SVB Financial in <1,800 words
"The key takeaway here is that SVB’s bank run was an IDIOSYNCRATIC event due to their unique client base — mostly Tech sector institutional clients with accounts far >$250k FDIC insurance threshold."
There have been a number of excellent articles covering the recent SVB Financial debacle (linked below). However, I was surprised by the number of people who reached out to request a TL;DR version of the events — prompting me to quickly bang this article out for those who want a quick recap.
TABLE OF CONTENTS:
1. CONTEXT: How The US Banking Sector Responded To The Interest Rate Hikes of FY22
2. SVB: “Yep, that’s me. You’re Probably Wondering How I Ended Up In This Situation”
3. Downfall: What Happened on March 10
4. Key Takeaway — And Is the US Economy Doomed?
CONTEXT: How The US Banking Sector Responded To The Interest Rate Hikes of FY22
As US interest rates spiked over the past year, the banking industry as a whole partially reassigned their bond portfolios from AFS accounting to HTM accounting.
AFS stands for ‘Available-For-Sale’, while HTM stands for ‘Held-To-Maturity’.
An AFS portfolio is marked-to-market under OCI, but allows the flexibility of bonds in the portfolio to be sold at anytime.
An HTM portfolio is not marked-to-market and is recognized under the amortized cost method — but bonds in the HTM portfolio cannot be sold without revaluing the entire HTM portfolio to fair value (FV). By design, you can’t sell bonds once they’ve been assigned to an HTM portfolio without consequence.
Banks profit by borrowing at the short-end of the yield curve and investing at the long-end. In any given bank’s bond portfolio, there will be a mix of short-dated and long-dated bonds to serve risk:reward optimization.
Hence, it made a lot of sense for banks to shift their long-dated bonds to HTM accounting as interest rates spiked — given that bond prices fall as interest rates climb.
The largest bank in the USA, JP Morgan, shifted 70% of its bond portfolio away from AFS to HTM at around the same time that SVB Financial did.
SVB Financial also did the same. In 4Q20, they held $30M worth of AFS securities and $16M worth of HTM securities (2:1). By 4Q21, they held $27M worth of AFS securities and $98M worth of HTM securities (2:8). This 2:8 ratio remained roughly the same until 4Q22.
SVB: “Yep, that’s me. You’re Probably Wondering How I Ended Up In This Situation”
Silicon Valley Bank (SVB) is the 16th largest bank in the USA — with $211B in assets as of FY22.
As the name implies, SVB’s client base is mostly made up of tech firms in Silicon Valley — many of which are fledgling startups, and some of which occupy the beleaguered crypto industry.
Many of these profitless firms have been burning capital with a ‘growth at all costs’ mindset over the past-decade, and FY22 has been a true funding winter for the sector — resulting in industry-wide cash burn and outsized redemptions of bank deposits at SVB over FY22.
Total Deposits at SVB fell by -12% or $25B from 1Q22 - 4Q22, from $198B to $173B. This loss of $25B in Total Deposits was partly offset by an increase in ST Borrowings of $13B throughout FY22.
While SVB’s AFS Securities remained largely flat from 1Q22 - 4Q22, the FV of its HTM Securities fell by -16% from $91B to $76B (-21% from 4Q21 peak of $97B).
SVB’s abridged balance sheet (4Q20-4Q22) At end-FY22, the ST:LT liquidity ratio of its investment assets was 2:8. For comparison, the same ST:LT liquidity ratio was ~1:1 for JPM, CITI & BAC respectively.
i.e. ST = Cash + AFS, LT = HTM + Loans
$86B of their HTM portfolio was held in securities with >10Y maturities (FV: $71B). Most of the remaining $5B are held in 5-10Y maturities — for a total HTM portfolio of $91B (FV of HTM: $76B).
SVB’s portfolio exposure is basically unhedged.
This collapse in their HTM portfolio happened at the same time that their Total Deposits was collapsing throughout FY22. In the spirit of prudent Asset-Liability Management (ALM), SVB sold $21B worth of AFS securities last week — which resulted in them realizing a $1.8B loss that hit their regulatory capital. Management subsequently concluded that the company would be best served with a capital raise.
Downfall: What Happened on March 10
On March 10, one of the leading US crypto lenders Silvergate Capital filed for a voluntary liquidation. This came hot on the heels of the FTX saga in Dec 2022, which has largely painted the crypto sector with a broad brush as a fraudulent industry. US regulators have also been clamping down on the industry recently.
On the same day as Silvergate’s filing, SVB Financial announced the aforementioned capital raise. Famed VC investor Peter Thiel advised its partner firms to pull their deposits from SVB, citing the possible risk of a bank run despite not having any deposits there themselves.
In reality, there was unlikely to have been any such immediate risk — in the absence of hindsight, depositors incorrectly associated the Silvergate filing with SVB’s capital raise and assumed that SVB was also in similar trouble.
Regardless, the damage to SVB’s credibility had been done, resulting in a bank run on SVB Financial. The bank run at SVB was a very IDIOSYNCRATIC event, due to several unique factors:
The majority of their client base came from the Tech sector — which had just been through a sector downturn and had been drawing down on their deposits over the past year in the absence of fundraising opportunities.
Perhaps due to SVB’s stature, clients may not have considered the bank run risk to be high. This resulted in an average SVB account balance of $4.2M — far in excess of the FDIC deposit insurance threshold of $250,000 per account per bank.
This compounded the bank run concern when it actually got running, since depositors had a legitimate worry that their money might go poof in the event of a liquidation — creating a feedback loop that exacerbated the situation.
On top of the bank run, SVB themselves had no ability to increase confidence in their balance sheet — since most of their assets were tied up in LT investments. Liquidating their HTM portfolio could have forced SVB to recognize a $15B loss overnight — which would have immediately rendered their regulatory capital nil and the bank practically insolvent.
