March 17, 2023
A drop in startup fundraising was the primary driver of SVB's collapse
So here's yet another take on the Silicon Valley Bank (SVB) collapse. Hopefully this will be somewhat unique though.
First, it seems the driver of all of this was SVB's deposits falling due to VC backed companies, in aggregate, burning more than they could replenish in fundraising. I hadn't heard folks mention this as the #1 issue here, but instead have heard the cause as "SVB was trying to sell treasuries that declined due to rates increasing." This was the second link in a three-link failure chain, and probably the least important. VC asset burn was the most important.
How SVB failed
- VC startups burned through cash and couldn't re-raise so net deposits fell
- This caused SVB to sell investments, which because rates had increased, lead to write-offs
- The sale & write-offs lead to VCs alerting their portfolio co's and caused a run on the bank
I think even if you'd remove #2 and #3, SVB was doomed. It was just a question of when and how, not if. I wouldn't be surprised if venture startups make up over 50% of their deposits and maybe even over 50% of their loans (need stat here and couldn't find it, so this is just conjecture).
SVBs deposits boom from $50B in 2019 to $173B in 2022.
SVBs deposits drop from $198B in March '22 to $173B in Dec '22
And apparently the deposits continued to fall from Jan to Mar '23 as VC backed co's failed to raise follow-ons.
As startups continue to burn through net cash for the next 1-2yrs, here's what could have happened to SVB:
- SVB's deposits could have easily fallen back to 2019 levels, from $173B to $62B. This would make the firesale and timing mis-match much larger than last week, and would make a failure almost inevitable.
- SVB's loans to startups (i.e. its assets) are high risk, highly correlative, and backed by weak collateral startup assets like computers and source code. As more of these startups fail in the ongoing correction, SVBs loan default rates would have jumped, eroding their assets, and possibly causing illiquidity.
Funds raised by VC startups fell from a high of ~$90B in Q4 '21 to ~$40B in Q4 '22 (~50% drop)
As a four-time startup guy and current "small business" founder, I'm sorry this happened for all involved.
The lesson: Any company who has a large portion of their profits tied to venture dollars could see a big profit dip in the next 12-24 months. I've thought this for a long time about Google (ads, cloud exposure), Facebook (ads), and Amazon (cloud). I didn't recognize how much more concentrated the impact of a startup correction would be for a bank like SVB. Another bank obviously at risk is First Republic (FRC) which a lot of startup folks have deposits and loans with.
Other companies at risk, to name a few: anyone who sells chips that fuel cloud services or crypto, certain CRM/project mgmt/analytics or other software predominantly used by startups, and companies tied to housing or commercial space in startup concentrated cities like SF, Austin, and Boise. Most of these companies have already sold off because of the perceived risk, but some more than others.
Other companies at risk, to name a few: anyone who sells chips that fuel cloud services or crypto, certain CRM/project mgmt/analytics or other software predominantly used by startups, and companies tied to housing or commercial space in startup concentrated cities like SF, Austin, and Boise. Most of these companies have already sold off because of the perceived risk, but some more than others.
Let's hope these other companies with venture exposure have been diversifying faster so they have a softer landing than SVB.