Schrödinger's Bank

As you may have heard, Silicon Valley Bank (SVB) was taken over by the FDIC on Friday morning. Or, more specifically, Silicon Valley Bank committed suicide last week with an assist from Goldman Sachs. This was both a sudden occurrence and a long time coming. We’ll start with the ending and then come back to try to explain how we got here. First, the ending: Silicon Valley Bank no longer exists as an independent bank. It technically got split into two parts by regulators, but that’s a detail that is unimportant right now. What is important is that insured depositors are of course fine. Their money at the bank is good and will be available for use Monday morning. What’s also important, and causing quite a bit of concern (rightly so) is what will happen to the uninsured depositors. (As a quick aside, equity owners will get nothing. Bond holders may get something, but that’s quite uncertain right now.)

So let’s discuss the possible outcomes for the uninsured depositors. First, who’s uninsured?  Basically any account with more than $250,000 in it. Which is almost all of SVB’s deposits, as it was the bank favored by a significant majority of technology companies backed by venture capital, as well as a large portion of the biotech industry. It was also the largest banker to the wine industry. What makes this situation particularly pernicious is that many of these technology companies, as best we can ascertain, do not have other banking relationships, and kept most of their funding at SVB.

This is causing a serious problem for these companies. As just one example of which we have firsthand knowledge, Rippling, a payroll provider for many smaller technology companies, was unable to process its payroll ACH payments on Friday because its account at SVB was frozen. So employees expecting their paychecks to be direct deposited Friday weren’t paid. As another example, a company that had their account at SVB was told that unless they were able to get a new bank relationship open by the end of the day yesterday, they would not be able to process next week’s payroll either. The angst these companies are feeling is real, as they don’t know when they will have access to their funds. 

Our current expectation is that at a minimum, most depositors at SVB will have access to about 50% of their uninsured deposit balances on Monday. This assumes that no other bank steps in to assume/acquire these deposit accounts this weekend, in which case almost all of their funds should be available. However, we would put the likelihood at about 90% that another bank or consortium of banks does acquire these deposit accounts, which have real value to another bank, by Monday morning. We know that there are bank CEO’s right now hoping they get the call to take over these accounts. So that should happen. We would expect that, given what the regulators did to banks like JP Morgan after the GFC in 2008 (recap: force/encourage strong banks to take over weak ones, then sue/fine them a few years later for the transgressions of the weak banks they rescued), that there will not be a full purchase of SVB. Instead, regulators will break it up and sell off the pieces, with the residual value, if any, going to ensure full depositor repayment (if the deposits aren’t acquired) then to pay off holding company bonds, then finally any remaining funds (we don’t think there will be any) going to shareholders.

Therefore, our opinion is that depositors at SVB will probably be depositors at a different bank on Monday morning, with full or nearly full-access to their funds soon. This is the logical outcome for a number of reasons. Banks want deposits. SVB has them. The FDIC needs to move them. Hence, this should be fairly straightforward to resolve, even though it is a big bank to rescue. We caveat our opinion with this, however: a lot of really illogical things have happened lately. More than once in recent years we’ve debated a situation in markets or the economy and said “well that XYZ outcome doesn’t make any sense, so it won’t happen.” And then it did. So we’re putting that asterisk out there. The FDIC as a government agency is actually pretty good at moving fast and not breaking things, which is a minor miracle when you think about it. But we have an administrative state that is hostile to all-things financial, big, and money making, so weird things could happen.

We expect Monday to be quite volatile again, with maybe more of a decline in the broader market as it catches up to the banks, with a rebound likely on Tuesday as the situation settles down and fear subsides. We were hedged well for the event on Thursday, and as these hedges increased substantially in value we sold them to lock in our gains. On Friday, as the selloff continued, we covered other shorts and bought some banks we like at prices we never expected to see again. We also added a significant broader market hedge near the close on Friday in case things are worse than we expect on Monday to protect our downside. We expect this situation to remain fluid and volatile for a few days, which may provide us with more opportunities to buy strong banks at great prices.

