The Bear’s Lair: The Strange Death of Silicon Valley Bank

Packard De Luxe Eight 904 Sedan Limousine 1932

Proper bankers drove these – the Packard De Luxe Eight Model 904Source: Lars-Göran Lindgren Sweden

Silicon Valley Bank (SVB) was at first sight a splendid 21st century replica of the 18th century Country Banks that fueled the Industrial Revolution – it was local to Silicon Valley, supremely attuned to the major industry in its locality, and easy for that industry to deal with. So why did it fail? The answer is complex and leads us to a number of reforms that are needed to today’s banking scene.

SVB embodied a central disconnect. It aligned its political and social preferences with those of its customer base, devoting great effort to the battles against climate change and for diversity, but that alignment proved to be incompatible with sound banking. The key problem was epitomized by SVB’s head of risk management in Europe, who ticked all the diversity boxes but, judging from her profile, appeared to have no interest whatever in managing financial risk.

That presumably explains the most elementary folly committed by SVB, of investing some eight times its capital in long-term Treasury bonds and then watching in stupefaction when interest rates went up and bond prices went down. They would have benefited from reading Kevin Dowd’s and my “Alchemists of Loss” (© Wiley 2010) which discusses the follies in Wall Street’s risk management, but that was published in 2010, when most of SVB’s key executives were still in grade school.

The philosophical problem of how to fit in with a billionaire customer base whose political and economic values are incompatible with competent bank management, is an interesting one. Suppose, for example, you are offered the chance to finance a coal mining venture led by Alexei Grigorevich Stakhanov (1906-77). Stakhanov has the best possible qualification for managing a coal mining venture: he was named a “Hero of Socialist Labor” for his success in 1935 in extracting 227 tons of coal in a single shift, more than 25 times the normal quota. One could easily imagine that a workforce led and trained by Stakhanov would be able to achieve prodigies of output from a suitably well-endowed coal mine, and that a coal mining venture led by him would thus have every prospect of stunning financial success.

There is just one problem however: Stakhanov is a convinced Communist, so reliably so that he was the inspiration for Joseph Stalin’s creation of the Stakhanovite movement. As a banker, one would have no difficulty in donning a flat cap, singing the ‘Red Flag’ and toasting the Dictatorship of the Proletariat in a suitably vodka-lubricated bankers’ lunch for the client. But surely it would be madness to go further to accommodate Stakhanov’s political eccentricities – attempts to allocate credit on appropriate Communist principles had a notorious tendency to end up in catastrophic losses, and your bank regulators would be most unhappy if you did any such thing.

If it would be unwise for Stakhanov’s bankers to adopt his political principles, surely the same is true if one’s client base consists entirely of wokies. Feed them tofu and falafel at lunch, by all means, donate modest sums to their climate change and diversity charities, but avoid at all costs hiring wokies for the serious business of banking. In this respect, SVB failed, and it is to be hoped that future banks with politically eccentric customer bases learn from its failure.

The second problem with SVB, unrelated to wokery, was that it was operating with 2023 capital levels when it should have been operating with 1928 ones. At a conservative estimate, 90% of its peak $172 billion in deposits were not covered by FDIC deposit insurance, which operates only up to $250,000 – this appears to have been the highest non-FDIC proportion of any substantial bank. In principle, therefore, 90% of SVB should have been capitalized like a bank operating before the 1933 institution of deposit insurance, in other words with 20-25% of assets represented by capital, compared with the 5% that is considered appropriate today.

1928’s banking was very different to 2023’s banking, and in most respects better. Because deposits were not guaranteed, banks had to operate with an ultra-solid balance sheet, and could not market themselves aggressively, because their depositors were always watching for possible overextension. To instill confidence, banks built Palladian marble offices and operated primarily locally, with a correspondent in one or more big money centers who could help them with such arcana as foreign exchange transactions. To ensure appropriate action by bankers themselves, they wore wing collars, which had the desired effect of holding the head in a stiff, non-slouching position, thus instilling conservatism and probity. The more senior bankers also drove substantial, well-made automobiles, typically Packards.

