Running an honest Bitcoin exchange isn't very profitable. Some exchanges don't charge commissions, so all rates are low. Coinbase doesn't charge customers commissions until they do $1m a year in business, which few do.
I don't think they intentionally cancel transactions as a way to make profit. You can see reports of them cancelling transactions when the price goes down as well but still refunding the full previous amount.
I also have an instant account so I can immediately get my funds with no chance of cancelling.
As to the accusation you can see is thrown around all over the place, that is possible, but my interest-free loan to them of several thousand dollars left a sour taste in my mouth regardless and it wouldn't be hard for them to make some money off of the capital, in BTC or not.
Coinbase is a payment processor acting as a market maker for their clients in which case they charge a spread.
If you try to build a pure exchange this isn't an option.
VCs and founders take chances all the time that a large and loyal group of users will eventually lead to an enormous revenue stream. No one really knew that search leads to a fantastically profitable ad platform. If an exchange is really good for a long time and transactions fly through at volume and cost per transaction unseen up to now, there are all sorts of possibilities.
I don't really understand bitcoin, so I could easily be totally wrong. That said, my understanding is that one of the big (or maybe biggest) advantages of bitcoin as a technology (regardless of the money supply set up and such) is efficiency, the kind of efficiency advantages that the web or bittorrent has over other communication technologies. If it potentially that eficant, it's good that transaction costs and rates are confusingly low, that lets volume be high. Let other companies build other parts of the ecosystem and let them have the advantage of near-zero transaction costs.
Basically, if bitcoin is going someplace, being a trusted exchange may very well pay off. I imagine that beyond a certain point of proven reliability and transaction volume, VCs will be willing to take some of the action off a promising horse.
Bitcoin's efficiency is "regulatory", like Uber; by transmitting money internationally without requiring KYC/AML compliance, and by putting all fraud costs firmly on the victim, it allows for nominally low transaction "fees". The fee ends up hidden in exchange spreads or covered by the losses of people who want to play with unregulated forex trading.
putting all fraud costs firmly on the victim...voluntarily. Ideally in a mature system with various options for mitigating risk cheaply. If someone wants the (currently a tight oligopoly) 'credit card insurance scheme' why not let them have it in a more naturally competitive market?
The spreads are the spreads and if transacting costs more then it does elsewhere no one will use it. Otherwise, some positive things might happen.
Anyway, what about people who want to play with regulated forex trading?
Bitcoin exchanges aren't free, right? They have to make their money. They will end up working almost identically to the way banks work. Or they will go out of business.
Bitcoin already has a transaction fee built it, its reasonable and value-added. I have no problem for paying for a service that gives me value.
I can handle myself in the real world, If I give money to someone who steals it from me? My bad, I'll eat my mistake. Don't steal from me and tell me its for my own protection.
Imagine you build a bridge for $1m and 1 person drives on it. Does that bridge cost $1m per user?
No, it's just an average metric. There isn't actually a marginal cost of $1m for every person that drives on that bridge. It just happens that the bridge is underutilized and that the average cost appears high.
Blocks are like that. If the block holds 0 transactions, the block reward is paid out. If the block holds 1 trillion transactions, the block reward is paid out. Either case, it's 25 bitcoins. In other words, a transaction does not carry a marginal cost in block rewards, as block rewards are paid out regardless of ANY number of transactions.
Transaction fees are the marginal cost, and those are pennies.
As for the block's capacity to hold transactions? That's interesting. If the $1m bridge could only ever move 1 person, then the cost of driving on that bridge was indeed $1m.
But blocks have quite large capacities. Currently 1mb, they're set to go to 20mb probably this year, and grow to hundreds of megabytes over time. Read about it from Chief Scientist at TBF on bitcoin [0].
The roadmap is that in 12 years according to the work that's being put in the protocol right now, you can fit a little under 400 million bitcoin transactions in a single block, using technology that's as expensive as the one that's running bitcoin nodes today. (cheap consumer grade home computers.) In 12 years, bitcoin will have had nearly 4 more halvings, meaning block rewards are by then only 1.5 bitcoin per block.
In other words, in roughly a decade bitcoin's block rewards will be a tiny tiny fraction of a fraction of a penny.
And even then, it's STILL not a 'cost per transaction'. It's not even an 'average cost per transaction', because the sender doesn't actually pay. The block reward bitcoins are created out of thin air, not paid by the senders of the transaction, which means it's a cost to all owners of bitcoin, regardless if any transactions are made, in the form of inflation of the money supply that devalues existing money by a tiny bit. Hey, just like say, every single currency in the rest of the world, whose money supply increases (aka everyone, including the dollar).
And this money supply is, unlike the dollar, which is printed every year in gigantic amounts, set to go to 0 by design, by software, and there's absolutely no way to change it except by consensus (moving to a new bitcoin blockchain).
[0] http://blog.bitcoinfoundation.org/a-scalability-roadmap/
Its a value redistribution by diluting all the existing holders of bitcoin. Bitcoins may be created out of thin air, but the value that makes them a reward is not.
> And this money supply is, unlike the dollar, which is printed every year in gigantic amounts, set to go to 0 by design
Which means that to continue to give miners an incentive to validate transactions, actual transaction fees will have to pick up the slack of providing value as the block rewards decline.
What's your point? I said so in my post. Maybe I'm missing something. We're both saying the same thing, which means that this value dilution is not actually a transaction cost. It happens regardless of whether any transactions are made. And it's similar to the dollar, in that daily/yearly new money supply dilutes the value of all dollars. Bitcoin isn't unique. It's only unique in that this new money supply eventually ends, while the dollar and any other currency keeps diluting forever.
> Which means that to continue to give miners an incentive to validate transactions, actual transaction fees will have to pick up the slack of providing value as the block rewards decline.
Agreed. And that can easily happen. Bitcoin is very tiny, it's maybe used on average like once a month by 1 million people. It can easily scale 100x in use case, meaning 10 million people use it every few days. It'd still be tiny. But if that'd happen, the price would likely be at least 10x higher. At that point you could literally cut out all block rewards, the transaction fees (in the pennies) would already be enough to pay for the miners. And seeing bitcoin, a global currency, scale to 10m users in the next few decades as mining rewards taper off, is a very very tiny expectation. If it can't even do that, then who cares if the economics work out, nobody wants to use it anyway. But if it scales because people want to use it, to 10m people or 100m people or even to the order of the billions of users that people like Marc Andreessen think it could be used by one day, then block rewards are completely unnecessary, and you can pay miners a shit ton of money on transactions that cost pennies.
And again, if it doesn't scale because nobody wants to use it, then who cares if block rewards go to 0 and suddenly transaction fees have to be $20 per transaction... It's like saying 'if by 2015 typewriters won't be used by millions of people, they will become super expensive'. It's completely irrelevant. Nobody cares. And if it does scale, you can keep transaction fees cheap and pay miners with larger volumes of transactions.