US central bank payment system down for 'hours'
bbc.com
bbc.com
EDIT: For clarity, I was referring to an actual online wallet, a wallet on a website, but it’s not too different to ask the same question about a software wallet on a computer-connected device that does automatic OS or wallet software updates.
Secondly, you can use computer hardware from before even the invention of Bitcoin, and then code the software on that hardware, developing the libraries yourself.
Third, you can use computers simple enough that it’s not feasible to hide any management engine trickery in it.
Fourth, yes, you CAN actually build your own computer, which I’m trying to do. :)
But to further your point, I still haven’t proven to myself the argument that the elliptic curve used by Bitcoin hasn’t been intentionally compromised like the NIST curves were. I take the word of the Internet for granted there... But I suppose I could prove that myself.
You can track the bitcoin reserves on exchanges here: https://cryptoquant.com/overview/full/247?window=day
Well, with different issues. FDIC insurance is probably better than whatever Coinbase has. You are also open to technical f-ups on Coinbase's end, which can move the price down if they are bad enough.
On the other hand, Coinbase cannot print BTC. This characteristic is what draws a lot of people to Bitcoin, beyond just "nobody can take my keys".
Agree it's a false equivalency, disagree that holding BTC in Coinbase defeats the purpose.
The value of cryptocurrency is inversely proportional to how functional the regular digital banking system is.
The actual user experience of SEPA is 2-5 business days just like America's oh so antiquated ACH system.
There are some very specific edge cases where SEPA instant exist, and that edge case may work for a large number of people that live in that walled garden.
That garden seems to be same country to same country transfers, where both banks have SEPA instant.
But to make it seem like the Eurozone or SEPA participant network has this solved has been very disengenious.
I frequently sent money to foreign countries and never experienced taking it longer than one day.
Oh, and it's free too.
This happens. Both are (most of the time) reversible in standard banking.
Fortunately the sum was trivial, but nevertheless ...
And the clearing feerate, which correlates to mempool queue depth, incoming transaction velocity, and cohort feerates, would increase dramatically.
But if there was a run to convert BTC to USD, the exchanges would not be able to cover the USD required. They'd limit or suspend withdrawals immediately. At some point someone would have to start buying with fresh USD for the exchanges to be able to pay out the sellers. With big enough news, there could be tens of thousands of dollars gap between bid and ask, which would take a long time to ease.
https://www.npr.org/templates/story/story.php?storyId=102384...
> But on the other hand, check these FDIC folks out. They know what they're doing. And every week they get more experience. In the 10 weeks since the FDIC took over the Bank of Clark County, 18 more banks have failed. That brings us to a grand total of 20 since the start of this year — a number that will likely grow tomorrow.
edit: The reason wire fraud is so popular is basically, because once the money leaves the account, the recipient has to agree to return it. You can imagine fraudster likely will not.
For the average person, I wonder if this means that if your mortgage servicer or your say student loan servicer does the same thing and inadvertantly pays off your loans that they then can't come back at you.
FDIC does not protect you from sending money to the wrong person. It only protects you from a bank going insolvent because it lent out more money than it has. That can't happen in Bitcoin, because your wallet doesn't lend your money out to other people.
Bitcoin wallets don't lend your money out to other people because they aren't corporations trying to make a profit. They are just wallets, doing a service for you.
We've come to rely on using banks when we want to send money to other people because USD alone doesn't know how to send itself to other people online. So banks have stepped in to fill that role. Now we can't even pay for things without a for-profit bank existing in the middle of the transaction. And that bank tries to make money by lending out your money to other people, and that gives it a risk of going insolvent, and so FDIC was created so that the government takes the risk of banks going insolvent.
If you send money to the wrong person, a bank may or may not be able to get your money back. It depends on the bank, and how well they know the other person, and what method they used to send the money, and how strong the threat of lawsuits is between the two parties. None of this has anything to do with FDIC, which becomes totally irrelevant when you use Bitcoin.
I'm not sure how to even understand this comparison.
FDIC protection ensures that if your bank fails, your money isn't lost; you can move it to another bank. This vulnerability doesn't exist with Bitcoin, so there's no need to protect against it (though obviously it has other risks that are mitigated by the presence of a bank in the first place, like loss of access due to poor secret handling).
If you use Bitcoin, then you remove the bank, and you don't need FDIC.
