3 ms·
> From whom? What are the actual rates? The rates are the cost of buying/selling bitcoin on both ends, let's call that 20 bps on each side, and then the cost o
by darawk 8y ago
> From whom? What are the actual rates?
The rates are the cost of buying/selling bitcoin on both ends, let's call that 20 bps on each side, and then the cost of a single bitcoin network transaction, which is about 66 cents at the moment, if you're willing to wait 6 blocks [1]. So, to send $1000 it'll cost you 40 bps + 66 cents = $4.66. By comparison, international wires at your local bank will run you around $50 all in [2]. That's quite a savings.
1. https://bitcoinfees.info/ https://bitcoinfees.info/
2. https://smartasset.com/checking-account/average-wire-transfer-fee https://smartasset.com/checking-account/average-wire-transfe...
> This is a big caveat for most people since “handle improperly” actually means “never use buggy or untrustworthy software, get mugged, etc. The benefits listed above don’t apply to most people enough to be worth that big of a risk transfer.
Sure, and those are real challenges of using cryptocurrencies right now. However, they're also extremely solvable problems and people are working on them.
- acdha 8y ago> The rates are the cost of buying/selling bitcoin on both ends, let's call that 20 bps on each side, and then the cost of a single bitcoin network transaction, which is about 66 cents at the moment, if you're willing to wait 6 blocks [1]. So, to send $1000 it'll cost you 40 bps + 66 cents = $4.66. Don’t forget to add in the currency conversion fees on either side since most people transfer money to use it. Walmart charges $8 for the same service. There is some potential savings here but it seems unlikely to argue that Bitcoin’s enormous costs can be justified by the number of people who need to regularly do international transfers and, if everything is lucky, save a dollar or two on the transfers. > Sure, and those are real challenges of using cryptocurrencies right now. However, they're also extremely solvable problems and people are working on them. “Extremely solvable” but unsolved a decade later suggests that they are not in fact that easy and will end up making the cost even less competitive. People have been preaching that Bitcoin will take over for years but even most advocates don’t use it for a non-token percentage of their transactions because it’s simply not financially sensible to do so.
- darawk 8y ago> Don’t forget to add in the currency conversion fees on either side since most people transfer money to use it. Hm? That would be factored in already. If I wanted to send money to say, India, I would buy Bitcoin with US dollars and then send it to someone's account on an Indian exchange, where they'd sell it for Rupees. The conversion is already baked in. > Walmart charges $8 for the same service. There is some potential savings here but it seems unlikely to argue that Bitcoin’s enormous costs can be justified by the number of people who need to regularly do international transfers and, if everything is lucky, save a dollar or two on the transfers. I'm not arguing that international wires are the only useful thing it can do. It's simply a counter-example to the idea that there are none. > “Extremely solvable” but unsolved a decade later suggests that they are not in fact that easy and will end up making the cost even less competitive. People have been preaching that Bitcoin will take over for years but even most advocates don’t use it for a non-token percentage of their transactions because it’s simply not financially sensible to do so. It's already solved if you keep your money on an exchange.
- acdha 8y ago> Hm? That would be factored in already. If I wanted to send money to say, India, I would buy Bitcoin with US dollars and then send it to someone's account on an Indian exchange, where they'd sell it for Rupees. The conversion is already baked in. Right, but if you're comparing apples to apples, you need to compare the cost of, say, going to Walmart.com (or a bunch of other places) with $1,000 and sending it to someone in India vs. starting with the same $1,000 and shipping it. Unless the conversion rates for gating in and out Bitcoin are always exactly the same you need to measure the whole cost. > I'm not arguing that international wires are the only useful thing it can do. It's simply a counter-example to the idea that there are none. Note that I wasn't saying that there no things which were technically possible, only that very few people have a situation where it's advantageous to use Bitcoin. A possible modest savings on international transfers does not seem like a viable foundation for a decade of big world-changing rhetoric. > > “Extremely solvable” but unsolved a decade later suggests that they are > not in fact that easy and will end up making the cost even less competitive. People have been preaching that Bitcoin will take over for years but even most advocates don’t use it for a non-token percentage of their transactions because it’s simply not financially sensible to do so. > > It's already solved if you keep your money on an exchange. … so, you get rid of banks by creating a new bank, putting a “not a bank” sign on it, and hoping nobody will notice a) how many people have lost money by trusting Bitcoin exchanges and b) how quick the community was to blame them for the loss and saying that they were fools to trust an exchange?
- janekm 8y agoThat's an overly rosy comparison of money transfer fees. To make it a little more fair, it makes sense to compare against a competitive money transfer service rather than typical bank fees... For example transferwise offer transferring $1000 to Euro for $6.86 total fees, at the mid-market rate, including the depositing fee. Depending on the bank networks on either end the transfer can be instant. In a way transferwise operate like an exchange only it's FIAT-FIAT rather than FIAT-BTC-FIAT. To compare you would have to take into consideration the bid/ask spread of the exchange on each end of the transaction for BTC, in addition to all charges on both ends. The bid/ask spread varies wildly from exchange to exchange. I'd be impressed if it were to cost less.