#156 - Silicon Valley Bank: An Autopsy
The First Bank Run of the Twitter Era
Hi 👋 - Over four decades, Silicon Valley Bank built a reputation as a great place for startups and founders. It benefitted from the tech funding bonanza of 2020 and 2021, with rapid deposit growth propelling it to become the 16th largest bank in the US. But managing rapid growth is hard for banks. Today, a look at the rise and fall of Silicon Valley Bank.
“Many individuals grew suddenly rich. A golden bait hung temptingly out before the people, and one after the other, they rushed to the tulip-marts, like flies around a honey-pot. Everyone imagined that the passion for tulips would last forever.” – John Kenneth Galbraith, A Short History of Financial Euphoria
Silicon Valley Bank 101
Founded in 1983, Silicon Valley Bank (SVB) catered to California’s burgeoning tech scene. It earned Silicon Valley’s trust by working with unproven companies that most lenders ignored. Providing white glove customer service forged tight-knit relationships with startups, founders, and VCs. SVB built a sterling reputation by offering a one-stop-shop for founders, including cash management, loans, wealth management, mortgages, and – this being California – vineyard finance1. Over 40 years, SVB went from a fledgling to the nation’s 16th largest bank, with around $200 billion of assets in January 2023.
What was built over decades crumbled in hours. On March 10th, this venerable institution and darling of Sand Hill Road imploded. Citing inadequate liquidity and insolvency, the California Department of Financial Protection and Innovation put SVB into receivership in an unusual mid-day announcement2. The Federal Deposit Insurance Corporation (FDIC) took control, marking the largest US bank failure since the 2008 financial crisis.
What is a Bank?
Banks take deposits and make loans. They generate revenue from the spread between the two. Let's say you deposit your paycheck into a checking account at JP Morgan. This account likely pays no interest. JP Morgan then lends this money out as a mortgage at 6% (or as a small business loan, or a credit card, or any other banking product) and pockets the interest and fees.
The US banking system operates with fractional reserves, meaning that banks only keep a portion of deposits on hand to satisfy customer withdrawals. The remainder can be lent out. If your paycheck is $1,000, JP Morgan might need to keep $100 in its coffers, but is free to do as it pleases with the remainder3.
Every bank borrows short to lend long. While many deposits can be accessed on demand, loans are paid back over time. You can walk up to an ATM and withdraw your $1,000 whenever. However, JP Morgan might have lent your $900 (they kept $100 on hand) out as a 30 year mortgage. There’s an inherent timing mismatch between deposits (liabilities) and loans (assets). No bank can survive if all depositors want their cash back at the same time. Generally this doesn’t happen. But sometimes it does. That’s why banking is a confidence game, and confidence is a fragile thing. As SVB shows, when depositors lose faith, things can go south fast.
Building the Jenga Tower
2020 and 2021 were fat times for tech. As lockdowns kept people indoors, digital businesses flourished. The pandemic forced people to change their behavior, meaning new customers were up for grabs. This growth, combined with low interest rates, led to buoyant tech valuations and heady fundraising.
What’s good for tech was good for SVB. During the pandemic, US banks took on record volumes of new deposits. SVB was the tip of the spear. Between December 2019 and the first quarter of 2022, its deposits more than tripled. Over the same period, industry deposits grew approximately 40%4.
Banks have three options for putting deposits to work: make loans, keep them as cash on hand, or invest in securities. This is where the peculiarities of SVB’s business model start to emerge. The company faced a lopsided supply/demand situation. 2020 and 2021 were go-go years for tech fundraising. Proceeds from IPOs, secondaries, SPACs (remember them?), and funding rounds gushed into SVB as deposits. Yet because startups were awash in cash and raising equity was a cakewalk, SBV didn’t have offsetting loan demand. Consequently, it largely invested its influx of deposits in securities. At the end of 2022, it had about $120 billion in securities versus $70 billion of loans5.
The next wrinkle was that short-term investments like Fed reserves or US treasuries yielded virtually nil. Yet unlike its startup depositors, SVB needed to turn a profit. So it went out the duration curve, buying longer-dated securities like Treasury bonds and agency mortgage-backed securities6. Many of these paid fixed interest7. Indeed, few other banks had as much of their assets locked up in fixed-rate securities as SVB did 8. This generated higher yields, but the tradeoff was increased sensitivity to higher interest rates.
At risk of putting you to sleep, there are two fundamental properties to understand about bonds. First, there’s an inverse relationship between a bond’s price and interest rates. Second, the higher the duration, the more the price of a bond drops as rates increase. Fatally, SVB gorged on long-term bonds while interest rates were at generational lows9. They top-ticked the market. This is like a chain smoker flippantly discarding smoldering cigarettes in a dry forest. As Bloomberg’s Matt Levine notes10:
In traditional banking, you make your money in part by taking credit risk: You get to know your customers, you try to get good at knowing which of them will be able to pay back loans, and then you make loans to those good customers. In the Bank of Startups, in 2021, you couldn’t really make money by taking credit risk: Your customers just didn’t need enough credit to give you the credit risk that you needed to make money on all those deposits. So you had to make your money by taking interest-rate risk: Instead of making loans to risky corporate borrowers, you bought long-term bonds backed by the US government. The result of this is that, as the Bank of Startups, you were unusually exposed to interest-rate risk.
Timber
There’s never just one cockroach. While higher interest rates were the main culprit behind the run on SVB, poor decisions, poor risk management, poor luck, and the peculiarities of its business model all contributed.
To tame inflation, the Fed started raising interest rates in March 2022. Since then, it has hiked nine times, bringing the Federal Funds Rate from 0% in February 2022 to 4.75%-5% in March 2023, the fastest pace since the 1980s. That’s bad news for long duration bond portfolios and SVB was sitting on $120 billion of securities acquired at the top of the market.
Generally, banks benefit from higher interest rates because extra interest from loans exceeds the higher deposit costs. This is another area where SVB was an anomaly, because the size of its securities portfolio far exceeded its loan book, as Robert Armstrong writes in The Financial Times11:
For most banks higher rates, in and of themselves, are good news. They help the asset side of the balance sheet more than they hurt the liability side…SVB is the opposite: higher rates hurt it on the liability side more than they help it on the asset side.
SVB was doubly exposed to rate increases. In addition to reducing the value of its bonds, they also spelled trouble for its depositors, many of whom were creatures of a low interest rate environment. Business models born during ZIRP looked uglier in the daylight of higher rates.
As rates increased, investors shifted focus from growth to capital efficiency, tech valuations plummeted, and the torrent of funding turned into a drought. Yet SVB’s depositors, staring down the barrel of slower growth, still needed to make payroll and fund their operations. Deposits peaked in 2021; after that inflows became outflows.
The situation was exacerbated by SVB’s concentrated deposit base. First, SVB had an extreme reliance on business deposits rather than retail deposits. Business deposits tend to be more shifty when there’s opportunity for better yields or a whiff of trouble. In comparison, retail deposits are stickier. When’s the last time you changed your checking account?
Second, SVB’s deposits were overwhelmingly uninsured. In the US, the FDIC insures deposits up to $250,000. Above this, there’s a risk of loss in the event of a bank failure. At the end of 2022, $152 billion of SVB’s $173 billion in total deposits – nearly 90% – were uninsured12. For comparison, about 50% of deposits in the US banking system are insured13. SVB’s average deposit balance was $4.2 million. Some companies, like Roku, a TV streaming service, held hundreds of millions of dollars at SVB while at least one – Circle, a web3 thingamajig – held billions.
Third, SVB’s depositors were part of a clubby community. Startups and their financiers are interconnected – everyone talks and tweets and retweets each other. Information spreads fast and can cascade quickly. Historically, bank runs are associated with long lines outside branches. SVB will be remembered by Tweets, many fear mongering in all caps. Returning to Matt Levine14:
Also, I am sorry to be rude, but there is another reason that it is maybe not great to be the Bank of Startups, which is that nobody on Earth is more of a herd animal than Silicon Valley venture capitalists…But if all of your depositors are startups with the same handful of venture capitalists on their boards, and all those venture capitalists are competing with each other to Add Value and Be Influencers and Do The Current Thing by calling all their portfolio companies to say “hey, did you hear, everyone’s taking money out of Silicon Valley Bank, you should too,” then all of your depositors will take their money out at the same time.
SVB’s concentrated customer base was influenced by an even more concentrated group of financial decision makers, fanning the flames15:
They [SVB] hadn’t realized the extent to which their customers were not separate entities with regards to their financial decision making. You might of thought you had some diversification there, but in fact, if all of these people are funded by the same few VCs and all those VCs are on the same WhatsApp group, then behaviorally, this isn’t 1,000 guys with $10 million each, it’s one big conglomerate with $10 billion.
