One concern is that depositors will realize they can get better returns on money market funds and withdraw from banks en masse.
The whole point of capitalization stress-tests is to determine that banks can withstand a certain amount of withdrawals.
It of course is going to be much much more likely on a bank that only has 3% FDIC - I have no desire to "run" on my bank because I am below the $250k limit, and even if the accounts were frozen for a week it wouldn't be that worrisome.
But if I were a startup with $100m at SVP, I would have been freaking out earlier this week.
Not only is there the 250k FDIC limit, but that's leaving money on the table. At least put it in highly liquid and stable securities of some kind so you're not bleeding it. And, of course, accounts can get locked out. Anything over a couple million should be split across two different institutions minimum.
Suppose you have 500 engineers each making on average $200k (some make above, same make below). That is about about 17000 / month or 8500 biweekly. It takes time to transfer money. What options do you have for storing the money that is immediately payable. Stocks are volatile. Treasury bills or bonds could work, but it would be a good idea to have it such that the money is deposited a day or two early so you can make pay roll.
500 * 8500 = 4.25 million, so even a medium-size startup simply using SVB for their payroll checking (due today, so they should have had it in by yesterday, and recall ACH processing times of 2-3 days) could easily have a few million sitting in a checking account as part of weekly practice.
Additionally, there is at least one person at the company whose full-time job is to handle this. Making a transfer every Tuesday or Wednesday or whatever is well within a workload. When I say "sitting in checking" I quite obviously mean more than two weeks by "sitting".
Also, I was mostly referring to series-A startups that have, for example, single digit millions in the bank. Keeping that all in checking is dumb. Larger companies that are moving millions each week can afford to have special arrangements with their bank or banks to limit liability independent of FDIC. A line of credit, perhaps, that is paid off a week or three later, secured by a bundle of securities held at the same institution, for example. I'm fairly sure a bank run can't touch those.
There are so many possible solutions to this problem.
Maybe it's because I've been burned hard by banks before, but I'll never trust them to not do precisely what SVB did here (in some fashion). Everything's great until it isn't.
What? The third-world banking system of the US often takes MUCH longer than that.
ACH transfers are also 24 business hours.
The check clearing thing is called Check 21 and it was a big thing back in 2004 when it went into effect.
https://www.federalreserve.gov/paymentsystems/regcc-faq-chec...
zero risk is not a thing
The problem was that this bank almost exclusively used long-term bonds to back short-term liabilities. This was extremely poor risk management on the bank's part, compounded by a truly idiotic decision to sell some of these bonds at a huge loss, necessitating an emergency capital raise (see Tuesday's press release).
The CADFPI and FDIC clearly agreed, since they took over the bank at the start of the day to prevent it from digging itself into a bigger hole. (Usually they wait until Friday EOD, and transition control over the weekend. WaMU was a much bigger collapse, and they actually waited several days and took the time to find an acquiring bank before taking it over.)
Once their HTM assets’ market price fell enough that rebalancing would force them to admit to insolvency by recomputing their reserves, they literally could not do anything with them except hold and pray that they make it, which may have just made things worse. Who knows how long they’d really been underwater
How would that have helped?
Once that 100B falls to 95B due to interest rate hikes they could sell them all and buy 95B of short term treasuries instead, insulating them from any further rate hikes
It's straightforward to reckon their exposure to interest rates: they had $90B in 10-year fixed rate bonds, so they lose $9 billion per % of interest increase. They must have known that a 4% increase in interest rates would put them underwater, but they did it (and were allowed to do it) anyway. It'll be interesting to learn about the process behind that decision.