Silicon Valley Bank, a cauldron of weirdness

In this post, I’m going to use the ticker, SIVB, that Silicon Valley Bank used before it was shut down and sold.

—-Perhaps the oddest single thing about the SIVB debacle is that it was a publicly traded company, whose detailed financials were publicly available both on the SEC Edgar site and, I presume, on the company’s website. So everyone with an interest in the bank could easily have understood that SIVB was unusual in two respects:

    –the average account size, including corporates, was about $4 million, with something like 90% of all deposits not eligible for FDIC insurance, and

    –the majority of the bank’s assets (i.e., the way it invested depositors’ money) were in long-dated Treasury bonds. To my untutored eye, the company didn’t have much of a conventional bank lending business and apparently saw no need to develop one.

    —-Most of the deposits came into SIVB during 2020-21. To make a 75 basis point spread between interest collected from Treasuries and the zero-ish interest paid to depositors, SIVB would need to buy 7-year–or longer–Treasury notes. In contrast to conventional, variable-rate bank loans, the interest payment on Treasuries is fixed. And the lender can’t have a loan committee meeting and thereafter demand immediate repayment. So the 1%-ish T note rate was locked in for a very long time.

    This strategy might have made sense during the deflationary1990s in Japan, or if you thought the pandemic might continue for another four or five years. Absent either, a 1% long-dated Treasury was about the worst thing you could buy.

    So SIVB customers were uniquely uninsured and the bank was in a poor position to meet withdrawals. All of this when the Fed was making it ultra-clear that rates were going to up by a lot.

    —-Another weird part is that neither the tech corporation VP finances nor the high net worth individual customers bothered to take a peek at the SIVB financials–or they did and didn’t care about the risk they were taking.

    —-The Fed and its San Francisco branch (the SIVB CEO was a board member of the latter) did care. Both apparently warned SIVB of the huge risk it was taking, and were ignored. Both now maintain that they could only advise and not force SIVB to comply. The Financial Times recently reported that the Bank of England was worried enough about the risk to the financial system that SIVB posed that it wrote to both Feds to express its concern.

    —-In late 2022, Kim Olson took the job of SIVB’s Chief Risk Officer, a position that had been vacant for some months (also weird, and arguably very revealing). Why would someone with her 30 years of risk management experience take this job, since even a back of the envelope calculation would have suggested that the losses on Treasuries (unhedged, apparently by SIVB design) from rising interest rates had wiped out all, more or less, of SIVB’s shareholders equity? At the very least, something had to be done immediately, in early January (again, easy to say now).

    —-this morning I happened to be watching CNBC when Joe Kiernan “explained” that the failures of SIVB as a company and the San Francisco Fed as a supervisor were understandable, because both have women in positions of responsibility. He forgot to mention Jim Cramer’s forceful endorsement of SIVB on CNBC at $300+/share in February.

    the first day of Spring

    pre-Spring housecleaning in the banking sector

    Silicon Valley Bank

    For Northern California entrepreneurs, having a Silicon Valley bank account was like sporting a Birkin bag or driving a Porsche–a clear signal that you’d made it.

    Unfortunately, it was also almost inconceivably badly managed as a bank. A huge portion of its investments were in long-dated Treasuries, bought when interest rates were close to zero and unhedged against the possibility of higher interest rates. This meant losses were inevitable as the Fed normalized interest rate policy as the pandemic waned. Warnings from government regulators and outside consultants about this risky posture, as well as the lack of management information systems, were ignored.

    Adding to risk, the bank was very highly vulnerable to the possibility of mass withdrawals on the deposit side of the balance sheet. That’s because the average account size was $4 million, meaning 90+% was above the $250,000 limit for being eligible for Federal deposit insurance.

    More than that, Jim Cramer of CNBC touted the stock as “cheap” at $320 a share a month or so ago.

    The house of cards came crashing down as the bank began to have gigantic withdrawals–reportedly $42 billion in a single day–and a belated attempt to raise new funds failed. The bank was closed by the authorities ten days ago.

    Credit Suisse

    CS has been the bad boy of Swiss banking for as long as I can remember. It wasn’t a case of plain vanilla incompetence in asset-liability structure the way it was with Silicon Valley Bank. It was more the company it kept–drug dealers, money launderers, fraudsters of all ilks. None of that did the bank in, though. It looked like it would be able to continue to limp along, until the announcement by its largest shareholder, the Saudi National Bank, that it wouldn’t/couldn’t provide more equity to help address potential liquidity problems. The result was a government-forced acquisition by UBS at a price that all but wiped out existing shareholders.

    One twist in the story: contingent convertible bonds. These are weird (to my mind, anyway) instruments that have been around for decades. Traditional convertibles are income instruments that are issued at a discount to prevailing rates, offset by an option for the holder to convert into stock at, say, a 30% premium to the issuer’s stock price at the time of the offering. So if good things happen, the holder is rewarded. Contingent convertibles, called “co-cos” or AT1s, do the opposite. They are issued at a premium interest rate, with an issuer option to convert them into common stock if the underlying business goes south.

    I’ve never understood the attraction.

    Credit Suisse had $17+billion of the se on its balance sheet. They’ve all been converted, wiping that debt away.

    where to from here?

    Morgan Stanley, which had predicted a US equity index fall of up to 25% during the first quarter and gradual improvement after than, appears to be saying that banking woes are the canaries in the coal mine or, to mix metaphors, the first of many shoes that will drop in the near future. My sense is that we’re much closer to the end of bad news than that.

    The big winners during all this have been cryptocurrencies and beaten down tech stocks. So far this year, the Russell 2000 (pure US earnings) is flat, the S&P (half US, half foreign) up slightly and NASDAQ up by about 10%. I’d be happy with a flat market from here, where stock and industry selection would be paramount.

    even more about Silicon Valley Bank (SIVB)

    I’ve been reading a New York Times article on SIVB this morning. The original version came out yesterday and was amended today to state that SIVB had gotten a call from the ratings service Moody’s was going to downgrade SIVB’s bonds to junk (the earlier version had mentioned other securities).

    What I take from this is that the bond community knew clearly what a risky mess SIVB was.

    The article also cites a Janney Montgomery Scott report on the stock that 75% of SIVB’s investment of depositors’ money in Treasury bonds (a category that made up the majority of the bank’s investments) was classified as “held to maturity.” This classification means that the bank would carry them on its financial statements at their purchase price, rather than marking them to market and recognizing the resulting gains or losses on the income statement. And the categorization would normally be used only for bonds that would mature at least several years in the future. This compares with the typical bank, which classified 6% of its debt portfolio this way.

    Why the enormous difference? Two possibilities come to mind: either most banks used hedging techniques that have been common since at least the 1980s to protect them from rising rates or they chose not to make the huge bet that SIVB did that interest rates would stay at pandemic lows for years.

    The fact of this brokerage report outlining SIVB’s unusual and aggressive bond bet suggests that professional equity investors also knew what about SIVB’s ill-fated bet.

    Who didn’t know the very worrisome data in SIVB’s financial statements and in bond and stock reports? Apparently lots of Silicon Valley startups, affluent California individuals, Roku, actors playing the role of investment professionals on shows like CNBC’s.

    more on Silicon Valley Bank (SIVB)

    A lot is now being written about the collapse of SIVB. The political left, for example, is blaming Trump, the political right the diverse composition of the board of directors. The latter, I think, is just crazy; the former a bit disingenuous. Cathie Wood of ARK has also chimed in with the, I think, totally bizarre opinion that the Fed is the root cause of the SIVB affair, having pushed short-term interest rates far too high.

    From my vantage point, it looks as if SIVB made a huge all-or-nothing bet that interest rates would remain at 2021 levels. It did this in two ways: by using customer deposits to buy enormous amounts of long-term Treasuries, and by choosing not to hedge against the possibility of higher rates. Dynamic hedging, a strategy that has been around for close to forty years, comes to mind as a more bank-like way to proceed.

    What strikes me as equally odd, if press reports are correct, that a company like ROKU would place close to half a billion dollars on deposit with SIVB without having taken a glance at the SIVB annual report–as did a bunch of companies in the Peter Thiele orbit.

    SIVB clearly saw the crisis coming as withdrawals mounted. One of the first actions we on the outside can see is the then-CEO selling a bunch of his stock. Then came plans for a private placement, to be followed by a public offering of preferred stock the following day, aimed at raising a total of $2.5 billion. The private placement fell through (my theory: SIBC put off the placement, on the idea that a successful offering would mean the placement could be made at a better price).

    In a JP Morgan note I read, the author makes the point that depositors took the enormous risk of putting their money with SIVB but were only compensated with a paltry extra 60 basis points in yield.

    The financial press coverage has, by invoking memories of 2007-09, been both uninformed and at least borderline irresponsible.

