Thank you @flowcarbon, this is an exciting proposal for the Celo ecosystem and a great real-world use case for Mento stable assets. Before diving deeper into some technical aspects, we wanted to discuss some important collateralization implications of the proposal.
As you may know, since CGP62 all Mento assets (cUSD, cEUR, cREAL) are backed 100% with USDC/DAI alone to minimize collateral risk (everything else, including CELO, ETH, BTC comes on top of the 100%). This means that for every $ outstanding, the Mento reserve holds at least an equal $ amount in USDC/DAI alone. (@tmoindustries: Mento is now a separate protocol with a separate team so it makes sense to call it Mento reserve instead of Celo reserve). The same policy would apply to the cUSD that is invested into the Centrifuge pool via the bridging setup.
This raises two follow-up questions:
- Where does the cUSD for the Centrifuge pool come from?
- How can the 1:1 backing mandate be upheld for the cUSD that is in the pool?
Mento protocol does not hold cUSD in its treasury, it only allows the creation of cUSD and other assets by providing collateral assets to the protocol. So for Mento to participate in the pool, it would need to create cUSD against the forward carbon collateral, e.g. in a Maker-like CDP mechanism. This would require some additional technical work (it would be incompatible with the current Mento setup), so it would not be an immediate option but in principle something that has been done before (see Maker).
However, and more importantly from our perspective, such a setup may raise collateralization concerns. As some cUSD would in this case be collateralized with forward carbon credits, Mento assets would no longer be 1:1 backed with USDC/DAI. This is potentially problematic because while the reserve would remain fully collateralized (or overcollateralized) by the forward credit pool, some of its reserve assets would then be substantially less liquid (the carbon pool has redemption windows every 6/12 months, limited secondary market). This may reduce Mento’s ability to absorb larger cUSD contractions because reserve assets cannot be liquidated quickly enough, and given the proposed transaction size ($7M compared to current cUSD circulating of $30-40M) we should not ignore this risk.
In light of this, we think that instead of creating cUSD against forward credit collateral through Mento, it would be better to stick to the current collateralization policy for cUSD and to identify another source of cUSD liquidity for the pool, e.g. the Celo Community Fund.
The Celo Community Fund will soon hold a substantial amount of CELO as ‘buffer collateral’ is transferred from Mento protocol to the Community Fund. The proposal for the first tranche passed recently. So one idea could be that the Community Fund converts some of its CELO holdings into cUSD via Mento or secondary market options, and to invest those funds into the Centrifuge pool. cUSD would maintain a 1:1 backing with high-liquidity assets and the Community Fund would earn cUSD yield over time, while at the same supporting carbon projects through the forward pool. It would also help with risk diversification as the Community Fund would have less exposure to CELO volatility and it would be beneficial from a TVL perspective as the CELO would be put to productive use, while CELO sitting in the Community Fund would likely not be included in TVL metrics (a concern that was raised here).
The downside is that converting a large amount of CELO into other assets puts pressure on CELO (either because CELO is sold into the open market or because CELO is transferred to Mento and subsequently rebalanced into DAI/USDC). So if this route is explored further, it might be beneficial to split the entire transaction up into smaller tranches and to perform transfers over time. We at Mento Labs would be happy to help the Celo Community explore this further.