Less than an hour before the California financial regulator closed Silicon Valley Bank’s doors on Friday, short seller Dale Wettlaufer is walking me through their financials, and laying out some metrics he’s been closely eyeballing for months.
“I’ve never seen a situation like this change so quickly,” says Wettlaufer, partner at the short-selling shop Bleecker Street Research, which opened its short position into SVB in January.
Wettlaufer wasn’t talking about the events of the last two and a half days—when a bank run ushered the California financial regulator into its offices to close it down not long after SVB said it was raising more than $2 billion in capital through a share sale. No—Wettlaufer was referring to the last two years.
In 2021, as the venture market soared to new heights, Silicon Valley Bank was flying high. The bank had long sat at the very heart of the private markets as a lender and banker to some half of the industry’s startup companies—not to mention as a prominent lender to venture funds, private equity funds, and a wealth manager to rich entrepreneurs. The bank has a fund of funds, investing in the likes of Accel or Sequoia Capital, and it invests directly in startups itself. The bank effectively touches every part of the private markets.
Riding on low-interest rates at the beginning of the COVID pandemic, startup funding soared to new heights as startups reeled in billions from venture capitalists and VCs raised enormous multi-billion funds. It was all a win for SVB, as many of those startups would park their newly-won funding at Silicon Valley Bank, then draw from it as needed. Venture funds would borrow from SVB as they waited for limited partner dollars to hit the bank. As Silicon Valley Bank required some of those funds to park money as collateral for those loans, SVB was sitting pretty.
Non-interest-bearing deposits at the bank soared to $126 billion in 2021, nearly double the $67 billion the bank held in 2020.
What to do with all that new cash on the balance sheet?
“In 2022, the liability composition changed so much and so quickly,” Wettlaufer explains as he walks me through his own models.
At the height of the venture boom, with the bank sitting on so much cash, its long-term securities portfolio grew from $17 billion to $98 billion, Wettlaufer explains. SVB invested that cash at the height of the market.
Now, interest rates aren’t zero any more, and that long-term securities portoflio is underwater by about $15 billion, Wettlaufer says, meaning that, if SVB had wanted to trade bonds within that long-term portfolio to free up capital, it might have had to recognize somewhere up to $15 billion in the unrealized losses it had reported at the end of 2022.
But another result of the newfound high-interest rates of 2022 was that Silicon Valley Bank’s own interest expenses would soar to absurd levels—both because of the Fed hikes, which necessitated SVB to up the interest rate it was paying to customers, but also because venture funding slowed down in 2022. That slowdown caused inflows of non-interest-bearing deposits at the bank to fall below outflows of those deposits. That’s because startups and other customers were burning cash. Because of this, the composition of SVB deposits changed drastically. Noninterest-bearing deposits fell $45 billion in 2022, forcing the bank to replace those with higher-cost liabilities than what had funded it previously.
As of Dec. 2021, SVB’s interest expense on its deposits was $62 million. By Dec. 2022, it was $862 million. By the end of this year, Wettlaufer was projecting it to be nearly $4 billion.
When Silicon Valley Bank posted its annual report at the end of last month, non-interest-bearing deposit levels were clearly deflating. And it seemed like those figures would keep falling. Wettlaufer was projecting non-interest-bearing deposits might go down to $40 billion, from $80 billion, meaning that SVB would have had to come up with $32 billion somewhere else to fill that hole.
“It’s insane. I’ve never seen anything like this,” Wettlaufer says, noting that the bank may have had to resort to selling anything it could, laying off staffers, or other means to improve the numbers.
All of that was before the panic set in. On Wednesday, the company said it had sold all of its liquid securities portfolio, recognizing a $1.8 billion loss, and that it was trying to sell shares to garner more than $2 billion in capital. The bank also revealed some updated projections: a decline of somewhere around 30% in net interest income from its 2022 outlook. On Thursday, venture capital investors were warning their portfolio companies to withdraw funds, founders were panicking, and it sparked a good old-fashioned bank run. By mid-afternoon Friday, California regulators had closed Silicon Valley Bank, and the FDIC had been appointed receiver.
“I’ve never seen a balance sheet crumble this quickly,” Wettlaufer says.
Bleecker Street Research, of course, has made out quite nicely from the demise of Silicon Valley Bank (The team won’t comment on how much they made off their short bet).
But “excited” would be the wrong word to depict how Wettlaufer and Chris Drose, founder of the short-selling fund, are thinking about it. As Silicon Valley Bank crumbled in a mere three days, it left the entire startup ecosystem frozen, and we have yet to see how far the panic will reach—and how many companies and funds will be impacted.
“You never short something thinking it will go into receivership,” says Drose, who called me shortly after the California regulators had shut down the bank.
While he doesn’t look twice at a company he’s shorted that has engaged in fraud or where stakeholders have taken advantage of people, Silicon Valley Bank was a real business that found itself in a difficult position, he says. This is different.
“This is not a contained event…. I don’t know what happens… I don’t know how it all shakes out, or at the end of the day where that money ended up and where it went,” he says.
And meanwhile, people who put their money in a bank, who trusted it was safe, have been left holding an empty bag.
“That is certainly not something to be celebrated,” Drose says.
This story was originally featured on Fortune.com
More from Fortune: