[wsxcde]

No, in terms of the overall cost of operation of the system, bitcoin is most definitely not cheaper. Transaction costs paid by the users are low because of the inflation subsidy paid out to the miners.

In terms of overall costs, the bitcoin mining network alone (i.e., ignoring the cost associated with running all the full nodes, the underlying network and so on) costs $1MM per day to process a measly 100k transactions.

[IkmoIkmo]

No, you're conflating two things.

The 'cost' of bitcoin the system, is in the inflation of its money supply. That's indeed about $1m a year. (let's not try to do an equivalent calculation for the many billions of dollars of value that are lost from dollar inflation per year).

But recognise that it's $1m a year whether there are 0 transactions or 1 trillion. It's not a transaction cost, it's not related to the amount of transactions. It's a fixed system cost in the form of inflation. Calling it a transaction cost is misleading straight away.

The actual transaction costs are a few cents or so on average.

Of course, it's still costing money, but not to the person making the transaction. After all, if I send you a bitcoin and you forward it, you pay 5 cents. You don't pay for any inflation, only bitcoin holders do, which don't have to necessarily be those who transact. Similarly I, at times, make payments in dollars on dollar-denominated financial systems. I pay transaction costs, which are completely separate from the cost of dollar-inflation which I don't incur as I don't hold my wealth in dollars (european here). These two things are not the same.

Do bitcoin holders lose value due to bitcoin inflation in the form of unrealised gains? Absolutely, but then also recognise it's the best performing currency of the past 5 years, or indeed last year, and that they don't really care. Further also recognise that in the code of bitcoin it's written that the inflation will permanently drop to sub 1% levels pretty soon, a level of inflation lower than any major currency.

Actual transaction fees will probably rise somewhat to compensate for mining rewards dropping over time, but not by much. For one miners are deemed to be overcompensated from the perspective of the level of security they offer, and if you scale transactions by 10x or 100x (which from an economics point of view is pretty trivial, although the political ramifications of bigger blocks aren't) you can keep the transaction fee really low while massively scaling fee rewards to miners, further compensating a drop in mining rewards. (and if bitcoin doesn't scale, who cares if it's economically sustainable, if it doesn't grow then it'll stay trivial and I couldn't care less).

[aback]

The cost to secure the system is orthogonal to the cost of processing transactions and rises and lowers proportional to coin price, providing a level of security commensurate with the market value of the thing secured.

The cost per transaction is simply a function of the rate limiter that is attached to Bitcoin. Bitcoin could secure many more transactions per second if the rate limiter was removed.

> a measly 100k transactions

As I type this the network is averaging ~12000 txns per hour, or 288,000 per day, FWIW.

[afreak]

12,000 transactions per hour translates to 200 transactions per minute or 3 per second. However, you're being generous here because what you're suggesting is what it currently does at best, when in reality it is anywhere between 1 and 3 transactions per second, meaning that it is doing anything between 86,400 and 288,000 per day--that's a large spread.

Visa claims it can do 2,000 transactions per second or 172,800,000 per day and Paypal suggests its maximum is 115 per second or 9,936,000 total per day [1]. At best, Bitcoin is able to handle 3% of the volume that Paypal is able to and it doesn't even register as a blip when compared to a single major credit card.

If we continue to read the link I cited for those numbers, we get these details:

>Let's assume an average rate of 2000tps, so just VISA. Transactions vary in size from about 0.2 kilobytes to over 1 kilobyte, but it's averaging half a kilobyte today.

>That means that you need to keep up with around 8 megabits/second of transaction data (2000tps * 512 bytes) / 1024 bytes in a kilobyte / 1024 kilobytes in a megabyte = 0.97 megabytes per second * 8 = 7.8 megabits/second.

>This sort of bandwidth is already common for even residential connections today, and is certainly at the low end of what colocation providers would expect to provide you with.

If we were to expect that everyone was to get onboard the blockchain, we would have to assume that we're all going to start sucking up 0.97 MB per second, or about 82 GB a day. This would mean that in a month, one would use just about 2.5 TB of bandwidth. If the average residential user were to want to transfer 2.5 TB for just a blockchain (ignoring modern life's pleasures like Netflix and other streaming services), you'd have to find an Internet service that would not impose a usage cap. As it stands with Comcast, if you exceed 300 GB, you'll get charged $10 USD for every 50 GB you use [2], meaning that a month's worth of transactions will cost you just about $450 on top of what you already pay for the privilege of having access to a blockchain.

This doesn't even take into account that transactions are replayed once confirmed so those numbers could end up even higher if the whole protocol isn't changed. Heck, I am not even taking storage into account which is a different problem all together.

The idea that blockchains are going to be the future is pure lunacy. It's either going to be limited to data centres that can handle this load or it's not going to be used at all. It has no future for the general public and is part of the reason why Bitcoin in itself is not practical.

[1] https://en.bitcoin.it/wiki/Scalability - Bitcoin's "official" wiki backs this up

[2] https://bgr.com/2015/11/19/comcast-data-cap-2015-bad-for-us-...

[aback]

First off, the numbers I cited were not theoretical, they were the observed network run rate at the time I posted.

Secondly, Bitcoin doesn't have to scale up to Visa levels, at least not in 2016. In 2016, it just has to scale up. And there's plenty of room to scale up.

Bitcoin has a growth plan that subsidizes the expansion of the network (via inflation and adoption) for many more decades. In two decades, the Moore's Law equivalent of a 1MB block size is a 16GB block size. Even without additional scalability, Bitcoin can process a lot of transactions with 16GB blocks.

[icebraining]

No, in terms of the overall cost of operation of the system, bitcoin is most definitely not cheaper. Transaction costs paid by the users are low because of the inflation subsidy paid out to the miners.

Now compare that to the subsidies in the form of bailouts to banks, and we'll see what's cheaper.

Yes, I'm being facetious, but my point is that if you want to include such costs on one side, you must do the same on the other.

[wsxcde]

Should we also then count the amount of money that are lost to bitcoins being stolen due to malware, poorly written private key generators, exit scams and so forth?

I think what I did was a fair comparison.