While Stablecoins Fight Innovation-Killing Bank Regulation, Banks Are Innovating
JPM Coin, Tassat, & USDF
JPM Coin
JP Morgan has an internal blockchain and internal stablecoin, JPM Coin (they really didn’t need to name it…), that can be used to make real-time payments between counterparties. JPM has also proven operational capability for tokenization, through repo contracts that can run intraday and charge interest by the minute. (Daniel Nielson has a good description of the mechanics here.) This is an improvement from the current situation in repo markets, where the shortest maturity available is overnight, and to actually execute and then close an O/N position takes at least two days.
Moreover, there are rumblings of JPM’s blockchain unit working with “real economy” type players, such as Siemens. While in early stages, this is progress toward significant blockchain paydirt: getting the physical goods scattered all over myriad spots on criss-crossing supply chains onto a shared ledger. There is especially economic value to be reaped here if you can provide payments on the same blockchain going the opposite direction from the goods. The catch here is that you need bank money on the blockchain. Suppliers, generally speaking, don’t want tether, Pax dollar, etc.; they want bank money. Luckily for JPM, bank money is their specialty.
However, JPM’s “Onyx” blockchain unit remains an intra-bank operation. Everything is done on the bank’s books—but on the bank’s blockchain, which allows for programmability and atomic settlement and real-time-ness and 24/7/365-ness. Because of JPMorgan’s size (nearly $4 trillion in assets) and because it’s a counterparty to everyone, this as big deal even though it’s only on an intra-bank basis. It’s JPM being a quasi-central bank again. (Notably, to the extent JPM gets securities and physical goods onto its blockchain, it’s leapfrogging what would be FedNow, the Fed’s upcoming real-time payments system, which is only for money flows.)
Tassat Extends the Tech to the Inter-Bank
Tassat is another interesting case. While not a bank itself, it essentially provides the JPM Coin service to banks. Their TassatPay product allows for banks’ business customers to make real-time payments to each other via blockchain-ifying their deposits. Even more interesting, Tassat recently unveiled “The Network” which is a group of over 100 banks that are going live with a private blockchain for their business customers to also make real-time transfers between/amongst them. So a buyer at one bank could pay a supplier at another bank in real-time for a Saturday delivery through a transfer of bank money. A deposit is simply turned into a digital marker—call it a stablecoin if you want—that is sent via blockchain to another bank on the Network’s blockchain. The end-user is never in possession of a stablecoin de jure; they only ever have a deposit, which is sometimes represented on a blockchain.
USDF – Redeem anywhere stablecoins are sold?
Last week, the USDF Consortium launched—inclusive of five community banks, for now. The Consortium, in its own words: “an association of FDIC-insured financial institutions . . . with a mission to build a network of banks to further the adoption and interoperability of a bank-minted stablecoin (USDF™), which will facilitate the compliant transfer of value on the blockchain . . .”
An interesting facet of USDF’s take on a bank-issued stablecoin is that the stablecoin is redeemable at any of the member banks. This implies that it’s not just a “digital marker” used for payments between accounts and/or between banks; rather it can live on the blockchain for a time, in the possession and at the direction of the depositor. The Provenance blockchain, on which USDF lives, appears to have lending functions and varying degrees of DeFi capabilities, as does Provenance founder and USDF Consortium backer Figure. Could a depositor thus move into USDF and then deposit their USDF into one of these protocols? This would lead to some funky accounting but could, in theory, avoid any double-counting.
While USDF has emphasized the regulation-compliant nature of its product, the nature of the liability is a bit fuzzy. Unlike the JPM Coin or Tassat’s The Network, which truly are just digital markers—i.e., a blockchain representation of a concrete deposit balance—the USDF coin resembles something of a clearinghouse certificate.
Or maybe worse. When the clearinghouse would issue certificates to a member bank facing a run, it took haircut collateral in exchange. It’s not clear that anything of the sort is the case with USDF. Thus, USDF is maybe more akin to a bill of exchange—with multiple signatories as potential obligors. This brings up a couple issues.
Will the FDIC insure USDF? Given that it’s a joint liability of all the member banks, if I’m a large corporate depositor, my incentive is to make a deposit at one of the banks, and immediately move the entire deposit into USDF. I now have a deposit that is the joint liability of several banks—much safer. From an insurer’s point of view, the FDIC should love this feature: to the extent insured deposits move onto the blockchain and are redeemable at any of the member banks, the FDIC would essentially get free reinsurance. The FDIC’s position in all five (at present) members of the consortium is thus made safer.
From a supervisor’s perspective, however, this is problematic. A consortium member bank is on the hook for potentially all the deposit liabilities of another member bank. Moreover, imagine the USDF Consortium hooks a big one. Say, Wells Fargo. (Mike Cagney, co-founder of Figure, which created the Provenance blockchain that USDF runs on posted that “every bank should join” the USDF consortium.) Wells Fargo has over $1 trillion in deposits. To the extent those deposits are able to be moved onto the blockchain and become USDF, they are a contingent liability of the other member banks. This comes with capital requirement implications, possibly unmeetable ones. A failure of Wells could gobble up all five of the small other member banks if they were in front of the FDIC. Moreover, given the aforementioned lack of transfer of collateral ownership, member banks are getting the contingent liabilities with no claim to the failing bank’s assets.
Put simply, a supervisor of one bank might need a clear picture of the activities of all the other member banks (to a greater degree than they already do). Worse yet: this picture could change on a dime during a crisis—i.e., instead of running from one bank into, say a government MMF, maybe I simply run onto the blockchain, creating possible liabilities for all the member banks. Good from a systemic perspective—an aggregation of bank balance sheets in crisis—but muddy from a micro supervisory perspective.
Decentralizing! One less entity
A benefit of all bank-issued stablecoin options is that there is no middleman like Circle or Tether needed to do the transfers. It’s bank money through-and-through (with a side of some weird accounting/transfers in the case of USDF). This means two good things. One, there’s not a quasi-MMF or weakly regulated money transmitter or sketchy offshore company involved in the middle of the payments. Two, there isn’t a huge demand on safe assets created by the desire to run payments on blockchain; this contrasts with nonbank stablecoins, which place demands on the safe assets/collateral market. These collateral demands from nonbank stablecoins bring financial stability concerns. But when it’s simply a blockchain payment of bank money, aggregate portfolio demands need not change. Banks can continue with their mix of risky and safe assets.
(These are early thoughts; Twitter/email replies welcome and encouraged! I turned off replies on here; I’m not starting a blog I swear.)
2/3/2022 UPDATE
The folks at USDF Consortium reached out and shared several helpful and clarifying insights. Among others:
Despite several reports that USDF could be redeemed at “any” Consortium member bank, the stablecoin is at all times attributable to a particular bank. More on this in #3 below.
Some of my concerns noted above are challenges for future iterations of the product, but, at present, an individual’s wallet is controlled solely by the bank. That is, an individual does not have a visible USDF balance that they control; the wallets just serve as a means to execute the transfer (aka, we’re basically back at the Tassat model, but on a public blockchain).
There may be a future in which there is this capability for the USDF to live on the blockchain as I discussed above. In such a case, the bank will still need to maintain some sort of control over the wallet. Imagine a case in which I have a $1,000,000 deposit which I’ve placed on chain in the form of USDF. I’m only 25% insured (de jure). As soon as I see a headline that the FDIC has taken over the bank, I’d want to move my USDF to another wallet I maintain attached to a backup deposit account at another member bank. The FDIC/bank would need a way to stop this transfer the same way they do with deposits at present; without it, we’re back at the wonky reinsurance situation noted above.