In comparison, every bank in existence would have trouble with this.

Its strange that the high profile mismanaged companies that happened to be dealing in crypto are influencing this kind of exclusion from Federal agencies now. But its also validating that these accounts that are moving will be handled with care, and whatever crypto service those deposits are supporting will most likely be able to function fine. Assuming these were predominantly US-based stablecoins. (Circle’s USDC used to bank there)

Leveraged illiquid organizations with losses cannot withstand a collateral call. Whether that's a bank or a crypto exchange. Alternatives that remove or prohibit the leverage would be great.

Fun fact, the Fed just denied chartering a new crypto focused bank, Custodia, that was planning to keep not just 100%, but 108% of customer cash on hand, thereby foregoing insurance, and make money from fees instead of lending customer money like a typical bank.

> The Fed’s eviscerating explanation for its denial of Custodia cites the crypto bank’s decision not to insure its deposits – instead, the bank proposed to be fully capitalized, holding $1.08 in cash for every dollar deposited by customers – which the FRB said could increase Custodia’s risk of runs and contagion.

https://www.coindesk.com/policy/2023/03/24/federal-reserve-s...

I suspect they aren't talking about runs and contagions on Custodia, but runs and contagions on all other banks if Custodia were allowed to exist. You can't have one bank totally safe from bank runs in an environment where every other bank keeps less than 10% of your cash on reserve. This breaks the system. Either everyone has to be fractional reserve, or no one can be.

That's a somewhat absurd assertion. There's nothing about paying a bank a fee to keep your cash in a safe deposit box which is 1) novel 2) attractive or 3) at odds with a large system of fractional reserve banking.

"You can pay some oddball special purpose depository institution in Wyoming to hold onto your money for you, but they're a bit dodgy and have a bunch of weird ties to crypto. And if they lose it, it's not insured by FDIC. Or you can put it in JP Morgan, and they'll pay you interest, and balances up to $250k are insured by FDIC, although realistically JP Morgan is very, very much too big to be allowed to fail, so the odds of those limits very relevant is almost precisely nil."

If that choice seems at all tricky to you, much less if your answer is "Custodia", you need to realise what an incredibly tiny minority that places you in.

Now, what is a threat to the banking system are narrow banks; for more on that see, eg, Matt Levine a few years ago: https://www.bloomberg.com/opinion/articles/2019-03-08/the-fe...

The Fed was pretty clear that they were refusing to allow this business model because it was, in part, too safe. :)

If the bank isn't insured, then depositors end up taking losses, which means that - certainly for people with less than $250k of deposits - this "safe" bank is strictly less safe than a traditional bank. (And realistically, almost certainly less safe for large depsositors too.) So why would you pay money for a strictly inferior experience as a depositor when a traditional bank would pay you?

And if the bank is insured, then you've got to cover the insurance fees out of the fees you collect from depositors. And the insurance is presumably either FDIC, or something which is some mixture or less generous and/or more risky (FDIC is backstopped by the US government; Joe's Discount Deposit Insurance Agency...isn't.). But even in the best case it's FDIC, in which case...what's the point? You're exactly as safe as a traditional bank, except you're refusing to make loans, so you have to charge your depositors fees.

> If this model were competing with traditional 10% margin banks, it might well be supported by stockholders due to its ability to withstand runs a priori.

It is true that while these models have no benefit to depositors, they nominally benefit shareholders, we are going to be wiped out in fewer cases. (Note, though, that a bank with this model can't neccessarilly withstand runs. Let's say a false rumour sweeps through that one of these banks has had money stolen from the vaults. People correctly realise that if this is true, and they're the last ones to withdraw money, there will be nothing left for them, so the race to withdraw first. Unlike a traditional bank, in this case all the deposits can be returned, but now you've got a bank with branches and staff but no depositors and thus no revenue. And the panicked depositors who pulled their money out and put it in some other bank will have no incentive to bring it back now that you've proven you could cover the deposits. Bankruptcy is likely inevitable at this point.)

But in any case, sure, given the same risk adjusted return, it's rational to always prefer equity in something like Custodia than a traditional bank! But since there's no benefit to depositors, there's no way a bank could make an equivalent return. And while plenty of investors do want low risk, low return investments, it's hard to imagine any of them taking a flyer on something like Custodia, which is likely to be the exact opposite of that.

None of this actually works.

If a manager runs off with 9% of the funds, or if other liabilities cause the bank to be insolvent a bank run is still possible and an orderly receivership would be needed to ensure that depositors were paid out before other creditors.

Deposit insurance is one of the mechanisms that the banking system use to pay out depositors swiftly (usually next business day after receivership). Traditional bankruptcy proceedings could leave depositors waiting for years for their funds, which is a reasonable risk to be concerned about.

Yeah. It doesn't sound like a great idea. Sure, the bank can charge fees for maintaining accounts. But what do you do if you suddenly have a large influx of deposits? To maintain the 108% ratio you'd have to raise additional capital, fine. But what happens when deposits fall? Presumably that would reduce the fees collected and thereby reduce earnings per share with no feasible way getting out of the situation.

> I suspect they aren't talking about runs and contagions on Custodia, but runs and contagions on all other banks if Custodia were allowed to exist. You can't have one bank totally safe from bank runs in an environment where every other bank keeps less than 10% of your cash on reserve. This breaks the system. Either everyone has to be fractional reserve, or no one can be.

Why? As long as my bank pays more interest on my (insured) deposits than a hypothetical full-reserve alternative, why would I switch to them? Why would anybody (below FDIC limits)?

Banks keep cash on hand inside vaults, on deposit at their local Federal Reserve Bank, "in transit" (stored at an armored car company), or in accounts at correspondent institutions internationally (primarily for settling debits and credits on international wires).

Cash in vaults has negative yield (the vault itself isn't free, nor is insurance for it), and if you don't like fractional reserve banking and/or counterparty risk, foreign correspondent banks don't really seem like a good way to eliminate it either.

[1] https://www.bloomberg.com/opinion/articles/2019-03-08/the-fe...

Now narrow banks - as proposed by, among others, TNB USA - would be effectively entirely safe and could pay interest. Efforts to set one up were shut down by the Fed because they would, in fact, be too attractive compared to traditional banks. See, eg, https://www.bloomberg.com/opinion/articles/2019-03-08/the-fe...

Yeah, it's an absurd claim and the Fed is correct in denying it. It's welcome to exist somewhere people will believe it has magic money printing mechanisms.

Heck, if it worked, the bank could put in $1m, suddenly have $1.08m, invest that in the same money magic mill repeatedly, and soon have all the money it wanted.

All this on the "promise" that nothing will go bad.... Sorry, the purpose of insurance is to share volatility, and it works quite well. The purpose of claims like Custodia is making is to skirt regulations that exist for solid, historically demonstrated, reasons, so they can play loose with people's money.

Not sure what that would look like. Maybe a fee on every deposit?

Occam's razor.

Ah, the mythical "properly run crypto custodian". Does such a fantastic beast actually exist?

I'm purely in tradfi investments now because, well, look at crypto..

Honestly I would expect the market to be won by an exchange that plays by the rules. The dodgy ones will outcompete in the short term, but they are prone to instant and complete destruction in a way that the well run ones aren't. Additionally, it's clear as day that regulation is coming, so those exchanges who are already compliant will have an advantage.

Nexo pulls out of US market (https://financefeeds.com/nexo-cease-its-business-operations-...)

So yeah...

This is no different than any hedge funds prime brokerage, which is the more appropriate comparison.

Both can lend out the securities that they hold but both should be able to call back and make their depositors whole.

Crypto has nothing to do with this as far as I can tell.

Which crypto custodians are those?

> In comparison, every bank in existence would have trouble with this.

That isn't true. Schwab just announced that they can cover 100% of bank deposits without selling a single security.

https://www.financialadvisoriq.com/c/3995094/511764/schwab_l...

Why sell a security that's only worth 74 cents on the dollar when you can use it as collateral to borrow 100 cents on the dollar, and if you can't pay back the loan then the public will bail your ass out?

Most crypto statements can be suffixed with "in theory".

This is all properly-run custodians. They don’t make the news because they shouldn’t. Institutions demanded custody for crypto because they wanted the safety traditional custodians provide.

In the old days, CFOs were trained on sweeping cash across FDIC-insured accounts and custodying Treasuries and the like. Apparently, that is now obsolete.

