How much bitcoin does it cost to maintain the Bitcoin network?

Let's go straight to the point: if Bitcoin is really a cheaper and more efficient payment processing mechanism, for-profit organizations such as large payment providers, they'll start replicating it and at least start using it internally to boost profitability. They didn't do it because it's not a bargain, in fact, maintaining a network like Bitcoin is much more expensive than maintaining a centralized payment system, such as running a MongoDB (a new distributed document storage database) on a PI server.

For them, the bottom line is that the marginal value of these centralized solutions must be greater than the cost of maintaining them (MV>MC), otherwise no profit will be achieved. It will be detailed below that Bitcoin does not meet this requirement.

There is a mistaken understanding – the Bitcoin network is cheap or even free – regularly will be mentioned in the Bitcoin community, which is in fact the same root as the looming issue of online transaction fees.

For example, Gavin Andresen (Gavin Andresen) recently wrote an article about how to increase the block confirmation speed:

It is asserted that "Bitcoin does not have economies of scale" means that the amount of resources that the Bitcoin network handles for a single exchange does not significantly decrease as the number of transactions increases – the translator notes "", theoretically this is correct: the bandwidth and CPU resources you need to process n exchanges per second are still required to be proportional to N. When Bitcoin becomes the world's second largest payment network one day, we may start to get into trouble.

Here are some rough calculations that show that our home broadband can support 15,000 transaction confirmations per second. Looking at it in a different way, today the world's first payment network system (VISA) deals with 212 million transactions per day, equivalent to 2,500 trades a second. During the busiest shopping season, their peak handling capacity is said to be up to 40,000 trades per second. My home Internet bandwidth is 90mb/seconds, a Bitcoin transaction is averaging 2000 bits (bit), so my home internet broadband can download 45,000 transaction data per second, exceeding the 10 times times the average transaction volume of visa per second.

I'm glad to know that I was at home. You can create a node and process data that exceeds the visa transaction size, which is better if you run a dedicated server in a data center. 15,000 average 250 byte (byte) size per second of transaction data, which means 7TB per month of data traffic. One of my hosting providers is charging $20 per month for each virtual server with a monthly traffic limit of 8TB, or in other words, you can handle the amount of transaction data at the MasterCard credit card level by spending $240 a year. (August 2014)

Andresen the above solutions to the scale of transaction data are innovative and feasible at first glance. But he only considered the cost of transaction data, not the actual pay of the miners for data security and processing of all transactions. Hash calculations (Hashing) are not free.

As we have discussed elsewhere, the network is not running in a well-intentioned environment – in fact it may be the opposite. As a result, real costs and real costs for each transaction using the Bitcoin network will fluctuate between $30 and $50. These costs are not directly visible to the user. The newly generated bitcoins actually bring inflation, which has depreciated the currency in circulation – so the price of these deals is inflation or the price dilution of bitcoin (for example, if there is no mining, there will be no new currency, but there will be no trading).

There are many objections to these economic descriptions, such as Antonis Bolamitis Antonis Polemitis does not endorse the way in which the actual cost of the network is measured and calculated, which he considers to be "self-correcting and without the fear of Bitcoin users" and " Do not underestimate the fact that everyone can dig a mine, as long as you are willing to spend some money. "

These are all fallacies. Polamitis's mistake is that maintaining the network consumes real money, and once the benefits disappear, users will have to face losses. By contrast, Visa is compensating for this cost by imposing a handling fee of less than 1% (see Richard Brown Richard Brown's latest article on this issue http://gendal.wordpress.com/2014/08/09/ a-simple-explanation-of-fees-in-the-payment-card-industry/).

The way to measure real costs is as follows: In an open, competitive market, the marginal income (or marginal value) of a product or service is infinitely closer to its marginal cost (MV=MC). Based on historical experience, we can almost see the same scene in every niche market with lower barriers to entry, and if this marginal income or marginal value disappears, the labor force will enter other markets. Bitcoin is no exception.

The cost of generating a new Bitcoin is also a bitcoin.

Or you can ask how much bitcoin it takes to produce a new Bitcoin in a fully competitive market.

For the sake of discussion, let us assume that Bitcoin is a unit of accounting, and that all costs (mining machines, site fees, management fees, electricity bills) are also denominated in bitcoin.

For example, miners (labor) are constantly "burning" a kind of goods (bitcoins) in a competitive environment to produce a common cargo (bitcoin). That is, they must constantly consume some value in exchange for the same or more value.

The end result is that in theory miners will spend a bitcoin to provide security and transactional services to the entire network while reaping a Bitcoin return – in the long run, the cost of generating (or creating) a Bitcoin will be equal to one bitcoin. Note: The difference between the two is called a mint tax, or in the language of Bitcoin, it is called "Block return".