As the Talking Heads would say, “So how did I get here?”  Like all good stories, this one has three Acts. Act I: Silicon Valley Bank was started 40 years ago to bank technology companies. By the 1990s it was public, and two of us at JCSD were bank analysts covering the company on the sell-side. John Dean was  a good conservative banker, and SVB built one of the strongest franchises in the industry. We were fans. But then things changed. New management came in, and success bred hubris. Eventually, management thought that they didn’t need the traditional bank investors that gave them their expansion capital and invested in them when they were small, and instead wanted only growth and tech investors. Management stopped going to investor conferences. Despite attending half a dozen conferences a year, we’ve never met the current CEO in person, only IR and occasionally the CFO. Cultivating relationships with bank investors was not worth the CEO’s time. On roadshows, the company actively avoided meeting with any hedge funds, even those that tend to be biased long bank stocks. But things were good for SVB, and their stock traded at a tech stock like valuation for awhile, so no one cared.

Act II: Covid happens, the Federal government floods the economy with cash, VCs raise record funds, and they invest those funds in startup technology and life-sciences companies at a record pace. Deposits surge in the banking industry as locked-in consumers can’t spend the money they are getting. Between Q4 2019 and the first quarter of 2022, deposits at US banks rise by $5.4 trillion, and due to weak loan demand, only ~15% is lent out; the rest is invested in securities portfolios or kept as cash. SVB’s deposits rise by 300% in three years. That’s a lot! Its stock price soars with the tech stock bubble, as it’s a tech bank. Management gets even more arrogant, as their skyrocketing stock price means they must be geniuses. Questions asked (by us) at a conference to the CFO about capital and asset/liability management are dismissed as irrelevant in this new era of extreme liquidity. And then mistakes are made.

SVB believes its own hype and makes a bet that everything will be awesome forever. In a reach for yield, they invest all this excess cash not in short-duration t-bills or fed funds, but instead in long duration MBS at quite low yields (while we believe it was 8-10 year duration at 1.87%, we have seen other similar but slightly different numbers – the exact number isn’t really important). This was back when interest rates were still low, but inflation was already ripping higher. The Fed was stating that inflation was going to be transient, so some banks took them at their word and went long on the asset side of the balance sheet. This was mistake number one.

Inflation wasn’t transient, and the Fed raised the benchmark, risk-free interest rate banks can receive from 0% to 4.50% in about a year. This caused the value of the long dated bonds SVB purchased to go down in value. (Bonds are priced as the net present value of their future interest payments. The discount rate is basically Fed Funds or T-bills plus a small spread). But this wasn’t the actual problem for SVB, as the bulk of their bonds were in their Held-to-Maturity portfolio, which doesn’t have to be marked to market. The assumption is that while rates go up and down, so long as the bank intends and does hold the loan to maturity it will be “money-good” and therefore not a capital issue. This is generally correct. The actual problem was self-inflicted, a result of management’s hubris and bad advice from Goldman Sachs.

SVB wanted to boost its net interest income on its other bond portfolio, the one that is Available For Sale. This was about $25 billion in size. Because it had bought the bonds in this portfolio when rates were low and now rates were higher, it could sell those lower yielding bonds and buy new, higher yielding ones and earn more interest income, which would increase its earnings and hopefully its stock price. But selling the lower yielding bonds would lock-in a loss on those bonds. It’s just simple bond math – a low yielding bond will trade down in price until its current yield matches that of newer bonds. And worse, while the lower price was reflected in its GAAP book value, it was not in its regulatory capital until it sold them. So despite this transaction (swapping low coupon bonds for higher coupon bonds) being economically a wash (you’d never do this in your own account, it wouldn’t change your returns), and detrimental to regulatory capital, Goldman Sachs convinced SVB to do it. And ego driven SVB management did it. This was mistake number two. At this point SVB was like Schrödinger's cat – both alive and dead at the same time.

“How did you go bankrupt?"

Two ways. Gradually, then suddenly.”

