> The big losses experienced by the bank are directly related to the surge in interest rates over the past year, as the company's US Treasury holdings were bought at a time when interest rates were still relatively low. Bond prices fall as yields rise.
https://markets.businessinsider.com/news/stocks/silicon-vall...
More general context:
- Banks are required by law to buy US Treasuries (UST). This regulation came about after the GFC.
- UST prices fall as interest rates rise
- the fall of UST prices in the last year is abnormally abrupt and deep
- banks are not required to "mark-to-market" their UST holdings if they plan to hold to maturity
- cash crunches can cause banks to sell UST before maturity, turning unrealized losses into real losses
- SVB joins Silvergate as a previously high-flying tech-related bank suffering a cash crunch and forced to liquidate bond holdings at a loss
It's hard to judge the scope of the problem that Silvergate and SVB might point to. What's clear is that unrealized UST losses on bank balance sheets can surface very quickly and lead to very ugly outcomes.
The brewing crisis is that the Fed needs to trigger a recession (with job loss) to bring down inflation, because the root cause of the inflation is that there are too few workers for the available roles in the current structure of the economy, and so the economy needs to be refactored to drop non-critical industries and inefficient firms. But that's going to cause a cash crunch, since laid-off workers start pulling cash out of banks instead of making it at their jobs. Plus many consumers are drawing down on their savings and going into debt now because of inflation. And it's going to happen right at the greatest velocity of interest rate increases, when Treasuries are at their lowest. So we're going to see bank failures on top of job losses, right as interest rates hit their highest.
IMHO we're already off the cliff, we just haven't realized it yet. It was going to hit in ~2024-2025 anyway as demographics started creating a labor shortage, but COVID accelerated it with a bunch of early retirements and supply chain snags.
Civilization-ending? How is that even remotely in the realm of possibility?
In practice, the cash still probably ends up at a bank, just in a different account. I don't think anyone is going to pull it out and start burying it in their back yard.
Yes. But they’re not required to buy long-dated, high-yielding, high-duration Treasuries (or MBS). Silvergate and SVB, out of incompetence or greed, optimised for yield, not liquidity, despite banking flighty depositors.
Just to underscore the point here, in the past year, the fed has raised rates a ton, and counterintuitively, AGG, an ETF tracking a bond index fund heavily weighted towards US gov debt (by necessity) is down 15 percent over the past 2 years[1].
You might naively assume a bond fund values would reflect interest rates but there is a lag as you wait to roll over old bonds into new debt at the new high interest rate, and until that happens you don't collect any of the extra interest. Even if you sold the old bonds to buy new good ones, nobody will buy them without a discount to make up for the low interest rate.
This is why you have the weird mark to market rules. A US bond _will_ mature at 100 dollars, but can rationally sell on the market below 100 dollars.
Is that counterintuitive? "Existing bond prices fall when interest rates rise" is pretty common knowledge I thought, and it seems quite intuitive to me. If I have a bond that matures in 2 years that only pays 5%, and I can buy a new bond, with the exact same characteristics, but which pays 10%, then if I sold my bond now I'd have to do it at a discount in order to give it an effective 10% yield.
In fact it's so direct that they are quite literally the same thing. The difference between the face value of the bond and the actual amount you have to pay to buy the bond is how you define what the interest rate is.*
* Yes I know subject to time to maturity and coupon and all that.
Personally I always advise friends/family who are considering bond funds to instead consider using Treasury Direct to create their own bond portfolio if their intent is to hold to maturity. Similar can be done with liquid corporate bonds through brokers like IB.
* Withdraw all their holdings, forcing the bank to realise losses in their holdings
* Buy shorts in the stock of the bank
* When the losses are announced, make lots of money from their short position.
As a more practical matter, you would need a very large sum to do this for even mid-sized banks. SV bank alone had over 200 billion in assets, so you would need at least ~10 in cash to make a significant dent in that. If you have that much cash in any bank, there are probably many options you could go for that promise bigger profits at less risk.
It's like everyone reads "tech layoffs" in one news article an then "VC bank default", conclude everything one is directly causing the other and there's a new dotcom crash
If they were to sell those MBS today to get cash, bank’s equity would be wiped out
Source: https://twitter.com/ragingventures/status/161582608803847373...
Given that the law allows them not to if they're HTM, you get the same result, just with some delay. Their deposits are short term, and will leave the bank if SVB is not providing competitive interest rates. They cannot afford to pay competitive interest rates if they've loaned out the deposits at a lower rate than depositors now expect.
Once depositors realize this, it's a classic bank run scenario. They know that without help, the bank goes bust, and there's no point in taking any significant risk with bank deposits that aren't even paying any interest, so they take them out. As more people withdraw, the chance of a failure increases, and more people withdraw, etc.
> $SIVB's HTM securities had mark-to-market losses as of Q3 of $15.9 b...compared to just $11.5 b of tangible common equity!!
> Luckily, regulators do not force $SIVB to mark HTM securities to market. But the bank would be functionally underwater if it were liquidated today. 5/10
[1] https://twitter.com/RagingVentures/status/161582609427121766...
They will hold onto these MBS with their "Diamond Hands" (r) and hope for Fed pivot.
Even if bank will go bankrupt and sold to another buyer - new owner will still have to hold onto these MBS
https://www.reuters.com/markets/us/feds-net-income-turned-ne...
https://thehill.com/opinion/finance/3886002-the-feds-trillio...
https://fred.stlouisfed.org/series/RESPPLLOPNWW
This is a strange world.
I mean that is literally the entire point of their existence.
But I think the Fed has been over-accommodative since 2008. Their reaction function to crises has been to lower rates/print money and then wait. They should've tightened much faster post 2008.
Did Silicon Valley Bank really screw up in 2021 by buying Treasuries if the Fed itself was doing the same? The Fed was doing QE and buying Treasuries in March-22 well into inflation.
I think the error was trusting the Fed to be a good steward of inflation. It is not. We're all learning the hard way.
In early stage land where valuations are the result of a fairly small consensus, it is plausible that SVB would have over-extended.
These losses aren’t related to SVB’s debt portfolio. It’s due to their deposits being flighty.
SVB banks start-ups. Start-ups are spending cash faster than they’re getting it from VCs or customers. That leaves SVB with fewer deposits with which to fund their assets, so they must fire sell assets, which isn’t fun to do.
They have capital ratios to maintain.
If the underlying assets (the assets backing the bank), move in value, then they need to provide extra capital from somewhere. This is them securing that capital base that they need due to the change in value of their current assets (largely US treasuries and mortgage back securities- this isn't really about the value of their tech portfolio).
0 - https://www.federalreserve.gov/monetarypolicy/reservereq.htm
My guess is that many startups are withdrawing their deposits due to this news. I wonder if SVB can cope with a significant bank run.
since there are legions of bureaucrats whose entire professional life revolves around this, and there are legal stress-tests to measure this, and the banking and finance world has a huge whisper network.. maybe one-off speculation is obviously pointless and also maybe manipulative in some way?
Just yesterday Silvergate Bank collapsed. It was the #1 bank in the US for crypto companies. The FTX fallout caused a tidal wave of crypto-related deposits leaving the bank and they were unprepared, apparently having invested the money in bonds that were deep in the red. The Feds stepped in and told them to shut down the bank and repay deposits before things get worse.
(A weird thing about Silvergate is that they bought Facebook's aborted Libra/Diem cryptocurrency tech just last year. It's like there's a curse on Libra.)
It is, of course theoretically possible that some crooks repackaged and sold a bunch of equities as a 'safe' investment instrument to a bunch of morons, on a truly gargantuan scale. But if that has happened, nobody has heard about it.
