FDIC Takes over Silicon Valley Bank - https://news.ycombinator.com/item?id=35096877 - March 2023 (737 comments)
"Silicon Valley Bank is the first FDIC-insured institution to fail this year."
Four failed in 2020. Four failed in 2019. None failed in 2022 or 2018. Banks fail fairly regularly, but it's the size of the bank and the types of deposits that are concerning in this case.
Most years seem to have a few failed banks, and a lot in the 2008-20011 era, but the past few years don't seem to have as many.
The story seems to be that SV put all their deposits into 10 y bonds in 2021. I'll use that as an approximation.
A 10Y bond will usually move about 8x as much as the underlying interest rate (it's called "duration"). So if SVB did nothing but buy these bonds and sit on them, then they would have lost about 36% on these (8 x 4.5% rate movement). That's a lot but also better than losing everything. And you'd hope the bank wasn't idiotic enough to take no precautions: no hedging of IR changes, putting all assets into just that.
My finger in the air haircut would thus be maybe 15-20% loss of deposits, if this simplified model has anything to do with reality. That's a lot but also means depositors get 80-85% of their money back.
It's possibly pushing the limits of my toy model here but a 10Y bond bought in 2021 is now only an 8Y bond, because time has passed. So instead of 8x you get more like 6x and the pessimistic haircut number is thus 29%.
That's not what happens. Let's say 100 clients each deposited $1 in the bank, and the bank loses $20, so only has $80 to pay out when liquidated.
Let's say half (50) depositors withdraw their funds early, they each get $1 back. So now the bank has $30 in assets and has to pay 50 people. Suppose 20 people demand withdrawals, and the bank pays $20 to them. Now the bank has $10 and has to pay 30 people.
I admit that I don't know when/whether a bank can actually refuse withdrawals in such situations, or when bankruptcy proceedings are triggered, but the earlier you withdraw the less likely you're going to lose money, and the later you withdraw the more likely you'll lose a significant part of your deposits.
Sure your calculus still works out on average, but I'm not sure that means anything and it's no consolation for the ones that reacted last.
We'll see what happens, but I would be surprised if more than 20% of deposits were withdrawn.
Depositors have been getting 100% of their money back. [and will still do if they are quick (or has the bankruptcy been declared already?) - edited: too late!]
The loss - that is an amount measured in dollars, not an abstract percentage - is going to be supported by a smaller number of depositors.
If 70% of depositors exit (which actually seems possible) the bank does not have adequate assets for the remaining depositors and the bank fails.
Well, not at the same time. No bank can do that.
I never traded cash bonds but would struggle to imagine it would be hard to liquidate. Sure, over a week, but surely SVB can't be a major bondholder.
If SVB really parked most of the cash in 10-year old treasury bonds that sounds.... questionable...
Not sure who told you that, its incorrect.
https://twitter.com/jamiequint/status/1633956163565002752?s=...
We’ve had 15 years of people getting billions for phone apps made in 12 months and forgotten in 6. Random individual software startups are valued more than basically the entire hardware industry under them. Bitcoin peaked out at 1.28 trillion and it still has no use beyond being a converter from dollars, to crypto, to dollars again with the occasional hand wavey “dude it’s totally completely secure and stable and everyone will use it in the future as a currency! Look how much I made by investing at the peak!” story.
I was out of work for 18 months during that period. Luckily I had a background in Animation and Architecture to fall back on and got a job designing tech companies when the crisis lifted.
2008 was worse though.
Will missing payrolls result in layoffs/resignations? Thus, increasing unemployment rate that the FED desires. If contagion doesn’t spread outside of tech/startup, does the government have any incentive to intervene? Maybe this is not too big to fail.
A sale seems like the most likely scenario out of this liquidity problem (insolvency). It’ll be dirt cheap and has to make sense to its buyer.
edit: Even with recent downturn in tech, there is still sizable value in having tech as banking clients.
So, in that light, losing these jobs would be like throwing away the kid with the bathwater.
I personally don’t agree with all the premises and conclusions I’ve enumerated above, but imo that’s how a mainstream economist (like one working for the FED right now) is most likely to view things.
I've also worked at both startups and large companies where problems with the bank or payroll software delayed payroll a day. In those situations there's typically overcommunication about the steps being done to get it resolved ASAP so employees don't walk out.
Q1 of 2023 has stabilized a bit and began with a market rally, but it could just be a dead cat bounce on the path to lower lows as the year goes on. Seeing SVB implode suddenly is ominous not so much because of them, but what it means for the industry right now. I was hoping that we maybe hit the trough in Q4 of 2022 but it would appear this year will be bleak and I sadly suspect we'll be seeing more layoffs in the second half of this year if the first half continues to miss.
I hope I'm wrong. We are about at the point where a bear market ends (around 9 months) but that's just an average. We could very well have quite a few more months or even the entire year and into next year ahead. High flying tech has been in a bear market now for over a year, although it probably needed a correction.
