Bailing Out SVB Was a Really Bad Idea
On CBS’s Sunday morning interview show, Face the Nation, Treasury Secretary Janet Yellen stated clearly that the federal government would not bail out the failed Silicon Valley Bank. Many of us took that to mean that there would be no bailout. But Yellen said in the same interview that the feds would try to meet the needs of depositors. Translation: there would be a bailout, not of the shareholders of SVB, but of depositors. This would include those whose deposits were above the FDIC-insured limit of $250,000.
The bailout is a terrible idea. It increases moral hazard. It creates uncertainty about the rules. And it suggests to participants in a market economy that if they have ins with the people in power, they will get special treatment. The bailout adds, in short, to what philosophical novelist Ayn Rand called the “aristocracy of pull.”
These are the opening two paragraphs of David R. Henderson, “Why Bailing Out SVB Is A Bad Idea,” Defining Ideas, March 16, 2023.
On March 12, former treasury secretary Lawrence H. Summers tweeted, “This is not the time for moral hazard lectures or for lesson administering or for alarm about the political consequences of ‘bailouts.’ ” Actually, it is exactly the time for all of that.
But imagine what would have happened if the federal government had stuck to the rules. Yes, there would have been pain. Yes, several Silicon Valley companies would have had trouble meeting their payrolls. But the reason that SVB was not counted as a “too big to fail” bank was that the federal government had judged that there would not be a system-wide run on banks if SVB’s depositors had taken a large haircut. The impact would almost certainly have been regional, not national. In short, “contagion” likely would have been limited. Then banks, depositors, and others would have thought much harder in the future about what to invest in.
The FDIC, Janet Yellen, and the rest of the feds who were involved had a chance to do something good for the economy: stand firm on existing FDIC rules. They blew it.
Read the whole thing.
Mar 17 2023 at 10:07am
I thought they were only guaranteeing money for the depositors and not bailing out the bank itself. From what I have read it looks as though it was more of a liquidity issue than insolvency.
Mar 17 2023 at 10:09am
You thought correctly, as did I. Notice that I said that in the passage quoted above.
Mar 17 2023 at 10:26am
Well, the lesson for shareholders and managers seems to be:
Manage your bank taking a lot of finantial risk. Being more aggressive that what good finantial sense suggest. Just be careful not to be the riskiest balance sheet in town. All you need to be fine, is a couple of banks going down before you get into troubles.
I don’t think that’s a good message to send.
If SVB had not been rescued, some (many) other banks would have gone done. That looks to me like a bail out of the shareholders and managers of those other banks. So Yellen can, maybe, say “we are not bailing out SVB”. But she can’t say “We are not bailing out some banks”
Mar 17 2023 at 10:30am
There was a real problem with losses; they had a lot of longer-term Treasuries whose value declines significantly with rising interest rates. The accounting issue was that these were “held to maturity”, meaning that they were valued on the books at par. If you believe that accounting, they’d be OK. But in a run context, they had to sell those Treasuries and accept a loss: The active market values those bonds at significantly less than par. In a mark-to-market sense, they probably were insolvent. Under other circumstances they could have dealt with the run by using those bonds as collateral at the Fed, and the message is now the Fed will accept such bonds at par. That’s a statement that the government really, really doesn’t want a run on other banks now, but is likely to lead to trouble later.
Mar 17 2023 at 10:51pm
No, they had virtually no treasuries. Almost 100% of their HTM portfolio was agency MBS/CLO. Which compounds the duration mismatch because rising rates significantly extend the duration.
Mar 18 2023 at 12:16pm
Rising rates extend duration?
Mar 17 2023 at 11:05am
They bailed out the venture capitalists who now no longer have to pick winners by choosing which companies to provide bridge funding for as SVB’s receivership process played out.
Mar 17 2023 at 11:18am
The stakes were much bigger than just SVB. There would have been bank runs everywhere.
Mar 17 2023 at 12:22pm
In 2018, SVB successfully lobbied Congress to change the rules so that any bank with less than $250b of deposits isn’t too big to fail.
Mar 18 2023 at 12:17pm
So did the Frank in Dodd Frank.
Mar 17 2023 at 3:29pm
The risk from this was overstated (other banks wouldn’t have had such a singular consolidation of uninsured depositors) and could have been managed other ways.
This was about helping VC’s – a primary donor class of the political party in control at the moment.
Mar 17 2023 at 3:50pm
Not by insured depositors. SVB was unique in having 90%+ uninsured depositors. Most banks it’s 50% or less, sometimes far less, right? If you have 50% or less you can always pay out the uninsured deposits with the insured ones.
Mar 18 2023 at 12:15pm
When the FDIC takes over a bank, are you sure it wouldn’t split the bank’s cash evenly among all depositors first?
Mar 17 2023 at 7:20pm
Where were the “running” depositors putting their money? In more stable banks, not under their mattresses. Yes, there would have been (and still may be) more runs, but it wouldn’t and couldn’t have been system wide as long as deposits were moving between banks, not out of the banking system.
Mar 18 2023 at 9:16am
George Selgin has some excellent work on bank runs. You should read some of it.
Why do you think there would have been bank runs everywhere?
Mar 18 2023 at 12:19pm
At least that’s what the depositors–and anyone else benefiting from another bailout–want you to believe.
If our banking system really is that fragile, then …
Mar 17 2023 at 5:22pm
You guys should let Canada regulate your banks. This is from OSFI, Canada’s banking regulator. Note the shock scenarios we test for: