I've been trying to figure out exactly what warning signs regulators supposedly missed in the months before Silicon Valley Bank collapsed, and I haven't come up with much. The $15 billion hole in SVB's bond portfolio was mildly concerning, but no more than that since it didn't affect SVB's books. The outflow of deposits seemed to be cooling down. Capital ratios were solid, as were leverage ratios.
But today I learned about this:
For Kathryn Judge, a professor at Columbia University with expertise in financial regulation, another “classic red flag” was the fact that SVB was so reliant on the Federal Home Loan Bank of San Francisco for funding, with $15bn of outstanding loans by the end of 2022 after no loans the year before.
“We consistently see that before banks get in trouble, they struggle to access market-based sources of financing and increase their reliance on the Federal Home Loan Bank System,” said Judge. “That does raise questions over how closely the Fed was following developments at SVB.”
Hmmm. This is from five days ago, but I haven't seen it mentioned anywhere else until now. Maybe I just missed it. I wonder how much of a red flag this really was?
Of course, I keep coming back to the same thing: At COB on Wednesday, March 8, virtually nobody was talking about SVB. Newspapers barely even covered their proposed share sale. SVB stock was up slightly compared to the day before. Analysts remained bullish.
It was only after the release of the mid-quarter review in late afternoon—which didn't really reveal anything new about SVB's condition—that anything happened. SVB's stock price tumbled in after-hours trading and the next day Founder's Fund began its infamous run, for reasons that remain veiled. And then SVB collapsed.
So what really happened? Everyone seems to think they know, but they mostly seem to be slaves to narrative. I continue to feel like there's something yet to learn about this whole affair.
Vulnerability to libertarians now goes on the list of things for regulators to watch for.
Can't say I really understand the puzzlement. The bank became insolvent largely due to losses in the bond market. The bank's insiders likely hoped no one would notice and they'd muddle through without depositors asking for their money back. Depositors did in fact notice and started withdrawing their deposits, resulting in the crash.
The delay between when the bank became insolvent and when the run started doesn't seem surprising; after all, apparently we don't expect depositors to notice anything since we bailed out even the largest ones on the grounds that we can't expect them to pay attention.
The puzzlement is for people who properly understand, versus people who do not.
The bank did not become insolvent due to bond losses, bloody hell. They were not trading book and they were not loosing in bond trading. Hold to maturity bonds don't lose actual cash becuase on trading you would lose - any more than you lose money if your house price falls below your mortgage so long as you do not have to sell it. The market price is immaterial to you.
The bank was not insolvent at time failure, it was however locked-up illiquid.
The failure came from a classic bank run on deposits.
Hard to make sense of this word salad, but maybe you're making an irrelevant pedantic point about the word "loss." Yes, of course the losses weren't realized until the bank had to begin selling the bonds to meet withdrawal demands. So what?
Well dimwit, the point is not irrelevant nor pedantic, it is fundamental to the actual point you ignorantly tried to make.
The business of the bank was solvent, profitable.
The US Treasury holdings were backing to deposits, not the bond trading.
Until the deposit run began the paper losses were irrelevant, every bit as irrelevant as that of a homeowner finances whose house is underwater but does not intend to sell.
The bank failed because of the outflows - of which the Treasuries were not the cause but the symptom.
Are you dense? If the bonds had not declined in value the bank would have been able to sell them and cover the withdrawals. They did decline, and therefore the bank could not cover them and subsequently failed.
The two of you are arguing over two terms in an “and” - the bank would have survived a “run” were it not tied up in illiquid assets. At the same time, the bank would have survived if depositors hadn’t (been) panicked into a run.
The chicken or the egg? Walking along the cliff is fine... until Peter.
But it's not a liquidity issue. The bonds are liquid; you can sell them any time. If they had not declined in value the bank could have sold them and covered the withdrawals. As it is the bonds had insufficient value to cover all deposits. That means the bank was insolvent.
Yes, had the depositors not noticed and had left their deposits until the bonds matured, the bank would have been okay. They just didn't get lucky that way.
No. SVB absolutely would not have survived the run, under any circumstances. By Friday, depositors were trying to withdraw 60% of the deposits SVB held. There is no bank in the world that could survive that.
Kevin looked at one of the "stress test" criteria a day (or 2 ago). This one was the ratio of High Liquidity Assets to monthly withdrawals. On paper, SVB was nicely set - minimum requirement was HLA to cover 70% of normal monthly withdrawals. SVB was at 105%.
I'm still waiting for the morning coffee to kick in but the Thursday run cost SVB closer to 80% of their HLA stash in a single day. That's not a "run" that's a major dam break. I
Lounsbury is a subject matter expert on how smart he is and how dumb everyone else is. He always focuses the conversation on this area of expertise.
I love his point that a company is always solvent if you ignore its liabilites and asset values. And that its just like being underwater on your house, if you ignore the banks liabilities.
This may not make sense, but remember that he is getting a bit outside of his area of expertise here.
Yeah it's just like your mortgage, that is if the bank could call it at any time on a whim and throw you out on the street if you couldn't cough up the balance. Perfect analogy on his part.
If the loan includes a “demand feature,” it may. https://www.consumerfinance.gov/ask-cfpb/what-is-a-demand-feature-what-does-it-mean-if-the-demand-feature-is-checked-off-on-my-closing-disclosure-en-1997/
Thanks, Lounsbury. I believe yours are sensible observations. I know I'm not an expert, though.
You previously pointed to an FT article that looked at the adjusted leverage ratio of five banks with the worst-performing stocks post-SVB.
The markets, it seems, had focused on adjusted leverage ratios and updated cash ratios to judge risk.
You said there was no risk of contagion. Are you implying that you know that no other banks have low leverage and cash ratios matching that of those five banks?
It is further that unless one dug in, the overall metrics of SVB were not bad - it was not until the deposit outflow started to accelerate that suddenly everyone became worried that maybe that pile of safe US Federal securities that were not supposed to actually need to be sold were going to be need to be dumped and they would not have the liquid cash to cover cash withdrawals.
Not insolvency. Not losses on bond trading. Losses triggered by unscheduled / unplanned sale of hold-to-maturity by depositors pulling out money on the fear of other depositors pulling out money. (of course had they not tried to juice their holding portfolio return with long-dated securities instead of more typical short-dated the stress would not have been as bad).
Bank run and the self-fulfilling prophecy of panic.
Something perhaps to worry about is that people heavily invested in crypto can profit from bank failures because a bank failure encourages people to buy crypto. They can look for a bank that is struggling and encourage a run. Bitcoin is up 31% this week after a couple of bad years.
The evidence that people buy crypto when banks fail is zero. That's the opposite of flight to quality, which is what most investors do when people are nervous about bank stability.
SVB was bank for a number of crypto companies. Signature actually held more than $10B in crypto as assets. Had the government not stepped in to protect crypto depositors (among others), there's a pretty good chance the price of crypto would have gotten clobbered.
If you want to find a corollary for the price of crypto, look at the Fed's balance sheet, which zoomed higher this week. Maybe crypto isn't an alternative to central banks but a major beneficiary when central banks inject liquidity into the system.
I wonder if SVB did something to anger Peter Thiel. He has already shown a vindictive streak.
Drum may wish to read FT Alphaville review of the SVB hedging issue
https://www.ft.com/content/f9a3adce-1559-4f66-b172-cd45a9fa09d6
"The first thing to remember is that SVB’s bond portfolio was basically in two different accounting buckets. At the end of 2022 it held $91.3bn in a “held-to-maturity” portfolio — bonds you plan to hold on to until they are repaid — and $26.1bn in an “available-for-sale” portfolio, which is marked to market."
[...]
Let’s take the chunkier HTM portfolio first. Securities in the HTM basket can be carried at their nominal par value, because the assumption is that they are being held until they’re repaid in full.
As the table below shows, most of SVB’s $91.3bn HTM portfolio consisted of very-long-term, agency-guaranteed, mortgage-backed securities maturing in 10 years or more ($56.6bn to be exact).
The creditworthiness of this stuff is extremely high, but it’s also very sensitive to interest rates (for bond nerds, the average duration of the HTM portfolio was 6.2 years).
Because of rising rates the actual market value of the HTM portfolio was about $76bn at the end of 2022, according to someone who has seen the details of the portfolio and shared them with FTAV — an unrealised loss of $15.1bn.
Yes, SVB didn’t have any hedges on this bit. But doing so would arguably be nonsensical. [discussion re hedging impact omitted] That means it basically becomes a directional bet on interest rates that flows straight into the income statement, something that most banks abhor.
[...]
That [hold shorter term securities to minimise interest rate sensitivities] is something SVB definitely did not do — since ca 2018 they actually added a lot of duration by piling into 30-year MBS. But in practice, not hedging the interest rate risk on the HTM was probably not Silicon Valley Bank’s biggest mistake.
However, let’s turn to the AfS [Available for Sale = trading] side. Unfortunately, here be dragons.
This is what SVB’s AfS portfolio looked like at the end of 2022. As you can see, it was mostly Treasuries. Remember, these are carried at fair value, ie marked to market. [chart - showing 26bln odd] That’s pretty big. For comparison, Credit Suisse ... actual AfS portfolio (of mostly corporate debt) stood at $860mn. [NDLR: recall Credit Suisse, one of the largest banks in the world]
The AfS bucket is definitely where most self-respecting banks lugging around a big portfolio of bonds will hedge their interest rate risk. Otherwise, the income statement would bounce around according to whatever the market does from one quarter to the next."
[more hedging discussion]
"Well it looks that weakening profitability in 2022 as the tech world made SVB do something really dumb. In the first quarter, it unwound $5bn of AFS hedges to book a $204mn gain, and in the second quarter it dumped another $6bn of hedges to lock in a $313mn gain."
So SVB dumped its hedges to juice near-term profits in a bet on peaking rates (one that it may have ironically achieved....)
However it was sitting on a huge amount of flighty and highly correlated VC related deposits that could run on little notice....
"Gormlessly, SVB had amassed a stupendous pile of uninsured deposits, almost entirely in just one industry that was burning through its deposits as VC funding dried up."
and outflows started....
and then Mr Theil made some phone calls one Thursday.
I understand that over 90% of the money on deposit at SVB was uninsured. This is much higher the industry average. I don't know if that percentage counted as a official "red flag" or not, but maybe it should. Uninsured depositors will surely run for the exit at the first sign of trouble
SVB Regional Peers as data in FT showed are on order of 30%
SVB deposit base was not only extremely jumbo oriented (more like95-97%) it was also extremely sector concentrated in one broad sector - which is extremely unusual.
I believe that the bank regulators can specify that a bank is a SIFI even if it not yet big enough to meet that definition. Based on previous positing's by Kevin, it is also likely that SVB would have had sufficient equity under SIFI rules.
Perhaps I am wrong about this, but unless bank regulators have the ad hock flexibility to redefine equity and liquidity definitions, it is unclear to me what any Federal regulator could have done prior to SVB's failure.
Of course the rating agencies could have sounded the alarm (and quickened the bank run) prior to their failure. It was also gross negligence, or complete fucking incompetence, that so much money was parked at one financial institution, like Roku with almost $500M deposited with SVB.
May founders fund shorted SVB and yanked
Perhaps that sometimes, shit happens?
So what really happened?
A timeline:
1. On or about 3/1: Moody's notifies SVB of pending downgrade
2. Later in the week: concerned that downgrade may erode confidence in company's finances, SVB consults with Goldman Sachs
3. Weekend of 3/4-3/5: SVB devises plan to boost value of holdings (it would sell $20B+ bonds, at a loss, and reinvest, and sell stock to cover funding hole)
4. By Wed, 3/8: SVB completes bond sale (at $1.8B loss) and preps for stock sale of $2.25B (inc. $500M to General Atlantic)
5. Wed, 3/8: Moody's downgrades SVB one notch (less than originally planned bc of planned stock sale to raise capital)
6. Thurs, 3/9: plan to complete stock sale before market open falls through bc SVB failed to complete paperwork to provide investors with 24 hours notice
7. Thurs, 3/9: word gets out, stock price plummets, big customers run on bank ($40B+ in withdrawals); General Atlantic & other investors back out of buying stock
8. Fri, 3/10: Calif. regulators close bank, appoint FDIC as receiver
How did word get out? Can't pin it down, but I recall hearing about a comment SVB CEO had made (prob. late Wed.) about the stock sale that was meant to assure a questioner that bank finances were OK. But instead the comment raised worries. By Thursday, everyone had heard of SVB's "troubles."
This article adds some detail to what went down the days before SVB was shut down.
https://www.semafor.com/article/03/17/2023/how-svb-lost-crucial-time-to-save-itself-before-its-collapse
Also:
So it's not the case that everyone thought SVB was sound at the end of the day Wed, 3/8.
We likely will never know why, but Peter Thiel at Founders Fund decided to yell l"FIRE!" in a crowded theatre and started a stampede.
My theory is that Greg Becker looked a Thiel funny or cheated at golf or something equally inane and that caused Thiel to pull the rug.
Did you see the movie "The Big Short"?
At this point, I have to keep open the theory that this was done by Peter Theil with intent. He's done enough in my awareness that I have to mark him as a malignant actor in many circumstances, his action with no exogenous context was the trigger for the run, and his funds got out clean. No proof, but no other cause is found, he actually pulled the trigger by his widely announced markdown, and motive is... uncertain but likely. He certainly drove the rest of the sector into a loss he avoided, and he's deep enough troll that he was Elon's mentor.
Can't say that it's more than a hypothesis yet, but I have yet to see any other cause for the run than Theil's decision to fuck with them.
There is no great mystery. When 90% of deposits are not insured, banks are subject to panic runs, whether they are "sound" or not. And this case, SVB was less sound than it was at the start of 2022 because of the rise in long-term interest rates, which can be attributed to the Fed. But holding long-term loans or bonds is and was not unique to SVB - this is what normal banks do. When the Fed raised rates in the 70's and 80's there was considerable damage to S&L's and banks, with many failures of both, though this was due to bad loans, not panic runs.
The high ratio of uninsured deposits, which does seem to have been fairly unique to SVB, should have been a red flag, but either this was not included in regulations and prescribed procedures or authorities just blew it. If Kevin or anybody else is really interested in reforming regulations and/or the applications thereof, they should look into detail into all of them, not just Dodd-Frank and the 2018 revision, and see what there is about uninsured ratio.
Good points. There's been a lot of talk of whether SVB would have passed or failed stress tests, as designed. Either way, that doesn't let regulators off the hook. The job for regulators is to avoid the kind of trouble we've see this past week. If a bank is operating recklessly, regulators need to step in. If a bank is operating within prescribed limits yet still wreak havoc, then regulators need to rethink how they regulate banks.
Two items that stand out to me:
1. High level of uninsured deposits, which makes a bank run at the first sign of trouble much more likely
2. Effect of rapidly rising interest rates, which erodes the value of bank holdings (true to one degree or other at most or all banks)
One takeaway is that the regulatory regime needs to change.
Another may be for the Fed to consider the collateral damage of its rate hikes.
Word today is 186 banks may be prone to similar risks as SVB. This crisis is going to be around for a while.
https://twitter.com/financialjuice/status/1636817207064768512?s=20
"Everyone seems to think they know, but they mostly seem to be slaves to narrative."
The "narrative?" Isn't this why there are regulatory agencies regarding banking? Isn't this the Fed's job, to keep track of such things across the banking system? Must it make the news before they do their jobs? Way too lenient here looking for excuses for extremely powerful agents tasked with preventing such things from happening. And when they do the opposite, like Volker, Greenspan, and now Powell (all Republican appointees), it points to a pattern of corruption and abuse of power. Powell should resign.
Can’t speak to the other two, but Powell was reappointed as Chair by a Democrat. And appointed to the Fed in the first place by a different Democrat.
Official Trump Presidential portrait revealed,
https://cdn.cronachemaceratesi.it/wp-content/uploads/2018/03/Asier-Sanz_Donald-Trump.jpg
I'm waiting for the official Trump portrait on Tuesday.
The most recent stress test did not include the possibility of interest rate hikes because they were afraid that many banks would have to mark to market, which would paint every balance sheet red. They should have sold all low yielding Treasury notes to the Fed and they likely would have avoided insolvency. They didn’t have a risk manager for the past year and a half. Assuming their bond portfolio was at par, instead of a sharp discount was the fatal flaw, but even the Feds don’t want to address because that open the conversation of mark to market, which they avoid like the plague. I am sure they are examining every bank balance sheet for discount treasuries. Will be interesting to see if this pauses rate hikes given what it does to bank liquidity.
"...The $15 billion hole in SVB's bond portfolio was mildly concerning, but no more than that since it didn't affect SVB's books. .."
No, if the books don't reflect reality it is reality that matters.
As a former bank examiner and current CFO of a multi-billion bank:
There is NOTHING AT ALL WRONG with borrowing from the FHLB.
Up until last fall, banks had far more in deposits than any bank wanted. So there was very little borrowing from FHLB. Loan volume picked up and people started to spend their savings and ALL BANKS saw their liquidity decline.
So banks did various things: most every bank saw a decline in their liquidity; some banks borrowed money from FHLB; some banks increased deposit rates; some banks increased loan rates; some banks did all of the above.
We were one of the banks that did all of the above. We had about 1/8 the loans from FHLB that SVB had but so what.
I am not familiar with FHLB SF but I assume it is very similar to FHLB NY. You pledge securities and they will lend you a certain percentage of the market value of those securities. This means the FHLB is always well secured and it also means you can get more liquidity from selling the bonds than borrowing from FHLB.
I am guessing that last week, SVB had borrowed as much as they possibly could from FHLB.
I don't want to keep going over the same thing but SVB had a liquidity crisis and didn't handle it well and since all their customers were comfortable with online banking, the run was easier to gain steam than at other banks.
NO BANK COULD SURVIVE the loss of deposits on Thursday.
I could also go into a long argument that stupid GAAP rules that make it appear that many assets are still worth book value when any idiot knows that they aren't.
One last point. Your 3% 30 year mortgage is still on your bank's book at book value even though the bank couldn't sell the mortgage for 90 cents on the dollar.
Yeah, like, I'd like to understand how upper management could coincidentally sell off their stock a week earlier.
Because they'd made SEC filings declaring their intent to sell those shares at least 30 days before they actually did.
As TMD says, that's standard practice these days that most US companies follow if their execs are smart (or well advised).