I work at a different bank. The rates charged to banks for FDIC insurance have been based on the assumption that the FDIC would cover depositor losses up to the insured limit. By choosing to cover all losses even above the insured limit, we have chosen to put the burden for paying for those losses on all of the other banks (and indirectly on those banks depositors). I suspect this means that you will not see the interest rate on savings accounts go up as much as it might have otherwise.
I'm not saying this outcome is terrible, perhaps it was the best solution for the system as a whole. But using an insurance fund to cover a kind of loss, the insurance was not sized to address is not a choice that has no impact.
If I were in charge of everything (perish the thought!) I would probably have insisted that the uninsured portion of the deposits take some haircut. If depositors had gotten back 90% or 98% of their deposits instead of 100%, it might have increased the chance in the future that institutions with 100 million+ bank accounts would pay more attention to the risk profile of the banks they choose to invest with. Banks are rewarded mostly in proportion to the risks that they take; having a force other than government regulation that pushes in the opposite direction can be very useful.
I'd also add that covering uninsured depositors isn't new behavior for FDIC, at least as I understand it. The mechanics of how it was done here are different than in previous instances.
And if we are going to expect them to be guaranteed in the future insurance rates must go up, not just to cover more things but to cover the riskier behavior it creates.
If we aren’t going to cover them 100% in the future then apparently it’s true that there are not just different classes of banker but different classes of depositors.
What's the difference between the FDIC insuring all deposits at US banks directly and US banks doing it themselves by forming a complete graph? (Other than there being a clear upper limit in the latter case, which is currently greater than $1b per account.)
But financially I think it's the same. If 400 customers each use 1 bank each, then a single bank failure means the FDIC needs to make whole one customer.
But if every customer put 1/400th of their wealth into each of the 400 banks, then FDIC has to cover all customers for 1/400th each.
The cost to us as depositors/taxpayers is equal.
If customers are only utilizing a single bank, and the FDIC will insure all deposits regardless of amount, they would need 400 times as much than would be necessary if the balances were swept.
> Insurance only pays out . . . if a bank fails
That's a good point. So one difference is that while the money is equally insured in both cases, the payout dynamics would change. Very roughly, the amount of a payout might be expected to go down in the cross-bank case (smaller account values, but then also more accounts per bank, so it isn't quite so simple), and the likelihood of a payout might be expected to go up (higher chance of failure with more and smaller banks). But this all depends on how interlocked the banks become; in the extreme they could end up functionally a single bank.
The first thing that came to mind for me is somewhat related: Spreading deposits across banks is relatively better for small banks and worse for big ones, since the small banks gain deposits and the big banks lose them. So you can definitely argue there's some advantage to keeping a lower insurance limit, although it gets murkier when we bring behavioral considerations and "too big to fail" into the picture.
This is new with SVB. I get the whole "250K minimum" argumemt, but this is the first where we are seeing major 10M++ depositors getting 100% guarantees.
I don't see issue with spreading money across smaller banks - other than perhaps they may not be able to assess risk as well as larger banks. But again, SVB.
I think one thing to keep in mind is that most bank depositors hold far, far less than the 250K guaranteed by FDIC. By a large margin. Id be surprised if the average was above 10K. There are $17T in deposits across the country [1]. The FDIC at the beginning of the year had $128B in balance as insurance to depositors [2].
[1] https://fred.stlouisfed.org/series/DPSACBW027SBOG
[2] https://www.fdic.gov/analysis/quarterly-banking-profile/inde...
The question is in what way(s) does it matter whether deposits are insured without limit in a single account, which is new, or with limits when spread across an arbitrary number of accounts, which isn't? If one costs 1x, why should the other cost > 1x?
It's a serious question. FDIC assessment rates aren't as simple as tax brackets and for example it's possible that the act of spreading deposits across more accounts in more banks would increase the fees paid into the existing system anyway.
It matters because only 1 (SVB) bank failed. If those depositors had their money swept across multiple banks this point would be moot.
Edit: If deposits are spread about multiple banks, the FDIC does not need to carry as large of a balance to cover the loses of any single bank, which results in less indirect fees to depositors of different banks.
If the deposit is 250k across 100 banks, and only 1 bank fails, FDIC is only on the hook for at most the 250k.
Now multiply that by number of depositors. It's RAID for banks. You're gaining fault tolerance by decreasing the impact of any particular failure.
Just wanted to say this is such a great analogy. Thanks.
Seems like having everyone use at least 400 different banks achieves one of the core goals of the FDIC guarantee - making small/midsize banks viable and preventing everyone from piling into the big four megabanks.
The regulatory regime is one that makes depositors whole when doing otherwise seems likely to cause a major crisis in the banking system. Which we've had for a long time. It merely seems inconsistent because evidence of "likely to cause a major crisis" differs by current possible crisis.
As https://www.bitsaboutmoney.com/archive/banking-in-very-uncer... explains in painful detail, the reasons why they likely concluded that there is systemic risk. But long story short, rising interest rates caused the banking sector to have $620 billion in unrealized losses. Unsurprising since the interest rate rise was *INTENDED* to make people lose money, making money more valuable relative to goods and services, which reduces inflations.
But $620 billion is substantially more than the $130 billion in the FDIC insurance fund. It is substantially less than the $2 trillion in equity in the banking sector, but both losses and equity are unevenly spread. Therefore there are banks under water, and others that are fine. But nobody is sure which are which. And given cash outflows from worried people, we were about to find out the hard way. And once there is a bank panic, even fine banks become not fine.
Like Wile E. Coyote, running off this cliff works fine until you look down. But we've looked down. And now the whole sector needs saving. Thus these actions.
They will go back to normal behavior once the crisis is over.
About 50 cents on the dollar.
The typical depositor in post-2008 bank failures (all tiny, until SVB and Signature) got about 75 cents on the dollar.
As far as I know, regulatory requirements require/encourage holding good bonds but if the FDIC is going to start treating gov/muni bonds of any maturity length as good as cash, then there's less reason to hedge against interest rate risk.
