Mar 16, 2023·edited Mar 16, 2023Liked by Dave Karpf
Excellent observations! One thing though, you make this observation:
"And that meant SVB — the “bank of startups” was in trouble. Its depositors were burning through cash, and it simultaneously had way too much money invested in long-dated treasuries."
This is absolutely true, but there's a bit of unjustified passive voice in the last clause. I was sitting next to my wife while she was listening to a New York Bar Association webinar for lawyers about responding to the failure of SVB and Signature Bank and the lawyer running it said that in the 3rd quarter of *2022* SVB had INCREASED its holdings in long-term Treasury bonds very substantially. Why? Because the interest rates on them were much higher than they had them. At a time when everybody was predicting that the Fed was going to continue to raise interest rates until inflation was strangled (it strangles inflation by strangling new borrowers *especially* the low interest dependent speculators that you're talking about here). SVB was juicing its earnings on an active and highly risky bet that interest rates were at peak when they bought the bonds and that VC's were going to continue active enough that their deposit base wouldn't erode very fast.
Lou Whiteman, one of my wife's favorite former colleagues as a finance journalist made this observation the day BEFORE the bank run:
"Duration risk is one of the most fundamental things to understand in banking. Every banker knows what it is. And yet, nine out of ten times when a bank fails it is due at least in part to duration risk. Silicon Valley Bank (SVB) failed due to duration risk."
These guys had one job--to manage their durational risk during a dynamic time of raising interest rates--and they chose to make this risky bet for the sake of propping up their revenue stream. They wanted to be heroes, break things, win big, instead of being boring bankers and have some lackluster quarters of earnings. Making the right choices, their stock prices would have sagged in late 2022 and 2023, but they'd still be in business.
Which just goes to show that you're right, and the culture of Silicon Valley has been toxic and worse than useless.
I am going to register a mild dissent from the position taken by Andrew Kadel below. I have no problem with the proposition that risk management at SVB could have been better, but I don't think it likely that it was egregiously bad. This view is statistical: according to the FDIC (https://www.fdic.gov/news/speeches/2023/spfeb2823.html), unrealized losses on securities in the American banking system in Q4 2022 were about $620 billion. That is down a bit from the previous quarter, but still about a quarter of the total capital of American banks. There are plenty of banks with better risk management than SVB but there are also plenty with worse. This is the "systemic risk" posed by SVB: it is not that it plays a significant role in the American banking system, but rather that it is an example of a large class of banks that are in serious trouble because of rate increases. It happened to be the first bank that people noticed and panicked over, and that is because of the special social factors in Silicon Valley that you have identified here.
It is worth noting that the risk of bankruptcy is inherent to the function we demand from banks. We want bank deposits to be money, perfectly safe and available on demand. But we don't want to pay for banking service directly; on the contrary, we expect to be paid. Under these constraints it is absolutely necessary for ANY bank to take more risk on its asset book than it sells on its liabilities. Maybe SVB took more risk than it ought to, but there is a lot of hindsight and selection bias at work here.
Finally, may I point out that the Silicon Valley techbros could have ensured themselves a life free of the heavy hand of the nanny state simply by rescuing it themselves, buying into SVB when it tried to raise capital back on Wednesday the 8th. They were looking for a little over 2 billion and the criticism at the time was that they should have "gone big" and asked for 5? That is, like, change in the couch cushions for some VCs and certainly no problem at all if several had clubbed together. Of course, collective action is not a techbro strength, but it was not beyond the means of some individual investors. Back in the previous gilded age, J.P. Morgan was able to stop the panic of 1907 by arm-twisting his fellows; I guess he must have been a commie.
Why didn't they do this? In my opinion: 1) they valued the opportunity to look smart to their friends above saving their money, 2) as mentioned, they are opposed to cooperation, and 3) they are willing to invest in a 1% chance of making a 1,000 percent return but not a 99% chance of making a 10% return.
As an Episcopal priest, I have a certain affection for that old scoundrel, J.P. who was a very devout Episcopalian. While it often took the form of noblesse oblige & patronizing condescension, he was definitely no libertarian--he believed in collective action for the good of the whole, though as one on the top of the heap, he regarded keeping the heap in good and undisrupted shape as of utmost importance. (On a note of personal interest in ~1910 Morgan and Bishop Lawrence of Massachusetts--the city is named after his family--decided that the church would be in better shape if old priests could retire, so they arranged to have the Church Pension Fund established, Morgan personally gave $1 million in 1914 dollars to the effort & twisted arms to get another $8m & arranged to have an 18% assessment of compensation written into the church constitution--which in practice can't be changed, so the fund that pays me better than I was paid while working will outlast the Episcopal Church)
Anyway being a robber baron doesn't require being an antisocial asshole, but here we are with Silicon Valley.
I agree that SVB may not have been the worst bank at risk management, but they were a big bank by deposits and assets. In their particular situation with the majority of their deposits, by design, being from cash burning startups and their prospects for new business being from highly interest rate dependent VC activity, it's less forgivable for them to ignore this and pile on more "hold to maturity" securities in the second half of last year. "Hold to maturity" securities do NOT have to be "marked to market" so the unrealized losses don't matter--AS LONG AS THE BANK DOESN'T GET INTO A LIQUIDITY SQUEEZE. There are a thousand articles from the middle of last year predicting that VC's would have to cut back and startups would have to burn cash in the coming months because of interest rates. But SVB, which could have afforded first rate risk managers, chose to ignore the conventional wisdom and pursue those profits. For a top 20 bank, they should have had risk management in the top 10% of banks, not resembling smaller regional banks.