FDIC is there to protect bank customers from a bank going insolvent. Banks go insolvent if they lend too much money out, and then all customers demand their money back at once.
Your Bitcoin wallet doesn't lend your money out. It can't go insolvent. You don't need FDIC for it.
Here's a clear example of why lacking this is Bad For Bitcoin: https://www.google.com/search?q=hacked+site%3Ahttp%3A%2F%2Fr...
Bitcoin is more like having a pile of cash. The FDIC doesn't protect you against a pile of cash in your house getting stolen. Keeping a large amount of cash has advantages and disadvantages and risks. If you aren't willing to understand and mitigate those risks, then don't do it.
Banks don't go insolvent because they lend "too much" money out. They go insolvent because the collateral on those loans is less valuable than they estimated. E.g. the assets are lower than the liabilities. That is the only possible way that any business can go insolvent.
So when you take a loan on a house, the house is the asset the money lent is the liability. Housing prices drop enough, people walk away from the loan, the bank goes insolvent. That is true if it has a big balance sheet or a small balance sheet as these numbers just scale out.
> and then all customers demand their money back at once.
That is applicable to the era portrayed in the Mary Poppins movie, but not in a modern banking system. Even in the early twentieth century "bank runs" was something people would talk about when what was really going on was asset deterioration and breakdowns in the capital market.
But Hollywood preferred the more intuitive story and also the story in which the survival of the bank was in the hands of common people's choices rather than in the hands of the capital markets, where it truly lives. Thus you get It's a wonderful life where a stirring speech to not withdraw money can actually effect the P&L margin of a bank rather than the less exciting story of whether the bank's cost of funding exceeds its cost of borrowing.
Today banks borrow at a low rate and lend at a higher rate.
This includes borrowing whatever cash they need to meet outflows. In fact some banks don't even accept deposits at all, they just borrow from capital markets and don't even deal with depositors. Most big banks are depository institutions and tap both depositors and short term funding markets.
All that matters is that the interest received from inflows is less than the interest paid on borrowing to satisfy outflows. It is all about making money on that spread. When banks can no longer make money on the spread, they go insolvent even if no one makes a withdrawal. If banks are making money on the spread, then withdrawals are not a concern to the bank.
Not quite. Yes, bank runs nowadays are very rare, but they can still happen, as exemplified by Northern Rock in 2007[1]
"On 14 September 2007, the bank sought and received a liquidity support facility from the Bank of England, to replace funds it was unable to raise from the money market. Reporting of this complex scenario led to panic among individual depositors, who feared that their savings might not be available should Northern Rock go into receivership. The result was a bank run – the UK's first in 150 years – where depositors lined up outside the bank to withdraw all of their savings as quickly as possible, particularly since many others were doing the same."
[1] https://en.wikipedia.org/wiki/Northern_Rock#2007_crisis_and_...
Source: Am Korean who has lived in the US and UK
I'm not personally aware of the exact details of Fedwire SLAs, but I understand it to be near instantaneous, and that page doesn't suggest otherwise.
But more fundamentally, currency and money IS a technology. It's use is to distribute power and express trust. But fundamentally it is a technology used for that purpose.
Cryptocurrency is ultimately about putting power in the hands of the people and trust in mathematics rather than central authorities.
Eventually the central banks and banking institutions as they exist now will be phased out and replaced with different paradigms that have some similar functions but more sophisticated technology.
edit: Almost makes you wonder if it was a targeted attack.
coincidence?
Costs at most about $35 for any size wire, shows up in a corresponding bank within a day. If the corresponding bank is not the ultimate destination, takes whatever the amount of time it takes for the end transaction to complete. In event of a bank mess up on a transfer, the wire gets rejected/credited back. For banks that do "online" FedWire rather than batched, wires show up and are credited within seconds.
What, exactly, is the problem?
P.S. Non-immediate transfer is a feature and not a bug.
edit:
"P.S. Non-immediate transfer is a feature and not a bug. "
That may have been true/acceptable before - back when time span between order and payment was greater. If it is a feature, it is a feature for the bank, not the user.
AML, OFAC will absolutely take time and no interbank transfer would bypass it in any US chartered bank as they are triggered during a match in the FedWire or any other system that uses the Fed.
KYC and fraud investigations are triggered by the originating bank, most likely because the account has been flagged. Until the account is cleared any transfers out of the account would trigger them and delay them.