1. which seemed pretty likely, I know people who over the weekend were planning to pull all of their money out of a regional bank and put it in one of the 4 super banks Monday morning)
Agency MBS average trading volume is $250B to $300B a day, it could absorb SVBs entire bond portfolio without breaking a sweat in a single day.
Of course the banks are only happy to have ANY justification to complain and not raise saving rates, and blame the Federal Government.
But there have been no indication yet that the FDIC will be drawing on the insurance fund to cover the depositors.
It's hypothetical. The FDIC is likely to recover a vast majority of the uninsured deposits through asset sales, people just want to be outraged.
At smaller banks, 60-75% of uninsured deposits are usually recovered by FDIC and those banks went out of business for bad balance sheets, not bank runs due to duration mismatch.
We've gotten so used to assuming every statement is spin on 'you're getting screwed' that people assume it's always the case.
(Note that I’m not trying to suggest that you were not harmed, I just want to make sure that I understand the source of the harm.)
1) Is this outcome better than (expected) panic on the banks and bank runs? There still seem to be runs going on e.g. First Republic. Are the markets “calm” now?
2) Because insurance fund is not designed for the task it is currently experiencing, how the gap will be plugged?
3) Could this cause depositors move money from smaller banks to larger ones, and then larger banks just lend this money back to small banks w/some nice profit?
The assumption that an operating company should understand fixed income pricing dynamics, have a POV on future FED interest rate moves, and dive into each potential banking partners' asset duration is an interesting one, especially when most of SVB's deposits were made during a period of perpetually falling interest rates, where duration mismatch helped banks.
Relying on customers to check their bank, besides being fairly impractical for all but the largest and most sophisticated customers, is just going to trigger bank runs as soon as customers get any whiff of a problem.
I don't see the problem with unlimited FDIC provided the regulations are congruent with that. If you want to go the private route for better interest rates and take on the regulatory burden yourself, you can always go with an uninsured bank. If you just want a simple place to stash a few hundred million, why make you jump through hoops?
What I don't understand is why start-ups were apparently keeping millions of dollars of capital in a bank account rather than government bonds or other instruments, given that the FDIC limit was well known.
No sensical person is concerned with the interest rate on savings, it is nearly zero and effectively negative.
I dont believe anyone should have a single account with 100 million dollars nor should banks allow that, but they do. Perhaps part of the problem is that the $250k coverage is a value that should adjust annually and coverage should be relative to each account as opposed to each account holder
This might have been true for the past ten years or more, it's not true now. You can easily find savings accounts now which offer over 3%, and it's going up as the fed continues to raise rates.
This still might not be the best investment when you consider the high inflation, but it's great for emergency fund type accounts that need low risk, high liquidity.
I cant believe this got downvotes...
But isn't it likely to cause more bank runs if depositors lost money? So in a real sense, many other banks were saved from going under, by assuring depositors that their money is safe, whichever bank they're at.
My understanding is that other banks have massive unrealized losses as well, due to the steep interest rate increases. So they're all kind of vulnerable.
How were you harmed specifically?
> you will not see the interest rate on savings accounts go up as much as it might have otherwise.
Bullshit. Interest rates for savings accounts are and have been an absolute joke. Are they going to become a more hilarious joke? Probably but seriously, who cares?
> I'm not saying this outcome is terrible, perhaps it was the best solution for the system as a whole.
Yes, it is the best solution, otherwise you'd have written a completely different statement if the bank run went viral and you would have been really harmed.
That's like asking how you were specifically harmed as a result of there being one more CO2-spewing pickup truck in the world. Bad things amortized over millions of people are still bad, even if the harm to any individual is too small to verbalize.
> "Insured cash sweep is a safe and convenient service that provides FDIC insurance on large balances while giving you access to your money, as well as the ability to earn interest. Choose between demand accounts, which offer unlimited withdrawals, and money market accounts that permit up to six withdrawals per month."
I've heard some claims that SVB was offering incentives to depositors who kept their funds in one lump account, is this true and if so what's the benefit to SVB from doing that?
What’s unforgivable is why these large depositors that didn’t tend this cash for short term operational needs didn’t have the funds in govt securities in an insured brokerage account at a trust company…
Where many venture backed companies got hung up though is that the terms of venture debt provided by SVB required the borrowers to keep use SVB as their bank…
Do you have some reliable source about it?
(NOT a "look at it logically" or "here's how my health insurance works, why would the FDIC be different", or "do your own research" or anything else that's some random internet comment - I'm looking for real meat about this claim).
Any losses to the Deposit Insurance Fund to support uninsured depositors will be recovered by a special assessment on banks, as required by law.
In other words, if the FDIC's current funds can't cover the bill, an extra fee will be levied on banks to make up for it.
The problem, as it were, isn't some lurking or hidden loss, it's arguably that these answers aren't currently available and we don't have the slightest idea in the public what will happen if profits result from the bailout venture. This shouldn't be a "heads-I-win-tails-you-lose" redux.
The only thing we can be certain of is shareholders are wiped out and general creditors will take a significant haircut.
I guess the FED could have bullied some banks into buying the securities at a higher rate, but the books themselves are transparent.
1. Illiquidity. For example, there might be few people anywhere in the world with the expertise to accurately value some weird loan to a startup. If those prospective buyers are busy or undercapitalized, then the market may become inefficient, with best bids well below FMV. If we wait for those expert buyers to research the startup's credit risk, to raise more money themselves, etc., then the bids will eventually come back up to FMV. This is the classic "It's a Wonderful Life" style bank run, and was an important dynamic in 2008.
2. Time value of money. Assuming positive interest rates, a dollar later is worth less than a dollar now. When interest rates increase, the NPV of a future cash flow decreases, so the FMV of the corresponding asset decreases. This isn't a market inefficiency; if you assume that money is lent at interest and no riskless arbitrage opportunities exist, then it's just how money works.