Excellent observations! One thing though, you make this observation:
"And that meant SVB — the “bank of startups” was in trouble. Its depositors were burning through cash, and it simultaneously had way too much money invested in long-dated treasuries."
This is absolutely true, but there's a bit of unjustified passive voice in the last clause. I was sitting next to my wife while she was listening to a New York Bar Association webinar for lawyers about responding to the failure of SVB and Signature Bank and the lawyer running it said that in the 3rd quarter of *2022* SVB had INCREASED its holdings in long-term Treasury bonds very substantially. Why? Because the interest rates on them were much higher than they had them. At a time when everybody was predicting that the Fed was going to continue to raise interest rates until inflation was strangled (it strangles inflation by strangling new borrowers *especially* the low interest dependent speculators that you're talking about here). SVB was juicing its earnings on an active and highly risky bet that interest rates were at peak when they bought the bonds and that VC's were going to continue active enough that their deposit base wouldn't erode very fast.
Lou Whiteman, one of my wife's favorite former colleagues as a finance journalist made this observation the day BEFORE the bank run:
"Duration risk is one of the most fundamental things to understand in banking. Every banker knows what it is. And yet, nine out of ten times when a bank fails it is due at least in part to duration risk. Silicon Valley Bank (SVB) failed due to duration risk."
https://fitsandstarts.beehiiv.com/p/theres-no-metric-stupid
These guys had one job--to manage their durational risk during a dynamic time of raising interest rates--and they chose to make this risky bet for the sake of propping up their revenue stream. They wanted to be heroes, break things, win big, instead of being boring bankers and have some lackluster quarters of earnings. Making the right choices, their stock prices would have sagged in late 2022 and 2023, but they'd still be in business.
Which just goes to show that you're right, and the culture of Silicon Valley has been toxic and worse than useless.
That's very well put, thanks.
“bank-run by idiots” is pretty great, tbh.
I am going to register a mild dissent from the position taken by Andrew Kadel below. I have no problem with the proposition that risk management at SVB could have been better, but I don't think it likely that it was egregiously bad. This view is statistical: according to the FDIC (https://www.fdic.gov/news/speeches/2023/spfeb2823.html), unrealized losses on securities in the American banking system in Q4 2022 were about $620 billion. That is down a bit from the previous quarter, but still about a quarter of the total capital of American banks. There are plenty of banks with better risk management than SVB but there are also plenty with worse. This is the "systemic risk" posed by SVB: it is not that it plays a significant role in the American banking system, but rather that it is an example of a large class of banks that are in serious trouble because of rate increases. It happened to be the first bank that people noticed and panicked over, and that is because of the special social factors in Silicon Valley that you have identified here.
It is worth noting that the risk of bankruptcy is inherent to the function we demand from banks. We want bank deposits to be money, perfectly safe and available on demand. But we don't want to pay for banking service directly; on the contrary, we expect to be paid. Under these constraints it is absolutely necessary for ANY bank to take more risk on its asset book than it sells on its liabilities. Maybe SVB took more risk than it ought to, but there is a lot of hindsight and selection bias at work here.
Finally, may I point out that the Silicon Valley techbros could have ensured themselves a life free of the heavy hand of the nanny state simply by rescuing it themselves, buying into SVB when it tried to raise capital back on Wednesday the 8th. They were looking for a little over 2 billion and the criticism at the time was that they should have "gone big" and asked for 5? That is, like, change in the couch cushions for some VCs and certainly no problem at all if several had clubbed together. Of course, collective action is not a techbro strength, but it was not beyond the means of some individual investors. Back in the previous gilded age, J.P. Morgan was able to stop the panic of 1907 by arm-twisting his fellows; I guess he must have been a commie.
Why didn't they do this? In my opinion: 1) they valued the opportunity to look smart to their friends above saving their money, 2) as mentioned, they are opposed to cooperation, and 3) they are willing to invest in a 1% chance of making a 1,000 percent return but not a 99% chance of making a 10% return.
As an Episcopal priest, I have a certain affection for that old scoundrel, J.P. who was a very devout Episcopalian. While it often took the form of noblesse oblige & patronizing condescension, he was definitely no libertarian--he believed in collective action for the good of the whole, though as one on the top of the heap, he regarded keeping the heap in good and undisrupted shape as of utmost importance. (On a note of personal interest in ~1910 Morgan and Bishop Lawrence of Massachusetts--the city is named after his family--decided that the church would be in better shape if old priests could retire, so they arranged to have the Church Pension Fund established, Morgan personally gave $1 million in 1914 dollars to the effort & twisted arms to get another $8m & arranged to have an 18% assessment of compensation written into the church constitution--which in practice can't be changed, so the fund that pays me better than I was paid while working will outlast the Episcopal Church)
Anyway being a robber baron doesn't require being an antisocial asshole, but here we are with Silicon Valley.
I agree that SVB may not have been the worst bank at risk management, but they were a big bank by deposits and assets. In their particular situation with the majority of their deposits, by design, being from cash burning startups and their prospects for new business being from highly interest rate dependent VC activity, it's less forgivable for them to ignore this and pile on more "hold to maturity" securities in the second half of last year. "Hold to maturity" securities do NOT have to be "marked to market" so the unrealized losses don't matter--AS LONG AS THE BANK DOESN'T GET INTO A LIQUIDITY SQUEEZE. There are a thousand articles from the middle of last year predicting that VC's would have to cut back and startups would have to burn cash in the coming months because of interest rates. But SVB, which could have afforded first rate risk managers, chose to ignore the conventional wisdom and pursue those profits. For a top 20 bank, they should have had risk management in the top 10% of banks, not resembling smaller regional banks.
Is anyone tracking where the deposits that were in SVB are going? Where did Peter Thiel put the money he withdrew?