― Ernest Hemingway, The Sun Also Rises

Act III:  They did it wrong. They sold the bonds first, locking in a regulatory capital loss of about $2 billion. In any two-legged trade, you always do the riskier, hard to execute leg first. That’s trading 101. You don’t do the easy leg first, then see if you can execute on the hard one. That’s just dumb. But that’s what they did. Mistake number three. Then, after selling the bonds, Goldman couldn’t raise the capital. Why?  Because Goldman isn’t a well-respected investment bank when it comes to depository institution investment banking. Well, of course there were other investment banks on the deal who are respected in the industry who could help fill out the order book right?  No.  As we’ve noted, SVB’s management is a) arrogant and b) dismissive of bank investors – you know, the kind of investors that can make a decision to inject a few billion dollars into a bank overnight. So SVB only hired Goldman and the investment bank it owns, Leerink, to run the deal. Leerink is a technology investment bank. Goldman doesn’t have deep ties to bank investors. If SVB had instead hired KBW or Sandler O’Neill as deal runners, we believe it’s highly likely this deal gets done. But Goldman didn’t even call us to see if we wanted to put in capital, and that’s all we do. And we weren’t the only ones they didn’t call. Mistake number four. That’s too many.

Dénouement: Silicon Valley Bank can’t raise the capital it needs in a timely manner. Its depositor base, which is tech savvy and therefore reading about these problems on Twitter, immediately opens their banking apps and wires out their money. Some prominent VCs advise their portfolio companies to do the same, a run on the bank ensues, and the bank fails about 36 hours after announcing its intent to sell bonds and raise capital.

So what happens now? There is a lot of chatter on TechTwit about how companies should move their accounts to the largest banks as they are safer than smaller banks. But SVB was a big bank!  Don’t do that! SVB was also uniquely exposed to large deposit accounts. Small banks are not. Smaller banks tend to have well diversified deposit bases across both commercial and retail accounts. SVB had almost no retail accounts. Moving to JP Morgan will not be a good idea for the vast majority of businesses. But as we mentioned above, a lot of illogical things have happened that we didn’t expect, so we need to be open to the possibility that people come to the wrong conclusion and move money from small banks to large banks. We’ll be closely watching for any developments here.
​
What should happen? This was an old fashioned bank run, a liquidity crisis that caused a short term solvency problem. The FDIC was created to prevent exactly this type of liquidity crisis from happening. But the $250,000 FDIC insurance limit only prevents retail investors from doing a bank run, and SVB didn’t really have that type of customer. There is currently no program in place to prevent a bank run by large corporate accounts – the kind that SVB had. Our solution would be that the FDIC insurance limit be raised dramatically, to say $50 million. One person who we suggested it to said that the FDIC can’t afford to do that – we replied that it can’t afford not to. And by raising the cap to $50 million, it dramatically lowers the need to ever have to use it to cover a liquidity run again, as there wouldn’t be a liquidity crisis to avert. Current bank regulations are already really tough on lending in order to prevent a credit crisis, which is really trying to prevent insolvency. By increasing FDIC insurance limits, they can greatly reduce the chances of another SVB type liquidity-driven failure happening again as well.


Best,

Jeff
 
DISCLOSURES

 

DISCLAIMER: The opinions expressed in this quarterly letter are for informational purposes only and should not be construed as investment advice.  The letter is not a recommendation of, or an offer to sell or solicitation of an offer to buy, any particular security, strategy or investment product.  Past performance of any strategies discussed herein is not necessarily indicative of future results.  The research for this letter is based on current public information that we consider reliable, but we do not represent that the research or the letter is accurate or complete, and it should not be relied on as such.  Performance numbers presented herein were prepared by JCSD Capital, LLC, and have not been compiled, reviewed or audited by an independent accountant and are presented for informational purposes only.  All performance estimates are subject to future adjustment and revision.  The information provided herein is historical and is not a guide to future performance, and any potential investors should be aware that a loss of investment is possible.  The views and opinions expressed in this letter are current as of the date of this letter and are subject to change.