Curve’s new stablecoin is a game changer. Here’s why.
Curve, the second largest DEX in the world, has released the first public iteration of their new stablecoin whitepaper. $crvUSD is designed to work around a few novel concepts, three of which are highlighted in the whitepaper: lending-liquidating AMM algorithm - LLAMMA, the PegKeeper and the overall monetary policy. While the whole model is not yet 100% figured out, @newmichwill is comfortable enough with the level of practicality of the work achieved thus far to publish it.
The main idea behind $crvUSD is to create a more sustainable stablecoin model with a smooth and gradual liquidation as opposed to the current end-all be-all model where liquidations happen in a relative instant. I will be going over the basics of $crvUSD, its model and the major concepts from the whitepaper as short and simply as possible.
Lending-liquidating AMM algorithm - LLAMMA
Arguably the biggest risk that a borrower faces when participating in a money market is their collateral dropping below the liquidation price. As the vast majority of DeFi borrow/lend protocols function right now, whenever the liquidation threshold is reached, the collateral is liquidated in a single event. A bad borrower having to pay the lender is of course a good thing in itself. However, a big problem with decentralized lending markets is that the liquidation of a collateral could be higher than the liquidity available to actually liquidate.
This is why all lending protocols have massive participation in liquidity pools (LPs), on Curve, Uniswap, Balancer, etc. Great for accessing liquidity, however they are greatly affected in periods of extreme volatility or in the case of a bad actor looking for an exploit. Both Mango and Solend have had this issue under different forms in the recent past.
This is where LLAMMA swoops in. The AMM participation is internalized with the LP tokens themselves being the collateral token. Therefore, following the whitepaper’s example, in an ETH/USD pool, as the price of ETH goes down, the LP gradually sells ETH into USD. When price moves back up, the opposite happens, and USD is sold for ETH. If ETH never goes back up, the LP has enough USD at hand to guarantee the debt.
More so, the AMM itself is intended to operate through concentrated liquidity. This means that the liquidity itself is concentrated within a range. When the proportion of the LP moves too much to one side, collateral will be completely converted in ETH or USD. This is opposite to how AMMs work normally, where the LP holder suffers impermanent loss by always getting less of the appreciating asset and more of the depreciating asset.
However, once deposit proportion parity is reached again, the loss is dematerialized. For LLAMMA, the loss could be permanent as @DeFiCheetah very well notes, due to the fact that its liquidity moves solely within a range. This is one of the details yet to be clarified in the whitepaper. Impermanent loss for a normal AMM is impermanent because if price level restores back to the level where LP deposited assets in the pool, there is no change in asset value. For LLAMMA, as the logic is reversed, this loss may be a permanent one instead. Details are not clear now.
If, however, this issue gets resolved, the model can be a game changer: Gradual liquidations imply a higher tolerance for risk, which in turn can mean more capital put into the market. More details will be put out in due time. At its base, it seems LLAMMA has taken inspiration from MakerDAO’s DAI stablecoin overcollateralized model.
PegKeeper
The PegKeeper is essentially an algorithmic market operations controller (AMO) that keeps $crvUSD value tied to $1USD. It does so through two actions:
If 1$crvUSD>1$USD, then the PegKeeper autonomous contract will mint uncollateralized stablecoin and deposit it into the stableswap pool. This increases supply, therefore causing the price to go down and regain the peg.
If 1$crvUSD<1$USD, then the PegKeeper will withdraw stablecoin and burn it. This decreases supply, thus increasing scarcity and causing the price to go up and regain the peg.
$crvUSD is quite algorithmic in nature, swapping between scarcity and abundance in order to keep the stablecoin…well…stable. This is not that foreign of a concept compared to OlympusDAO’s or TOMB’s tokenomics from last year. While those projects have a bunch of issues, I always argued there is merit behind this style of monetary policy and I think it’s great for crypto that a project like Curve is looking at fundamental math applications to make their best decisions as opposed to just doing what is popular.
Monetary Policy and implications
While there are still many details yet to be figured out and released publicly, it seems for now that $crvUSD will effectively be a somewhat algorithmic stablecoin with an internal AMM that creates an overcollateralized treasury that helps it keep its peg under volatile market conditions. The model is relatively similar to FRAX, however with some notable differences: Where $crvUSD will mint more stablecoin to decrease its scarcity, FRAX does so by taking out USDC from its Curve pool. FRAX also uses its treasury as the main tool to keep its peg, whereas $crvUSD controls its own supply.
The way that monetary policy works in this model represents a very interesting opportunity for LP participants: As $crvUSD is minted, users are charged a floating interest rate. Note, that even though the mint is uncollateralized, the stablecoin itself appears to be implicitly collateralized by the liquidity in the stablecoin pool. The system is at its most profitable when it is stable. Thus, when $crvUSD is >$1, the interest rate decreases in order to incentivize users to borrow. This increases supply, thus price goes down back to peg. The opposite is the case if $crvUSD<$1: The interest rate goes up, forcing lenders back and decreasing supply, thus regaining the peg. Welcome to the world of protocols being de facto Central Banks.
Curve really has the potential to break into new ground with its novel approach to stablecoin tokenomics. I sincerely believe that models such as Curve’s, FRAX, DAI or any other overcollateralized stablecoin, especially those heading towards decentralization, will do good not only to crypto, but eventually to the real world. A more fair economic system is desperately needed and I am pretty sure it will be born out of the degen brains of DeFi builders and apes.