The other problem was that SVB’s investment base was extremely top-heavy in the long-end of the yield curve — it basically only consisted of Cash, AFS securities (ST & LT), HTM securities (LT) and Loans (LT) — and the ST:LT liquidity ratio was 2:8 as we saw earlier. A bank with a more diversified investment base — or even one which was more diversified duration-wise — might not have sparked the magnitude of panic among depositors which ultimately resulted in the bank run.
The bank ended up being put into receivership under the FDIC, and the Federal Reserve just announced that it would hold an emergency closed-door meeting to explore possible remedies.
WHY WAS RISK MANAGEMENT SO CONSPICUOUSLY ABSENT AT SVB? Here’s my own hypothesis:
In an environment of easy monetary policy (prior to the interest rate hikes), SVB somehow managed to convince its depositors to deposit money with them in excess of the FDIC threshold (i.e. as unsecured lenders) — in exchange for higher-than-market deposit rates. This might have sounded palatable with a pointed “What’s the risk of a bank run at SVB? Zero.”
However, doing this would have increased the cost base of SVB’s deposits — which subsequently required management to look for higher-yielding investments to allocate deposit proceeds towards. In classic TINA fashion, this could have meant reaching for yields at the longer-end of the yield curve in an unhedged manner — not too dissimilar from others buying 100-year tenure bonds.
Once it became apparent that interest rates would climb persistently, it might have dawned on SVB’s management that they could soon be forced to entertain withdrawals and recognize unacceptable losses on their investment portfolio. In the meantime, it might have sounded like an easy out for management to follow the industry status quo and switch to an HTM accounting treatment for their LT fixed income investments — this could have initially been a kick-the-can-down-the-road strategy, in hopes of stumbling upon a better solution later. Unfortunately, that solution didn’t come until the risk blew up in their face last Friday.
Key Takeaway — And Is the US Economy Doomed?
The KEY TAKEAWAY here is that SVB’s bank run was a very idiosyncratic event owing to their highly unique client base — which i) were largely institutional clients in the beleaguered Tech sector, and ii) mostly held account balances far in excess of the FDIC deposit insurance threshold of $250,000.
Most US banks have largely retail clients with individual account balances far below the $250,000 insurance threshold — and are not subject to the same degree of bank run risk. The screenshot from the JPM report (linked above) emphasizes this as the main contributing factor behind SVB’s bank run.
While most banks with a diversified asset base & a fully insured depositor base might have safely avoided a bank run despite allocating the bulk of their capital towards illiquid LT investments, this was not the case with SVB:
SVB’s client base was largely from the Tech sector — many of which were likely short on cash flow and much more panicky about losing their deposits than the average business. On top of that, 97% of SVB’s deposits by magnitude were uninsured.
Given that management knew in advance that their client base was unique, they could have easily spotted the risk of a bank run in advance and chosen not to assign HTM accounting to their LT securities. A competent risk management team could have sidestepped this bank run risk from light years away.
For instance, if SVB had assigned AFS accounting to their LT investments instead, the FV losses on their investment portfolio would have been much more visible starting from last year. It may have led to some depositors withdrawing, but it would have given SVB plenty of time to react to capital deficits and rectify the situation.
The real reason for SVB’s failure is due to the wholesale panic for the exits which led to its depositors redeeming their accounts en masse. No bank could survive that — and luckily very few banks will need to given how unique SVB’s depositor base was.
Many market commentators have been projecting that the collapse of SVB spells certain doom for the US Banking sector — with possible contagion effects to the rest of the global economy (alá 2008). Actually, systemic risk is completely negligible due to:
SVB Financial may be the 16th largest US Bank by assets, but a power law dynamic resides within the balance sheets of the US Banking sector. There are at least 5 large US Banks which could swallow SVB whole like a bitter pill and simply be fine the next day. While the fallout from SVB will be unpleasant for the US Banking sector, it is unlikely that there will be any contagion risk to the rest of the banking sector — much less the global economy.
This is to say nothing of the Federal Reserve itself, which can literally print money that is practically free relative to SVB’s tiny asset base of only $211B. Fed officials have successfully dealt with much worse tribulations during the 2008 financial crisis, and they are amply experienced to quickly restore market confidence from the fallout of such a relatively tiny bank (inclusive of the possible future collapse of First Republic).
There is actually historical precedent for this. The Fed could pull another Bear Stearns by asking a big bank like JPM to acquire SVB Financial’s equity for pennies on the dollar + a government guarantee — in exchange for making depositors whole, given that the Fed’s priority is for an orderly exit for SVB.
Notably, the Bear Stearns acquisition in 2008 also involved a “facility to ensure an orderly process of deleveraging as the bank sells off its less-liquid assets” — which is exactly the same problem that SVB is currently dealing with (i.e. one of a liquidity rather than a solvency nature).
Secondly, take a look above at the $2T Reverse Repo balance on the Fed’s balance sheet (RRP). RRP acts as a cash drain on the US economy — and has a similar effect as interest rate hikes (i.e. reduces economic growth).
Even if the big banks don’t step in to bailout SVB, managing any potential contagion risk will be trivial for the Fed — given the measly size of SVB’s $211B of assets relative to the massive RRP balance of $2T.
This was what Powell was referring to when he mentioned that the balance of risks lies with further economic tightening in January — in the event the economy goes into a recession, they can unwind the RRP balance and flood the economy with cash (i.e. QE).
In summary, there is absolutely no risk of any rate action whatsover — pertaining exclusively to the recent SVB developments. If anything, hotter inflation in the wider economy remains the much more pressing risk for markets.
Personally not familiar with the banking sector. Do majority of banks hedge their HTM portfolio and SVB was an exception or this is not common practice to begin with?