Let me ask a dumb question: If money exchange is the main problem that Bitcoin solves, then why isn't the problem of money exchange attacked directly instead, i.e. through some other less volatile classical stores of value like gold or stock? Why is money exchange a problem anyway and why couldn't a classical wire transfer solve the problem?
The article doesn't say anything about these things and the underlying principles other than that it's "suprisingly difficult". Instead it goes right into proposing a solution based on Bitcoin and making big picture comparisons with network protocols that probably sound very smart to the right audience.
Given the amount of agreement and hype that the article has received so far, I feel a bit like an outsider right now questioning what is proposed. I think I just don't get it. Is everybody just a lot smarter than me or does the community miss to address some basic questions here?
The answer to basically all of your questions is that to transfer money, you need to move money. This can be basically a promise of money, as with wire transfers, or physical, as you'd do by transporting bullion or cash.
The problem with using a promise to move money is that you have to trust whoever's promising. That works alright if there's a central authority, like a bank, but less well if you don't want to trust the authority, don't have access to the authority, or are unable to comply with the rules of the authority. Also it's kind of a shitty experience, as is. (You wouldn't pay for all of your Amazon purchases with wire transfers, I'd imagine.)
The problem with moving physical money is that it's difficult to do over long distances or en masse.
In order to solve the problems with existing money transfer, then, bitcoin needs to be able to do without a central authority and without using a physical representation of money. Which it accomplishes by allowing you to store money - not just a representation or promise thereof - digitally.
Thanks for your answer. However I still don't fully understand the problem I'm afraid. How is the trust required to move money cross borders different to the trust required to have an account with money at your bank? That acct is just a "promise" of money as well isn't it?
Also, according to your response a domestic money transfer would suffer the same problems but these seem to work just fine (at least much better than international transfers).
Why does a mechanism that works just fine in between bank and customers and generally domestically just break down when country borders are involved?
> Also, according to your response a domestic money transfer would suffer the same problems but these seem to work just fine (at least much better than international transfers).
Certainly they work better, but "just fine" is a much stronger claim and one I'd disagree with. I don't find myself spending money by wire transfer hardly ever, eg, and when I do it is a much larger affair then using cash. See also andrewla's comment here [1].
> Why does a mechanism that works just fine in between bank and customers and generally domestically just break down when country borders are involved?
Because banking systems and the regulations surrounding them are different in different countries, basically, and that adds lots of complications to an already extremely complex system. The bitcoin protocol is not different in different countries.
This is absurd. I spend money by wire transfer all the time. A good 80% or so of my ebay shopping would be wore transfers to merchants in Hong Kong via Paypal. This is a big enough issue in Australia that we've had retailer associations complain about it on occasion. Overseas money transfer is not at all difficult on a consumer level.
> This is absurd. I spend money by wire transfer all the time. A good 80% or so of my ebay shopping would be wore transfers to merchants in Hong Kong via Paypal.
It's worth noting that EFTs (Electronic Funds Transfers) between domestic banks in Australia is nearly always free, and that in your case you're not transferring money overseas at all.
You are transferring money to PayPal Australia, who is then communicating to PayPal Hong Kong that all is well - so that PayPal Hong Kong can pay the merchant. Both entities can transfer to/from local bank accounts because they have explicitly set up B2B interfaces (and met regulations) that allow them to do so.
In this case, PayPal is the bank, and they are a bank seeking to specifically facilitate money transfers between countries. You could not, for example, transfer money to a merchant in (e.g.) Siberia - they wouldn't be able to get that money out of PayPal into their regular bank account, because PayPal has no local presence/connection to the banks there.
The article explicitly notes that PayPal is offering a very similar service. They argue that the open nature of Bitcoin is a competitive advantage to the network in the long run.
From the article: "This would not, of course, be the first global payments network. One obvious comparison is with PayPal. The fundamental advantage a Bitcoin gateway ecosystem has over PayPal is that it’s open".
I can't reply the comment below so I will reply to this one.
XorNot: Please correct me if I am wrong.
USD is not a relevant comparison. What matters here is how money gets moved. When money is transferred within the country, a central bank (Fed Reserve, Bank of England, Reserve Bank of Australia) keeps an account for each local bank. Which means that when 'Alice' of Wells Fargo sends money to 'Bob' at Chase, the central bank actually debits and credits the banks. And then several times a day the banks will settle with each other and net out the differences.
But this is not possible internationally, that means banks must each other have direct bank relationship and have an account with each other. And if 2 banks don't have direct relationship, then they have to go through intermediary banks which they do have a relationship.
Enters Bitcoin, what bitcoin provides is not a reserve currency, what it provides in this context is a global ledger. Instead of the Commonwealth bank of Australia (CBA) needing to have a direct account relationship with Wells Fargo (WF) bank. CBA can just simply send bitcoin to WF.
Internationally banks simply have to be able to acquire sufficient foreign currency holdings. Which is trivially easy because every competent reserve bank on the planet has enormous foreign currency holdings.
But this is all irrelevant to Bitcoin - which is marketed by its advocates a consumer currency, not an international system of exchange between institutions. Institutions have no need nor desire for such a thing - it is a saturated marketplace, with literally thousands of avenues of exchange, of which Bitcoin is a particularly poor one.
Which circles back to my original point: foreign currency transactions are very simple for anyone ranging from consumers to medium or large size businesses, barring tax issues (like not paying a lot of it). Bitcoin does not solve a problem not already solved for centuries by the banking system - this is literally the thing that got it started way back with the Knights Templar.
I have a feeling that international remittance is not as simple acquiring FOREX reserves. Often times you see international transfers from an Australian bank to a US bank taking 4 hops in between over peering banks. The process is complex.
Because unlike domestic transfer where the central bank helps keep a ledger between banks, there is no "global ledger" internationally. So banks have to resort to "correspondant banking" which is why you see so many hops in international transfers.
Bitcoin provides such global ledger. I don't think FOREX is the issue here.
Instead of going through correspondent banking (many hops), 2 banks can simply send bitcoin to each other and immediately net off.
Could you clarify if my understanding is incorrect?
EDIT:
Here is a document on how VISA handles international payments
2.3.3 Clearing and settlement procedures
...
Settlement is not carried out through Base II; Visa merely provides the data to allow settlement to be
carried out. For settlement in US dollars, Chase Manhattan Bank, New York, acts as the settlement
bank. For multicurrency settlement, Chase Manhattan Bank, London, acts as the settlement bank. All
members may hold their own settlement account with any other financial institution, such that all
requests for funds or payments are ultimately settled through the correspondent services of domestic
clearing and settlement systems.
...
Remember that many (most?) recipients of international remittances don't have bank accounts or access to payment cards. Hence why Western Union et al have to maintain extensive agent networks.
The thing to remember is that Visa authorisations are done in near real-time but settlement between issuers, Visa, acquirers and on to merchants is done over the normal banking system, as far as I know.
I don't know enough about Visa's design to comment authoritatively - but my working assumption is payments from issuing to acquiring banks in-country are done net and through visa - i.e. one net payment per issuer into visa and one net payment out to acquirer. e.g. see the last page of this doc: https://usa.visa.com/download/merchants/visa-core-principles...
What isn't obvious to me is what happens in international scenarios - you ask a great question.
Transacting in US dollars accomplishes the same thing, with far less volatility. In fact, transacting in any commonly exchanged currency does (US is the global reserve for this reason - you can exchange US dollars for any currency on the planet - notably Chinese currency - easily).
Moreover, there's no benefit here: dealing with separate Bitcoin services removes any end-user guarantees. If I send money from Australia to England, Paypal Australia has to deal with the Australian government and the British government to obey consumer law. With a Bitcoin exchanger, once the BTC is transferred you're at the mercy of whatever local exchanger you use at the destination.
Paypal wants to stay in both countries, which means there's an end-to-end legal protection for both parties.
The open-nature of Bitcoin is irrelevant. The USD is pretty damn open.
These are good questions. Someone should write a primer on money transfers and cryptocurrencies (someone with better answers than my best guess, included below.)
My hunch is that the trust required to move money is very similar to the trust required to have an account with money at your bank, but the main difference is you pay for that trust in different ways.