KYC-type and fraud investigations could also be triggered by the receiving bank if the account is either flagged or the activity is matching certain parameters - it is less likely but possible. Should that happen, no matter what is the method that was used to credit the account is likely to trigger it.
> For those 20% the banking system as it is right now, is a problem.
That makes it 1 out of 5 wires. It is extremely unlikely to be the case.
> That may have been true/acceptable before, where time span between order and payment was greater. If it is a feature, it is a feature for the bank, not the user.
It absolutely is a feature to the user - it allows for an easy rollback of journaling mistakes.
>> That makes it 1 out of 5 wires. It is extremely unlikely to be the case.
I don't think I could honestly qualify 1 out of 5 with adverb extremely. Even unlikely feels like a stretch. Let me give you an example, if you submitted your taxes online and before you submitted it, it gave you a chance percentage of an audit, would you roll the dice at 20%?
>> It absolutely is a feature to the user - it allows for an easy rollback of journaling mistakes.
Having dealt with those mistakes, I still maintain that it is there cover the bank.
There's no such thing as 20% of international wires that get delayed. American brokers do hundreds of thousands of international wire transfers every day. If 20% of those were delayed, it would collapse the global financial system.
"P.S. Non-immediate transfer is a feature and not a bug. "
Since we are playing with words, non-immediate transfer is a delay and it is at odds at the other statement you have made:
"If 20% of those were delayed, it would collapse the global financial system. "
It wouldn't, because it didn't. But clearly that part doesn't matter, because entities and individuals affected tend not to be well-heeled and so unlikely to be able to force the bank to take some sort of action.
All that said, I am done discussing this.
The delay in the international wires that are a feature ( 1-3 business days ) have nothing in common with the 20% arbitrary delay of unknown duration that you have promoted. That's the 20% delay that would collapse the world financial system.
That is what I want.
And if you really tried to use credit/debit card for those purposes, you quickly understand their limitations ( amounts come to mind ).
This is very much so like the entire cargo/freight "Hyper-Loop" thing, or Rocketship cargo/freight, or even the Supersonic Aircraft thing... a solution in search of a problem.
There are very few scenarios were someone legitimately (and more importantly, frequently) requires sending a package to the other side of the world within minutes. Would it be nice? Perhaps... but just as Concord found out, after the novelty wears off, few actually want to drop $10K on a one way ticket.
With huge amounts of money being transferred, the delays are in fact a feature. It gives you time to cancel if there was a mistake... and it gives banks opportunity to sniff out fraud/laundering, that the money is in-fact real and not pending other transactions, etc.
I've just never heard of a scenario where someone needed $50k instantly sent anywhere for anything. When dealing with that amount of money, in a non-reversible/permanent system... being careful, deliberate and slow is pretty much what you want.
Even Bitcoin transfers are not instant...
maybe not $50k, but "deposit paycheck on same day rent is due" is a fairly common problem that people have. the person has the money to pay rent, but it might not clear before their rent check bounces. there still exist landlords that only accept payment via check.
I agree in general. wires are fast enough for most truly large transfers of money, if only because the recipient is usually willing to be more flexible with such a large amount on the line.
If you're bouncing rent checks because it took too long to clear, you have other issues and immediate withdrawal isn't going to save you anyway.
Besides, if you're cutting checks of the same amount every month, you can setup BillPay (free at most/all? banks) and let your bank handle writing/mailing the check on time.
I'm talking about people who barely make enough money to cover their monthly expenses, so yeah that financial situation is hardly ideal. someone in this situation definitely does not want to set up autopay against their checking account.
it's a sort of death by a thousand cuts for people in this situation. you generally have to pay rent in advance (ie, you pay that month's rent on the first of the month). you get "paid" at the end of a two week period, but then it takes about a week for your employer to process payroll. if your employer doesn't do direct deposit, you now have a check in hand that takes about two days to clear. that two days isn't the bulk of the problem, but it can definitely be the straw that breaks the camel's back.
Two things:
1. For paper checks, this has been addressed via Check21 which took effect at the end of 2004. It enforces funds availability for even for non-local checks and/or bank drafts.
2. Writing a check against an account that does not have available funds has all the characteristics of check kiting. Check kiting is illegal.
