It's totally incredible that Circle, owned in big part by Coinbase, a HN unicorn which always tried to do things legally, ends up probably losing $1bn (about 30% of of $3.3 bn) due to good old regular banks doing what they do best: making customers deposits disappear.
Now Circle is sitting on $30 bn in short term US treasuries backing USDC and I don't think they're forced to give the yield back to people holding USDC, so Circle has to be minting something like, what, 1 billion yearly now that rates went up?
So they may be able to get away with eating a $1 bn loss.
Meanwhile the tether/Bitfinex fraud in the Bahamas is continuing to do just fine and its Circle/Coinbase/USDC that happen to be in trouble.
> What would you estimate is the total value of customer deposits that banks have loss in failure over the last say 30yrs?
Well of course if we consider only those where loss actually happened and not those where the losses have been mutualized by bail-outs, it may not be gigantic. But, I don't know, in Cyprus hair cuts made $8 bn of customers deposits vanish.
That's just one I know from the top of my head without needing to google.
Would you be willing to make a honest try to compare quantitatively the percentage of deposits lost in regulated western banks vs money lost in failed crypto exchanges, stablecoins etc?
I think the difference must be many orders of magnitude to the benefit of regulated banks.
Percentage wise I would bet crypto is way riskier.
In gross terms, crypto is tiny compared to the dollar/euro banking system, so it's likely to be a much smaller number.
Interestingly, for those who use crypto as-designed, the percentage lost to bank runs is almost nil since you 'hold your own coins'. It's people who leave their money on crypto exchanges that lose their coins.
> It's people who leave their money on crypto exchanges that lose their coins.
People who hold their own coins lose them as well. It was estimated that 3.7 million bitcoin have been already lost irreversibly [1]. That's a staggering number. It's roughly 20% of the entire supply.
They equate “lost” with “not moved in 5 years”. That’s a strangely short cut off — I would use 10 years instead. It’s completely possible for coins to sit around for 5 years in cold storage addresses without being touched.
It's funny because just this week I moved some BTC from a wallet I hadn't touched since 2013. And I used it to pay for something pretty mundane. In retrospective, the idea of someone whipping their "10 year old wallet" just to pay for something trivial is a bit amusing to me.
I think you are selective here. There are lots of coins that have lost value permanently (FTT, luna etc), I also remember reading that there are lots of coins whose network has simply shut down due to lack of interest (no source here, though)
there was also a very uncommon event in '01, '08, '22..
.. at some point, we're gonna have to step back and take a look at how we're doing business. Something about our whole way of organizing society is busted.
Aside from the crypto scandals, the insider dealing scandals, the money laundering scandals, the price and index fixing scandals, and the corruption scandals, there have been 562 bank failures since 2001.
Which is probably well under 0.1% of total companies that failed over the same period.
The thing is failed banks don’t mean depositors lose all their money. The point of banking regulation is fail fast so banks are taken out when they still have 98% of their depositors money not 8%. People with under 250k are made whole and people above that have already been compensated for their risk by higher interest rates.
It doesn’t always work perfectly, but in many cases a failed bank just means people come in on Monday and there is a new sign on the building with little else changed.
Appending scandal to an event isn’t having the persuasive effect you’re looking for. Finance is also the industry which helps us collectively coordinate our resources, grow our wealth for retirement, and facilitate risk taking. Every industry has issues.
I have personally been bailed out by a government when the bank I had a savings account in failed, and its countrys insurance scheme refused to pay. I have had fraudulent charges made to my bank accounts.
On the other hand, the keys to my ethereum remain safe on my hardware wallet, and none of my ethereum has ever moved without me telling it to. Perhaps my experience is atypical, but I am surely not alone.
Sure, and I didn't answer the question. I think it's hard to answer the question in a fair way. Banks and financial instututions fail in various ways from small to big more than most people think. On the other hand, they are 'too big to fail' so their mistakes get fixed by governments out of the public purse.
These options are not currently available to cryptocurrencies, but that's a social choice.
I tend to think that the scamming aspect of cryptocurrencies is massively overstated. There are lots of ways to be taken in and scammed, but that's true in the fiat world too. A small amount of care goes a long way in both worlds.
But you seriously believe that banks are generally better at completely losing deposits than keeping them? It’s just hard to see any evidence-based person believing that.
And how much money was lost in those failures by depositors?
And don't fall back on 'they were bailed out'. They were bailed out by other banks buying their assets + liabilities, and by the insurance payments they were making. They weren't getting a blank check from the treasury. That's the system working.
Who bailed out the crypto losses? Some of the losers got made whole, but many, many, many are realized losses.
> And don't fall back on 'they were bailed out'. They were bailed out by other banks buying their assets + liabilities, and by the insurance payments they were making.
Is that true? I'm not sure it is. The U.S. government gave hundreds of billions of dollars direct to banks and other institutions. The total actual cost has been estimated to be around $500 billion, despite claims that the loaned money was all paid back.
Let me correct that for you. "The total actual cost has been estimated to be around $500 billion, though to be frank the loaned money was all paid back."
That quote appears nowhere in the linked MIT Sloan article or the paper it summarizes, so please don't post such misleading quotes.
As I already mentioned and from the paper's abstract:
> Those conclusions stand in sharp contrast to popular accounts that claim there was no cost because the money was repaid, and with claims of costs in the multiple trillions of dollars.
So to be clear, the paper claims there was still a cost of $500 billion even though money was paid back.
FDIC can insure up to $100k (before 2008) and up to $250k (post-2008), so that's at most one person can get back, the rest puffs and it's gone.
What crypto losses you're talking about? The ones from centralized custody operators that hold your money in arbitrary way rendering your balance on web page in crypto currency or something else?
> FDIC can insure up to $100k (before 2008) and up to $250k (post-2008), so that's at most one person can get back
This is wrong for two reasons:
(1) The insurance limit applies per owner per ownership class; its possible to have accounts in more than one ownership class, and thus more than $250k insured.
(2) Recovery of the insured amount is guaranteed by the full faith and credit of the US government. Recovery of additional amounts is possible, and FDIC will make an effort to make it happen, but that's no guarantee.
No, it is the minimum , maximum, and only insurance the FDIC provides. Banks can and do get private insurance beyond that.
And stop acting like any uninsured money evaporates in a bank failure.
Actually probably just stop talking about banks. This whole thread is an embarrassment of “smart people with strong opinions about things they have zero knowledge about”.
We had foreign banks who advertised that as a strength and sent their profits to their countries. But when there was a crisis they said we are just a local bank with the same brand as a US bank and won't give the customers their savings back.
In practice since 2008 even if the theory is that it’s up to 250k in practice banks get acquired from the FDIC basically on the unspoken condition that all deposits are maintained at 100 cents on the dollar
Its not “unspoken”, its the overt goal to find a bank to take over that will provide that.
But SVB is a much bigger bank with a much larger share of uninsured deposits than a typical bank failure. If the problem extends beyond liquidity and really is an insufficiency of assets by any significant share, its going to be hard for a no-loss takeover to be facilitated.
If I recall correctly, Cyprus “doesn’t count” because those depositors were making something like 9% interest, and therefore they were morally closer to investors. I don’t have a citation at hand...
There was also the Icelandic bank collapse, which was multiple banks totalling tens of billions of euros. I personally lost a large chunk of savings in that, and was eventually bailed out by my own government (i.e. my own tax money).
Coinbase isn't in trouble because of this. Coinbase is in trouble because they're trying to run a legitimate cyrpto exchange, their chances of success are the same as an abstinent whorehouse.
>Now Circle is sitting on $30 bn in short term US treasuries backing USDC and I don't think they're forced to give the yield back to people holding USDC, so Circle has to be minting something like, what, 1 billion yearly now that rates went up?
What makes you think they aren't susceptible to the same type of run? They now own treasuries with higher yield and lower prices.
They've been convicted of fraud. They had hundreds of millions seized from bank account of criminals which they helped launder money for. They cannot operate in NY anymore and the NYAG fined them. I think Canada banned these fraudsters too?
They also sold Moonstone bank to SBF/FTX which illegally siphoned money out of the US to the Bahamas.
Nothing smells of fraud as much as tether/Bitfinex. Not even Madoff.
I don't know. Centre is publishing the individual short term US treasuries numbers of every single short term US treasury they own. They've got $9.8 bn at various US banks, minus potentially the loss on $3.3bn they may have to take now but even a $1bn loss and that'd still make USDC backed 0.975:1 or something. Not taking into account the insane yeld they're making on the $30 bn of short term US treasuries they have (for their clients, but I think the yields are for them).
That compared to a bunch of shady people in the Bahamas, deeply tied to the SBF/FTX fraud...