If you write a remittance, you trust it will be remotely delivered upon request (ie, immediately). That trust is ensured by an organization that has access to ready capital in many locations (which involves some opportunity cost, Western Union could just be pooling all that money and investing it). You pay for that trust through fees.
When you deposit money at a bank, you trust that they will return it to you at any of their branches at some point in the future. That's a very similar sort of trust. Yet here, you really pay for it by foregoing the opportunity cost of lending your money to strangers. Though they pool your money with the money of others to smooth risk, so they're getting a better return / less risky return from lending than you could get on your own. But you're really paying through the difference between the return you would earn by loaning it out and the interest you earn. You're paying that gap.
Similar problems occur in domestic money transfers, so domestic money transfers still have fees. However, I would expect that establishing trust with international money transfers involves dealing with multiple currencies (possibly some of which are being inflated by a government), magnifying the costs. I would expect the fees would tend to be higher.
(Western Union doesn't suggest this is the case. Sending $1000 instantly seems to bounce between $86 and $95 no matter where I send it, domestic or international. They may be making some money by setting exchange rates, I'm not sure. Also, they sometimes gave me wildly outlier fee quotes, so I'm not sure those are their actual prices, or how stable they are, or if there's not a bug in the website. For comparison, World Bank says remittances average around 8.14%: http://remittanceprices.worldbank.org/en )
Bitcoin offers some opportunities to bypass some of the required trust, possibly resulting in drastically lower fees. (You still have to trust the network won't implode though.) That said, I don't want to suggest remittance services are gouging anyone. I have no doubt it's costly to set up an international trust network with cash on hand all around the world. But I think there's an argument to be made that the infrastructure for a cryptocurrency scales a bit more easily than the infrastructure for a Western Union. (On the other hand, ensuring there are buyers and sellers of bitcoin in whatever two cities you're using as endpoints isn't trivial either.)
That World Bank link above talks about the "5x5" goal of reducing remittance fees by 5% (from 10%) over 5 years (beginning in 2010). Work anywhere in the developing world or on development economics and you'll get a sense of how critical remittances are to developing economies (often swamping the impact of foreign aid). It's conceivable that many humanitarian and development goals might be hit if we could use technology to lower barriers to easier money transfers.
You might be interested in the Hawala system[0] of money transfer. It has existed for hundreds of years, and allows efficient international money transfers with no central authority.
Hawala has fascinated me to no end, as essentially an anarchistic peer-to-peer web-of-trust banking system. It has existed for a long time, and keeps functioning even where traditional banking systems have broken down. As cool as Bitcoin is, I have more hope for a system akin to Hawala than some proof-of-work based system. Guerilla banking that depends on having more (computational) power than your enemies isn't too realistic. PGP'd remittances are much more cryptographically robust, and depend just on the one factor that any real monetary system relies on anyway: human trust.
If your money transfer model includes gateways, then trust is involved. You have to trust the gateways (local banks or exchanges) to make the conversion.
The gateways on each end would also want to take a fee.
Add to that the (MASSIVE) cost of compliance globally, and the costs of the model described by Stripe may not be much better than that of existing services (TransferWise, Western Union, etc.).
It's just an absolutely mind-bogglingly large, long, and costly endeavor, and the end results may not be worth it.
> The answer to basically all of your questions is that to transfer money, you need to move money.
Is that true? I everyone around the world is moving money around, then most trades can be covered by not moving any money around, just between people in the same country.
If I want to move $1000 to France, and someone else wants $1000 from France, we just swap money, and two people in France do the same.
What is left is keeping track what everyone has put in, and what they have taken out.
> What is left is keeping track what everyone has put in, and what they have taken out.
And that's exactly what Bitcoin is; a secure digital ledger. It's a way to securely and globally record "person A has 10 BTC, gave person B 5BTC, then person B gave persone C 3 BTC" in a way that doesn't allow person A to simultanously have given those same 5 BTC to person D (who, given the anonymity of the internet, may also be person A).
That's all Bitcoin does; it provides a secure way to record such transactions without having a trusted third party who must trust to increment and decrement the right accounts in the right way.
So is that not an impressive solution - if Alice wants to move 100 USD to France and Bob wants to move 100 USD out of France (let's say both want to buy a new textbook) then one could envisage a peer to peer money transfer system.
