Capturing Arbitrage in Carbon Commodity Pools

Note: I wrote this post as a Medium draft and circulated it to Toucan people when the mass-rebalancing event happened a while ago. I initially did not publish it, as I didn’t feel the need to post this kind of criticism publicly since this was very specifically aimed at sharing my thinking with Toucan. Since then, Toucan has made steps in a similar direction to what I outlined. Toucan has asked me to share this thinking publicly.

On to the actual article!

With the recent situation around 1:1 targeted redeemal in Toucan’s BCT pool I’ve felt motivated to share a bit about what I would personally change if I could wave a magic wand and get exactly what I wanted. I’ll preface this with some general context around commodity pooling and why I don’t feel 1:1 targeted redeemal is a healthy mechanic for the protocol. Then we’ll have a chat about N:1 redeemal mechanisms and how they can align the ecosystem to want the same thing.

Why I disagree with 1:1 targeted redeemal

Since I’m writing this to Toucans, I think we all agree — commodity pools are a kickass invention. They make a fundamentally fragmented, infinitely differentiable market liquid. Because commodity tokens are a liquid asset they also provide potential investors with a way of taking a long position on the carbon market. They work as an index. Functionally this is part of what Klima does when it buys up a large amount of BCTs — invests in an index.

While a currently mainstream way of thinking seems to imply that the expected use-value of a commodity pool is it’s worst credit — I would instead propose that the value of a pool is it’s best credit. The credit with the most general, universal use value. This would be like a super fresh carbon credit from a good project applying a good methodology. Anyone would be happy to use this to offset their footprint. There would be no backlash or optics risks. The purchase is easy to justify to shareholders and nice to show in marketing material.

As we have seen the value of BCT go up to about 9$, we saw some credits with quite good use value enter the pool. The pool’s price will however go down from 9$, since there is potential to bring credits on-chain at that price and capture quite a good spread. But now that the value of the price goes down — there is arbitrage value in taking that nice 9$ credit out of the 5$ credit pool.

So now we get to the crux of the problem — who gets to benefit from this arbitrage and why? I would argue that when you implement 1:1 targeted redeemal, the ideological position you take is that ‘exclusively the arbitrageur deserves to benefit from arbitrage value because they got there first’. They extracted excess use value from the system. The use value of all other BCTs decreases because they no longer have the same option. The use value moves down to the next best credit and so-on.

This mechanic in my opinion creates an unhealthy spiral within the commodity pool that actively incentivizes quality raiding, inevitably collapsing the use value for the remaining BCT down to a point where no credits in the pool are attractive to people that would like to retire inside of BCT. If you are first — dollar dollar bills, yo. If you weren’t the one arbitraging — tough luck. I feel this is a net negative to the system leading to the exact problem Toucan faced. ‘Who benefits from this arbitrage and why?’ resulted in a controversial answer with no payoff to BCT holders at large. This was by design.

Toward N:1 redeemal mechanisms

If 1:1 implies that all arbitrage value rightfully belongs to the arbitrageur because they got there first — N:1 simply implies that first and foremost, other stakeholders must benefit by a sufficient margin before the arbitrageur gets paid. Say for the sake of example that Toucan charged a 20% fee for redeeming specific credits out of the pool. I picked the 20% fee arbitrarily. It can even be dynamic based on properties of the specific credit. For now let’s go with 20%.

So for each 1 BCT taken out of the pool — the person redeeming pays a premium of 0.2 BCT. This 0.2 BCT can now be distributed to the ecosystem. Let’s have a look at potential stakeholders who could be rewarded with this premium. It can be a combination of all of these.

Burn — Similar to how EIP-1559 reduces the supply of ETH, there is no reason why this premium couldn’t simply be burned. In addition to reducing supply (hopefully short-term increasing price), this would slowly increase the use value of the ‘last BCT to be redeemed’ because some undesirable credits would be left in the pool, never to be used. These would be the HFC-23 credits. The same could be achieved also by directing this 0.2 BCT to retire HFC-23 directly, though this implies more human intervention to target specific credits.

Stakers — Theoretically you could incentivize people to buy and hold BCT (what Klima is already doing) — creating a direct incentive for taking BCT out of circulation by awarding them with more BCT in the future. This theoretically improves the price of BCT by restricting circulating supply.

Bridgers — The people that originally tokenized the desirable credit can be incentivized to bring more desirable credits onto the chain by paying them each time someone specifically wants a credit they brought over.

Developers — Devs gotta’ eat. This is one way Toucan could create a revenue stream to pay for ongoing development and maintenance.

Project Proponents — Finally, though quite hard to pull off, Toucan could direct some BCT to project proponents directly in order to incentivize them to bring more of their credits onto the chain themselves. This would require a human component which makes me personally not very sure this would be easy nor worth the investment in the near-term.

The point is: I think all of these parties are inherently more important to the ecosystem than the arbitrageur and should be compensated first. I see no reason why the arbitrageur should be capturing all the value — especially if markets get more liquid over time.

Headline in Toucan’s “Introducing Carbon Pools” docs

A commodity pool’s minimum criteria will not account for the range of credits in the pool. People still have preferences for credits in the pool, so it’s not a perfect commodity. If there is an appreciable difference — the ecosystem should benefit from the arbitrage first and foremost. The arbitrageurs are a service provider.

1 Like