Notes on the token engineering of Curve.finance, Convex Finance, and Solidly

Curve.fi + Convex, and Solidly has made some very interesting design choices for Automated Market Makers and I wanted to share my notes on my own understanding of the repercussions of those choices based on my research.

Some of my thoughts and analysis may be incorrect, mis-informed or misguided so please provide feedback and corrections as that will help me improve my own understanding and also of any future reader’s understanding.

I’m assuming that readers already have some familiarity with the token engineering of curve.finance and Solidly otherwise this video provides a very good overview of that topic.

Here is an excellent video that goes over Convex finance.

Ofcourse, Andre’s blog is a great resource on Solidly. The below video provides an excellent overview of the curve, convex and Solidly designs. https://andrecronje.medium.com/

Background: At a very global level AMM’s are two sided marketplaces which are usually winner-take-all systems.

As with any classic two sided marketplace, once they become fully functional, more liquidity begets more traders which in turn begets more liquidity, thus leading to a winner-take-all scenario. This implies that in the future there is likely going to be only one major AMM that is going to be the mainstay of Defi or atleast one major AMM in each L1 chain. Of course to ultimately end up with only one major AMM across all chains, it is important to have technologies to centralize liquidity such as distributed Automated Market makers that Starkware and Loopring are working on https://medium.com/starkware/damm-decentralized-amm-59b329fb4cc3 and Vitalik’s theory that the future will be multi-chain but not cross-chain should materialize.

However, As it stands today, we are currently in the midst of a two horse race on ethereum to take the crown of being the winning AMM in the long run and Solidly is getting started on fantom.

Here are some interesting points regarding the design choices of Curve, Convex and Solidly .

1)Inspite of it’s technical advantages when swapping Stable coins in Curve V1, given that Uniswap was already established at the time Curve got started , it was important to bootstrap liquidity by providing token emissions to liquidity providers within the curve ecosystem. The trade-off is that you are suffering token dilution for growth but this seems like a good trade-off to make for an up and coming protocol. Andre’s solidly has mitigated the dilution problem by providing an equivalent token emission for existing locked token holders every time there is a token emission to incentivize the liquidity in the pools. Also the ve(3,3) philosophy of solidly might result in fewer token emissions and so would likely have a much higher priced token and it will be interesting to see if this can provide enough bootstrapping power during initial stages to get liquidity on to solidly. Given Andre’s and Daniele’s Innovation track record and following and looking at the way liquidity moved into Fantom for the Solidly airdrop , I donot think bootstrapping liquidity for this project when getting started is going to be an issue but once it gets going it will be interesting to see the effect of this mechanism on the ability to bootstrap liquidity.

2) The idea of giving a boost to Liquidity providers who have locked VeCrv, encourages LP providers to lockup up their curve to get a boost. I think that this ends up being somewhat like an airline miles loyalty program

but for liquidity providers and can lead to loyalty from liquidity providers due to sunk cost fallacy. A liquidity providers incentive is to lockup the smallest amount of protocol tokens for the longest amount of time so as to lockup the minimum amount of capital. Once some number of protocol tokens are locked up, a liquidity provider would much rather provide future liquidity to Curve as they may already have already lockedup VeCrv that would get them a boost.This could very well become a significant moat for an AMM.Solidly also provides a 2.5x boost to liquidity providers who lock their tokens and so the same incentive system is in place within Solidly as well .

3) The locking mechanism is not ideal for short term market participants who may just want to provide liquidity or hold curve until the next market cycle top and so they have to be served by a secondary protocol like convex finance in the case of curve. Convex finance has created a market for curve holders and liquidity providers by sharing the boosted rewards between them and locking up the Curve behind the scenes. Convex’s mechanism has enabled convex token to disaggregate the yield and the voting power of curve as described here https://tokenbrice.xyz/crv-wars/. The vote locked convex token now represents curve voting power in a very optimal manner and this led to the accumulation of convex by Daos as shown here https://daocvx.com/ and this again is a significant for moat for Curve. Solidly has made this much more easier without needing a secondary protocol like convex by issuing these voting rights as NFTs which can be traded and can be acquired by protocols that want to incentivize emissions. However, Andre confirmed in the above frog nation podcast that because there are boosted incentives there is a still a possibility for a Convex like platform for Solidly that shares boosts between locked token holders and liquidity providers.

4) Currently half the curve is voting locked while rest is not locked and approximately half the voting locked curve is through Convex while the rest is held by others as shown below https://dune.xyz/Figue/Understanding-Curve-Wars

Given the above stats it seems like there are some LP providers who directly want Fees and while some LPs want to be invested in the protocol. Solidswap explicitly provides this option to the liquidity providers without again needing a secondary protocol like convex.

5) When a protocol’s token is distributed as incentives it makes a lot of sense to have voting-escrowed “locked” version of the protocol’s token to be used for governance not only due to security reasons to reduce flash loan attacks but also because the protocol fees can be provided to locked token owners based on how much of a die-hard believer they are in a protocol based on how long they are willing to lockup their governance tokens. Also locking the tokens will lead to more curve tokens being taken off market which is good for investors and this also makes emissions more valuable as there is fewer curve tokens to go around. Not to mention that this will also encourage long-term thinking in terms of the Dao’s voting decisions and fees are rewarded to those curve holders who are long term investors. Solidly also follows the same model and I would assume a number of other protocols will take up this model as well in the future and as expected we are seeing that in the lending space VFat has used the same model with hundred finance.

6) Coupling the token emissions to VeCrv holder votes has resulted in the bribe ecosystem. Getting bribes for VeCrv can actually serve as a nice loyalty program for long-term investors in the protocol as they are delighted to get bribes every week for their votes. This has the drawback of the emissions being uncorrelated with where fees are being generated and this has led Andre to propose that locked token holders in solidly should only get the fees for the pools for which they want to incentivize emissions. Solidly approach seems like the most efficient approach to allocating emission resources.

Here is a summary of the analysis so far:

Even though, I have loved and used Uniswap way more than I have used Curve. If I had to choose between Uniswap or Curve in winning the AMM war in ethereum then my bet would be on Curve as the sheer innovation that has been done by Curve in terms of the incentive mechanisms that it has built and the fact that the LPs donot need to be active with CurveV2 unlike Uniswap V3 and are rewarded in a simple manner will probably get me to bet on Curve.

I would love to see Andre’s solidly come to ethereum and enter the race with Curve and Uniswap but maybe it is better for Andre and Daniele to get it started on Fantom which is more of a home turf. Also, it would be interesting to see how well Solidly works with SpiritSwap, SpookySwap and Beethoven on Fantom and whether those protocols would start or even sharing pools inside Solidly. I think it will come down to the market efficiency gained by the alignment of the emissions to protocol fees and the fly wheel effect of such a closed loop as implemented by Solidly.

As I said before, these are some of my off hand thoughts as I tried to learn about these protocols. Please help me improve the article and my understanding with your comments and feedback. I’m super thrilled about all the innovation happening in the AMM space as the ability to exchange assets is such a fundamentally important part of the foundational infrastructure for a free, inclusive, decentralized financial future.