Wouldn’t they have been holding large amounts of treasuries from before rates went up? If so then they would be sitting on large mark to market losses, just like SVB, no?
I was confused. Circle and Coinbase created Centre and Centre's website say Centre created the USDC stablecoin. Circle and Coinbase each own 50% of Centre. I don't know if Coinbase also owns a part of Circle or not.
Wow, converting USDC to USD (fiat) on Coinbase is paused for the weekend, "while banks are closed"
USDC is currently trading at .996 on CoinMarketCap (although some exchanges are reporting .98, .97, or even .95 -- although that's mostly smaller exchanges for now)
Either way, people are taking a discount to get out of it and switching to USDT (now at a $72b market cap) or Bitcoin (hence the price rise)
And Binance paused auto-conversion of USDC to BUSD
So nominal price should be $.9175 on the dollar presumably.
Add in another couple bips for the uncertainty.
Edit: It would be actually hilarious if USDC failed because the banking system. I doubt it will happen, you'd have to see further contagion from the other banks Circle uses for their dollar reserves. I figure the tipping point for a bank run is quite high?
I don’t think it’s remotely plausible that uninsured SVB deposited are just gone, $0 on the dollar.
Absolute worst case, they might have lost 10%, or $330m, so .99175.
It’s not like SVB was a Madoff scheme and the money is gone; it became insolvent because the current value of its investment vehicles couldn’t support 100% withdrawals.
The person who said SVB's situation is "idiosyncratic" is an employee of Wells Fargo - of course he would say that. Generally, banks run on confidence on the system - if any of the bank regulators or any of the major banks even flinch a little bit in front of the media right now, 10 more banks would probably fail tomorrow.
If there's a general panic in the market and absolutely everyone demands cold, hard USD - no percentage will save them.
Let's say you're the person who's speaking for Wells Fargo at that moment - you absolutely don't want to give the public any reason to test whether 28% is safe or not. 3 months later, when people are no longer panicking, and cooler heads want to start discussing what're the safety measures to put onto banks - maybe. But seriously not right now. Let's say you quote the 28% number to say your bank is maybe safer than SVB - but the moment the reporter asks whether 28% is safe enough or not, or how about another bank that have more or less, you'd probably realize you shouldn't have said too much already.
So, at this moment... if you're a spokesperson for a bank or a bank regulator - "We're safe. SVB is a special case" is pretty much the only appropriate response right now.
Because held-to-maturity treasuries don't have to be marked down unless you start selling them, and because if you hold them to maturity, they will be worth exactly what they say on the tin.
SVB's problem is that it can't hold them to maturity, it has to sell them today at firesale prices, because they already sold everything they have that's not marked held-to-maturity.
Yes, it could, but it’s not especially meaningful to say that the same thing that happened to X could happen to Y if Y were tbe same as X.
Seriously, read Matt Levine’s stuff about SVB. It was a different bank with different customers, different Fed regulations, different depositors, and a unique asset portfolio problem.
35 year old dies BASE jumping. 34 year old BASE jumper says, “yeah he was taking on too much risk — that’s way too old to be BASE jumping.”
56% of assets in long term fixed rate loans is fatal. 26% is also fatal. The latter is not in a position to criticize the former’s excessive risk taking. A 5.6% hole in your balance sheet (20% loss on 28%) is still very bad for an institution that needs to satisfy demand deposits.
The different customers is probably the most important difference really, because it'd be harder to cause a run vs the VCs telling their portfolio to get money out en masse on the same day.
But different investment strategies present different risks. For a bank with a high correlation between deposits and low interest rates, having more investments with strong alpha to low interest rates presents more risk.
So, 28% is less risky than 56%, and Wells Fargo’s depositors have very different profiles than SVB, and WF is subject to the Fed’s liquidity rules that SVB was not.
So, while there is no such thing as risk-free anything, they are such totally different animals that the only reason to act like WF is lying / making PR noises not based in reality is the underwhelming observation that they both have “bank” in their name.
It’s a very low quality, knee-jerk, low-effort comment. That’s all.
(I really dislike WF so if the new HN ethos is to make wild unfounded claims just because we don’t like a company, I guess I could get on the bandwagon.)
I would say your comment is the low quality one, when the thread just sketched out[1] why WF’s “safe” level ought to be similarly fatal, and you’re just restating the obvious with a lecture about how 28% is safer than 56%, which was never in dispute. And for the kicker, you threw in an aside, which I hope is a joke, that we should say false things just because we don’t like the target.