Long-term readers of this column will remember that, when the fad for “dress down Fridays” appeared around 2000, and the first disasters of “funny money” were manifesting themselves, I called for an abandonment of casual dressing, sloppy monetary policy, aggressive bank lending and wild stock market investment, and a return to “wing-collar and Packard” capitalism. I still believe in that ideal, and SVB is a fine example of why it would be preferable. Yes, 1928’s banking ended in disaster, but that was mostly the Fed’s fault (as today) hugely exacerbated by the economically suicidal policies of the Hoover administration. Wing-collars and Packards were not to blame, and solid bank capitalization prevented the Great Depression from being even worse than it was.

SVB’s mistake therefore came in 2021, when its deposits roughly doubled to the peak figure while its share price was high enough to give it a market capitalization of over $50 billion. At that stage, it could easily have raised an additional $30 billion of share capital, to give it a total of over $40 billion. Had it done so, the $16 billion in unrealized losses on its Treasury bond portfolio between 2021 and 2023 would have been seen as unimportant, and even the extraordinary $42 billion of deposit outflow on March 9 (which would probably not have happened, given the extra capital) would have been relatively easy to manage. (Incidentally, J.P. Morgan knew how to deal with a run on the bank; you give everybody their money, but in $1 bills, counting which slows the rate of withdrawal – you would not have got $42 billion out of the Morgan bank in a single day when J.P. was running it! Dozy Silicon Valley, having to do everything electronically!)

Another solution to the excess deposits problem of 2021 would have lain in deposit pricing. SVB appears to have paid as much as 5% on some deposits, well above prevailing interest rates. From my days on various ALCO Committees, this is what I used to term “Soviet Steel Company Banking” where the cost of inputs (the interest paid on deposits) exceeds the average return on outputs (loans and investments). It is a guaranteed recipe for loss generation, but is quite common in banks where the marketing people plea for special treatment for their favorite customers. The secret in banking as in no other business, is to quell the marketing people and keep them in their subordinate place.

The natural tendency now, given the failure of SVB (and that of Signature Bank, which appears to have been put into administration to discourage crypto-currency traders) is for further bank consolidation, with all deposits guaranteed and the vast majority of banking concentrated in no more than 4-5 banking behemoths. The Biden administration’s dream of a Central Bank Digital Currency could then be used to further this centralization, de-platforming cash and forcing all transactions to be done by book entry. Naturally, this extreme centralization and loss of anonymity could be used to prevent disfavored individuals from any independent economic existence at all. The Biden administration would doubtless deny any such intention, but given wokie dominance, it would become irresistible.

SVB was a defense against that eventuality, and its disappearance will damage the banking freedoms of all of us. As this column has noted previously, the ideal banking system is that of Britain in the late 18th Century, a multitude of small local Country Banks, operating within the framework of a Gold Standard, which prevented government from almost all monetary meddling. In that system, a local entrepreneur with a good idea, even of working-class background, could get introductions to suppliers of capital and short-term banking facilities. The banks would not be universally capable, but the banks in each town would have a good specialized knowledge of the businesses that were important to that town.

Silicon Valley Bank, had it been run on a properly conservative “Wing-collar and Packard” basis was an example of such a bank – in that case serving the extraordinary growth nexus that is Silicon Valley: venture capitalists, private start-ups and entrepreneurs. Its specialist knowledge, network of contacts and understanding of start-up needs all appear to have been incomparable, potentially providing a banking intermediary that was both hugely economically powerful and immensely lucrative. The bank’s wokie staffing, incompetent risk management and gross undercapitalization, all of which combined to make its bankruptcy inevitable, are thus an enormous pity. It is to be hoped that a “Wing-collar and Packard” successor can emerge from its ashes.


(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)