief description of how all of these SIPs will help to address these two pressing issues at a protocol level, before we launch back into further product development.
- SIP-85: Ether collateral will provide a much needed lever on the Synth supply. There is currently no direct mechanism to increase the Synth supply other than lowering the collateralisation ratio of the network, and this has very limited impact while introducing significant risk into the system. The introduction of sUSD borrowing against Ether will enable ETH holders to mint sUSD when sUSD is trading above $1 to help reduce the premium. Given that demand for sUSD continues to grow, it became increasingly obvious that this was one of the most impactful changes that could be made in the current environment. This also opens the possibility of sUSD borrowing against BTC tokens such as renBTC and tBTC in a future release.
- SIP-86: This change was required to support new data types following the transition to Chainlink, completed in the Pollux release. Specifically it allows for consuming gas price data critical to enabling incentives for Keepers required for the Synthetic Futures launch.
- L2 migration Phase A is the first step on the migration path to Optimistic Ethereum (the Layer 2 solution). It will be an incentivised testnet aimed at alleviating gas costs for small SNX stakers. More details coming soon.
- SIP-83: Debt pool snapshots will resolve two issues, the first is the fact that the current debt pool does not scale well as additional Synths are added. This is due to how the debt pool is calculated and it has an impact on Minting, Burning and transfers of SNX. We have thus been hesitant to add additional Synths for a while now as gas prices increased. This change will thus allow us to add significantly more Synths to the protocol, but it will also reduce the baseline costs of all of the functions above, which means that the staking process will become much cheaper on L1.
- SIP-TBC: Once SIP-83 is live we will be able to re-enable all of the paused Synths from SIP-84 that were paused to lower the impact of the higher debt pool calculations due to using a network of price feeds from Chainlink rather than a single centralised Oracle.
Both of these releases are direct responses to increased gas costs due to Ethereum congestion. Some of the changes are stop-gaps while we transition to Optimistic Ethereum but included in these two releases is the first step towards L2 Synthetix. This is something which will not only vastly improve the current experience but significantly increase our ability to deploy new solutions that have been prohibitively expensive to deploy on L1. This hybrid approach to L2 will likely take us through to the end of the year, but it represents the near- to mid-term future of the project.
Post Deneb, we have several releases planned which will bring new functionality and complement the upcoming front-end releases of Kwenta and staking.synthetix.io.
The next release is Mimosa, which is both a tasty cocktail, a bright star and the launch of the Synthetic futures tesnet competition. We now have a backlog of around ten SIPs in various stages of implementation, so we expect to see a very fast release cadence over the next few months with potentially smaller releases of only 1-2 SIPs as they are ready. This is enabled by the vastly improved deployment infrastructure we invested time in earlier in the year.
Without recapitulating the entirety of the state of Synthetix post, we will be heavily focused on several major themes to the end of 2020:
- User experience improvements and protocol growth
- Continued transition to decentralised governance
- Composability improvements, specifically virtual token bridges for AMMs
Over the last several months the market has awoken to the potential for decentralised finance, specifically the power of tokens as coordination mechanisms for trustless financial infrastructure. This has been in spite of significant friction from high gas prices and other UX issues. I strongly believe that the transition of much of DeFi to the OVM will be an inflection point where DeFi is no longer a whale game, but is accessible to anyone to both build on and use. We are not there yet, but we are getting much closer. In the meantime we must optimise for the L1 network we have.