[fsckboy]

no bank is going to give carte blanche to large customers to run the bank, they can't, and even if they promise it... As with all businesses, like a restaurant wants you to enjoy your food so you come back, the bank wants to provide the services the customers need. But everybody can't have everything especially all at once, and banks can't provide liquidity that has dried up.

sure, customers will periodically get mad and take their business elsewhere, but it's a fever dream to think that elsewhere is any different.

[quesera]

> no bank is going to give carte blanche to large customers to run the bank

If the funds are in a DDA (Demand Deposit Account a/k/a standard US checking account), the bank is legally obligated to honor outflow transfers of any size, at any time.

(Not physical cash withdrawals, although they have to make timely arrangements to satisfy these too)

Banks can make DDAs unattractive for large balances (e.g. no interest), but I'm not aware of any limitation on a customer's right to access their funds held in a DDA.

If banks could bend these rules, bank runs would never happen.

[fsckboy]

> the bank is legally obligated to honor outflow transfers of any size, at any time

they should have legally obligated SVB to remain solvent!

[quesera]

> they should have legally obligated SVB to remain solvent!

They did!

When SVB failed to meet those regulatory obligations, the bank was seized and the depositors were made whole.

So the system worked, right?

[tripletao]

The FDIC and Fed made policy changes in response to the SVB's failure--the FDIC is insuring all the SVB's deposits, including those >$250k, and the Fed is allowing all banks to borrow more than the FMV against certain assets that lost value when interest rates increased.

Without these changes, the SVB's depositors would have had access to maybe 50% or more of their uninsured money immediately, and maybe 90% or more eventually. They'd maybe have been made whole eventually; but the FDIC's inability to find a buyer over the weekend suggests that the SVB's assets weren't obviously greater than its liabilities to depositors, so maybe not.

The SVB's depositors have been made whole now only because regulators intervened with emergency policy changes to rescue them. That might have been a good idea, since it stopped contagion; or it might have been a bad idea, since it encouraged future risk-taking in anticipation of a similar ad hoc rescue. It certainly wasn't any kind of rules-based system working, though.

The SVB had been insolvent on a mark-to-market or NPV basis since around September. Accounting rules on bonds they intended to hold to maturity allowed them to ignore that, but didn't change economic reality.

[quesera]

FDIC did not need a policy change to insure more than $250K per, that was a predefined option in existing policy, available if the bank failure was judged to have risk of systemic contagion.

I'm no economist, but I think the (1-year) window for banks to borrow against the full face value of government bonds and MBS assets is interesting and probably reasonable. These are not risky investments, just illiquid. The Fed will get their money back. Providing liquidity to the system is part of the Fed's job.

[tripletao]

While alive, the SVB deliberately avoided designation as "systemically important", in order to avoid the corresponding regulatory burden. Then as soon as it failed, its depositors got beneficial treatment under a "systemic risk exception". I'm not saying that was necessarily the wrong choice; but do you really think that's a coherent rules-based system?

> These are not risky investments, just illiquid.

It's not a question of liquidity. Similar assets trade with tight spreads, at prices very closely predicted by a textbook NPV model. The price just went down when interest rates went up, exactly as expected. There's no significant uncertainty in the price. Waiting won't make it go up, except in the same sense that waiting turns $100 in Treasury bills into $104 a year from now.

The Fed is making an undercollateralized loan. If a bank fails with such a loan outstanding, then the Fed will lose money. Interest rate risk is as real as credit risk or any other risk, and this would be a real economic loss.

[randomname11]

Well the fed already has the money which was used to purchase the bonds that they’re are now lending against. So I think that this amounts to giving some of that bond principal back early but slightly less in total in the form of interest on the loan.

[dwighttk]

>it might have been a bad idea, since it encouraged future risk-taking in anticipation of a similar ad hoc rescue.

From what I understand (could be wrong) the discouragement to risk taking is that the risk takers were wiped out in this case, only the depositors kept their money.

[tripletao]

The SVB's shareholders have indeed been zeroed. The shareholders of other banks will benefit from the undercollateralized lending program though, in proportion to the amount of bad interest rate risk they took.

I don't think these are necessarily bad decisions--it's reasonable to make an example of the worst offender, and then help the rest survive to prevent systemic contagion. It's absolutely not "the system worked" though, except to the extent that the system is regulators making stuff up on the fly.

[renewiltord]

We moved our stuff out of First Republic. Everything but $250k. Happened without a hitch. No one was mad.

[yieldcrv]

actually SVB and First Republic filled this niche of low friction and convenience.

so it was a perfect storm, for them.