[1] Yes, my comment wasn’t explicit, but if you put in some effort and read the context or a sibling comment, you’ll see the justification.
The percentages don't matter. Banks set aside a certain amount of capital for the only purpose of absorbing losses. A bank that is well-run will have allocated the right amount of capital in proportion to the risks that it has taken.
The same could have been said of SVB. And, surprise, they didn’t set aside a loss buffer equal to 5.6% of deposits either, just to cover that part of the portfolio, because that’s not feasible.
The job of any bank including SVB is to manage financial risk. If they can't do that, it means they can't do their job, and they deserve to go out of business.
What do you think your comments are adding to the discussion here? You’re going in circles and I don’t see the coherent point. The original criticism was that WF criticized SVB while still holding a sufficiently dangerous fraction of their portfolio in the same assets.
If you’re not speaking to that criticism but just giving vague generalities about how banks need to manage risk right, then you don’t need to make a comment at all.
Edit: Hit my comment rate limit so...
Sorry, I don't see the answer -- again, you were just speaking in vague generalities about how "banks need to manage risk right". Doesn't get to the argument you need, which is "this why WF can suffer a 5.6% loss on its loan portfolio but still satisfy demand deposits, and why it's a difference in kind, not degree, from an 11.2% loss happening to SVB".
The point was never that the percentage doesn't ever matter at all, but that both WF and SVB hold way too much in fixed rate long-term loans.
So if we're really going to go there, I think the substantive points went over your own head. When you're ready to say something more informative than "it's a bank's job to manage risk", I'll be ready to learn from it and appreciate the insight you're bringing.
I was answering your question: "The percentages [of dangerous assets that they hold] don't matter." But apparently the answer went over your head. Never mind.
> Do you think it's possible that someone working for Wells Fargo might actually know something about banking?
Replace Wells Fargo with SI or SVB and you have your answer (which is, it isn't how much they know about the economy, but what they're really cooking with it).
> One obvious question is: If you are “another, healthy bank” working through this weekend to buy SVB and assume its deposits, how much would you pay for the assets, which were worth $212 billion in December? 8 I am pretty sure the answer is higher than $8 billion, the amount of insured deposits: The FDIC will not be on the hook for the insured deposits. The $15 billion of FHLB advances are also quite senior and will presumably be no problem to pay back.
> I would also guess — not investing or banking advice! — that the answer will also turn out to be higher than $188 billion, which is the total amount of deposits plus FHLB advances. I say this not because I have done a detailed analysis of SVB’s assets but because it seems bad for the FDIC to wind up a big high-profile bank in a way that causes significant losses for depositors, including uninsured depositors. There was a run on SVB in part because there hasn’t been a big bank run in a while, and people — venture capitalists, startups — were naturally worried that they might lose their deposits if their bank failed. Then the bank failed.
> If it turns out to be true that they lose their deposits, there could be more bank runs
Part of me wonders what exactly were the toxic assets that pulled SVB under, and my gut feeling can't stop thinking it is that SVB is more exposed to startups than your average bank.. and that the next dominos, so to speak, will be startups seeking lending from banks.
Mass layoffs, finance pulling out, feels very dot com bubble bursty to me.
What do you mean by "toxic assets"? My understanding is that the assets were long-term treasuries issued in 2021. They'd declined in value, as expected, because of interest rate increases that occurred afterwards. SVB is now unable to sell those bonds without losing money, and so would have to wait until maturity to get their money back.
Yeah, this feels very dotcom 2.0 ish. Capital has been drying up a lot over the last year or so and this is the type of thermonuclear event that flattens industries. I'm glad I have a good amount of savings.
USDC has already depegged. Game over for USDC. We won't know what happens in the real world until next week, but this weekend is going to be interesting for crypto.
Bets on which crypto exchange is going to shutter next?
While also being ~160% collateralized. I’d link more specific figures but the makerdao site, like most crypto projects, doesn’t make those figures easy to find.
Disclaimer: somewhat motivated reasoning since I’ve left a standing order for DAI at 0.995 on an exchange, which was just filled this evening, in the expectation that DAI always returns to parity.
Assuming USDC is backed by the money it says it is, then it going to zero is incredible news for anyone who buys it post de-peg right? If the assets are there, and you buy USDC at 5c on the dollar, you’ll make 95c even if USDC has to entirely liquidate?
