Just because crypto hasn’t triggered a systemic collapse doesn’t mean that it won’t be the perpetrator of quiet banking crashes. We would do well to remember that the 2008 calamity came shortly after the collapse of small subprime-mortgage finance companies. These would have been proverbial dead canaries had anyone looked down the mineshaft. And, even as the U.S. subprime crashes formed into a single, torrential crisis, bank regulators confidently foretold no systemic impact because they comfortably believed that no bank had undue exposure to high-risk mortgages. So bank regulators still say now when it comes to crypto and let’s hope the outcome is different this time. However, bits and pieces of bank wreckage are already to be found in FTX’s rubble and may well surface as the crypto tide continues to ebb. No bank shipwrecks have emerged, but some of the wreckage has the look of a sizeable hull.
The most tantalizing bit of banking wreckage is a super-tiny Washington State bank which FTX appears to have surreptitiously acquired. As the New York Times reported, one of FTX’s affiliates last March invested more than double all the capital previously held in Farmington State Bank, doing so in a carefully-structured way to avoid triggering legal control thresholds. The bank is the nation’s 26th smallest and, after this generous investment, it deposits went up about 600 percent from its initial $10 million level via four new accounts. Sill more intriguingly, Farmington’s crypto ties via shadow owners appear to go back to 2020 even though the bank then and before FTX’s acquisition had negligible assets and only three employees.
The undeniable appeal of a tiny corner of the banking system? As John Heltman pointed out in the American Banker, acquiring a bank gives you a key – small though it seems – that unlocks the magic door into the global payment system. Neither FTX nor Farmington’s former crypto pals seem to have used this power, but little is still to be known about what they did for whom. At the least, this is dodgy, as is the Fed’s current payment-system access policy.
I’ll turn back to the policy questions this raised by a stealth-crypto bank, but it’s important to note another way in which crypto has wormed itself into regulated banking even though the banking agencies thought their charges immune from contagion risk. As Bloomberg reported, several insured depository institutions of middling to even regional size were more than happy to accommodate crypto companies with giant deposit accounts. For example, Silvergate offered a payment service that topped $1 trillion in cumulative payments processed earlier this year. Perhaps it’s thus unsurprising that crypto deposits are said to be ninety percent of the bank’s deposit funding at the end of September. Bloomberg notes that these deposits were invested in long-dated and often held-to-maturity securities, putting the bank – which has $155 billion in assets – in more than a bit of a bind.
Signature Bank is another that went big into crypto. The $114 billion bank had about a quarter of deposits from the cryptosphere, and several other mid-sized banks are also seeing significant risks as crypto funds evaporate. One of these banks, Provident, has 58 percent of its loan book in exposures to crypto miners and other entrepreneurs. This bank once pronounced that “old banking is boring.” So it may be, but staying thrilled isn’t the best way to run an institution that takes other people’s money with backstops provided by taxpayers on the assumption that money won’t come from volatile deposits flowing to God knows where for who knows what.
Thanks to all the boring rules, it may well be that all these crypto-vulnerable banks manage deposit outflows and credit losses with aplomb. I certainly hope so, but these cases reinforce two critical policy conclusions.
First, the Fed must get its payment-access act together and go beyond the like-it-when-a Reserve-Bank sees-it rule on payment-system access eligibility. The Bank Policy Institute suggests an access-eligibility process akin to the BHC-charter process with prior notice and comment and this would indeed be better than the Fed’s opaque construct on a question of systemic scale and with profound competitive clout.
It seems clear that the FTX/Farmington transaction was carefully constructed to avoid a change-in-control notice and it thus might not trigger any filings under new rules because the bank already had payment access. As with so much else about this transaction, the extent to which this structure was deceptive will come out in time. For now, though, it’s even clearer that Fed payment-access processes are easy to hijack and ages behind the times.
Second, the crypto calamity makes it clear that deposit concentrations may pose as much risk to bank solvency as the single-counterparty credit exposures long targeted by bank regulators. Liquidity rules are posited on diverse funding sources because all but the smallest banks have long had them. There seems to be no reason to rewrite the liquidity rules to mandate single-counterparty constraints, but there’s a lot of reason to ask why examiners didn’t notice funding concentrations with posed acute run risk.
To be sure, FDIC insurance slows down the race to the exit for any depositor not in urgent need of its funding if its deposits are relatively small or it believes – not unreasonably from past history – that something will salvage any bank of size. However, shame on a banking system that allows banks to survive only due to access to FDIC insurance, the Home Loan Banks, Fed discount window, or other taxpayer backstops. Insured depositors must be made whole, but banks should otherwise pay for their sins via either liquidation or severe sanction if only the moral-hazard imprimatur can save their skins.
Fourth quarter results will tell us how much damage crypto really has done to vulnerable banks assuming all of them stay liquid and solvent till then. Should a happy fate befall all these banks and none fails, that’s all to the good. But, it still won’t answer vital questions about how these banks got so close to the edge and what kept them from going over it. The answer to this question is at least as critical as that to how the overall cryptoverse should be governed going forward.