Stablecoins on Unstable Chains | Gordon Liao - Circle

[Music] good morning everybody thank you guys for joining us today on this beautiful Friday morning next up we've got Gordon from Circle talking about stable coins on unstable chains right thank you so much for the introduction hi everyone my name is Gordon l i r research at Circle and I serve as Circle cheap Economist today I'll be uh presenting a a short presentation on stable coin on unstable chains so first let me start what defines stable in stable coin so often times when I ask a question the common answer I get is price stability on exchanges that is on uh say coin market cap or some uh dexi what is the price of that stable coin is it close to one is it not close to one but that's actually not the way I think about it at least I I like to think about it uh I think the more fundamental feature of stability is stable coins must meet one toone redemption in Fiat in the primary market with the issuer in real reasonable time so what does it mean one to one redemption in Fiat means that you actually have to have banking access banking real access and in the primary Market means that it's not just trading on secondary Market exchanges sometimes are IL liquid but rather it is actually with the issuer directly I think that's fundamentally what defines stable in stable coin rather than some price theories that's a factor of Market liquidity at a time so now let me get into the technical part so the necessary but not sufficient condition I think is needed is uh there's no double spending of course this is a fundamental issue for all tokens but I think for stable coins is a little bit unique here one reason is uh many of the stable coins and also R tokens are issued um man chains usdc for instance are issued and uh on 18 blockchains and uh there's a chain proliferation which could lead to risk proliferation uh many of these are evm compatible chains but then there are also nonm chains and some of these are relatively new and some of these are not do not have as long a history in addition I think having cross chain transfer as well is important in utility functions but it could actually introduce quite a bit of risk so I think stable coins are multi-chain stable coins are only as stable as the wi case chain that it is deployed on to some extent there are ways to medicate it and I'll talk about that in a little bit uh but for instance Crossing transfer uh of stable coin on chain a to chain B and if there is some real work on chain a uh that happens uh you could end up with double meting so what can go WR uh we've seen many of these uh small incidents in the past there are Forks uh chain splits the D hard fork in 2016 was the largest but that was a decision that's made consensus uh by by the community as well there were temporary splits uh doing the building upgrade due to get errors uh some some of were Prem merg chains and then there were uh now recent proposals of reversing for instance the bit hack of course uh I think there's a gr line there of what type of hack uh should be reversed and what have should not be reversed so I don't think it's the best way for Community to necessarily propose and figure that out uh without some sort of formalism realate works actually happen quite a lot real works on l2s on test Nets on Beacon chain 20122 uh 20 22 uh we actually see more than our fair share of great works uh as a state issuer because we work and partner with many of the l2s and l1s and in the testing phase we experience this you know for instance there was actually a 45k uh rework recently on the test net for a chain that recently deployed and those sort of stuff happens and it does worry us as a stable coin issuer and of course there are chain hals and outages and but that's more common that's probably less of a issue in creating double Mint type of issues so the notable patterns we're seeing is you know client software bugs configuration issues with uh ndes as well as Network stress but I'll focus on client diversity I think client divers is lacking in the space it's because while ethereum has has multiple clients uh L1 has multiple clients many of the l2s don't necessarily have multiple clients and many alternative l1s don't necessarily have many clients so you end up with the same blockchain firm deploying the same code and a small set of validators deploying the same code and uh sometimes those code that's supposed to actually do the validity proof or is not actually doing the validity proof we're not challenging uh optimistic rollups necessarily so those are issues that we are concerned with um another thing that my cryptographer is telling me that you know there's more or more of these blackbox cryptography used on the edges to for optimization purposes that's not well documented so the Primitive stays to be very public but you have more of these uh smaller incremental improvements on cryptography that's not well documented and that creates more trust in teams rather than proofs and cold another issue is stim finality uh for stable coin there's an Institutional use as a payment usage institutions want to make sure settlements are final and irreversible both from a technical but also from a legal perspective so here I think C coin ISU could play a key role and this is normative not positive it's not necessarily everything that cirle does but this is a normally statement from myself on what stable coin issuers could do uh Fork detection so issuer could run phone to monitor consensus of course what you do when there you detects a fork there's a fork choice I think stable coin issuers have an important role to play in ensuring the FK choices adhere to law and requirements in the Fiat space and principles for financial Market infrastructure which might be sometimes different from what the community chooses and I think but that it creates an issue inherently because if stable coin issuer end up choosing a say minority fork for whatever legal reasons that certainly could create confusion so it's not the best thing to do either um providing backup finality and Escape hatching I think that's another uh possible way uh I certainly will show examples of that and then U providing Capital cushion in case of irreversible loss and that's the ultimate Safeguard I think in many instances so is H patch uh what are they well their emergency exit mechanism so example of this is dydx last year decided to uh basically stop their L2 that's built on uh I think Stark X Stark kns infrastructure uh so that led to I think third party actually building uh if I remember like LTB building something like a escape hatch based on the merco proofs uh that leted to $40 million of withdraw safely uh to L1 ethereum based on those proofs but you need to have those proofs in place so the rups need to have the right proofs and for L1 chain deprecation that's more difficult because not necess it's harder to implement a Escape patch for L1 onto another L1 I think the ultimate Escape patch is to Fiat and this is now being required by law in some jurisdictions uh that is you had to meet the onetoone redeemability I part in your bank account for the Fiat value of the stable coin in which it represents um that is challenging of course with some sometimes uh blockchain house and conflicting final States and proofs so that's not a perfect solution either but it's one step in a way of providing that stable guarantee I'll talk a little bit about maintaining an offchain Ledger for legal finality this seems to be uh somewhat redundant but it's it's actually uh sometimes necessary uh so what ISU could do uh which is to store transaction records offchain maybe with some latency after a certain number of uh blocks and this could serve as a legal finality backup this is similar to the concept of transfer agents in the AR blaz space if you know about how buildo the Black Rock uh uh token works is they have a transfer agent which is securitized and that sort of Distinction legally of a separate database that's stating this set of transfer are final I think that provide at least some legal guarantee it's not necessarily what stable coin issuers are doing today but perhaps it's some area that of exploration we could also build a code base for checking updating States separately from the no client code base so that requires a lot more work for every single no to do or for every single chain if you're a large issuer but I think that's another way of guaranteeing that uh the transactions are actually what they are finally I speak a little bit about Crossing transfers and why they are both risky but also areas of opportunity there are areas of opportunity because St coin issuers have this unique ability to not just do the lock and Min bridge but also do the issuer burning on the source chain and minting on the destination chain type of transfers uh using uh messaging protocol in between so Circle has a uh transfer protocol ccpp but the challenge is if there are rate works on a source chain affecting a burn transaction it could lead to double Min or double spend so how do you deal with this how do you get faster speed yet uh some way of ensuring uh there's not double meting uh so circus cctp V2 for instance uh which we recently announced allows faster than finality across transfers and how we do it is we set a global allowance of capping allowes across all the CCT chains for inflight uscc that are in transfer and This Global allowance is effectively absorbed by a reserve Capital buffer in case of devilment what are capital buffer Capital buffer is basically uh uh Access Capital on the equity side that makes up in case their asset Falls be liability so we'll have some framework there and check it out lastly just to mention I think blockchains might not be the most stable not all blockchains are stable but stable coins certainly can with this unique properties thank you so much