For non-checks ACH credits settle before ACH debits or any other debit transactions.
Or, if I write a check, and they deposit it on their bank's phone app, nearly always the funds are available immediately.
$15,000 Euros is a lot of money... I honestly wonder what problem was solved by mandating that amount of money must be transferred, cleared, and available for withdrawal within 2 minutes. It seems to me nobody has a legitimate reason for that, unless you're a Nigerian Prince.
What problem did it solve and why is it considered a good thing?
And I've certainly sent somewhat smaller amounts for stuff like "Hey trip is on, we booked that holiday villa, €1000 for single rooms, €800 each if you're sharing, let me know if you can't pay this week".
I'm sure this rule would have been fine if it was €10 000 rather than €15 000 but how often do you want politicians to re-authorise? If it was €1000 by now people would be annoyed at the size limit already.
I don't have one on my keyboard unfortunately, otherwise I'd happily use it where appropriate!
> And I've certainly sent somewhat smaller amounts for stuff like "Hey trip is on, we booked that holiday villa, €1000 for single rooms, €800 each if you're sharing, let me know if you can't pay this week".
Your scenario doesn't seem to have been impacted by 2 minute mandatory transfer times in either direction.
> I'm sure this rule would have been fine if it was €10 000 rather than €15 000 but how often do you want politicians to re-authorise? If it was €1000 by now people would be annoyed at the size limit already.
Sure it's an arbitrary amount and they had to pick some amount I suppose. Even $10K is a lot for an ordinary person - potentially life changing if lost in some unfortunate instant transfer with the wrong number entered. That, of course, is a human problem (typing wrong number or whatever), but the delays we're talking about provide a safety cushion to cancel transactions in those scenarios.
If I try to give money to Victoria Smith 12-34-56 0123456789 then the backend will notice that's not Victoria Smith's account, and tell me I need to figure out what I did wrong. Maybe the problem is Victoria married, and is now Victoria Jones, you can clear that up. Maybe I got the number wrong. Maybe she's Victoria Smyth or Smithe and then it'll suggest this was just a typo using some heuristic. Or maybe it's the account of a scammer named Michael Black and the scam fails because the name doesn't match.
This was a bit awkward for my ISP because their name has an ampersand in it, and so of course this means some banks end up believing their name is "Andrews & Arnold" and similar stupidity. But it's been bedded down now a bit.
This approach leverages a reference for the payee that existed but was previously not checked, plus KYC checks already in place, it's hard these days to open accounts for names that aren't yours or with no apparent credit history for an organisation.
(b) cheques have been deprecated and stopped being accepted in the EU in 2004.
It solves the same issue as paypal, cash, venmo, etc, except the receiver can immediately use the money for any and all purposes, without fees, without yet another third party app.
This in turn allows new third party apps to be built on top of the new infrastructure.
I get a few transfers in the EU per month and sometimes I am lucky if it is same day (if done in the morning) but in most cases I get it the next day.
Heck, in same country, I have two bank accounts and transfer between them have never been faster than 4-6 hours. (Same bank is seconds though)
If your bank doesn't support it yet, the old 24h limit applies.
The more I do that - the more I want to go off into a corner and plant shit in the ground and stay the hell away from people. They are super dangerous. And honestly, I can shape the world exactly the way I want by doing it that way.
In all other efforts where people are involved, trying that for 42 years and it hasn't worked well yet.
There have actually been a few times where non of my bank cards (chase, bofa, credit union) work for cash withdrawal and I've used a Bitcoin ATM to get cash until I could get something figured out.
That's simply not the case as someone who authorizes or validates wire transfers for the companies would be able to confirm.
> There have actually been a few times where non of my bank cards (chase, bofa, credit union) work for cash withdrawal and I've used a Bitcoin ATM to get cash until I could get something figured out.
There have been several orders of magnitude more cases where standard bank cards worked in the ATMs and bitcoin ATMs did not.
I’ve never had a Bitcoin ATM fail, but I’ve had normal ATMs fail all the time even in first world countries like Japan, HK, Singapore.
Well there's your problem. Where is that $35 heading? For a $35 transaction fee, I'd expect my money to arrive instantly and to work reliably. Sounds like that's pretty far from the truth.
