Visa confirms Coinbase wasn’t at fault for overcharging users
techcrunch.com
techcrunch.com
It was probably just stupidity.
More on common razor heuristics:
Stupidity should be assumed. Malice demands evidence.
But it comes down to intent. Impute intent only when indicated. Which when you think about it is only fair. It’s self-evident Visa fucked up, but a claim it was malicious must be substantiated or it’s just trolling.
It’s a philosophy which many people find beneficial, to try to shape your thinking in a way that you set a higher bar before believing something negative was done intentionally rather than unintentionally.
The inverse (converse?) is also be true — when lacking conclusive evidence either way, assuming positive things are done intentionally and not unintentionally.
Another way to think about it is; what’s the cost of being wrong? It’s as simple as giving the benefit of the doubt.
I am fine with that. I am not going to assume "good intentions" by default, but case by case as I see fit. Even if my biases get in the way (like with any untrained human). If there is a non-zero probability of somebody intentionally causing something, i.e. it cannot be counter-factually refuted, then it has to be handled as one possible explanation, however unlikely it is.
Haven't you ever seen how people cook the books for businesses to appear flawless? Ohh, that small mistake? Just an error, don't look too much into that! Let's have a drink and here is your ticket to a strip club to entertain you! Have fun!
ergo, we must accuse all others of whichever crimes are not literally logical impossibilities.
And then I saw this comment: "Thats weird. I have a mastercard and was charged twice. So how is it Visas fault."
– Alfred North Whitehead
https://en.wikipedia.org/wiki/Occam%27s_razor:
"In science, Occam's razor is used as a heuristic guide in the development of theoretical models, rather than as a rigorous arbiter between candidate models."
Looking at Universal Induction:
https://en.wikipedia.org/wiki/Solomonoff%27s_theory_of_induc...
Isn't Universal Induction non-computable? What logics does it work in and which ones are beyond its reach? Is it applied purely to countable set of all programs or could it extend to real-world as well? One would expect it holds in precisely structured logics like first-order predicate logic, so that's not a surprise.
"Assume positive intent" is not going to help you when your money is gone.
This one maybe wasn’t there fault but they certainly have a checkered past.
The centralized system you refer to here is the status quo (banks, fiat currencies, etc.)
For all practical purposes, you can't bootstrap a new decentralized monetary system without depending on an existing one -- value needs to be transferrable between the two entities until most of the world has migrated.
This isn't a failure of decentralized systems -- it's just the cost of any large-scale transformation with huge powerful incumbents.
This is equivocation bordering on intentional dishonesty. GP is saying that without engaging the incumbents, the decentralized system would have absorbed his money and given him nothing in return. That is in no possible way a win for decentralization.
Crypto proponents assert that the world is better without centralized currency/money systems. Yet when push comes to shove, people have been forced to engage with the centralized systems in order to force the people behind the decentralized systems to do the right thing.
I don't see how that follows, given that the failure was at VISA, a centralized system, in its interaction with Coinbase, another centralized system. The decentralized cryptocurrencies continue to work as designed.
Well then they aren't very well designed, are they... I mean, if they don't interface with the real world in a workable way, who's at fault?
These problems occur because you still have to deal with "legacy" currencies. If everything would be conducted in cryptocurrencies, you wouldn't have any of those.
Although I'm not convinced that this is a likely future. I'd rather expect banks to offer cryptocurrency accounts with conversions. Ironically, another centralized solution but at least a regulated one.
Yes. And then I expect that someone figures out that it is not actually necessary to transfer my crypto to my coffee vendor for cup of coffee, but as we both trust the bank, we just tell the bank to move some crypto from my account to the coffee vendor's account.
Then, I expect that someone figures out that it would be cheaper (well, actually bank would be willing to pay to borrow my crypto) to keep my crypto in bank if I let the bank to lend that forward for interest.
Finally, I expect that someone figures out that the raison d'etre of crypto, fixed monetary supply just broke down, as now banks can "create" "cryptocurrency" at will.