In practice, however, as Jonathan Levin (Jonathan Levin) put it, inequality is on the way: large mines always have better, more efficient mining equipment than others. For them, it takes only a fraction of a bitcoin to produce a bitcoin. This is a bumper harvest for them (like a bumper crop in agriculture). However, mining is a 0 and a game, and only a few participants will get a bumper harvest – in fact, most of the marginal participants will not have a good harvest because they are always unable to receive the best mining equipment on a priority basis. Again, unlike Polamitis's conclusion, to enter the mining industry in 2014, barriers are more than "spending money" – a huge capital investment and a good relationship with miners.

So, for the miners, since the Bitcoin network (and all its derivatives) returns 99.8% of the money from the mint tax, if the compensation disappears, the miners will leave (as is the case for all other networks, since few people pay the transaction fees). This is the real cost, but few people admit it and don't know what the reason is, people always assume it doesn't exist.

Peter Tode (Peter Todd), a bitcoin program developer, also found this problem, he said:

Bitcoin is not an effective payment system. It replaces a centrally managed data center with a huge team of miners and thousands of copies of blockchain data that are constantly converting electricity into heat. That's why each transaction costs more than $30 and requires a constant appreciation of the bitcoin to subsidize it. In contrast to existing systems, the meaning of Bitcoin's existence is that it has freed itself from the shackles of government management, but the cost of this freedom is immense.

Robert Sams (Robert Sams) and Ray Dillinger have discussed the issue in the past.

The process of digging a mine is equated with the private printing of a banknote:

It is an unnecessary (and useless) trait for people to spend money on the right to bid for a banknote, which will always guide the cost of printing money to the face value of a banknote.

Similarly, Sams wrote 8 months ago:

Miners (those in the industry who choose to mine) will continue to work only if the expected yield on the mine is greater than the cost of electricity spent on the mining equipment. For the mining industry, there is no access system, the industry's balance mechanism from the Bitcoin agreement in the difficulty of adjustment. If the price of Bitcoin climbs and mining is profitable under existing difficulties, more miners will pour into the industry. After 2016 blocks have been excavated, the protocol automatically increases the difficulty, again making the expected yield of mining = the cost of mining. The same process is reversed when the price of Bitcoin declines. In the process of creating cryptographic electronic money, marginal cost = marginal gain (MC = MP).

"Note: MP stands for marginal productivity, it is interchangeable with Mr (marginal revenue) or MV (marginal value)"

As I mentioned for the first time in May, the short-lived mint spreads that miners (private print makers) can see are always being narrowed down until marginal gains = marginal costs (MV=MC). In contrast, in a country's Fiat Mint system, marginal gains are always greater than marginal costs (except for coins).

This is not a policy endorsement, I am just trying to explain the economic truth and the motives behind the mining industry.

Conclusion

At first glance, the idle proof mechanism (Proof of idle) seems to work as a solution (see the two short reviews of Andong Porotinski (Anton Bolotinsky)

http://www.ofnumbers.com/2014/08/18/downplaying-statistically-possible-double-spending-risks/), but it seems to only avoid "marginal gains = The emergence of marginal costs may also create another problem, in which the industry's large mines may collude and further oligopoly, leading to a complete inability of newcomers to enter the industry; there may also be another problem, as core developers are employed by large pools and miners, The core protocol for Bitcoin in the future is unlikely to develop in the direction of reduced energy consumption (a "captive regulatory capture").

The last part of this article is obviously derived from the author's speculation, but its deep economic mechanism is indeed long-standing and can not be ignored. I do not think Bitcoin can be used for any other purpose. As Todd said, and I've discussed it in several publications, Bitcoin is not a competitive means of payment, but it is, or at least has its own original purpose, to help build a distributed trust system. Coincidentally, Bolamitis has an article describing some of the possible applications of Bitcoin (Http://ledracapital.com/blog/2014/3/11/bitcoin-series-24-the-mega-master-blockchain-list).

In summary, the significance of Bitcoin lies in building distributed trust. The Word trust (or some of its variants) appeared 11 times in the text of the Genesis paper of Bitcoin, appearing 1 times in the diagram and appearing 1 times in a reference. It is said to be the true passion of Nakamoto. In order to build a distributed trust system, he has to design a network that consumes more resources than a "trusted" centralized network. I'm not saying it's "bad", but the design is not greener or more efficient than running the MONGO database on the Raspberry Pi (Pi box), and the latter can run stably even in an untrusted environment.

Reprinted from: http://www.sailingnet.net/sailingnet/4971