I'm sure there are a hundred reasons why it's not that simple and it needs to cost a lot of money, but fundamentally, you're decrementing a number in one bank and incrementing a number in another. I can see why people would want that to be cheap, fast, and reliable given that it's a pretty simple thing to accomplish on the internet.
The bank charges the service fee. It ranges from $0.00 to about $35.00
> I can see why people would want that to be cheap, fast, and reliable given that it's a pretty simple thing to accomplish on the internet.
International wire transfer is, at most, a settlement service for payments, not a payment service.
Tone deaf in what sense? What tone should she strike? BTC is comically inefficient.
Crypto will have to grow all these other encumberances before it can become something ordinary people can use instead of a normal bank.
Also, Bitcoin could drop a transaction if a fork occurs and you were tracking the one that fizzled out; that is why people wait for roughly 5 confirmation blocks before proceeding and that takes roughly 1 hour.
At currently network congestion, a high priority tx (next block) is 102 sat/vB ($6.95). A low priority tx is: 53 sat/vB ($3.61)
Fees depends on network congestion, check mempool.space for current estimated fees and survey of recent block fees.
https://bitinfocharts.com/comparison/bitcoin%20cash-transact...
GP's question was ambiguous about which fork he was referring to.
Also what the hell is up with cheques and using your signature for card payments? I can only remember these from my early childhood. It’s so... backward. Why hasn’t it been improved? What’s been blocking progress for all these years?
Killer flaw of overly binary thinking.
You can get free settlement on a check under certain thresholds in about a day if you deposit it via an ATM at your bank. You endorse it, stick in the machine, it images the check, does some OCR, and turns it into an electronic transfer. Banks often let you take a picture of it with your phone via an app. It's pretty common now.
So that's a lot of infrastructure to support something that seems pretty backwards, but anecdotally works well enough. Ink signatures are just really "sticky" in American commerce, I guess it gives regulators warm fuzzies.
As far CC, it's gone largely chip-pin, but signature is a backup if comms goes down.
Keep in mind that in the states there's a prevalence of credit vs. debit, and it's pretty lenient to dispute charges as a credit card user, merchants get the short end of the stick there. Cloned CC being used to make phony charges was never something that you were liable for. A bit of a pain to deal with, but not a financial risk. So there's little to drive adoption except merchants and the CCs who want to reduce fraud to increase overall profits.
Edit: in order for it to get "so slow its stopped" more than half of the miners would need to go offline and not rejoin for hours.
Bitcoin costs 50-100x that: $10-15-20 for "small number hours" settlement.
A Bitcoin transaction with a fee of $0.20, might settle after several days, but would more likely expire from the mempool before confirmation. And the really fun part is the unpredictability!
ACH isn't perfect. But Bitcoin is completely inappropriate for the common ACH use case.
Except when you realize the mempool has a complete record of every pending tx, unlike the central bank whose system is completely down and cannot and will not have a record of any of these transactions and it's services cannot be reached.
A part of me wants to be elated, but another part realizes that the end of US empire may allow the CCP to continue unabated, and that has incredibly horrible consequences given all the sabre rattling and their focus on Taiwan after illegally annexing Hong Kong and enslaving Xianjing.
The West will continue, and the lack of the US hegemony may allow it to prosper in other ways, but this will be an incredibly rough patch as it loses its status on sole Super Power, I don't think the CCP's China qualifies as a super power as it cannot unilaterally impose the world to operate on its own standard in anything but trade reliance, something it can and should lose. As creditor it cannot really do much except in a close sphere of emerging African countries, of which Bitcoin is actually flourishing, or the Russia/Iran block, and Iran is a hotbed of Bitcoin influence with state sponsored miners.
I hope the US takes the lead once more without the need for overt militaristic, police-state tactics as the petro-dollar loses its gravitas, I just don't want to be here when it happens.
Also, how does borrowing and interest work with something like Bitcoin or other distributed ledgers?
You sign some legally binding agreement, they send you the bitcoin, you pay it back in the agreed intervals + interest over time. If you don't, the legal system gets involved.
I'm wondering if there is a smarter way to issue loans through something like smart contracts as another comment mentioned. Where we perhaps remove the need for large single institutions to hold large amounts of capital for the purpose of loans. Otherwise we would then be entering in a very similar structure of what we have today and the inherent drawbacks.
We could imagine mediating that transaction through blockchain something something, but degree of risk pooling - centralization - is the essential feature of the transaction