The SVB's problem was #2. There's a liquid and orderly market for their assets, trading at prices close to those predicted by a simple NPV calculation; that price is just lower than the SVB wished. Hold-to-maturity accounting allowed them to ignore that, but that didn't change the economic reality.
The FDIC's current funds were $128.2 billion on December 31 2022
SVD had ~$200 billion of deposits December 31 2022 - but obviously had significant withdrawals prior to being taken over. Likely < 100 billion remained.
Beyond the special assessment, they almost certainly need to raise a new assessment to cover $250k+ deposits. Full insurance can't be on a discretionary basis.
For the Treasuries, this is known but not public. For the MBS, a theoretical value is known but not public. (The federal government has to sell these securities. It doesn't make sense to announce the holdings so they can be front run.)
They state clear as day that they had 91 billion of securities, currently worth 76 billion in their hold to maturity portfolio.
In their avalable for sale (AFS) portfolio, they had 28 billion, currently worth 26.
They break it all down by asset type, MBS, treasuries, foreign debt, ect. They break it down by the duration, eg <1yr, 5-10, 10+ yrs
There is no mystery. I dont know why people keep saying this.
Check out page 124 onward if you are curious: https://d18rn0p25nwr6d.cloudfront.net/CIK-0000719739/f36fc4d...
Yes, categorically. No, not specifically. 5 and 10 years produce different answers, particularly with current convexity. (It is fine if you're trying to get broad-grained answers.)
Potential buyers over the weekend got a list of CUSIPs. The public does not get that until ex post facto.
We still have a very good understanding of the "delta" and the magnitude of the loss, if not with crystal clarity.
Nearly all of the unrealized loss was in securities more than >10 years. They had 15.1 billion in unrealized HTM losses. 14.7 billion of this had a maturity >30
9 billion is agency MBS
2 billion in agency collateralized mortgage obligations
2.3 billion in commercial mortgage backed securities, and
1.2 in municipal bonds and notes.
They actually have relatively few treasuries, and none in their HTM portfolio. Unrealized losses on treasuries are ~1 billion, And mostly treasuries from treasuries less than 12 months ( contrary to what most people report).
I really recommend actually looking at their filing. It is extremely detailed.
https://d18rn0p25nwr6d.cloudfront.net/CIK-0000719739/f36fc4d...
That doesn’t matter, because they’re federally backed. But it’s a crucial difference to appreciate less than one week after a run.
I'm not willing to do it to win a hn debate, but there's enough public information to estimate the size of their hole within a billion or two.
Claims that they were evasive or hiding things in their financial reportings are off base. Claims that it would be impossible to get an idea of their Financial losses are also off base.
Nobody (in this thread) claimed this? And I disagree with the “within a billion or two,” though do think within a few tens of billions is accurate. The problem isn’t the specificity of the filing per se but the delay; that isn’t the balance sheet anymore.
It is clear from the statement that we are almost entirely talking about 30 year MBS purchased in 2021.
3% 30yr MBS were about $105 throughout 2021. These were about $89.2 at the end of 2022 (10-k filing), and about $89.4 today.
You might say, "but they may have put it 30yr MBS!". The thing is, you can validate this by the ammount they lost from 2021 to 2022. It also doesnt make much of a difference if they bought slightly different rate MBS at 2.5% or 3.5%, as they track pretty closely. Lastly, MBS value havent changed much since the time of the 10-k filing.
To get "a few tens of billions" difference, they would have had to tripple their losses on MBS in a time when MBS value has been stable.
As long as the rates on the bond are > than the rates SVB is paying its depositors it will be fine.
Buy Banana futures while it's still cheap.
Since 1933.
It always has been.
How much extra are they taking from your bank to cover the delta between what the FDIC will recover and what they have to pay out?
There are all kinds of legal and practical reasons that this isn't really a fair comparison, but again, it's not really about the specific policies, it's about a sense of where the government's priorities are and its flexibility seeming to only bend in one direction.
I wouldn't be surprised to see e.g. Signature Bank shareholders challenge these actions in court.
The only party that made out like bandits is the SVB management that piled on the risk in the first place -- but investors are ultimately responsible for letting them do that and investors have been punished.
While there wasn’t counterparty risk with those assets, there was duration risk. And their mistake seems to have been not selling those the instant the Fed publicly committed to killing inflation with higher interest rates. It should have been clear to them that their exposure to duration risk was rising, and they needed to restructure back in 2021 or early 2022 to mitigate that.
The actions are borderline criminal. To avoid a $36M hit they literally bet the bank. This was a step beyond regular incompetent mismanagement.
From the article:
In late 2020, the firm’s asset-liability committee received an internal recommendation to buy shorter-term bonds as more deposits flowed in, according to documents viewed by Bloomberg. That shift would reduce the risk of sizable losses if interest rates quickly rose. But it would have a cost: an estimated $18 million reduction in earnings, with a $36 million hit going forward from there.
Executives balked. Instead, the company continued to plow cash into higher-yielding assets. That helped profit jump 52% to a record in 2021 and helped the firm’s valuation soar past $40 billion. But as rates soared in 2022, the firm racked up more than $16 billion of unrealized losses on its bond holdings.
Throughout last year, some employees pleaded to reposition the company’s balance sheet into shorter duration bonds. The asks were repeatedly rejected, according to a person familiar with the conversations. The firm did start to put on some hedges and sell assets late last year, but the moves proved too late.
I love HN because many of us actually listen to each other and debate in good faith, helping each other sharpen our views.
Glad you found it helpful.
I agree that it is important to argue I good faith and help each other sharpen our views. Thank you for providing an important resource in the discussion.
While I still feel the instinct for vengeance or rage at SVB is counterproductive given that you showed evidence that the leadership of SVB knowingly, willfully, and with reckless disregard for the consequences, decided to invest in more long term securities than they were advised by their own team does justify some of that instinct.
That being said I still feel the system worked as expected. I think, given human nature as on display with SVB, the regulatory environment should go back to the stronger standard but I also think getting bank failures to 0 isn’t a laudable goal. Some failures will happen and as long as we eliminate the motivation for a run on the bank the system should allow for a degree of risk and failure.
So part of the problem is that SVB had a reasonable-looking balance sheet of HTM bonds, then had to sell some at market, which flipped their entire portfolio to MTM and destroyed their balance sheet.
It seems like SVB was perhaps a little more exposed to interest rate risk than others, and had a pool of depositors that were more likely to withdraw significant funds in lockstep.
All bank risk assessment regimes measure both credit and interest rate risk. Measuring 1 in isolation is idiotic for now obvious reasons.
[0] https://www.bloomberg.com/news/articles/2023-03-13/svb-failu...
> that was one of the lowest risk assets they could invest in and still get enough
Was it? They had enormous deposit inflows and were struggling to scale, so their costs should have been undersized by default. They really ought to have been able to survive off the pennies that weren't in front of the steamroller.
Otherwise you are totally right that it should have been no problem to cover the shortfall by selling off some of the treasuries, even though they would have had to take a haircut on them thanks to the Fed jacking the rates so fast.
Earning a nice bonus last year is a reasonable consolation prize, but I'd wager most execs would rather have had a lower bonus and the ability to continue to manage an operational bank through 2023.
Not really. They were sold 1 month after the 10-Qs were filed, which is shorter than the holding period most reputable banks require for their executives, and the 1O-Qs had only that one sale in them.
They may have "made out like bandits" in taking advantage of equity holders, and perhaps without duty of care to depositors... but all that is true regardless of the subsequent actions. They did not "make out like bandits" because of the Government's actions. And I think that's important, given the criticism levied against the "bailout".
> Earning a nice bonus last year
Why do you think this was limited to last year? I suspect they made risky moves again and again and again and got paid out again and again and again.
Edit: after reading this article posted by lordfrito below I stand corrected. SVB executives knew the risk and took it anyway. But not for personal gain but to maximize firm value as it allowed higher profit which increased the valuation (so yes they benefited personally, but to a greater extent than just a few million in bonuses).
https://www.bloomberg.com/news/articles/2023-03-13/svb-failu...
Disclaimer: was one of them, though not nearly as aggressive as I should have been.
Does that have anything to do with this article at all? First few lines of this article: Banking is a confidence trick. Financial history is littered with runs, for the straightforward reason that no bank can survive if enough depositors want to be repaid at the same time. The trick, therefore, is to ensure that customers never have cause to whisk away their cash.
This article is about the possibility of the total loss of confidence in the banking industry leading to a run on a system that can't handle it. I understand the context you meant when you said things like "the system works, why is everyone upset", but I find those a pretty poor choice of words regardless with this much fear circulating.
It seemed self-evident to me, based on the explicitly stated limit on FDIC insurance, that if you had an amount of money over that limit, you really need to have a plan to deal with that risk, and people who failed to do so should suffer the consequences of their poor decisions. As things stand, the people who did spend the time and/or money to provision for that risk have suffered for it.
I think what many people are having a hard time with (myself as well, sort of...) is how the rules were changed out from under everyone in yet another example of how the rules don't apply to the politically connected.
They worked together to put out a joint press release, and Biden gave a down-to-earth, rough-and-tumble speech about protecting depositors and kicking the failed executives and bad-luck shareholders to the curb, because this is "how capitalism works". It was an unusually blunt attempt to preemptively push back at the perception that this guarantee of FDIC-uninsured deposits will be branded a 'bailout'. (I predict that this attempt will fail and this will widely be perceived as a 'bailout' in casual and political discourse, which is the exact forum at which they've aimed this message.)
After 2008, the public gained awareness of the consolidation -- both forced and emergent -- that occurs in response to these sorts of crises. Public opinion views these outcomes unfavorably, because they seem unfair and irreversible, albeit no palatable alternatives have emerged that are acceptable to both to the public and government and industry incumbents.
[1] https://apnews.com/article/silicon-valley-bank-failure-depos...
Buyers of banks in the last round got screwed. Dimon has been vocal about this. They inherit a string of liabilities that take a decade (or more) to resolve. Add to that, the number of private equity buyers for a $200bn bank is limited.
"Simply".
WaMu - acquired, depositors got 100 cents on the dollar
IndyMac - 50 cents on the dollar
Silver State - 11 cents on the dollar
Depositors have not always been made whole in the past. Calvinball has certainly been played in the past, for IndyMac the FDIC limit was retroactively raised from 100K to 250K.
That's what people are pissed about. The Calvinball rules.
And we know how that works out, if you're in the in group, you get paid, and if you're not in the in group, you get fucked.
As Black Flag once sang, "We're tired of being screwed. Revenge!"
And depositors to their dismay are learning they are about as much loved as Wall street bankers, corporate execs and billionaires. More than any particular moral deficiency I think people are finding a general lack of self-awareness common among SV startup founders infuriating.
Exactly this. I even read a comment from such a founder saying essentially: "Why are people so angry, don't they know I oppose brogrammer culture?" As if brogrammer culture were the meat of of the reason why people are sick of the hypocrisy of the capital class, or even American startup culture specifically. Totally out of touch, completely clueless. Utterly tone deaf.
Particularly, the decision to change the rules in the middle of the game and make depositors over the FDIC limit completely whole again is clearly an unfair favor to the rich. Normal people don't get to have the rules changed mid-game in their favor. If all the depositors were merely semi-wealthy commoners with only $300k in their accounts, they would have only gotten $250k back. Nobody would expect the rules to be changed in that scenario. But if you're much richer than that, then the rules are apparently just guidelines. It isn't fair and that's why people are mad. Anybody confused by people being mad is completely out of touch, and voicing that confusion is only going to make people even madder.
No, they would have gotten $250k + ~90% of the balance. Maybe even 100%. Annoying but I wouldn’t care enough to take to twitter about it.
To explain why: it’s an overall loss of a few percent. The same as the daily fluctuation if I had kept the money in the stock market or whatever.
That was amusing.
"S..See! A normal person!"