I guess, assuming the assets are there, I don’t see why/how it would drop below that, or be finished, other than some volatility due to panic
stETH also depegged at some point, and I made thousands of dollars by buying it at a deep discount. Then the peg was restored even though there was nothing backing it apart from locked ETH in staking pools. The USDC de-peg is pretty similar.
Why wasn't Circle using deposit networks? Did they take out private uninsured deposit insurance?
Circle's competitor Paxos is using IntraFi Network and Reich & Tang Deposit Solutions for BUSD according to their reports (https://paxos.com/wp-content/uploads/2023/03/BUSD-Monthly-St...) and has obtained insurance for most of their balances in excess of FDIC insurance limits.
The good news is that USDC has 74% of their assets in short dated US treasuries (weighted average duration is 36 days). Even if they had to sell all of their treasuries TODAY, they'd only be looking at a ~0.4% haircut (assuming an efficient market for treasuries, which should be close to true).
Definitely not a good situation, but they're in a lot better position than SVB.
I've also been under the (possibly incorrect) impression that the majority of Circle's dealings are with exchanges like Coinbase — parties who are probably willing and able to play some inside baseball in terms of dealing with duration mismatch and waiting out the feds.
> I've also been under the (possibly incorrect) impression that the majority of Circle's dealings are with exchanges like Coinbase
Circle belongs to Coinbase AFAIK.
EDIT: ah I was wrong, after reading the comment below I googled and Coinbase+Circle created the "Centre" venture and they're 50/50 in it. And Centre is the company which launched the USDC stablecoin. So Coinbase doesn't own Circle but owns 50% of Centre.
Also, while USDC is like a deposit, legally it isn't. So they are under no obligation to produce USD in exchange on demand. They can (and have) shut redemptions. So they could just wait 36 days for those treasuries to mature instead of taking even that small 0.4% haircut.
You have to have liquid dollars in inventory to manage USD redemptions. Banks have to keep 10% of deposits on reserve at the Fed, so $3.3 billion in cash out of $40 billion is actually on the low side.
I think the market is stupid on this one. I would buy USDC on a discount if I had the liquidity on hand. I have been wrong before, but this looks like an overreaction.
Not necessarily. As more businesses start exchanging USDC for USD with Circle the ratio between the reserves and USDC will get worse and worse. If a bank run happens 3.3 billion will become a big percentage of the total market cap and USDC will go to $0. The value of USDC right now is a combination of how much of that $3.3 billion will be recovered, how likely people think a bank run on USDC is, and predicting what actions Circle will take to try and save this.
If you borrow money from a bank, you have to cover your loan with something. If you fail to give the money back, they liquidate those assets and recover their money and interests.
If you are borrowing the bank through deposits, why isn't the bank forced to cover the loan so you can recover your money if the bank fails to pay you?
If the bank pays interest they make that money by loaning money to someone. They don’t pay you to warehouse your money in the bank. With this is the assumption that they can’t immediately give you your money back because they don’t have it.
Checking accounts are different so the presumption is that money is just sitting in the bank doing nothing because they can’t predict when it will be spent.
Of course this isn’t how banks really work because reasons.
If circle had decided to stop using svb usd reserves and put it all back into a stablecoin even if it depegged > 1.00 wouldn't that have been the winning move?
They obviously have more than 1 banking institution, so its not like they had no options
What stopped them from exiting SVB via their own currency ?
Are you asking why circle didn't back USDC with USDC?
It's back with dollars in bank accounts to fund redemptions, sadly $3B of which are in SVB. The rest (~30B) is in short term US treasuries (36 days average duration)
Why would they not use their own USDC reserves to buy more USDC ?
I suppose I dont know how the choice of what amount to keep in reserve is made, and who is deciding on bank institution(s) . Because, even if the price of the USDC temporarily go over 1, it would be fine considering the situation ?
Now Circle is sitting on $30 bn in short term US treasuries backing USDC and I don't think they're forced to give the yield back to people holding USDC, so Circle has to be minting something like, what, 1 billion yearly now that rates went up?
So they may be able to get away with eating a $1 bn loss.
Meanwhile the tether/Bitfinex fraud in the Bahamas is continuing to do just fine and its Circle/Coinbase/USDC that happen to be in trouble.
Because of an actual bank unrelated to crypto.