On March 10, Silicon Valley Bank (SVB) imploded. The US bank was the 16th largest in the country, with US$209 billion ($282 billion) in assets, and a key cog in the wheel of the innovation start-up ecosystem. Its implosion was the second-biggest bank failure since Washington Mutual’s collapse at the height of the 2008 global financial crisis. The way I see it, banks in the US, as well as the entire tech ecosystem — the key driver of American growth for the past 15 years — will never be the same again.
Founded in 1983, SVB was unlike any other bank in America. Its depositor base was one of the most concentrated of any financial institution on earth — mostly venture capital firms (VCs), their limited partners (LPs) and their investee firms — tech and biotech start-ups. It was the beneficiary of most of the money that large VCs were ploughing into start-ups. More than 50% of Silicon Valley start-ups did business with SVB and it had a hand in 44% of VC-backed tech IPOs last year. Among its early clients was Cisco Systems, the darling of the tech boom in 2000. Big depositors included streaming device maker Roku, video gaming platform Roblox and aerospace manufacturer RocketLab.
The bank was also the biggest lender to startups, which do not just raise a ton of money from VCs in funding rounds but also borrow a ton along the way. A loan to a pre-revenue startup that has not even begun testing its business model is the riskiest loan any bank can make. That said, one in a million start-ups ends up becoming the next Google.
SVB was also more than just a commercial bank. It also owned an investment bank, which three years ago bought a biotech-focused investment bank. It had a growing private banking business to cater to the nouveau mega rich in the Silicon Valley. Moreover, the bank to the VCs was a VC itself, taking stakes in start-ups it did business with through its US$9.5 billion VC fund — which also owned stakes in VC giants like Sequoia Capital and Marc Andreessen’s a16z. It lent money to young founders, including those running promising early-stage startups, like mortgages for mansions or vineyards, or loans to buy sports cars, using their overvalued start-up stakes as collateral.
Yet last week’s implosion was not due to any FTX-like fraud, but lack of proper risk management. Put simply, SVB collapse was about asset-liability duration mismatch. Banks make money by taking short-term deposits and lending it longer term. They pay depositors a pittance and charge borrowers more. The difference is their net interest margin. Make no mistake, banks make a lot of money. In 2021, JPMorgan Chase made US$48 billion in net income. It all works well until customers lose confidence in a bank or indeed the entire banking system. Don’t let anyone tell you that good banking is about stellar balance sheets — it is actually all about confidence. The moment that confidence is lost, it’s game over.
While depositors can ask for their money back at any point, banks themselves have far less leeway selling long-term instruments — treasuries, government-backed mortgage securities, or even loans. If a lot of depositors ask for their deposits back at the same time, the bank, in most cases, is toast. That’s what happened at SVB.
See also: First Citizens weighing takeover of Silicon Valley Bank
Gradually, then suddenly
On the social media megaphone, anyone can trigger a bank run with a few random tweets. If something smells like smoke and you scream “Fire!” there will be a stampede even if nothing is actually burning. Here’s how it all unfolded: A bunch of big VCs in the valley decided that it was no longer smart to bank with SVB. A handful of customers yanking their business normally makes little difference to a bank with a diversified asset and deposit base. But SVB had substantial exposure concentrated in 25 or so large VCs in the valley. Each of those 25 VCs was in turn invested in 40 to 60 start-ups, which accounted for close to 25% of SVB’s total deposits.
VCs and start-ups are nothing if not extremely tech-savvy. Everyone is on Twitter all the time. They do their business digitally, making most transactions through their iPhone apps. One big VC, right-wing billionaire and Donald Trump crony Peter Thiel, urged everyone on Twitter to withdraw all their deposits from SVB immediately. That led to huge withdrawals — not just from the VCs as well as their investee companies but their employees, friends and acquaintances. In a single day, US$42 billion or 24% of SVB’s US$175 billion in deposits were withdrawn. Everything, everywhere, all at once collapsed, like that bank run scene immortalised in the classic movie It’s a Wonderful Life.
See also: Silver lining to the SVB crisis
Though that movie was released long before I was born, I did witness a real bank run, on Standard Chartered, in Hong Kong in 1991. As I walked to my office one morning, I saw queues outside its branches. Rumours had spread that the bank’s licence was being withdrawn by UK regulators. The Hong Kong authorities kept the branches open until late, promising depositors all their money back. It took two days to calm the depositors.
For its part, SVB took deposits from the VCs, then deployed over 90% of that money into treasuries or government-backed mortgage securities. These are, of course, the safest instruments any bank can hold. Here’s the twist: SVB was in long-dated instruments which cannot be cashed at short notice. If you sell them in a hurry, you might get 90 cents on the dollar.
After ratings agency Moody’s indicated it was ready to downgrade SVB’s debt to “junk” status on Feb 28, the bank sought to raise US$2.25 billion in new capital, including US$500 million from giant VC General Atlantic. Regulators had been asleep at the wheel. They knew that SVB was holding treasuries with three- and five-year maturity periods, but holding them until maturity is no good when 25% of depositors want their money back immediately. SVB’s portfolio was not mark-to-market. The US Fed has raised interest rates from zero to 4.75% over the past 12 months in the fastest rate hike cycle in history. When rates rise, bond prices fall. The part of SVB’s bond portfolio that was easiest to sell was at the time worth around 92 cents on the dollar.
SVB’s banker, giant investment bank Goldman Sachs, advised that in the interim SVB should sell some of its treasuries. Goldman itself bought US$21.4 billion in government securities from SVB at around a US$1.8 billion loss. Goldman also charged SVB US$100 million in fees. Now that two-year treasury yields have fallen from 5.1% to 3.9%, government-backed mortgage securities that Goldman is holding are worth a lot more. All told, Goldman could walk away with nearly US$1 billion in profits and fees while SVB has been wiped out.
Yet without the capital raise, SVB was essentially bankrupt. The market had gotten wind of that, SVB stock plunged and VCs like Thiel who had been key beneficiaries of the bank’s ecosystem started pulling out their money. The Twitter feeds of VCs exploded, everyone rushed to the door and the bank, in the end, did not have enough money to keep paying depositors.
As for the blame game, aside from SVB’s risk management issues, look no further than the regulators. Before the fiasco occurred, only up to US$250,000 in deposits were insured by the FDIC, so small depositors did not have to worry about the bank going belly up. Only 2% of SVB deposits were protected by the FDIC. In trying to rescue SVB, the regulators gave out wrong signals. They could have told SVB: “We know you have long-dated government securities that can’t be sold quickly. We will take those assets and help you secure a cheap loan so you can pay the depositors who are queuing up.” That would have ended the bank run, and regulators taking over SVB bonds would have made US$1 billion in profit instead of Goldman Sachs. Yet the Fed, FDIC and Treasury Secretary Janet Yellen dragged their feet, which led to depositors yanking out 24% of all SVB deposits.
Sink your teeth into in-depth insights from our contributors, and dive into financial and economic trends
From now on, depositors and borrowers will flock to bigger, safer, too-big-to-fail players like JPMorgan, Citigroup, Bank of America and Wells Fargo. Small banks are toast because the government is unwilling to say it will do whatever it takes as long as they have good assets. Confidence in small banks has been shattered.
There are far too many banks in America — 4,150 or so at the end of last year. Don’t be surprised to see a rapid consolidation among US banks over the next decade. The vast majority of the small banks in America will either merge with larger, too-big-to-fail players, or fail.
Moreover, if regulators are now implicitly guaranteeing all deposits, even those over US$250,000, the cost of that guarantee will continue to rise. It is the small banks that will pay the FDIC to insure all of the deposits. They would rather merge than lose depositors to JPMorgan or pay higher costs. To stem the tide, some smaller banks will be forced to raise deposit rates, which means lower interest margins for them.
Sure, small community banks with a single branch in a small town and no ATM might still thrive, but everyone in between the big four banks and the community banks is now vulnerable. Small banks are well capitalised and their deposits are backed by the safest US treasuries. The problem is that many of them hold long-dated instruments that can be sold quickly only at a loss when rates are rising.
Why didn’t regulators see the red flags? The Dodd-Frank Act of 2009 allows stress-testing of banks to make sure that if one is failing, it doesn’t bring others down as well. Until 2018, banks with more than US$50 billion in assets had to undergo annual stress tests. The Trump Administration pushed Congress to raise that limit to US$250 billion. That allowed SVB, and Signature Bank — the other bank that went belly up last week — to evade the mandatory stress tests. If the two had been stress-tested, and their bond portfolios were mark-to-market, they would not have to be shuttered and required a backstop for depositors.
Changes in how banks operate aside, there will be an even bigger shake-up in the start-up ecosystem and the larger tech sector. More of the smaller tech companies will likely fail or be forced to merge with bigger players because they can no longer access cheaper capital from specialised banks like SVB. Fewer start-ups will get funding from VCs, who themselves now have to bank with the biggest banks. Fewer start-ups will be able to access bank debts now that SVB is out of the picture. Why would Citigroup make a loan to pre-revenue start-up? Why would Wells Fargo give a mortgage to a start-up founder for a mansion or vineyard? Moreover, start-ups with dodgy business models will find it hard to get funding, and those with funding will find the IPO process or the exit for their founders and big shareholders becoming a big issue.
That means less innovation and indeed slower economic growth in the US. Whatever the failings of SVB, the start-up ecosystem and VCs, until now it has helped build a robust engine of growth for the US economy. That engine is now smaller, burdened with more regulations and driven by larger, more risk-averse banks. Surely that can’t be good for the tech sector and the US economy as the country tries to crawl back up from the hole that it is currently in due to persistent rate hikes of the past 12 months.
Assif Shameen is a technology and business writer based in North America