(I understand that Big tech is also doing layoffs because they massively overhirred during the pandemic)
I look to Sequoia because, of any VC fund, they are the most likely to be able to pull off a takeover of SVB. I am still gobsmacked that they invested hundreds of millions in a company with no board or CFO (FTX)
However, sad to see what looks to be another government caused implosion.
- First, via the massive government stimulus COVID19 SVB got enormous amount of deposits in 2020/2021. They have to put that money to work, and the traditional lending it back out was not going to work, the market was already flooded with money from the government. So they sadly locked themselves in those MBSs.
- Then the gov finally reacts to their own stimulus and zero interest rates programs and they rapidly raise interest rates in attempts to lower inflation.
- And as a result, the MBSs that SVB holds are underwater in mark to market. Investments into startups fall, so no more influx of customer deposits into SVB. And finally, due to the difficult market (and VC market), all the startups are drawing down their deposits rapidly through their current burn rate. And SVB ends belly up.
Banks are in the business to manage risk.
The sad part is this risk management and investment process is concentrated in a very small group. Most of the bank employees are very good, had no idea what was happening, and some will lose their jobs as their equity goes to zero.
Let's see what happens. If only SVB goes belly up, then yes, only SVB was the one that couldn't read the tea leaves of the government's actions. But if SVB is just the first of more banks that will go belly up, then perhaps it is not just an error on SVB's part.
But they used to have a sheet they posted on the wall near the tellers - this was in 2007-2009 at the height of that particular financial fraud collapse (lehman, etc)
The sheet showed how much the bank was making on over draft fees - and every time I was in the bank I would check that number - and it was stunning to see as deeper we got into that recession, how much that number increased.
They were making about 500k per month on _overdraft_ fees alone...
That said a lot about the state of peoples finances alone.
In 2022. Up till the end of 2021, they kept saying inflation is "transitory". According to reports SVB bought a bunch of 10 year MBS in 2021.
SVB is obviously incompetent in believing in that Fed BS about transitory inflation -- wait but should they have believed in the rate hikes as well then?
The market bet against the fed, assuming fed was all talk no action. Wrong move.
Most of the market is still anticipating a fed pivot / fed put and / or tax-payer bailout. Time will tell.
You mean the one that would have been caused by the government's draconian lockdowns?
Good thing the government came in to rescue everybody...
Could they not have cut interest rates and take the dip in profit with shorter duration bonds? Was there really no sensible way to prevent this?
Businesses fail all the time from things outside their control, but I've yet to see any indication that's the case here.
That's putting it lightly. Powell has been saying over and over again "more inflation, more rate hikes." Even giving us approximately how many they are planning on doing in a given timeline. I don't understand why so many people think it's a good idea to prepare for the opposite of what the government is literally telling us it's going to do. That does not seem like a smart bet, but maybe that's just me.
You can blame the government for some stuff but at some point companies have to own the decisions they make in the face of not just evidence, but also clear statements, that run counter to how they are operating. SVB made their bed in a large way here.
SVB didn't implode on their own, Thiel decided all the startups he backed should withdraw their deposits. Then other VCs followed the trend.
SVB was what, $2B short on their balance sheets, out of >$200B assets? VC initiated bank run is the only real issue here.
Waiting to see how this all benefits Thiel.
They were going to have to say something sooner-or-later. I think the biggest part is that before they could raise capital, they needed to be honest about their books or risk securities fraud charges.
This is wrong. DETR, the Nevada Unemployment Office, gave out 1.4B too much [0]. I am absolutely sure this is not an isolated case. The absolute level of corruption disguised at absolute incompetence is too high to let the government off that easily. 2020-2021 was the biggest wealth transfer from the poor to the ultra rich in the history of the world. The stimulus didn't even have a clawback attached to it.
[0] https://www.ktnv.com/13-investigates/detr-sent-out-1-4-billi...
Fed hasn't raised fast enough and is looking at +.50% next meeting.
Sure, keep the lender of last resort; but maybe let the market determine interest rates through money markets, with incentive-control through judicious laws, instead of by fiat.
https://www.bloomberg.com/news/articles/2023-03-10/us-regula...
The fed has a huge advantage in that they can just print money, buy assets that deliver 1% returns and when those assets mature delete the printed money. The fed is basically leveraged to infinity.
edit: At the very least, FDIC should issue a statement guaranteeing beyond the $250K/depositor limit sooner rather than later in order to stem some of the outflow.
If the natural incentive of the bank's assets isn't enough, its literally the job of the FDIC to add that incentive. That's what the FDIC does. "We'll pay you $100M to continue operations of this bank" this happened dozens upon dozens of times during 2008. The insurance component of their mandate is an absolute last ditch "holy shit actually they had nothing it was all smoke and mirrors" tool.
It won't come to that because, well first of all, regulations don't allow banks to get to that point anymore, but also: SVB is actually a solid bank, who made a couple fuck-ups, caused a panic, and their customers are generally more agile and cutthroat than industry average. There are so many tools they, the industry, and regulators have to stabilize things that if you walked into the meetings that are happening about this right now and said "we should raise our insurance guarantees above $250k" you'd be laughed out of the room.
