Thiel : SVB :: Potter : Bailey Building and Loan?

Why would Silicon Valley's biggest venture capitalists want to start a run on their own bank?

I’m still getting up to speed on the collapse of Silicon Valley Bank, but here’s what it looks like to me so far based on my reading:

SVB was actually a fairly boring bank at heart. They made loans to legitimate companies, took deposits from legitimate customers (often the same companies), understood their clients’ needs well, invested their capital in relatively safe assets, and kept their capital ratios healthy. It just operated in a decidedly not-boring corner of the economy: Silicon Valley.

Because of their expertise in the sector, SVB was in a good position to benefit from the boom in tech startups before and during the pandemic, leading to a massive expansion in their deposit base. When that bubble popped thanks to rising interest rates on top of the usual late-cycle dynamic of over-saturation, SVB faced a substantial and accelerating drawdown of deposits. That meant they needed to sell assets. They had plenty of assets to sell, being well-capitalized, but these assets had faced substantial losses due to the backup in rates, which meant that sales hurt their balance sheet and their earnings. The company tried to raise a bit of capital to shore up the balance sheet, but when a bank raises capital that’s usually a sign of serious problems. A ratings downgrade badly hurt the stock, and it was not unreasonable for both investors and depositors to worry, and seek reassurance or even hedge their exposure to the institution.

So far, so normal. If you were a bank focused on the Texas oil patch in the early 1980s, you probably ran into trouble as the decade went on and oil prices collapsed, because the companies you were lending to would start going under. Understanding an industry is a good way to grow within that industry, and as you become a brand you’ll start to get even more business just based on that brand. But it will leave you poorly diversified, and that leaves you exposed if the industry you are concentrated in begins to turn. SVB’s management deserves to be faulted for not anticipating the extent of potential withdrawals, and keeping adequate liquidity on hand to fund them. And SVB’s portfolio should clearly have been better hedged against interest-rate risk, though it’s worth remembering that the accounting rules around hedging are complicated; I don’t think you can hedge held-to-maturity assets, for example, and most of SVB’s bonds were characterized as such (which is another thing SVB might legitimately be faulted for).

From where I sit, though, the bottom line is that such a large institution with such a substantial asset base should have had multiple possible avenues to explore to get itself out of the squeeze it was in. Any of them would have been painful, but so long as the brand remained valuable, and so long as there was no fraud, I wouldn’t think it would be that hard to find a potential buyer or investor who could have stabilized the situation.

But then something new happened. Major Silicon Valley venture capital firms—the folks who invested the companies that deposited their money in SVB—told their clients to pull their money out before the bank collapsed. And they didn’t just tell their clients privately—they tweeted out the news to the world. As a result, there was a massive digital run on SVB, which led quickly to its insolvency.

Bank runs are extremely hard to stop once they have begun, because they cause precisely the situation that depositors are pulling their money out in fear of. No bank can survive a massive run, because no bank has the liquidity available to pay out to all of their depositors at once. So if you have your money with a bank that is facing a run, you’re basically trying to guess how big the run is going to be and at what point the tiller will be empty, because if you still have your money there when that happens you won’t be able to get it out. It behooves you, then, to join the crowd rushing for the exits—but if you do that, you’re only making it more likely that the bank does go under.

So if you care about saving a bank, the thing to do is not to cause a run in the first place. Particularly if you’re another large institution with substantial exposure, the thing to do is work privately to shore up the faltering institution—on terms favorable to yourself, of course, but also on terms that protect the system as a whole.

Why, then, would the major venture capitalists who are deeply invested in the industry that used SVB as its main bank have decided to cause a run? That’s the question I’ve been asking myself since this all started. Unfortunately, their behavior suggests that those venture capitalists thought they stood to benefit as much by the collapse of SVB as they might have by saving it.

At first glance, that might sound perverse, because the failure of SVB will be very bad for startups, and not just because it could drive some of them into bankruptcy if it takes them too long to get back their deposits and they fail to make payroll. Tech startups are extremely volatile businesses, frequently running for years with massive negative cashflow. Their prospects are extremely difficult to assess. That makes them risky clients for banks, most obviously on the lending side; most banks would be leering of trying to properly value the assets of these startup firms, an essential input to determining how much to lend against said assets. But it even makes them risky as depositors, since they may move money in and out very quickly in large volume. The collapse of SVB will only make future banks less-willing to take tech startups on as clients, because on top of all the other risks they will now view those clients as overly willing to join a bank run.

