Qualified Custodians, a sledge hammer to crack a nut?
The need to “do something” was clear after the FTX debacle, and the response has been swift from the SEC around the need for qualified custodians to cover crypto as they do in traditional finance.
I wrote a blog on Smart Contract Counterparty and Custody Risk in April 2022, prompted by some thinking about how institutions can trade in decentralised exchanges and who their counterparty is, but importantly the issue of custody and a public ledger (DLT) raises many questions.
To summarise my thinking of last year. There was a heated discussion about “unhosted wallets” in the context of AML concerns and the idea that an “unhosted wallet” was a conduit for criminals. I made the argument in an earlier blog about the term wallet being misleading as we do not move tokens in an out of a wallet or maintain a balance that can only be seen by the wallet holder. Rather in the blockchain setting the wallet is the securing of the private key, there may be additional features such as being able to see your balances and transact but importantly the balance is held on the public ledger for all to see.
The use of a term like wallet causes confusion as does custody, again unless you send your tokens to a third party where you are in effect transferring tokens from your address to another there is only one ledger which is public. What we would see on a block explorer is the transaction from an address to another, the debits and credits at the third party is done off-chain to ensure they know who has what on the pooled address.
In traditional finance assets are moved around and there is a huge effort done daily between institutions to reconcile positions to check that their ledgers are correct and each ledger is on a different platform.
I asked the question
Would the term guardian or keyholder be better terms to avoid the ambiguity of what is meant by custody?
The issue of qualified custodians is super important for Centralised Exchanges who hold their client assets in a small number of addresses and manage them using omnibus accounting to account for each client position. The need for segregation of client assets is beyond question, we have seen what could happen if controls are lax. There is a reason why clearing and custody are separate in traditional finance and it is not down to unnecessary “red tape” or some traditional legacy. Qualified custodians are highly regulated and have the purpose of ensuring that the funds are secure and correctly looked after.
Now for the tricky part. What if you don’t want to adopt an omnibus model and require your clients to send you funds or tokens for you to look after them and account for them? There is the world of decentralised exchanges, lending protocols using smart contracts that do not require you to send your tokens to a central entity, rather you look after them yourself.
In an institutional setting self-custody is a non-starter even if it is a multiple-signature address. The inconvenience of requiring your customers to sign every transaction if you adopted a self-custody solution is also a non-starter, they expect a level of service such as best-execution which might be time sensitive. You also don’t want that call with your client who has lost their keys and can’t find where they wrote down the recovery phrase and explain to them that they have just locked themselves out of their portfolio forever and that they will get no sympathy from the tax authorities.
Is the only route to bring in the intermediaries and ensure they are fully licenced? Isn’t that against the whole narrative of decentralisation, namely that we can transact peer to peer without the need of intermediaries.
There are solutions that resolve these issues such as smart contract wallets which give the autonomy of the non-custodial approach taken by Vectis wallet and there is an emergence of other smart contract wallets that are being rolled out or being worked on.
Given that smart contract wallets address the concerns of key recovery, segregation, and security the function of a qualified custodian is largely met, and yet there are key differences on the approaches taken.
As most people transact through centralised exchanges the scrutiny of segregated client funs is welcome but as the sector evolves with innovations such as smart contract wallets and are more comfortable in using decentralised protocols I would argue that a blanket requirement for qualified custodians in a distributed ledger is using a sledgehammer to crack a nut.
