SVB

The Power of the Uninsured

Here is key phrase of the FDIC press release from March 10, 2023 concerning the closing of Silicon Valley Bank (SVB):

“ …Uninsured depositors will receive a receivership certificate for the remaining amount of their uninsured funds.”

Wait, what, did they say uninsured depositors? That is not supposed to happen.

So what should have been a classic example of how the provision of deposit insurance prevents a bank run turns into a story of a run. But this run happened in a world deposit insurance — exactly what the basic theory tells us cannot happen.

The key here to keep in mind is that there is a limit on deposit insurance of $250,000. If you have an account with a balance exceeding that amount, then you are exposed to a run. And evidently run they did!

Remember the Theory

Our theories of banking, building on the Diamond-Dybvig model, discussed in an

Bank Runs or Not:

earlier post, argue for the stabilizing role of deposit insurance. That argument rests on all deposits being insured.

So normally, when thinking about banks, we imagine that individuals wishing to have deposits in excess of $250,000 will split accounts across banks so that each account is insured. Well that seems not to be the case. After all, if you are running a business managing the flow of cash across accounts can become a nuisance if not expensive. And why bother if the financial system is stable.

Until it isn’t and then you have to move fast in order to protect deposits that are not insured. And so they did …. and we had a bank run. This was, as far as I can see, a problem of liquidity not insolvency. As usual though, the run did not come out of thin air but was triggered by “news” about SVB.

So how far off was the theory based on the assumption that all depositors will be insured. The FDIC seems like it is not saying .. yet. I did read an article stating that more than 85% of the deposits were not insured. This does not mean 85% of the depositors, just 85% of the deposits. It can be that there are a large fraction of depositors (households and businesses) with deposits under the $250,000 cap, leaving only a small number of depositors who have chosen to be exposed to the risk of a run. But if that small number have significantly large deposits, then their withdrawals can be enough to cause a liquidity crisis.

Evidently they did. And their actions had big effects to the extent that the financial operations of this bank mattered for business activity.

¹

Is there an easy remedy?

The $250,000 limit is not cast in stone. So isn’t just raising that limit a natural solution?

There might be a few arguments in favour of a limit. The first has to do with a cost of deposit insurance: who monitors the banks? If you have funds in a bank that are not insured, then you have an incentive to watch carefully what the bank is doing. Once insurance is in place, those incentives are gone. They are supposed to be replaced by the government in a regulatory role. Right.

A second has to do with commitment.

²

In order for deposit insurance to work the government must be totally committed to its provision in the case of a run. So if the limit on deposit insurance was lifted, in order for it to ward off self-fulfilling runs, everyone must believe the government will meet its promises in the event of a bank failure. If there is doubt, then magic of the promise of deposit insurance is gone.

Now, here it gets a bit tricky. If the only reason for a failure is a run, then if everyone believes in the government provision of deposit insurance then there will never be a bank failure.

But should everyone believe that the government will provide this insurance? A strong argument in favour of this would be that, after a run, the government will have an incentive to provide this insurance.

So we need to think about the incentives of the government after a run. And here, finally, is where the limit comes into play. With a very high (or no) limit on deposit insurance, then if there is a run the government would be, under the provision of deposit insurance, making transfers to households and businesses with large accounts. These transfers, of course, have to be funded by taxes.

So, and this depends on the tax system, it might be that providing deposit insurance entails a redistribution from relatively poor households (Main St.) to some wealthy households (Wall St.). This is something that a government might actually choose not to do. If so, unless there is an ironclad commitment to pay deposit insurance, raising the limit too high may not be an effective way to avoid bank runs.

For deposit insurance to work its magic it must be credible. So the response to this event will have to be more complex than a simple extension of deposit insurance.

³

1. The effects on the depositors with deposits under the cap seem minimal as they were will made whole by March 13.

2. In the spirit of total disclosure, this is an area of my research with Hubert Kempf.

3. I guess this means I do not believe in government promises unless they have an incentive to fulfill them.