    The only party that comes out looking good in this mess is the Federal government, whose quick and decisive action has hopefully nipped a possible panic in the bud.

    Silicon Valley Bank (SIVB), and why I’m not a 100% fan of financial stocks

    not a fan, except for sometimes

    Yes, I have owned bank stocks. In fact, I have a small position in Bank of America (BAC) at the moment, on the idea that in a time of rising interest rates banks increase their lending rates much faster than they raise the rates they pay to depositors. So the spread they make widens. This is conventional wisdom without not much more to it.

    My biggest issue with bank stocks is that they are opaque, even at the best of times. What they do (apart from their role in helping set overall central government monetary policy) is simple: they take in deposits (i.e., borrow) from individual and corporate customers, and then lend these funds, typically for much longer periods of time, and at higher rates of interest than their deposits, to individuals and businesses. Try to go much deeper–who are the customer?, what are the biggest loans? what are the riskiest/safest?, and there’s virtually no relevant information publicly available.

    As a result, professional investors tend to look at bank stocks in an old-fashioned way, that is, based on the company’s per share “book,” or shareholders’ equity, value. But that’s pretty much all a non-specialist in bank stocks can do is.

    The idea is this: shareholders’ equity is the accumulated profits of the firm that can be used for future investment. A stock that trades at 1x book is expected to product average profits. One that trades at 2x book is expected to make twice what the average bank can make with its capital. …and so on. At one time, for instance, Goldman Sachs was the star of the brokerage industry and traded at 3x book. As the GS star has waned, that multiple has contracted to 1x.


    First, remember that while I have been a professional investor for over a quarter-century, I’m not a bank expert. Having said that,

    this is a really weird situation. How so?

    –in late February, just days before the stock imploded, the company filed its 10K for the 2022 year. Its accounting firm, KPMG, issued the following unqualified opinion of the financial statements contained in it.

    “We have audited the accompanying consolidated balance sheets of SVB Financial Group and subsidiaries (the Company) as of December 31, 2022 and 2021, the related consolidated statements of income, comprehensive income, stockholders’ equity, and cash flows for each of the years in the three-year period ended December 31, 2022, and the related notes (collectively, the consolidated financial statements). We also have audited the Company’s internal control over financial reporting as of December 31, 2022, based on criteria established in Internal Control – Integrated Framework (2013) issued by the Committee of Sponsoring Organizations of the Treadway Commission.

    In our opinion, the consolidated financial statements referred to above present fairly, in all material respects, the financial position of the Company as of December 31, 2022 and 2021, and the results of its operations and its cash flows for each of the years in the three-year period ended December 31, 2022, in conformity with U.S. generally accepted accounting principles. Also in our opinion, the Company maintained, in all material respects, effective internal control over financial reporting as of December 31, 2022 based on criteria established in Internal Control – Integrated Framework (2013) issued by the Committee of Sponsoring Organizations of the Treadway Commission.”

    There’s more–commenting, as I read it, on the difficulty of assessing the quality of an idiosyncratic loan portfolio, but nothing serious enough to merit qualifying the audit opinion.

    —-looking at the financials,


    –cash and cash-like securities are $40 billion

    –long-term securities, carried at cost, are $120 billion, for a total of $160 billion, or about $200 a share.

    liabilities + equity

    –deposits are $173 billion, of which $92 billion is non-interest-bearing

    –common shareholders’ equity is $16 billion.

    The key things to notice here are that the long-term securities (predominantly Treasuries, I think) are not marked to market, and that if marking them to market today resulted in a loss of as little as 15%, shareholders’ equity would be completely wiped out.

    —-$200/share in net worth?

    Given that large inflows came into SIVB in 2021, when interest rates were exceptionally low, and $80 million+ went into Treasuries that year, it’s difficult to fathom how investors could have overlooked the possibility that that $200 a share in net worth might not be a solid number.

    If for most of 2021, as was the case, one would have to reach for a 7-year Treasury to get a yield of 1% and a 10-year to get a yield of 1.5%, the reported interest income for 2022 suggests something like a 50/50 split between the two. If so, this implies a loss of market value of these bonds of around 20% if they were sold today rather than held to maturity.

    This isn’t rocket science. It’s just drawing inferences from publicly available government filings. Yes, my numbers may not be 100% accurate, but they suggest that the company could easily have been flirting with negative net worth for some time. If so, management was running a huge risk by not raising capital once it became obvious the Fed was going to boost the Fed interest rates very substantially. And it left itself open to the run on the bank that developed as depositors (belatedly) worked this out.