Presumably they were onto something, as they jointly received the Nobel prize for economics last year for this contribution.
Investing involves risk. Sometimes that means losing money, even if you are the bank.
The reason a $100 par bond paying 2% sells for (I don't know, say) $87 when interest rates are (I don't know, say) 5% isn't that the original bond is impaired. It's that the same $100 buys you a bond that pays 3% better, so nobody will buy the bond without a discount.
But the bank doesn't normally sell the bond to begin with. That's why people say banks "borrow short and lend long". What the bank is supposed to do is hold the bond until it matures and is paid back in full. The only reason SVB can't do that is that all its depositors simultaneously demanded their money bank, so it couldn't wait the bonds out. But other institutions can do that waiting.
It will be a loss for the bank, but that seems better than total collapse. Banks are ultimately companies that take calculated risk to make money, if you can't afford to take an occasional loss then you shouldn't be in a risk based business.
The thread here asks: "who's paying to cover SVB's uninsured depositors?". Isn't that the answer?
Maybe enough depositors will leave money in the SVB at below-market interest rates that it will earn its way out of the hole. The FDIC has given depositors a special incentive to, since by guaranteeing all funds they've made the SVB the safest bank in the USA. If the depositors don't, then the FDIC will take the loss, and socialize it over all participating banks.
Per my other comment, the HTM accounting is a distraction. That accounting was compliant, but accounting doesn't define reality. The holders of long-term bonds take a real economic loss when interest rates increase, regardless of whether they sell. This may seem unintuitive since the cash flows don't change, but it couldn't be otherwise--if the bond is worth par, then why aren't any buyers willing to pay that?
As long as the holder does not intend to sell the bond, believes that is more likely than not going to be a position where it isn't forced to sell (to generate working capital etc), and there is no likelihood of a credit loss, then the bond is not OTTI.
The subjective assessment of whether you're "more likely than not" going to be forced to sell the bond is the pivot on which this whole thing tilts. It's probably a good question whether a simple balance of probabilities is really where that standard ought to be.
If you're willing to lock the depositors up for 10 years and pay them back when those assets mature, sure. But that probably wouldn't be seen as an acceptable way of making those depositors whole.
Someone's got $100 of deposits with you today; you're holding a 10-year bond that will pay $102 over the 10 years but currently trades at $87. Yes you "can" "pay" "them" "back" eventually, but what if they want to pull their deposit today, perhaps to buy a bond like the one you were holding? If you do the accounting based on today, they're entitled to $100 and you only have $87; if you do the accounting based on 10 years' time, they're entitled to $115 and you only have $102; either way there's a shortfall.
A distinction between held to maturity assets and market priced assets makes sense though, there should be some consideration in the calculation about term though for sure. The question of that though seems very complex to answer.
Further: the "market value" thing here is complicated. The reason there is separate available-for-sale and held-to-maturity accounting for bank assets is that, in the ordinary course, the assets are held --- the only reason you sell them is because of extrinsic distress. There isn't anything wrong with the agency MBS portfolio SVB had; they're worth less because if you have to sell them in Q1'2023, they compete with even more attractive bonds and are discounted accordingly. But if you just hold them, they pay back dollar for dollar, and that's what the bank normally does anyways.
Granted, it's sometimes hard to establish market value. In the case of assets with genuinely low liquidity, the market value is somewhere above the bid and probably below the ask. Accounting in this case comes with error bars, but it's not really justifiable to approximate the value of an asset outside of this range.
Here's an example. Suppose a hypothetical bank receives 100 dollars in deposits at an interest rate of effectively 0. They use 85 dollars to buy a bond that matures in 1 year that pays 90 dollars, and keep 10 dollars in cash as reserve requirement. They pay themselves the extra 5 dollars as bonuses. They have met their fractional reserve requirements, and according to your scheme, their holdings should be valued at 100 dollars. Are they solvent?
All bonds get held to maturity by someone (unless they default, but that's not the problem here). The FMV of the bond is determined by the value of the bond's remaining cash flows to that person; so if the FMV went down, then that should be a clue that value was fundamentally lost, regardless of who's holding the bond.
Interest rates take a random walk from now until maturity.
Under fair-value accounting, the balance sheet value starts at fair-value (obviously), then gyrates, but tends towards face value, and reaches it at maturity, due to time decay of bond premium. As you said yourself, every bond eventually matures in the absence of credit risks and fair-value can't indefinitely diverge from face value.
Under amortized cost basis accounting, the balance sheet value starts at fair-value but then increases every year until maturity, at which point it is also face value.
Surely you acknowledge that these are the same? They both describe the exact same cash flows.
That's the loss that took down the SVB, and it's a real economic loss. If the SVB's depositors behaved like the simplest textbook model, then as soon as short-term Treasury rates increased, they'd insist on correspondingly increased interest on their deposits. The cost of that extra interest would be the SVB's loss--the cash flows in from the bonds stay the same, but the cash flows out to the depositors would need to increase.
In reality, I understand that depositors are generally "stickier" than that, leaving their funds at their existing banks even when bank interest rates increase slower than Treasury rates. That gave the SVB some hope that they could earn their way out of the hole, passing the loss slowly to their depositors over time by paying below-market interest rates. That's the behavior that HTM accounting roughly models. The SVB's depositors had no economically rational incentive to accept that though, and they didn't.
... but SVB expected to have positive net-interest-income with increasing rates. Look here:
https://www.sec.gov/Archives/edgar/data/719739/0000719739230...
A 100bp increase in interest rates would've increased net-interest-income by 1.8%; a 200bp increase by 3.5%. This is typical for banks, because lending rates go up more than deposit rates as the base rate increases, increasing interest margin.
So the idea that this is about NPV of future cash flows is not right.
Of course most of the money left in a self-sustaining bank run, not directly for the reason above. There's a quite rational reason for someone to start that run, though. That reason wouldn't exist if the bank were well-capitalized on a MTM or NPV basis.
