Thread by @calchulus

1/n a thread on @SushiSwap's new plans for v3 and its implications for the ever-increasing defi competition in the smart contract world of dexes:
The keys for users are the implications for liquidity provision, capital efficiency, and composability

2/n . First off: one-stop shop: integrating the new swap into Bentobox and more, which naturally adds a wrinkle to battle the Aave & Balancer combo of allowing LP assets to also be borrowed out in a lending pool. medium.com/balancer-proto…

3/n. This is particularly useful for allowing assets to be borrowed, flash swapped, or generate other business lines and fee accrual outside of just direct, first-layer trading. It can also be very helpful for margin and other projects (see: Kashi).

4/n Liquidity provision is changing, starting with the addition of an ERC1155 for an LP position NFT, rather than a 721 like Uniswap, such that the ERC20 tokens could be contained inside the NFT. Personally, I don't love making all LP tokens NFTs, but 1155's are at least useful

5/n Where things get more interesting is the new Trident element and the factory that allows for different types of pools to be created, namely for correlated/stable assets (like Curve), concentrated LPs (like Uni), and non 50/50 pools (like bal)

https://twitter.com/Dogetoshi/status/1417466049260294147

6/n One thing to note - its stable pools can support up to 32 assets, which is the most of any AMM-style dex currently. This could be very interesting for LPs as indices and smart treasuries play that Balancer previously attempted, but could also mean higher complexity.

7/n but the design can support future pool classes. For example, maybe algo stablecoins or rebasing assets may make sense to treat in a method different from the above categories.
More edits can be made in the future by whitelisting new pool types in its flexible factory.

8/n its new router, Tines, allows for "horizontal" trades - similar to what Balancer launched in its v2 vault design such that trades that previously needed to be routed between multihop pairs now can be traded directly, reducing gas costs & slippage incurred on various pools.

9/n Some more interesting notes: whitelabelled pools for other teams, and an interesting TWAP implementation for a decentralized oracle.
Note: @impossiblefi also made a similar tradeoff to strip out this storage-heavy oracle design from univ2 to save everyday users on gas.

10/n Sushi will make use of this TWAP feed for its Kashi lending platform, which is already live on Polygon, where Sushi is seeing nearly 30% of its active addresses lately (data credits: @0xUrsa)
July 19th Sushiers on Polygon: 2280
July 19th Sushiers on ETH: 5396

11/n overall, these features give sushi some very interesting plays from a technological standpoint - it'll be interesting if its multichain approach can help send more of these features into other nonETH chains that have fewer swap options and more Univ2 clones.

12/n The last note is that Sushi purposely made this under the GPL-3 license, which allows for forking, unlike @Uniswap's more restrictive business license.
Long live open source software!

13/n personally will be interested if we can implement some of our logic from our Impossible v2 swap and see if they can be tweaked to adhere to potentially become a @SushiSwap factory-compatible pool.
medium.com/impossiblefina…

14/n If Tine router works as a mini-@1inch / @matchaxyz for Sushi pools - it might make sense for teams like @impossiblefi to also deploy pools inside this standard and our own swap, to allow for more arbitrageable trades with the gas/slip savings of Tine on multiple venues.

15/n the two biggest takeaways from this is that
A: there's incredible room to innovate together with teams like @SushiSwap and reap the benefits from it as it continues to ship open-sourced products with the community.

16/n and finally, cefi exchange tech doesn't evolve as quickly as defi exchange tech does. Cex liquidity from a theoretical perspective will quickly lose to composable dex liquidity if all the product-driven teams in defi work to create megasolutions of capital efficiency.