A site allows bob and Alice to find each other, agree a shared amount to transfer, agree to which end point they will onwards transfers (hmmm this might be the breakdown point) and then record that in the block chain
I think oddly there is still an enormous amount of trust involved - trust that Alice will complete the final important step of giving the money to Bobs silver haired grandmother or whatever
You can actually achieve the same thing with Coinjoin if all involved parties agree. Instead of a bunch of individual transactions, they collectively generate and sign one.
I am returning to this thread a bit late, but let me try to answer your questions:
> If money exchange is the main problem that Bitcoin solves, then why isn't the problem of money exchange attacked directly instead, i.e. through some other less volatile classical stores of value like gold or stock?
"Money exchange" is a poor way of phrasing the problem that Bitcoin solves. Satoshi called it "Peer-to-Peer Electronic Cash", which is a better way to think of it: currently, without Bitcoin, the only way to transfer money to a peer without relying on a third party of any sort is to hand deliver a wad of cash. Bitcoin enables that transaction to happen electronically, for free. I don't see how it would be possible to enable those features (no need for third party, free international transfer) while using classical stores of value, and as far as I know neither does/has anyone else.
> Why is money exchange a problem anyway and why couldn't a classical wire transfer solve the problem?
This is basically asking "why is the trust thing so important?" That's a pretty heavy question, but one simple answer that comes to mind is that trust is expensive, and economically inefficient. The capital involved in mimicking the behaviour of a company like Western Union is a small fraction of what they're worth by virtue of being Western Union (ie. having people trust them) and the fact that they meet regulatory requirements everywhere they operate. That brand is only valuable and those regulations are only necessary because transferring money used to be impossible without them.
Now, what Stripe is suggesting (and I happen to agree) is that there may always be trusted third parties and possibly some level of regulations involved to protect the consumer, but the level of trust and the amount of safeguards will ultimately be tempered by the enormously reduced barriers to entry (and hence vastly increased competition), potentially to the point where consumers will have almost the same variety of choice with their payment providers (ie. replacing credit card) as they currently enjoy with email. There are still a lot of technical hurdles to work out before that can become a reality, but that's the dream I'm excited about and the real value I see in Bitcoin.
Classical wire transfers have two things that people object to: high fees and anti-money-laundering legal requirements.
The problem with making it "easy" to move money is that fraud becomes easy as well. There was an article on HN the other day about someone who wired $4k to a fraudulent AirBNB host. The credit card system skews towards assuming that the merchants are fraudulent and the customers mostly honest. Bitcoin goes in the other direction: it's entirely caveat emptor, there's no fraud protection or recovery at all. If your key is used to send bitcoins, by you or someone else, they're gone to that address. This allows for the low transaction fees but creates a trail of people who've lost money, whether on mntgox or localbitcoins or random internet purchases.
(I've been playing with the description "uber for banknotes" for a while; the disruptive evading of existing regulated systems is important to both Uber and Bitcoin)
Bitcoin allows for escrow transactions, where a third party can act as a referee in case there's a dispute in the transaction. It has the important property, absent from current escrow setups, that the escrow party never holds the money: they have the power to accept or reverse a transaction, but nothing else. It limits the trust commitment towards the escrow service.
I imagine escrow being done by the likes of Airbnb, in the case of the service. It won't be error free, but should stop the absolute frauds. In those cases, the escrow service is already paid for: it's the market that facilitated the transaction.
As for the money being kept in escrow, that is the definition of escrow. The alternative is the scenario of credit card payment processors, which keep the money for a period, and retain a fixed percentage of sales turnover, as a guarantee for refunds. From the merchant perspective, the average payment delay should be shorter using bitcoin than it is with credit cards.
The answer for why Bitcoin instead of Gold or stocks or whatever comes down to trust and logistics. Take Gold for example you have to have to either ship the gold around which is expensive and slow or have a trusted third party store the gold and handle netting out the transactions between entities or accept delayed payment and trust that the counter party will deliver the underlying asset. Bitcoin does not require trust in anything except the math underlying the system.
> Bitcoin does not require trust in anything except the math underlying the system.