When someone tries to cash out, exchanges might need for someone to put some money back in to avoid a liquidity crisis.
I believe this is especially true for tether-based exchanges, which Coinbase is not.
Then again, this is just my guess and I can't provide anything to back it up. Maybe it's just due to a buggy cowboy-driven codebase and awful customer service. Time will tell. If it's a liquidity issue, exchanges will eventually go belly up, if it's a codebase/customer service problem, they might get to a point where it's actually fixed and things just work.
Except whole companies sell at a premium to the sum of their parts [1].
[1] https://www.investopedia.com/terms/a/acquisitionpremium.asp
Basically, all ACH transfers work via a FTP system. Your bank uploads a transactions file at the end of the day. That transaction file is then sent to a server that then connects to a central bank.
The next day, that central bank does an FTP transfer to the other bank's system, balancing out the transfer.
This process can take 3-5 business days.
But, once the money has been "transferred", it can still be rolled back for up to 90 days after the transaction has happened. This makes it really difficult to say when you've "received" the money for real.
It gets extra scary when the money on one side can't be rolled back at all (ie the cryptocurrency).
I'm sure at the scale Coinbase is working at, they're running into terrifying numbers, and want to stay as covered as possible. The customers don't really understand the time scales of traditional banking, and here we are.
The interface between the two makes for some "interesting" edge cases.
https://engineering.gusto.com/how-ach-works-a-developer-pers...
Yes there are some dispute/reversal windows that I have never had experience with, but if you are just transferring money between your own accounts this should not be a serious issue.
I believe in this case your bank 'trusts' you. With banks that I have an existing relationship with I can move money overnight. When I open an account at a new bank it takes days for the money to be available for use. Presumably the new bank does not trust me yet, and thus makes me wait for all of the confirmations to occur.
You are mistaking ACH for wire transfer (or are not aware of the good faith credit your bank provides). There’s three main ways to send money between accounts:
- wire transfer with immediate delivery
- wire transfer with next day delivery (likely what you are mistakenly thinking is ACH)
- ACH which takes the previously mentioned several days (for making sure there isn’t a return for reasons other than stolen account). ACH is essentially “writing a check, sans paper”.
The first two are essentially direct bank-bank transfers. The latter goes through a clearing house as described and has a clumsy (in modern terms) process with lots of unneeded delays.
Banks will many times credit the money to your account much faster, but they do that for two reasons.
- they have an existing relationship with you and know your risk profile
- they sometimes also send an electronic request to the other bank confirming fund availability. Not all banks support this, but when they do, generally banks will overlook some extra risk from the previous bullet (as in, you don’t have available funds to claw back).
Happy to provide much more technical details if you insist, but Google could trivially do that for you.
> Anyone claiming there are large inherent delays in that setup is full of baloney.
No need to be derogatory towards others stating facts to you.
But yes please provide those "technical details" whenever you have a chance.
https://www.frbservices.org/assets/financial-services/ach/09...
So your bank is being optimistic and giving you a credit, assuming that no return will happen. A pessimistic or cautious bank would really have to wait several business days to ensure no returns happen before crediting your account.
[1]: https://engineering.gusto.com/how-ach-works-a-developer-pers...
Ask anyone on the street when they can get the money from their paycheck direct deposit. Unless they just opened their account, they are not waiting days.
That’s because payroll isn’t sent the day of your payday. Go ask the person who handles payroll at the company you work for, they don’t submit it the day of payday (or even usually the day before). It’s normally submitted several days in advance.
Not to get in a pissing match, but you are the one confusing unrelated things. There’s a reason I’m getting upvoted and you are being downvoted.
1) for credits, the RDFI doesn’t place holds on the funds (except in very unusual circumstances). So your apology here is flawed.
2) read the link provided by @ad_hominem, it very clearly and eloquently explains the time lines involved, by a company doing payroll. Better yet, read the whole series of posts they wrote. (Part 1/2 are linked at the top of the article).