The big difference is that you pay for the privilege of using a banks custody services. That’s why most people don’t use them for cash, it’s equivalent to putting cash in a safe deposit box, it depreciates even in deflationary times.

Signatures crypto clients were not doing that. They were loaning money to the bank in order to achieve yield. No matter who it is you loan money to, you have counterparty risk, banks just happen to be one of the most regulated debtors in the world and depositors have unprecedented creditor rights.

No crypto firms have that same level of security.

They also didn't cause Signature to suffer a run so severe they couldn't even say for sure what the size was of their deficit at the Fed: https://www.fdic.gov/news/speeches/2023/spmar2723.pdf

You're right, Elizabeth Warren was just inciting panic talking about a rank run when there wasn't one in order to create one

They lost 20% of their deposits in a matter of hours on the Friday of the SVB collapse which left them with a negative balance at the Fed. When the Fed and FDIC asked them how many withdrawals were queued up for Monday morning and how much money they needed to borrow from the Fed window to stay solvent, the bank couldn't tell them because their systems weren't capable of doing so. Absolute clown show.

https://www.cooperkirk.com/wp-content/uploads/2023/03/Operat...

https://www.federalreserve.gov/newsevents/pressreleases/othe...

And in your mind it’s reasonable that someone tanked Silvergate and Signature to send this message? And their investors and lenders are just cool with that?

The FDIC isn't a bank. Signature was largely bought, but "Flagstar's bid did not include about $4 billion in deposits related to Signature's digital-asset business." The FDIC is telling depositors they couldn't find a buyer for to find a new bank.

> they pose a threat to the future hegemony of a government-backed digital currency

FedNow is not a CBDC. It's a modern set of payment rails for interbank transfers.

According to whom? (It’s not true.)

I would maybe believe the argument that the buyer of the bank didn't want to take on the depositors, if it weren't so convenient for the Fed. As it is, I have trouble believing the FDIC didn't pressure the buyer to make that distinction between depositors as part of the sale process.

> FedNow is not a CBDC.

No, but it's the clear first step towards one.

When you bid on a banks assets and liabilities, you get a bunch of spreadsheets. These describe the bank's assets and liabilities, including its depositors. You're allowed to say "not that branch" or "no depositors matching these parameters."

> what makes them fundamentally any different

They're riskier. Silvergate banked FTX. That was a fraud, and now Silvergate is gone. Signature banked less crypto, but when Silvergate began failing Signature started to slide. Flagstar doesn't want that problem. (I imagine there are also AML/KYC concerns. Signature's crown jewel was its high net worth clientele. That gels with NYCB [1].)

Another risky type of depositor are financial institutions. It's not uncommon, in bank takeovers, for deposits from certain types of financial institution (e.g. hedge funds) to be excluded.

> it's the clear first step towards one

This takes imagination. They're separate tracks. FedNow competes, most directly, with Zelle. It's an always-on modernization of Fedwire, a faster ACH. In any case, the zeal for a CBDC has cooled with crypto's fall and the advent of universal deposit insurance.

[1] https://en.wikipedia.org/wiki/New_York_Community_Bank

I don't want to sound like that guy in the SVB thread who ended up advocating for Sharia law, and I've got to go to bed, so I'll leave it at that.

I would also implore you to consider that closely related (quasi-)government agencies like the Fed, FDIC, SEC and Treasury are all capable of maliciously coordinating with each other in their timing of enforcement actions and agendas. And maybe they're just doing their job of protecting the dollar, but IMO it doesn't make it any less obviously corrupt.

One, correlation. A bank with one $1bn hedge fund is riskier than a bank with a billion $1 individuals. This is the part regulators focus on. (One oversight with SVB was not classifying VCs as quasi-financial institutions.)

Two, reputation. As we've seen, there is a premium to being able to say you're clean of crypto when e.g. Binance or Tether go under.

Three, cost. KYC and AML, for one. (It's why many banks won't open accounts for gambling businesses.) But also how often they use revenue-generating versus free services. If you're constantly wiring in and out your entire balance, you're substantially more costly to your bank than someone who deposits leaves it alone.

Finally, core competence. Flagstar/NYCB focus on a niche. They have built their processes to deal with those customers' demands, trained their staff to be familiar with the, and hired people who like working with them (and with whom they like working). They don't want a bunch of crypto deposits any more than they want aerospace manufacturers'. It's a different business.