When the FDIC shut down Washington Mutual in 2008, it had lost $17 billion in deposits over two weeks. In contrast, SVB lost $42 billion in a single day16. Bank runs used to be governed by physical limits like how quickly a teller could count cash or the speed at which an ATM could be refilled. This also bought time for bankers and regulators to calm panicked customers17. Smartphones and apps make this obsolete. Bank runs can now happen at warp speed in a few keystrokes.
SVB’s securities portfolio and peculiar deposit base provided a tinderbox that combined with rising rates created a conflagration. To accommodate mounting withdrawals and bolster its balance sheet, on March 8th, SVB sold $21 billion in available-for-sale securities18, recognizing a $1.8 billion loss. It also announced plans to raise capital. This was meant to boost confidence, but did the opposite. If selling $21 billion of bonds generated a nearly $2 billion loss, what did the rest of its securities portfolio look like? Was there sufficient capital to absorb further losses? Encouraged by their investors, many depositors didn’t want to wait around to find out, and so the drain on deposits accelerated. An institution built over 40 years evaporated in 40 hours.
If you’re finding this content valuable, consider sharing it with friends or coworkers. 🔥🏦🔥
For more like this once a week, consider subscribing. 🔥🏦🔥
More Good Reads and Listens
Marc Rubinstein of Net Interest on the demise of SVB. Marc’s appearance on Neckar’s Minds and Markets podcast is also a good listen, particularly the discussion on growth and TAM. As usual, Matt Levine’s coverage of the SBV snafu is informative, entertaining, and written in plain english. The Economist on how deep is the rot in America’s banking system? The Wall Street Journal on how bank oversight failed. Lastly, more autopsies from Below the Line: Casper, CNN+, Fast, and MissFresh.
The Wall Street Journal, ‘We Never Thought a Bank So Successful Could Collapse So Fast’, March 18, 2023.
These announcements typically take place after the market closes, often on Fridays, to give regulators and policy makers a weekend to work on a rescue.
The amount of capital required to be held by banks is heavily regulated.
Net Interest (Marc Rubinstein), The Demise of Silicon Valley Bank, March 10, 2023.
Matt Levine for Bloomberg Opinion, Startup Bank Had a Startup Bank Run, March 10, 2023.
Matt Levine for Bloomberg Opinion, Startup Bank Had a Startup Bank Run, March 10, 2023.
Interest on bonds can be fixed or floating. For fixed rates, interest rates are constant throughout the life of the bond. For floating rate, interest rates can vary over the bond’s lifetime. Floating rates are often quoted at a premium to a benchmark like LIBOR or SOFR. If a bond is priced at LIBOR + 100 basis points, as LIBOR moves up or down, the interest paid will move in tandem.
The Financial Times, SVB is not a canary in the banking coal mine, March 10, 2023.
Until March 10th, SVB CEO Greg Becker sat on the board of the Federal Reserve Bank of San Francisco, presumably giving him courtside seats into discussions about rate increases. The Fed generally eschews surprises and had telegraphed the possibility of higher rates well in advance of its first raise. This makes SVB’s actions head scratching.
Matt Levine for Bloomberg Opinion, Startup Bank Had a Startup Bank Run, March 10, 2023.
The Financial Times, SVB is not a canary in the banking coal mine, March 10, 2023.
Net Interest (Marc Rubinstein), The Demise of Silicon Valley Bank, March 10, 2023.
The Economist, How deep is the rot in America’s banking industry?, March 16, 2023. The total liabilities chart shows the US banky system’s insured versus uninsured deposit mix in Q1 2022. It’s roughly 50/50 and contrasts sharply with SVB having approximately 90% of deposits uninsured.
Matt Levine for Bloomberg Opinion, Startup Bank Had a Startup Bank Run, March 10, 2023.
Odd Lots Podcast, Dan Davies On What Brought Down Silicon Valley Bank, March 14, 2023.
The Wall Street Journal, How Bank Oversight Failed: The Economy Changed, Regulators Didn’t, March 24, 2023.
Banks can hold securities as available-for-sale (AFS) or held-to-maturity (HTM). Both have different accounting treatments, which are beyond the scope of this note. Marc Rubinstein provides a good overview of AFS, HTM, and bank balance sheets, in his note on SVB’s collapse.
Cogent and insightful as always, Kevin. Well done.
"Yet unlike its startup depositors, SVB needed to turn a profit." 🔥🔥🔥