In general, try to understand that your personal experiences might not reflect the “way things work”. There are folks on HN who have generally “been there, done that, have battle scars to prove it” (like me in this specific case) for many topics and you’d be wise to keep an open mind rather than keep fighting against them when they are trying to help you understand.
And using ACH delays as an excuse for why people can't get money out of Bitcoin quickly-- the original context of this thread-- is completely disingenuous.
I must have missed where you said that. You said I was likely mistaking ACH for wire transfer because mine were happening that fast.
The point stands that "inherent delays" in the ACH system are negligible, on the order of hours.
I reacted strongly to the person claiming otherwise because it was presented as an excuse for people not having access to their money quickly. If people are being told they can't get their money because of "ACH delays", "check is in the mail", "clogged tubes", or whatever, they should not be afraid to call BS!
Which has been repeatedly called out as wrong with proof in this thread, but you won’t listen. Have an enjoyable life!
Bank can revert ACH in case of a book keeping error (there was no ACH).
It can revert ACH in case of a fraud (someone at the bank committed fraud). ACH was not originated by the customer.
It can revert ACH in if it gets a court order.
It cannot revert ACH because Sally did "oops".
Source: SVP of compliance of a very large bank sitting next to me.
> It cannot revert ACH because Sally did "oops".
Quite a few of the defined ACH return codes[0] seem to be loosely in the “oops” you said wasn’t possible.
a) first response b) bank-to-bank journaling based on the bank to bank agreement.
But dont look at me - just try doing it and see how successful you would be.
However I still think it's accurate to not think of the funds as fully clearing for several days given the nature of the protocol, and that regulation simply forces the banks to be "optimistic" in giving a next-business-day credit.
Per Title 12 → Chapter II → Subchapter A → Part 229 → Subpart B → §229.10b(1), in general a bank shall make funds received for deposit in an account by an electronic payment available for withdrawal not later than the business day after the banking day on which the bank received the electronic payment. [0]
So reading that there are several observations.
- first, they say in general, emphasis theirs, and don’t define exceptions explicitly.
- next business day, so that means they are perfectly allowed to be two days delay for a payment initiated by Wednesday night, or up to five days for one initiated Thursday night (Thursday send, Friday received but held, Monday being next business day). If Monday is a bank holiday, then it could even be six days!
- in (2)(i) they further go on to state Payment in actually and finally collected funds. There are a number of reasons that could be delayed.
Ultimately, saying an ACH should always be by next morning is just flat wrong and no amount of experience with your own personal transactions negates that.
https://www.ecfr.gov/cgi-bin/text-idx?SID=b82e479da1df29eaa4...
I work with merchants and most are overnight as long as it hits the appropriate timeframe, some do experience 3-5 days delay, typically 3 though.
The WSJ linked is paywalled but the quotes are
“…these large transfers move in steps. Banks have slowed down the process further to reduce the chance of fraud… [the bank] sends transactions in batches during the day to an automated clearinghouse, which sorts them and moves them to the receiving bank in a matter of two to four hours… In many cases, the receiving bank gets the transfer the same day."
So delays are inserted on purpose. Certain banks don't introduce these delays, or have minimum delays, you'll know it when you're using one. USAA and Fidelity are two I've used that don't delay ACH transfers, those have always cleared ACH transfers next day. I even get my paycheck a day early when deposited into one of those accounts.
The other option is your bank keeps a tab of "transfers that are trusted."
This is because they are crediting you the money before they actually have it confirmed. See my reply below to GP, they are doing this because they are more risk tolerant and/or you are a less risky customer.
These delays seem to be caused by the slow US (or international wire?) system.
Often it feels like overreacting is the only way to extract a response from SV companies. Reminds me of the upwork story a few days back.
Otherwise they were just silly people venting on internet forums. That certainly seems like a fairly standard form of overreaction.
Don’t get me wrong, I’m not defending this behaviour. I’m just more interested in if there’s something that SV firms are doing that’s exacerbating it.