The SVB's expectation in that 10-K is a model of depositor behavior, predicting that depositors will leave their funds at a MTM-insolvent bank earning below-market interest for long enough for the SVB to earn its way out of the hole. That prediction obviously didn't come true. So doesn't that just mean their model was wrong, and shouldn't be relied upon elsewhere either?
What I am confused about is - if everything went "according to plan", then what did happen? Is it really all peter thiel's fault? Surely someone as smart as him saw something that made him do what he did, given that it was a pretty massive thing to do.
Of course they don't! Look at some charts of newspaper advertising revenue over the past few decades. There's one word that best describes it: apocalyptic.
Where did all that advertising revenue go? Google and Facebook!
This is what I'm mad about. FDIC insures to $250k in normal cases. It should not have been used to insure depositors for their full deposit amounts here.
You should learn that you the moment you put a deposit in the bank, the funds become the property of the depository bank. As a depositor, you are a creditor of the bank.
People are mad because the rules were changed in the middle of the game to serve the interests of a select few (mainly VCs and the startup crowd).
Those supporting this bailout seem to have some of the least knowledge on how banks work.
Putting a ceiling on FDIC insurance is effectively an outdated idea that doesn't work.
Take the example of a company that keeps payroll in a cash account. Let's say that company has 100 employees. Should the FDIC treat the account as belonging to 1 person or 100? If you say 1, I say you are irrational.
Companies with treasury departments already know this. They can put money in money market funds, CDARs, cash sweeps, or any other vehicle to protect their cash. There are multiple ways to hold cash with very low duration risk that does not involve putting it in a bank.
FDIC is not an outdated idea. It is just the reality of the current financial system because it would require an excess of $20 trillion dollars to insure every deposit in the banks.
Why, I do not understand.
I'm fine with banks being not for profit institutions run by the government. Allowing people to safely store their money is baseline civilization. If banks are private, the government is going to have to back them up because you can't have the operational accounts of nearly every business in the country getting wiped out randomly.
> Equity is getting zeroed out. Management was fired. Depositors were made whole almost immediately. SVB's assets are apparently not impaired; SVB would have held them to maturity had the bank run not happened, and now somebody else will instead.
Part of the problem is that the system that enabled them to end up in this situation is the erosion of Dodd-Frank. The systemic risk to depositors isn't going away. If pissant SVB (relative to it's contemporaries) can lobby congress effectively imagine what other banks are up to. Speculation? Sure you can say I'm speculating. But the apple doesn't fall far from the tree.
> Meanwhile: the point of the FDIC system is for customers not to have to do this kind of risk assessment themselves.
The issue of course is that the total balances required the FDIC to dip into special capital reserves in order to make the bold faced lie the taxpayer won't front this.
Anyone who knows the surface level details of a bank know that these FDIC "loans" are effectively collateralized by the taxpayer. Banks pay an assessment. With what money? The depositor's money. A perfect example of a hidden tax.
> But SVB is gone, so it's not much fun calling them out. I feel like people are flailing looking for someone else to blame.
Credit Suisse is in big trouble and getting a bailout. Several other banks have collapsed in the wake of SVB. The only people not worried have their heads buried so deep in the sand only their feet are showing. Calling Chicken Little because you believe it was only SVB and not a massive market level problem suddenly beginning to show it's head is not a very effective argument.
I'd ask you to consider the economy that allowed these levels of capital to even exist. Years of ZIRP and near-ZIRP allowing effectively free money. As it stands, the mainstream media currently blames the fed for this and implores it to once again lower rates. The problem of course is that there has been no sign of stoppage in market speculation and we are only now starting to see VCs really tighten their belts. History doesn't repeat itself but it often rhymes and terrible, borderline predatory, VC funding practices begin to approximate NINJA loans in the limit. There's no reason to believe it's just SVB and there are plenty of reasons to believe we have very serious economic concerns ahead of us. Only difference this time is the criminals responsible will be wearing Patagonia.
That being said, I could be wrong and not aware of the specific Dodd-Frank policy that, if followed, would have made SVB safer.
The fed doesn't need to lower rates necessarily, it could simply allow all member banks to exchange low interest rate long term bonds for new higher yield bonds and pay the Fed for the spread with a loan. That would reduce the liquidity risk if the member bank needs to sell some or all of its bond portfolio on short notice to fund depositor withdrawals, it would allow the Fed to hold the low rate securities to maturity while being fairly compensated by member banks.
Edit: after reading this article posted by lordfrito below I stand corrected. SVB executives knew the risk and took it anyway. But not for personal gain but to maximize firm value as it allowed higher profit which increased the valuation (so yes they benefited personally, but to a greater extent than just a few million in bonuses).
https://www.bloomberg.com/news/articles/2023-03-13/svb-failu...
Maybe we should admit Congress will never repay the national debt and simply have the Fed purchase new federal debt issuance. The current complicated charade just pays banker bonuses.
Duration risk is specifically tested for in larger banks. The stress tests SVB successfully lobbied out of would have saved them.
If you haven’t lived through a couple of these things then it’s perfectly understandable.
Back in ‘98 there was a huge monetary problem going on in SE Asia but pets.com could take a loss on every sale and make it up in volume. Everything was fine until it suddenly wasn’t.
In ‘08 cracks were starting to become obvious but housing prices never go down, keep selling $500k houses to someone making minimum wage. Everything was fine until it suddenly wasn’t.
Today you have massive layoffs in the tech sector but the CEOs are just trying to appease activists investors, nothing to worry about because tech companies never fail. That Dot Com Bust? Well, that was Web 1.0 and we have it all figured out this time, nothing to worry about. Everything is fine…
My understanding is that SVB would have met the Tier 1 capital requirements even without the 2018 revisions to Dodd-Frank, for the reason digitaltrees said: The bonds it purchased are considered highly liquid and safe.
If that were the case, they wouldn't have had to raise emergency funding last week. The assets were indeed impaired. Hiding the true values via AFS accounting treatment doesn't magically make the dire circumstances sustainable.
