Silicon Valley Bank: Depositor Frames of Reference
We take a shot at a simple asset coverage test for uninsured depositors at SVB as we await more current info.
Liability Management Team on site.
The threat to uninsured depositors in the Silicon Valley Bank meltdown is gripping the markets and is even more important to many employers and employees who might not make payroll. Some companies may have to file Chapter 11 to avoid any commitments to the bank (or out of revenge), and that will compound the problems.
As everyone watches the headlines over the weekend, we thought we would at least try to reassemble some of the high-level numbers that flowed into this disaster as SVB grew dramatically from the end of 2019 and then across COVID and the wild world of the IPO boom. Then came the 2021 spike in growth stocks and anything with a hint of cache (Peloton, Carvana). The magnitude of the multiplier effects of this bank crisis are somewhat reflected in how the UST market responded as discussed in yesterday’s commentary (see Silicon Valley Bank: How did the UST Curve React? 3-11-23).
The recent attempt to execute a major financing as part of a strategic overhaul obviously failed, but not before it did more harm than good as the sale of “substantially all” of the AFS portfolio at around $21 bn simply ran out the door on a first come first serve basis. That story remains to be told.
In looking at the chart below for the end of 2022, we can see that AFS line is gone, and the deposit line has shrunk by some number that we do not know for sure. Let’s assume $21 bn although we do not know what came in or went out before then in recent days. We assume SVB ran out of cash on the Prisoner’s Dilemma breakout of last week. Then we have the FDIC insured deposits and questions on the timing there this week. That will trim the deposit number, but the challenge is on.
We take a simple (very simple) approach to looking at the trend line in the key assets and most important liability in the public (and private) eyes right now – the depositors. We keep this analysis very narrow and warn that there is of course a lot more to this exercise. For presentation visual sake, we look from 2018 to 2022 at the main events of the securities portfolio time series (Available-for-Sale or “AFS”, Held to Maturity or “HTM”), the loan portfolio, and how that all frames up relative to deposits.
The chart shows the AFS line, and that balance had more than tripled since 2019 before SVB sold “substantially all” of it in March for $21 bn. The big leg up over the past three years was in HTM securities which climbed well over 6-fold from the end of 2019 or three calendar years. As a footnote that just got smaller, SVB had reclassified $8.8 bn in AFS as HTM in 2021. There were no redesignations during 2022, but we can assume everything is for sale right now to meet depositors’ needs. So far, the FDIC and company have been quiet on the topics.
The net loan portfolio line of $73.6 bn is the hardest to measure since the longer this drags on (I am referring to days, not weeks) the more it will decline in value. SVB detailed their credit scoring system in their footnotes and have very rigorous disclosure by category. For arguments sake if we haircut the loans by some number between 10% and 20% you can play “fun with crisis math” at home.
If we deduct the AFS and use the fair value of the HTM portfolio at year end (approx. $76 bn) and haircut the loan balance by 15%, the news is at least not that bad for the uninsured depositors. We still need to adjust the deposits for the $21 bn in AFS sold, apply the fair value of the HTM securities at year end and then we can get to a 90% area recovery. If we haircut for the insured deposits, you can even get to full coverage. The problem is the loan portfolio and uncertain valuation. The haircuts on loans will look more like Paris Island than Silicon Valley if there is not a lot of movement very quickly.
This is just a simple cut. There will be other claims in restructuring depending on what form it takes and how soon. Like everyone else, we’re on standby waiting for more information.