Some thoughts re: carbon credits as collateral:
Members of Curve Labs has been actively researching carbon credit/offset tokens for usage as collateral for over a year now, starting from the initiation of the C02ken project. Our ultimate aim is to create regenerative currency as described in this blog post by myself.
Our thinking on using credits as collateral has fluctuated quite a bit in this timeframe, and there are a number of design considerations that need to be applied:
-
On one hand, using credits as collateral suggests increased demand for these credits, as it reduces the secondary market supply. This, ceteris paribus, increases the price and consequently reduces the incentives for polluters to pollute.
-
On the other hand, credits represent a financial artifact that is meant to be retired. This means, essentially, carbon credits are meant to be burned at some point in their life. There is no instance of a collateral being used in DeFi that is implicitly intended to be burned at some point in its lifecycle.
Navigating this design tension — the social utility gained by increasing credit price vs. the intrinsic aspect of credits being burned — requires some careful navigation. Effectively governing the amount of carbon credits in the reserve – and the type of these credits – will require a more activist approach than simply setting a value (.5%, per your proposal). What could/should such an activistic governance look like?
- The first and most important issue is that what you are proposing is a de facto monopoly in support of the cMCO2 token. This creates a bottleneck and a reseller monopoly. A more appropriate artifact to collateralize would be something of a carbon credit index — likely some sort of n-dimensional bonded surface (i.e. Balancer LP tokens). To emphasize my point here:
- To initiate the nature backed assets portfolio purchase $500,000 worth of existing MC02 tokens. The purchase of the token will be done in an over the counter transaction, directly with the issuer. The support team is evaluating current possibilities to safely store the token.
Given that the outstanding liquidity of the MCO2 token is $1.2m USD, you are effectively suggesting to purchase 40% of all liquidity outstanding in a single OTC deal. This illiquidity rate is a substantial risk to the integrity of the reserve that can be effectively mitigated through an index product. If I can be frank, this $500,000 purchase is a great deal for MO2, but a dangerous deal for CELO token holders – the other means of offsetting risk here is to have an OTC swap arrangement that runs both ways, but I very much doubt that MO2 would be willing to buy the reserve’s bags if the price of their token collapsed.
- Building off of this, the second issue is that you suggest that the quantity of carbon-credit based collateral be governed by the reserve. This negates the design principle of subsidiarity: the community is not qualified, nor should it microgovern such a small concentration of reserve assets. I would propose that in addition to a carbon index — to negate the reseller monopoly you are proposing — that the governance federalize and a well qualified group of carbon experts be responsible for governing the index weights. Essentially, the .5% weight you are proposing would remain governed by the Celo holders as a whole, but the concentration of carbon-based assets in that index would adjust according to its own governance. A simple hack for governing the index is to assign its governance to LP token holders of the index itself.
Here is somewhat of an architecture of what I’m describing, borrowed from the Klima DAO, which Curve Labs researcher and C02ken founder Raphael Haupt is helping steward:
And an embed of the architecture itself:
Now, moving onto some more general issues…
“as more assets get tokenized in the future, the partitioned reserve mechanism allows for the reserves to include real assets. This is helpful from a stability perspective, and also allows for natural-capital-backed means-of-payment currencies (for example, currencies backed by forestland), where the growth in demand for those currencies will increase the amount of natural capital backing them.”
You are making a number of problematic assumptions here:
The types of natural capital assets you are describing here are tangible ecosystem assets and geographies, which implies that you are imagining ownership tokens. But carbon credits are not tangible assets; they are an abstracted representation of a change of ecological state. That is: you do not need to own ecosystem assets to sequester carbon and receive a proof-of-sequestration; you need only to get a certifying party in the carbon value chain to rubber stamp your project efforts. Returning to my point, though: Celo’s mission is to empower and work with last mile users. Many of these users live in environments of low social trust, without functioning or heavily biased legal systems. Ownership tokens do not function in low social trust environments as they need a functioning and crypto-literate legal system to retain their value. If the legal system does not uphold the true owner of some extent geography and its associated natural assets, then there is no way to redeem the tokens. Now this in itself is not an issue so long as you are working with ecological state tokens, and not ownership tokens. The difference I discuss at length in the Kolektivo white paper under the “Natural Capital System” section (pg. 35):
Comparison table embed:
Adding uncorrelated assets to the reserve further diversifies technological and price risk. Adding tokenized carbon credits to the reserve is another step towards a nature backed currency.
You are assuming that MO2 tokens are uncorrelated to the crypto-market as a whole, but I can easily demonstrate that is not the case. Here is the 90 day MO2 price over time:
I do not even have to calculate the coefficients for each crypto-asset for you to see the obvious correlation here. The reason MO2 is correlated here is simple to understand: most of its liquidity is against USDC. Most of USDC’s liquidity is weighed against other cryptocurrencies. If you want a truly uncorrelated asset, it needs to be introduced to a different liquidity environment.
In general – and in conclusion – I’d like to say that the thing I find most alarming here is the two assumptions you have made regarding the liquidity of MO2 token and its correlation to the other reserve assets without offering any sort of viable mathematical proof or tokenomic simulation. This looks and feels like an inside deal to me and seems absent the rigor required for understanding risk from a strictly quantitative point of view. I believe a protocol needs to be established that includes simulation and analysis for any and all assets added to the reserve.