It doesn't matter that prices might recover in the future. I'd argue they might not -- and if anyone believed otherwise they would buy up the assets at inflated prices (why arent they?!) SVN rolled the dice, made bets, the value is way down and...they didnt have enough money to allow customers to withdraw money. That is a fail. They needed to raise a lot of cash, they didnt/couldnt raise enough. That is a fail.
Further, they underwrote tons of LoCs for startups which are underwater due to down-rounds. That isnt a temporary impairment, that is a permanent impairment.
Their customers are burning funds (as most VC-funded companies do) and VC funding is down, so declining balances via continued withdrawal is the natural state they need to support (even in the absence of a bank run.)
This is absolutely false -- there is no guarantee of this. Agency securities pay more than treasuries because there is a risk of default (never in aggregate, but pass-thru component cashflows i.e. individual homes.) Further, there is a risk that upon default, the home isn't of sufficient value to account for the lost principal. If anyone has doubts of the potential for default of agency securities, the mortgages underneath these bonds are all publicly displayed monthly: https://singlefamily.fanniemae.com/applications-technology/f... and you can see defaults also.
>> The only impairment they have is that they pay less interest in the interim than other available bonds, because their rates were locked in before the interest rate spike.
This is a third of the story.
Second third of the story: they may less interest than advertised due to defaults/delinquencies
Third third of the story -- and most important: their value has gone down, so in the immediate term, the bank depositor cannot withdraw money (because the bond cant be sold at book value.) It is absolutely not OK to tell bank depositors to wait 10yrs while a bond bays them back little by little. Bank depositors should be able to withdraw money at any point they want.
This misinformation needs to stop.
Note: AFS treatment reports the BOOK VALUE not MARKET VALUE
You can read this: https://corporatefinanceinstitute.com/resources/accounting/a... to understand AFS vs MTM treatment.
Losses are reported, definitely not clearly. Losses are hidden under "Other Comprehensive Income." If you want to be transparent about it, then it should show up in the Profit and Loss. You'd choose MTM accounting treatment. Not choosing MTM accounting treatment is literally hiding the losses into a vague bucket of income.
The idea that this was somehow hidden needs to die. People knew they were vulnerable to a run not due to some shady rumor Network, but because they reported exactly what they were doing plain as day.
They were making low but positive interest profit on those assets. They simply locked up too much money in a long-term investment to withstand a coordinated Bank Run. This was a real mistake, but it is not at all about secret accounting.
Anyone who can do simple arithmetic could determine that they were at risk in a bank run from their published documents. Most of the world simply did not care because they did not think the Run would happen
This "somebody" is the government aka the central bank putting these bonds on their balance sheet. This is a new form of quantitative easing.
If the Fed did take them, it would be a drop in the ocean. The FED is already holding 2.7 trillion dollars of underwater mortgage-backed securities they bought.
They have six trillion dollars of other securities they are holding.
Nobody can do a bank run on the Fed and they control the interest rate, so despite their colossal unrealized losses, they can wait it out.
Months of work undone by loaning to banks which made risky investments betting against high interest rates.
Any uptick on this graph is equal to printing money which causes inflation. They can wait it out while we collectively pay the cost of this increased money supply.
Bet on every number but 0 on a roulette wheel
Not 0: you and your investors make 3 billion this year
0: you and your investors lose your 20 billion you have invested, and the government bails out your depositors who kept 200 billion with you
This stylized bet is a good deal for the investors and management and bad for the government. Sometimes investors lose everything but it's still a very good bet in expectation. This stylized example is a case of "privatized gains, socialized losses".
Then the question is: was SVB reckless? They could have been less reckless by covering their interest rate exposure, but the fed has an equity to deposits ratio requirement, and getting any equity to invest requires a return. IMO they should have either diversified their business or stopped opening new accounts for tech companies because when depositors are uninsured and concentrated in the same industry, that is risky.
So Management has to invest in something and it has to have some interest. I would love to hear an investment thesis that would have been able to deploy over $100 billion in new capital during the low interest rate 2018-2021 time period that wouldnt have been ill prepared when rates drastically increased in 2022-2023.
First SVB was bailed out by FDIC funds which all banks pay into.
Second, to say 'privatized gains, socialized losses', you are assuming that banking is like gambling, with no value being created through the banking process.
Even if banks were being very very safe, they would still make money by lending out deposits. (Whether that is good or bad for society, is another question, which I would argue the answer to would be bad).
Honestly, it looks like in a year or two, the Government will make money off of this because as soon as interest rates come down the securities will go back to book value.
The real winner here is Goldman, since they bought the bond portfolio from SVB that triggered all of this at a discount and can hold to maturity and interest rates may need to come down or a broader asset exchange program implemented to stop any contagion, so those bonds will return to book value sooner than expected.
There are literal products to insure money in excess of 250000. But people who are getting bailout now were not using those products. They were not paying for insurance in excess of that.
The system did not worked purely as designed. The system socialized loses of well connected rich people.
Government will get back the number of dollars equal to the par value of those bonds. Inflation between now and then, however, will mean that in real terms there will have been losses.
And from where this money is coming from you think? From you and me and everyone else because costs of doing business are transferred to clients, so to the whole society because almost everyone have a bank account.
Great point. To rephrase a bit, they lost money...and then kept doubling down by not cutting their losses (?hoping things would turn?) They finally tried to do something about it, but it was too late to matter.
Had a good chuckle at that part in particular. I think it's the plan that a lot of people are having a hard time with.
But the bailout that people are complaining about is for all the other banks that aren't SVB. There are many insolvent banks out there that would otherwise have had to raise capital at punishment prices this year. Those banks are unambiguously better off with the Fed taking their underwater collateral at par, and this is a clear subsidy to (non-SVB) bank shareholders.
Many of the takes further complicate by implying that they had no assets. Which just feels like lying at this point.
Sure, if everyone had just waited for the 10 year bonds to mature to access their funds their bank was in perfect shape.
—edit—
Assuming they could come up with enough money to pay the over market interest rates on deposits while also seeing their money flows reversing because of VC capital drying up.