A regional bank is choosing to stick with its area of specialty. There remain banks who specialize in crypto, and differentiate themselves on that basis.

> it just seems morally wrong to me

Only one component is related to withdrawal probability. The others, like service cost, fraud risk, et cetera, are business decisions. It's odd to say a bank choosing a specialty is wrong. (Or vice versa, that a bank should be forced to serve clients it finds morally repulsive.)

That said, you and I see eye to eye when it comes to natural persons: a bare-bones bank account should be entitled to. In the era of cheque-writing, there was a necessary credit component to account opening. That's no longer necessary. (For businesses, it would require dismantling our AML system.)

A bank's entire business model consists of borrowing money for variable-length terms from depositors, and lending it out at fixed-length terms to borrowers.

This question is absolutely the bank's business. In fact, it is the most critical question that a bank needs to be asking itself every single day, if it wants to stay solvent, and actually be able to honor customer withdrawals.

SVB, and its shareholders have been wiped out because it didn't have a good answer to that question.

Alas, the people who dared to think from those first principles are now becoming unbanked from, or denigrated as criminals by, the very system they sought to replace. Perhaps they were onto something.

Folks doing honest research are fine. Those scaling glorified slot machines are having a tougher time.

I think it's quite likely that any society that wouldn't adopt it would be overtaken in economic productivity by a society that did.

> Alas, the people who dared to think from those first principles are now becoming unbanked from, or denigrated as criminals by, the very system they sought to replace. Perhaps they were onto something.

They are generally onto entrusting their money into organizations and investments ran by, or successfully attacked by criminals, which is why crypto money has a disturbingly frequent tendency to go poof.

The moral of the story, here, is that if you have an ideological reason to... Somehow stick it to the man and stay out of the insured banking system, you're going to be in real bad company.

(Also, isn't it kind of weird how quick crypto was to re-invent sub-prime lending, through coin staking, and how quick exchanges are to gamble with and lose customer funds?)

The goal is to do as much damage to crypto due to the risk against the dollar and prior to the FedNow rollout. If the bank gets hurt in the process then that's cool b/c it teaches other banks not to do business with crypto.

Regulators scrutinizing banks who deal with "high-risk" customers while not defining what risk means ex ante is absolutely a problem. This is not that. It's, first, an entire bank being turned off, not specific accounts. And second, a regional bank with a long-running specialty choosing to buy accounts that gel with that specialty.

The fact that a retired ranking Senator is blabbing to the papers versus filing a lawsuit should say enough with respect to his conviction in his claims.

But keeping wire transfers and ACH expensive and/or painful seems like a terrible way of achieving that.

edit: Never mind, I can't read. That said, no one should listen to Frank about why he's not responsible for his bank blowing up.

Timeline:

Nov 11: FTX/Alameda file BK.

Nov 28: Sen Warren, retired Sen Marshall send letter to Silvergate requesting all records relating to FTX. BlockFi files BK.

Dec 5: Deposits fall from $12B to $3.8B. Silvergate sells treasuries at $718M loss and take out FHLB loan for $4.3B to remain solvent and fully offramp depositors.

Dec 31: Silvergate reports 4Q loss of $1B.

Jan 5: Judge in FTX BK orders seizure of FTX assets at Silvergate.

Feb 8: DOJ announces investigation into Silvergate role in FTX AML/KYC failures.

Feb 14: Sen Warren-friendly Citadel reports 5.5% stake in Silvergate. In June Ken Griffin, CEO, announced he might like to be a crypto market maker.

Mar 1: Silvergate announces delay in annual report, says may not be able to stay in business.

Mar 2: Coinbase, Galaxy Digital drop Silvergate.

Mar 7: Silvergate says forced to pay $4.3B FHLB loan early which was the final nail in its coffin.

Mar 8: Silvergate announces liquidation although still solvent with assets exceeding liabilities of over $1B. Still, shares drop to $2 and a $48M market cap.

Mar 10: FHLB issues press release specifically denying it forced early repayment but "no comment" on whether usually automatic roll-forwards granted routinely were specifically excluded for Silvergate despite oral assurances they would be granted.

Mar 24: Silvergate stock doubles on rumors Citadel wants to acquire assets.

Mar 31: Sen Warren seen smiling while reading article on crypto bank collapses.

You be the judge...