Historically, Curve pools have been using DAI/USDC/USDT for base pool implementations as those were generally considered to be the most stable and most liquid assets. DAI however has not grown at the same pace as USDT and USDC have despite benefitting from deep liquidity due to its presence in the Curve 3pool.
3pool is used as the base pool for the vast majority of factory pools on Ethereum and all side chains and L2s. This means the tokens in the pool benefit from deep liquidity which can be made in a variety of metapools. This means all the large second generation stables like FRAX, UST, MIM are indirectly driving liquidity towards DAI when they pay to drive liquidity to their pool.
Metapools are a key part of the Curve protocol mainly because they prevent fragmentation of liquidity (which is being driven by the gauge system and among other things Votium and Convex). For example, thanks to UST being a metapool, you can swap UST to USDT/USDC/DAI in the same pool and that liquidity is still available in the 3pool to be suitable for extremely large swaps which are valuable to Curve.
As a result, when creating the pool factory on Ethereum, we prevented the creation of plain pools with USDT/USDC/DAI to avoid the needless dilution of the 3pool as it had historically been the pool making the most volume and the most fees. However that has come to change because of several reasons: other AMMs proposing 0 or 1 bps swaps, the PSM allowing to swap between USDC and DAI at no fees as well as the huge growth of a new generation of stable coins.
Abstract:
This proposal would allow the addition of a new base implementation of USDT/USDC which would become the defacto implementation which would make Curve and other protocols using Curve less reliant on DAI going forward.
As there has been renewed demand for plain pools without certain assets, it seems like a good time to propose this change which will give a welcomed option to protocols planning to build on Curve as well as those who wish to not rely on DAI and MakerDAO.
4Pool is an option of UST/FRAX/USDT/USDC which has also been added by popular request.
As a reminder, Curve is built with immutability at its very core which means pools, as they are deployed on mainnet cannot be upgraded or killed.
2pool would be good to offer as a basepool, the marketcap of USDT/USDC dwarfs the other stables, and while DeFi stablecoins rise and fall, those two will likely continue to scale and see wide adoption.
4pool should also be offered as a basepool, UST/FRAX and the other protocols pledging support for this pool (redacted/ohm etc) make up a significant amount of protocol ownership over Curve. They appear to be willing to heavily incentivize this as a basepool. If they’re willing to do what they can in their combined power to turn this into the default basepool, they should have that option available to them. It’s much riskier to Curve allowing this to become the default basepool, but it’s important for Curve to remain neutral and let tokenholders dictate its future.
2pool base implementation makes a lot of sense: DAI has had a good time freeloading off of Curve (meanwhile their public community has done nothing but shit on Curve) and has not expanded in a way that is appreciable for today’s needs. We expect that most swaps in the 3pool are between USDT and USDC anyway (needs to be verified). 2pool helps Curve expand much further and not get dragged down by DAI’s slack (DAI is still a cool stablecoin tho). For reference:
Having stable-2pool implementation would allow other algo stables to expand as well, upending DAI’s dominance. This is something we look forward to: stablecoins are a very interesting financial engineering attempt at damping the volatility of an asset to zero (depending upon use case), and there is no one way that’s the best (they must all be encouraged in this grand DeFi experiment). Alas, DAI (being as cool as it is, regardless of how toxic its community is) has not managed to grow as much as centralised and regulated stablecoins have.
On 4pool implementation
The combo with unproven algo stablecoins in their nascent stage is … questionable. UST does plenty volume for reasons that are not organic (either ponzi-buying BTC from CEXes or trying to repeg the pool after replenishing Anchor’s yield reserves). Frax on the other hand has been dumping CRV and calling it profit, while announcing to the world that they’re aligned with Curve. Curve can live without FRAX, the other way around is questionable.
While we wonder if something like 4pool is worth it, it should be given a shot regardless. 2pool would make Curve a bit more neutral. People who want DAI can always go the 3pool route (that will not change). People who are a bit more risk on can choose the 2pool-FRAX-UST route, if it provides more yield.
With 4pool metapool, the question is: who’s going to build it? What about fee burners? will veCRV holders receive 3CRV or 2CRV?
Finally: who’s willing to build curve from scratch?
I think 2 Pool USDC/USDT is a great idea. Stop supporting 3 Pool and let DIA stand on its own. Curve has helped build DAI liquidity and demand for far too long. A USDC/USDT Meta Pool will be a good foundation for any DAO to make combinations for a 3 pool or a 4 pool.
I concur that 3CRV as a sole stable coin metapool is unfairly incentivizing Dai liquidity without incentivization to veCRV or vlCVX holders. Addition of a 2pool or 4pool implementation is logical.
One question that arises is: Would it be possible to pivot payments to veCRV into one of these other metapool tokens down the line if needed? i.e pay revenue in 4 pool or 2 pool?
IMO a USDC/USDP/GUSD metapool would be a valuable addition - gives an option for risk averse LPs who don’t want tether exposure. Decentralized stablecoins could pair with this pool to have reliable linkage to fiat, while minimizing tail risk (eg from Tether or DAI insolvency risk in current 3pool).
Also think this would help lift up GUSD and USDP/PAX (both high quality stablecoins but with limited adoption), which benefits Curve strategically in the long run - a greater diversity and more even balance of stablecoin market share should lead to higher stableswap volumes.
This quickly becomes a existential question about whether or not Curve is intended to be a legit defi/decentralized system or just a facade for fiat. I’m definitely in the true-defi camp (although it bugs me that DAI now has non-defi assets). At what point do we determine whether we are investing long term in honest money or still captured by money being inflated away at will by people who do not have our best interests in mind.