False. It requires trust in the liquidity of bitcoin. If you can't convert bitcoin to local currency, it has no value. Bitcoin's fragility is in its ability to act as a holder of value, as repeatedly show by speculative price swinging (sometimes without apparent cause).
Gold, for all its faults, has a huge track record as a value store.
Beside the fact that it already exists, I think the best argument would be that the BTC network didn't require coordination between financial institutions or a design committee.
You're not alone: all the people who bought Bitcoin and holding it use Bitcoin mainly as the best store of value in human history. All fiat currencies are being printed like crazy (CHF was an exception, but it changed a few years ago)
Could you explain that? Compared with traditional currencies, Bitcoin's high volatility and unknown long-term risk profile make it look like a terrible store of value to me. (And, for similar reasons, very attractive to speculators.)
Volatility should never be important when selecting a store of value. Gold and silver has much volatility compared to USD and hamburger price, still if you sold a cow 5000 years ago in Egypt for silver, you could buy at least 0.1 cows from the same amount of silver today. Silver doesn't lose more than 0.01% of its value/year (still the volatility is much higher than that).
It's actually more important to understand why USD and CHF are so bad store of value. The best explanation is by Mike Maloney:
https://www.youtube.com/watch?v=iFDe5kUUyT0
Volatility matters a lot for a value store, because the point of storing value is being able to convert back to something you actually value later on. If you put it in something highly volatile relative to whatever you actually care about, you may end up losing a lot of the stored value.
I can't reply to your comment for some reason, so I reply here.
I bought Bitcoins more than 1 year ago. If you look at Bitcoin in terms of years, and as a long-term store of value, Bitcoin is not volatile: the value is increasing until total adaptation.
And I don't know anybody else who bought and held his bitcoins who's complaining :) (I'm not counting people who had Bitcoins held at MtGox, as they didn't have Bitcoins, just IOUs for Bitcoins)
That you bought something and had it go up is great for you, but it does not mean the volatility is low. Indeed, if it went up a lot, it means volatility is high. Volatility is the inverse of stability.
A currency can be as volatile as it wants from day to day, but if over a very long time the value appreciates consistently then wouldn't that make it a good value store? You're hedging day to day against a long term win, surely?
Nope: long term appreciation may make it a good investment but a poor store of value.
Lets say you have two magical safes, and you put $100 in each one.
The first safe gives you the real-dollar value of what you put in, less $0.50 per-month. So a year later, you get $94 worth of inflation-adjusted dollars no matter how high or low inflation has been over the course of the year.
The other safe gives you an additional dollar for each day you keep the money in it, but there is a 70% chance all but $10 catches on fire when you open the safe. So one year later you have $465 or $10, statistically averaging $149.5 .
The second safe might be worth gambling on as statistically you get almost a 50% return on investment: but if you need to be certain the value you put in is maintained, the first safe is the better option. Therefore, the first safe is a far better store of value.
Gotcha. I've been thinking a lot about BTC recently, but I'm useless when it comes to economics and there is a lot of "faffing" around it. Thanks for the information. :)
Bitcoin has a relatively fixed supply, but that doesn't mean it's guaranteed to appreciate over time.
Bitcoin is only valuable because people choose to find it valuable. One example of how it could crash in value is if a competing cryptocurrency starts to gain momentum, and people jump ship, or if a major flaw is found.
There are many plausible scenarios for bitcoin going to 0.
The maximum amount of each cryptocurrency is hardcoded into their design, but nothing stops people from "printing" new cryptocurrencies. There's bitcoin, litecoin, peercoin, darkcoin, namecoin, primecoin, and of course dogecoin... Gimme a few hours and I could cook up xiphiascoin in your honor.
And when something happens that makes people migrate en masse from one *coin to another (perhaps a security flaw in the reference implementation), you might wake up to find that your old coins are now worth less than a Zimbabwe dollar.
The article doesn't say anything about these things and the underlying principles other than that it's "suprisingly difficult". Instead it goes right into proposing a solution based on Bitcoin and making big picture comparisons with network protocols that probably sound very smart to the right audience.
Given the amount of agreement and hype that the article has received so far, I feel a bit like an outsider right now questioning what is proposed. I think I just don't get it. Is everybody just a lot smarter than me or does the community miss to address some basic questions here?