[chii]

> who should have been the optimal holder of their risks?

they _produced_ more risk (by holding long maturity bonds that lose value as interest rate grows). This risk was not something that is inherent - they could've chosen not to do that with the large deposits from the pandemic money growth.

There's noone who can be the optimal holder of the risk that is produced this way, because there's no value on the other end - SVB is taking the full value already (the interest payments on said long bonds).

If someone were to hold that risk, SVB would have to pay out premiums that would surpass the interest income they receive.

The alternative is for society (aka, the central bank) to hold that risk. But this just means socializing the losses but privatizing the gains - something i'm very much against.

In the end, SVB was the optimal holder of the risk (that they produced for themselves). And they can't actually hold that risk - thus their failure.

[hbarka]

What about First Republic Bank?

[chii]

I know less about FRB's failure. It was likely due to a domino effect from SVB's - specifically, FRB has a high uninsured ratio of deposits (they service rich people).

The FDIC has announced that they will not do a repeat of what they did for SVB - insure the full deposit amount rather than just the $250k. Therefore, anyone with a large deposit in a small bank is going to want to move their money out into a "too big to fail" bank.

Unfortunately for FRB, this is what happened to them. No bank can survive a real run, no matter how carefully balanced they are with risk (after all, they _do_ take on some risks in order to make a profit).

In my opinion, the FDIC's announcement of what they will not do (insure the full deposit, even if above the $250k limit) after doing it for SVB, while have good intentions, is what backfired.

They should've just lied, and said that they'd do it for another bank, if there's a need to; this would've stopped any fear of a run, and thus stop the run before any more dominos collapse.

[flagrant_taco]

> They should've just lied, and said that they'd do it for another bank, if there's a need to; this would've stopped any fear of a run, and thus stop the run before any more dominos collapse.

While this may have prevented FRB, that's a very dangerous game to play should the bluff get called.

I'm strongly opposed to the idea that those given the power and authority to control or markets, as best they can, should world that power by lying to us. Lying because they think it's the best thing for us or because they don't think we can handle the truth is a slap in the face to the very trust that empowered them to begin with. Our leaders do this often and it's such a slippery slope - it either works and you feel emboldened to lie again or it backfires and we're all worse off.

[chii]

The thing is, this white lie is what keeps confidence levels high, which is what prevents the run.

By merely suggesting that a bank can fail, and that the FDIC is not going to bail out high depositors, they paradoxically _cause_ the run. After all, the people who took the money out just merely redeposited it back elsewhere (that they trusted more).

The white lie is better than a loss of trust which lead to an actual problem. And the FDIC could actually lie without lying by putting in vague words and misdirect people - such as saying things like "if necessary". In fact, people in society today believe plenty of white lies already - what's one more?

[flagrant_taco]

While I totally agree that is how the system is designed, that's also the fundamental issue I have with it.

If we have such a fragile banking system that those in charge are expected to lie to us to keep people from seeing the fragility, we have to rethink the system.

> In fact, people in society today believe plenty of white lies already - what's one more?

That feels like a bit of a slippery slope, selling people on one lie shouldn't justify telling another. It also means first defining what a white lie is, and who gets to know the truth to decide whether it's acceptable or not.