This kind of assumes that the risk matrix of an executive is singularly indexed on the long term viability of their institution. But the short term gain of bad behavior is still in full effect. Bonuses for the years up to this crisis have already been paid and were probably inflated based on the banks over performance due to its riskier posture.
And the consequences have been softened. There's a very good chance that the people responsible here have had their guilt assuaged by the reduction in impact. They are probably less likely to become the kinds of pariah that they probably should because while we should always consider decisions in the context they are made, humans seem to always adjust their assessments to final consequences.
I'm in agreement that the decisions here on the part of the government are probably the wisest in this context. But this crisis does hint that perhaps we need to reconsider the structure of this system a bit.
1. Their reputation. How much less likely is it that a board of directors would think twice before hiring them to be a steward of shareholders' assets?
2. Their egos. How much less likely is it that people will be willing to invest time delivering projects whose value can be wiped out by poor risk management in the same way that SVBs has?
One member of the SVB c-suite was the CFO fr Lehman in the run up to that catastrophe. So, BOD don’t appear to care. They keep on hiring each other, making massive mistakes, but walking away with $$$$ in bonus money.
Most people are unaware they are loaning money to a bank when they open a bank account.
You've effectively said that bank deposits are now risk-free, meaning that the government is back-stopping 9.2 Trillion of deposits (40% of all deposits).
Can banks still provide a yield for these guaranteed deposits? Are they still able to loan out these deposits? What are the new capital requirements for these deposits, are depositors allow to take their money out when a bank run is happening?
We don't know
Then add in some vocal VCs' hypocritical stance on bailouts coupled with Surveillance Valley's overarching hypocritical stance on freedom, and here we are.
It seems that in this day and age of instant communication and social media mobs, even three days is too long for the exact fate of deposits to remain unknown. IMO the right way to proceed is to calmly raise bank capital requirements, create a few new tiers of FDIC coverage (eg coverage on accounts between $250k and $10M is funded from assessment on accounts between $250k and $10M), and institute criminal penalties for executives of banks that go bust beyond their capital buffer (otherwise nothing reigns in TBTF accounts that have too much variance to be absorbed by higher FDIC tiers).
The upset seems centered around the perception that rules were changed ex post facto to protect political donors.
I get that impression from HN, from newspapers of all ilks and biases as well.
69% of SVB employees' donations were to Democrats over the past three years. <https://unusualwhales.com/news/svb-donations>
The bank donated $74 million to Black Lives Matter. <https://nypost.com/2023/03/15/svb-donated-73m-to-black-lives...>
Hell, even with just layoffs the outrage on this site has been deafening. I can see shareholders and employees of SVB rightly being pissed off. It's not your place to tell them they should not be.
I mean... yes it is. I'm against zeroing depositors of failed banks, because (for better or worse) we've decided that banks should work like restaraunts and you shouldn't have to do a complicated risk assessment about how safe one is before deciding to do business there. But shareholders are a different story. If you invest in a company and they do dumb things and lose your money, that's at least somewhat on you - you're supposed to know what the company is doing before you invest, and potentially push for management changes if they're doing stupid things. Insulating shareholders from the bad decisions of their companies is an utterly unacceptable degree of moral hazard.
Does that resonate with you? Or do you still have no sympathy for all those people?
Re effective system: Maybe dinosaurs had to swallow rocks to digest their food, but we don't have to maintain this practice just so dinosaurs can keep existing. The system needs to be deprecated in favor of better tech that takes it out of the hands of dinosaurs. Legacy banking and gov/political class need to be replaced by a better solution.
Which is...?
They have all the data. If I was CEO of a bank I'd want to be able to get up in the morning and have some idea how much risk and what types of risk my bank was assuming. Especially in a dynamic environment of Fed interest rate changes. I would think they would be doing it all the time. Isn't that what computers do? Simulate scenarios like - What does our bank look like if the Fed raises rates to %2 etc. It makes me feel like they truly just don't want to know so they can do whatever they want.
They're super involved, requiring a full-time department to prepare for and run. That's a multi-million dollar recurring expense a bank with tens of millions of dollars of profit may not be able to afford.
But they're not.
The next banks all get to keep their equity because the fed is putting out a lifeline (the new BTFP facility) so that they won't be zero'd out the way SVB was.
The surprising part is, if we were willing to do this, why not just revise the Fed's discount window rules? They discount from market value. Why not change that to face value for Treasuries?
I mean, this comment almost wrote itself.
The issue isn't that too big to fail happened, because its immaterial after the fact. Its that little punishment was actually done for bad management, and not only that, the situation is left worse because we have a banking concentration problem.
If you look at the number of banks chartered after 2008, its dropped to negligible amounts going into the business, and regulation has only been increasing. You have to lie to get a charter because no reasonable person would accept the personal liability without something in it for them. The requirements are that onerous.
Everything is now so big it will certainly fail, and that's what people are angry about. There is plenty of evidence over the past 100 years (and longer if you go further), that as sector concentration goes up, so does corruption, frauds, and other crimes that are largely based around deception at our loss. It becomes easier to increase the scope, and get away with it when setting up dominoes to fall (so you can profit on event's you manufacturered).
The fed aren't doing there jobs, and worse, it looks like they could never meet their original charter to begin with. They aren't government, they are private bankers.
So they try to justify bailouts as a way of saving the system, and really its just acting as a wealth transfer to the elite rich whose pockets they are lining via a money printer at the expense of the public taxed by inflation.
Equity is getting zeroed out. Management was fired. Depositors were made whole almost immediately. SVB's assets are apparently not impaired; SVB would have held them to maturity had the bank run not happened, and now somebody else will instead. A bank made bad risk management decisions and got zeroed out; all the right incentives not to do that again are there. Meanwhile: the point of the FDIC system is for customers not to have to do this kind of risk assessment themselves.
It is remarkable how badly SVB managed to fuck this whole situation up. But SVB is gone, so it's not much fun calling them out. I feel like people are flailing looking for someone else to blame.