But if you think in terms of the relative position of the VC firms themselves, regardless of the outcome this leaves them more powerful. Venture capital firms make money by assessing the prospects of those high-risk companies; if everybody else is scared of doing that evaluation, then they set the prices. On top of that, they have proven that they can destroy a landmark financial institution of the Silicon Valley ecosystem; anyone else who wants to operate in that ecosystem will be that much less-likely to ever want to get on their bad side. Meanwhile, if they get the government to retroactively insure SVB’s uninsured depositors by threatening a larger financial contagion, they’ve further demonstrated their clout while also saving businesses in which they themselves are invested.

Of course, Peter Thiel and the rest of the VC crowd are already incredibly powerful and influential; they hardly needed this incident to demonstrate their power. But so was Mr. Potter in It’s a Wonderful Life’s Bedford Falls, and he still wanted to be even more powerful, more influential, more dominant, which is why he was eager to use a run on the bank as an opportunity to take over George Bailey’s building and loan. Indeed, like Mr. Potter, they may prioritize dominance over the prospects for the economy or even their industry. Or they may be convinced (as the fictional Potter undoubtedly was) that whatever is good for them by definition is good for their industry and the economy. They may even have been convinced that they were getting ahead of a disaster by telling the firms they funded to pull out, and that they were doing a public service by telling everyone publicly. It doesn’t ultimately matter. They’re not naive. They knew they could be causing a run, even if they weren’t trying to cause a run. The fact that they were willing to do so strongly suggests that they didn’t think a run would be particularly bad for them.

Where the analogy breaks down, of course, is that the folks who ran SVB, unlike George Bailey, thought they were the same kind of people as Thiel et al. They thought they were allied innovators creating a new world together. They lobbied together against bank regulation and fulminated against the takers of this world on behalf of the makers. They imagined that the big VCs would stick by them when times were tough as they stuck by their startup depositors in tough times. They never imagined the big VCs would actually start a run on them. And they are still sufficiently beholden to them that, I suspect, they will never admit that this is what happened.

So what is to be done now?

The FDIC has a process for dealing with insolvent banks, and one option is just to follow that process and not worry too much. If that process moves too slowly for a number of tech startups to survive, Silicon Valley’s VC firms can step up and be their lenders of last resort. Back in 2008, some critics of the bailouts said that they should never have been done at all, and the chips should be allowed to fall where they may, even if a depression resulted, and this time it strikes me as very unlikely that we’d see that kind of contagion.

But maybe there’s more risk than I think, and we should look to bail out SVB’s depositors. Before we do that, though, we should think about another set of critics of the 2008 bailouts: that, while necessary, they rewarded bad actors, both individuals and institutions. In particular, the bailout of AIG allowed Goldman Sachs—which had been a particularly malevolent actor in putting investors in mortgage-backed derivative products they themselves believed to be toxic, while hedging themselves against any residual risk from those products with AIG—to escape largely unscathed. The line between bailing out depositors and bailing out investors is a particularly fuzzy one when the most important people calling for a bailout of depositors are both investors in those depositors and the folks who caused the run on the bank in the first place. So if Thiel et al want Uncle Sam to do that for them, maybe they should be prepared for Uncle Sam to become their competitor, taking equity stakes in their companies alongside them. I doubt they want that.

If contagion is the real risk, though, rather than the decimation of the startup space, then the most important thing to do is to look at the next domino to fall rather than at SVB. I’m sure—I certainly hope—that the government has been talking all weekend with First Republic Bank and making plans for a possible run on that institution. Looking to the future, if they want to prevent a bank from getting in SVB’s specific situation, the government can certainly increase the limit for deposit insurance (and increase fees concomitantly), can look to tighten regulations around interest rate risk now that the magnitude of expected rate moves is so much higher than it has been for decades, and could even look at imposing diversification requirements for depositors. That all strikes me as fighting the last war, but that’s kind of how regulation works, and maybe that’s part of what prevents the next war from being like the last in the first place.

But we also need to look specifically at the role the prominent VCs played and continue to play in this crisis. This Twitter thread argues that what was unique about this run was just how fast it developed because of Twitter, and suggests that banks need to start worrying about “social media risk”—that is to say, the risk that some viral rumor suddenly causes a run—and how to manage it through better communications. That sounds nice, but we’re several years in to the proposition that the answer to misinformation and disinformation is better information, and I think we now know that while that’s true in the long run, on the tight Twitter timetable it’s not much of an answer at all. Also, there were real problems at Silicon Valley Bank; even if they didn’t threaten the institution’s imminent demise, good but accurate communications could not make them go away. The real point, though, is that this viral rumor has a known origin. If Peter Thiel wanted to know what was going on at SVB, he could have picked up the phone instead of tweeting his suspicions. I guarantee you, they would have taken his call.

I don’t have an answer to how to deal with the power of malevolent influencers in the social media ecosystem. I’m not sure there is a good answer that doesn’t cause more harm than good. But that, I think, is the most important question to come out of the collapse of SVB.