The FDIC doesn't give a shit about stemming outflows from SVB. The FDIC cares about the system, not SVB's bottom line or stock price. Customers should be able to withdraw their money. If they're even involved at this point (and that's a huge if), their primary concern would be to make sure customers still can get money out, not changing the incentive structure or adding limitations toward getting that money out.
This is a liquidity problem (so far). This isn’t because loans went bad or anything, just that they needed cash quick and had to fire sale things (the things they took losses on were US Treasuries. Mostly.)
They still have a big loan book that is probably worth more than their liabilities.
A large player with liquidity could get those loan assets on the cheap if they agree to provide the short term capital liquidity (a reassuring name also probably stems any bank run risk).
Also the FDIC is pay in pay out insurance. The commercial accounts at SVB are in the hundreds of millions and no payments were made for those accounts. The FDIC probably won’t help.
In theory they could hold that shit to maturity but in practice they can’t because some folks want money now and all depositors will want higher interest on deposits eventually to match market rates.
How is that known currently? Edit: As opposed to a solvency problem?
That said, if I'm an SVB customer...do I want to stick with the bank that's had this issue? Even so, would I want to bank with the acquirer?
is there precedent for that?
Again, it would be better to have a comment from someone who banks for a living, but back in 2008 when the FDIC limit was $100K, ISTR Sheila Bair making a statement to that effect (IndyMac?). One large issue was that companies would regularly be above the limit in order to run their payrolls, and so got caught in a bad position. That started companies with other banks worrying about where they were parking money, contributing to a systemic issue.
At this point, does SVB have positive brand value anymore, or are they negative?
> At the very least, FDIC should issue a statement guaranteeing beyond the $250K/depositor limit sooner rather than later in order to stem some of the outflow.
The FDIC chose instead to shut the bank in the middle of the business day. They apparently don't see this as inciting a systemic event, where other uninsured deposits at other banks will now begin flowing out.
That imho is not playing it safe, given some of the price drops in the other banks today.
I think odds are good that we see that this weekend.
Edit: aaand FDIC couldn't wait until COB
https://finance.yahoo.com/news/fdic-closes-silicon-valley-ba...
One reason that it was done in the past was because it saves the taxpayers money by bailing out a single meltdown rather than something systemic.
But you should recognize: The taxpayers back up the FDIC, but singleton meltdowns are paid out of the insurance fund, which is paid into by the banks themselves. As long as that fund does not get exhausted (ie a systemic meltdown) it doesn't cost the taxpayers directly.
They’d have to I assume to value their new books.
If yes that’s a lot of valuation haircuts.
What advantages did SVB provide that other brand named bank didn't? Wouldn't it have been safer to put your startup's money in Bank of America, or JP Morgan, etc? I've never heard of SVB until the crash.
SVB would be lenient to the entire portfolio in exchange for access to the best performing companies.
i.e. Company A is a Series A start-up that needs non-dilutive financing to bridge itself to the next round. SVB would be willing to lend when no one else would. In exchange, VC’s would introduce SVB to much healthier growth stage or pre-IPO companies that need a banking partner.
The lending keeps money coming in (repayments), makes money (interest), etc. - so how was SVB supposed to make money and maintain liquidity without much of that?
But it is a good lesson in money management…something about eggs and just one basket?
It seems like it would be better for VCs to keep the money, invest it (in, I don’t know, the S&P500), and then pay out to the companies on a monthly basis or some other terms.
I can’t fathom how there are hundreds of companies, each with 10s of millions of dollars, just sitting in a zero interest bank account?
Given this, it's pretty attractive for a buyer. The buyer just needs to provide some capital to bring the reserves back into compliance.
To be fair, the government may find value in shoring this bank up. That's a different story. The bank will be essentially nationalized at that point.
How do these keep blowing up?
It's alright if things fail or lose value. That's a natural part of markets. Taxpayers money going to bail out banks, otoh, is pretty much what started Bitcoin.
"Chancellor on the Brink of Second Bailout for Banks" - still encoded in the genesis BTC block.
Silvergate's failure doesn't affect BTC on paper, yet Bitcoin did go down.
SVB's failure doesn't affect the USD on paper, yet the dollar is down today[1].
All currencies, fiat or otherwise, are affected by catastrophic events in their ecosystem. The folks having a good day today are the precious metals.
---
[1] Although to be fair, there are a myriad of factors at play in FX, whereas Bitcoin probably has few. However, "bank failure" and the US Treasury announcing they are monitoring some small and regional banks, doesn't fill one with confidence.
They'll probably be in line with anybody else and being foreign makes enforcement and lawsuits so much more expensive and difficult.
I guess that's a price they were willing to pay to get more favorable conditions in SV in the past as opposed to whatever they could find locally.
My guess is the feds come in today and do it. They love to close banks on a Friday and work it through the weekend
"Two ways. Gradually, then suddenly.”
― Ernest Hemingway, The Sun Also Rises
They say treasuries are as good as cash, but seemingly it is not entirely so.
