Bitcoin Forum

Many of the miners already know most of this stuff, but it's still a worthwhile read.
Not only does this develop a model for the intrinsic value of bitcoins based on cost of production, but also cautions that with increasing efficiency lowering the cost of production (i.e. lower watts per GH/s), the value of bitcoin may also decline. The block reward halving in late 2016 may have the effect of eliminating all but the most efficient miners.

https://www.researchgate.net/publication/272022750_The_Decision_to_Mine_for_Bitcoins_A_%28Computational%29_Labor_Theory_of_Intrinsic_Value

Bitcoins are produced at a fixed rate and are not consumed. The cost of production has no effect on their supply or demand, so it has no effect on their value.

Take a look at Stellar, NXT, and Mastercoin. These three coins have the 2^nd, 5^th, and 8^th highest market caps, and yet their production costs are 0.

Value and price are different things. Supply and demand forms price. Cost of production forms value. Read the paper it discusses the difference.

Very true, to the end, it is the electricity cost decide bitcoin's value. After a certain degree of consolidation, only the miners without equipment cost (means the equipment cost is already paid back) can stay in the network and mine. Any new equipment investment will cost too much and have too little hope of ROI due to extremely low income (barely make back the electricity), unless they are much better in efficiency. A 20% increase in efficiency will not motivate the upgrading due to equipment cost

At beginning of 2012, bitcoin exchange rate is $5, hash rate is 8TH, for a general GPU miner at 2MH/W, the whole network consumed 4MW

At beginning of 2015, exchange rate is $200, hash rate is 320PH, for a general ASIC miner at 2GH/W, the whole network consumes 160MW

So the exchange rate jumped by 40x and the electricity cost also jumped by 40x

However, since the coin per day now is only half of 2012, it can mean two things: Either the current price is too low and should bounce back to $400, or the daily coin supply on market has not changed too much despite the daily coin generation has been cut by half (Due to more sell-off by centralized mining farms)

Bitcoin is consumed, many people buy some coins and hold it for at least 10 years, and in these 10 years they are disappeared from the network, just like a house bought

Of course from a long term perspective they might get spent some time later eventually, but just like central banks' gold reserve, they could also accumulate more and more and never get spent in generations. A large amount of reserve can serve many other purposes, security for example

If there is a demand for the coin, people will always select the method with lowest possible cost to get coin, which will make the exchange rate very close to the mining cost

I read your paper. It is concise and well-written. You do a good job of describing the economics of mining.


You claim that the intrinsic value of bitcoin depends on the production cost, but nowhere in the paper do you actually present any arguments to support your claim.


These are interesting statements because they contradict your claim that the intrinsic value is the marginal cost. You imply here that if the cost rises beyond the break-even value, then the miner should stop mining (because it exceeds the value of the product), yet if the value rises to equal the cost (as you claim), then there would be no reason to stop.

I'm a bit skeptical on whether the next block reward halving expected to take place in late 2016 may have profound effect of eliminating all but the most efficient miners but I'm very sure the following halving will really make the difference. I've given myself another 5 years for bitcoin.

I don't want to make this a microeconomics lesson -- but very briefly:
-the arguments for cost theories of value exist and go back to the time of Adam Smith, there is no need to restate centuries of theory in a small paper on a specific topic, if you are interested you can educate yourself there are a ton of resources for that.

-along the same lines, it is basic econ. theory that marginal cost = marginal product = price. All 3 equal each other in a competitive market. Marginal simply means the cost of the last unit produced. So it is the production on the margin that matters and not what has been produced in the past, for example. The break-even price is not the market price, the market price will fluctuate all the time do to imbalances in supply and demand.

-the value will only rise or fall based on the average cost for the network. That means there will always be some producers with lower than average costs that will earn excess profits and some who operate at a loss. Over time, the ones operating at a loss drop out of the market and the average efficiency tends to increase with those out of the picture. This will LOWER the value of bitcoin, not raise it since the average cost of production for the network will have decreased.

It probably will, and thats bad because we allready practicaly only have industrial mining, eliminating little guy.
And that also means majority of money will go onto mining equipement instead market cap, so once again the man selling shovels is going to earn more than the one digging gold.
imo it was best when there were gpu-s only, ever since asics were introduced its not the same.

Sorry. I wasn't clear, or perhaps my point is subtle. You claim that the value depends on the cost of production. I disagree -- the cost of production depends on the value. You claim that as the efficiency rises, the price will fall. I claim that as the efficiency rises, the difficulty rises and the cost of production is unaffected (assuming the price is constant).

No no. The causation is certainly cost of production confers value. Competition amongst producers will create innovation and lower the cost of production. This is classic economics and it occurs in every competitive commodity market. The difficulty rising does not affect the rate of production- the rate of production will always be one block every 10 minutes on average.

In the Bitcoin industry, increased efficiency does not lower the cost of production because competition causes the difficulty (and thus the cost of production) to rise. Miners with higher efficiency will increase their hash power, which results in a higher difficulty and a higher cost of production. Miners effectively bid for a fixed quantity of bitcoins. The factor that limits their bids is the price of the bitcoins they produce.

Producing bitcoins is not like producing any other commodity. Name another commodity in which the total production quantity is fixed regardless of production cost.

Yes and no. I see what you're getting at--

the value of bitcoin = $cost of electricity used per day per GHs/how many bitcoins you can mine each day per GHs.

If the difficulty increases, the denominator gets smaller meaning the value will go up.

If the energy efficiency increases, the numerator gets smaller and the value will go down.

So both forces will be at work, true.

But as the energy efficiency increases, the 'break-even' difficulty also rises- meaning that the efficiency factor is more influential on the price than the difficulty. If the efficiency increases twofold, the breakeven price drops 50% AND also the breakeven difficulty rises 2x.
In reality the difficulty increase will serve to stabilize the fall in value but it will not outweigh it.

Your understanding of economics is quite bad. Leaning on the thoroughly debunked labor theory of value to calculate the "price" of something that is literally not sold is so far gone that I'm not sure where to throw the rope so that it might be within your reach.

The most salient points;

1.) Bitcoin mining is paid completely through inflation. Nearly any fee above zero is acceptable to miners. There is no "price" for mining.
2.) "Prices" are a function of supply and demand. Labor, work, and difficulty have exactly nothing to do with the price of a product except insomuch as it inhibits supply.
3.) There is no such thing as “intrinsic value”. There are only intrinsic properties that are subjectively valued by people.

Detonating the logic bomb;

According to your paper, the price of a product gravitates towards it’s production cost. You essentially argue that as the cost of producing Bitcoins approaches zero, the price must also approach zero. This would be true if you totally ignored the supply side of the Supply and Demand graph.
Since the amount of bitcoins increases asymptotically, the cost to create more bitcoins always necessarily approaches impossible, and by your own premise, also approach infinite price, not zero. QED.

If you adhere to mainstream neoclassical economics then yes you are absolutely right :)
If you adhere to a post-keynesian or keynesian point of view then cost of production = value in competitive markets.
I am not suggesting a Marxian labor theory of value, and again do not confuse value with market price which is determined by supply and demand. Market price and value may never meet, it is only a point of gravitation, and this is only from the perspective of miners who produce bitcoin and not necessarily from the perspective of any other user of BTC.

As for your second point, ultimately you are correct: Once the block reward is tiny, the value will become very high. But if the efficiency becomes very low that will counteract that factor as well. But even after the last block is mined, there will still be electric consumption to verify transactions in order to earn tx fees. Or at least that's how it's supposed to work.

The thing is, if you look at historical price data and compare it with the model, it does a really good job-

This;


Directly contradicts this;

Not necessarily, I am only saying the tendency is to a falling value, not that it will be the case.

Consider iron ore extraction. Every year, it becomes harder and harder to extract iron ore from mining operations since every ton of ore is marginally more difficult than the previous one (one average). It is either deeper in the mine, or it is coming from less productive veins or there is less percentage of iron in the rocks being mined. All the easier ore has already been mined. So every year the difficulty increases, and yet the world gets a fairly steady supply of iron ore -- the cost of production has gone up.

But technological progress as also made it cheaper to mine for the difficult ore over time. Competition between ore producers in order to get a competitive advantage invests to develop machinery and technology to obtain the ore faster and more efficiently. Even though the ore is more difficult to mine - the same ton of earth has less ore in it to extract - it is still cheaper to produce that difficult ore than it was to produce the easier ore using yesterday's technology, which may now be obsolete. Mining the difficult ore with picks and shovels would be fruitless.

What I am saying is that technological progress leading to greater efficiency in mining may outpace increases in difficulty, which on net will reduce the cost of production.

The price of ore in the world market IS the cost of production. Why? because there are many competing ore producers each trying to undercut the next guy. If your cost of production is say $1000 per pound of ore and mine is $900, I will start by selling it at $999. Somebody else that can produce for $900 will come in and offer $998. I will react with $997 and so on. The dance will go on until $1000.01 is reached (in theory of course). And what of demand? Well if there isn't enough demand then the price will drop below $900 and I will stop producing. If there is more demand, we will all be incentivized to expand our operations to produce more and that will lead to making our operations more efficient. It all stabilizes.
Bitcoin production is the same --- BUT this assumes that producers are selling what they mine. And that is a big But. If miners hoard it won't work.

The paper models a theoretical process, not reality. Does the cobb douglas production function or the solow-swan growth model accurate describe reality? of course not.

So you're saying that the cost of production does not determine price. Price controls production.

In which position is the horse in relation to the cart? You're not being very consistent.

Supply and demand controls price. Cost of production controls value (or call it break-even level)
As for the carts and horses, this is a huge matter of debate amongst economists - whether prices are set supply-side vs. demand-side.
Supply siders say the horse is production sets prices initially.
Demand siders say the opposite.

Your paper states that value (as defined by you) affects price, but you still have not described how. What is the mechanism?

Generally, a change in the cost of production affects the price because it shifts the supply curve. Is this your line of reasoning? Bitcoin production is different because the cost of production has no effect on the supply curve.

That's an impossible debate. You can't have a price a-priori demand or supply, so they are both demonstrably wrong.

If you don't have any supply of a thing then you can't have demand for it, and no market clearing price could exist.

If you don't have any people that demand a thing, then the supply doesn't matter, and no market clearing price could exist.

If I come up to you and want to buy your shoes, is the price determined by the supply of your shoes, or is it determined by my desire for them? The correct answer is that neither of those answers are correct. We would arrive at a price at which we both feel we have gained in the bargain, or no transaction takes place. That process is informed by objective factors, including the properties of the shoes, the general market demand for them, and so on, but it is nonetheless also based on subjective valuations of both parties (Or groups in a more macro sense.) that evades neat mathematical algorithms.

But that's a red herring. You seem like you're not willing to defend your assertion in the abstract of your paper that the value of bitcoins will likely trend toward zero as the cost of production approaches zero (Due to increased efficiency in mining methods.), so have you conceded that this idea is wrong?

Perhaps the break-even price is better thought of as a price floor for an individual miner:

If the value is 100 and the market price is 102 I can produce for 100 and sell for 102 earning a profit
If the value is 100 and the market price is 98, I stop producing.
If somebody else can produce at a lower cost they will continue to produce and sell until it reaches their break-even.
If nobody else can produce at a lower cost everybody will stop producing, and the difficulty will drop making it once again profitable to begin mining.

That is the mechanism.
It doesn't predict a price, rather it sets perhaps a lower bound.

If it costs next to nothing to produce on average then yes competition will drive the value down.
Right now if the average efficiency suggests a price around $200 but you get extremely cheap electricity from your own hydro-plant out in the backyard, then you will earn excess returns. If everybody gets extremely cheap electricity because they also have a hydro-plant out back then each will compete with each other to drive the price down. If it costs all of us $1 and the price is $200 I will offer $199, and then my neighbor will offer $198 and so on until it approaches $1.

If you don't understand the simple mechanism of competitive markets, then I can't help you understand much else.

Here's your competition: https://www.youtube.com/watch?v=K8kua5B5K3I#t=498
 8)

Miners will compete with each other to mine the limited number of bitcoins available to them. As long as the cost of mining is below the price of the mined bitcoins, they will devote more resources to mining more bitcoins. The result is that the cost of production as a whole will rise to the value of the mined bitcoins. In other words, it will never cost only $1 to mine $200 worth of bitcoins because somebody will be willing to pay $2 to mine those bitcoins.

Furthermore, newly mined bitcoins are only a small part of the market. The market behavior of miners will have only a small effect on the price of bitcoins. Even if miners gave away there 3600 BTC/day for free, it will have little effect on the the price of the remaining 200,000 BTC/day.

You've assumed your conclusion, also called "begging the question".

"Right now if the average efficiency suggests a price around $200..." So you already know the market clearing price and therefore anything that follows can't help us understand where the price "begins". The down bidding of someone with more efficient energy production in an existing market of prices doesn't help us understand where the initial price comes from; You wrote that it either begins with supply or demand. I wrote that it begins with neither.

To illustrate my point; In a chain consisting of two links, A and B, which link initiates the "Chain"?

Obviously the answer is neither. The concept of the chain arises in relationship to both links at the same time. It is the same way with market prices. Prices arise in a bidding process with two or more people with perceived (Real or imagined) supply and demand.

Emphasis mine.

http://mises.org/library/how-should-prices-be-determined

http://www.cavemannews.com/Supplysidevsdemandsideeconomics.htm

Stopped reading there and realised it's worthless, thanks for saving my time.

If mining cost nothing, then the cheapest way to get coin will always be mining, no one will buy, the cost of zero will bring the coin value down to zero. In fact, if the demand rises and there is not enough daily generated coins to satisfy that demand, the mining cost will always rise due to competition. Mining cost is a good indicator for demand, since mining usually is the lowest possible cost to get coins

Value comes also from the fact that it is difficult to do something. Many people spend millions to just win the price of a tournament, psychological motivation is a great part of the value foundation, and for bitcoin many enthusiasts mining at a loss just for fun

In fact there are lots of feed back loops in the valuation of bitcoin, it is even driven by religious behavior, can't use a single profit based model to analyze it. What is the profit to be your own bank? What is the price for realize a dream of freedom?

I appreciate your feedback and criticisms. I really do ! That is the whole point of inquiry and understanding.

I have made edits to address many of the concerns raised including referring to the cost of production as an intrinsic value, which is a shaky assertion at the least.
Instead, I wish to draw attention to the break-even models put forward as well as the fact the cost of production based on energy efficiency is what will largely determine the objective decision to mine or not mine. I still assert that as efficiency increases it will provide downward pressure on price, but acknowledges this is only a tendency and there are a number of other subjective and technical issues at work including the propensity to hoard mined coins instead of selling each one immediately in the market and the linear rate of block production, once every 10 minutes.
Also as an aside, I am presenting bitcoin production as commodity production and not as money issuance--
You can check the link at the beginning of this thread for the latest version.

How then do you explain the fact that while miner efficiency increased by a factor of 10,000 from 0.1 MH/J to 1000 MH/J, the price also increased by a factor of 10,000 going from $0.10 to $1000? Shouldn't the price have decreased by a factor of 10,000? You have provided nothing to support your assertion, but even if what you say is true, the effect must be so small that it can be completely ignored.

This is a very interesting phenomenon worth debating

In traditional economy, people tends to think that more efficiency means more production for given resource, and more production will bring more wealth. When many people first learn about bitcoin, they focus on how to produce it from a manufacturing point of view. Thus comes so many companies trying to build miners and mine coins. For these people, the exchange rate of bitcoin makes it a sell-able goods, they make bitcoin just like make an iphone

Then they learned that no matter how efficiency they are, the daily coin generation is fixed (the quick rise in difficulty and miner efficiency just showed how much excessive production capacity there is in today's society). Then mining becomes a tournament, it is getting harder and harder. As a result the reward gets more expensive, those who paid so much resource to get his reward will not easily sell his coin at a price level lower than his cost, it just does not worth it. They'd rather hold it and wait until the next reward halving to sell at a better price, this is especially true when bitcoin's long term potential can be as high as moon

This could be that for a long while at the beginning, bitcoin mining was NOT a competitive producer's market. There were also many less miners than now.  Until GPU rigs were cobbled together and set to mine in a dedicated fashion, CPU mining did not perhaps extract much more electricity than the computer was already using to run. Many have attributed the introduction of ASICs to the sudden spike and subsequent drop in market price.
Another, smaller factor was that the block reward was twice as large as it is now.

The model put forth only describes reality to any degree if bitcoin mining is a competitive process and that requires a great many competing producers.

And again, a tendency is just that. Real market prices may fluctuate violently around that tendency, and in fact may never get there.

How do feel about the fact that production costs can actually follow demand?

Look at gold, for instance, as more people store it away arbitrarily as a store of value, they arbitrarily increase the price of the item, and as this price increases the amount of money companies are willing to spend to extract an OZ of gold increases commensurately.

This shouldn't happen, accorded to your assertion that the price of something is informed by it's production cost. If the price of a product overshoots the production cost (Because of ignorant consumers.) then the price should gravitate back down. It doesn't.

Consider also that the production cost of gold can be arbitrarily high. There's no reason we couldn't be spending $1,000,000/oz for gold. This could be accomplished through highly inefficient or efficient methods, it makes no difference.

The only conclusion one can come to is that production costs do not inform the price of a product. Price is a function of various factors, not the least of which is the subjective valuation of it's consumers.

So, getting back to the subject at hand, it's impossible to say what the price of bitcoin should be based on it's production costs alone, even if I granted that production costs have any effect on bitcoin's price whatsoever (It doesn't, since "production costs" are simply the redistribution through inflation of the currency at a fixed rate.).

Agreed, if you see the updated paper it pays a lot of heed to subjective valuation, the fact that people hoard, sentimental or speculative reasoning etc. The low cost producer will earn excess profits if subjective demand drive the price up. What the cost of production ends up being is a lower bound, so let's say the opposite happens and there is very little market demand for bitcoin - the price will drop as there are few bids. After time, miners will just stop mining instead of incurring daily losses. Eventually supply will meet demand at whatever low price it may be. Obviously the supply of bitcoin itself is regulated at 25 per 10 minutes or so currently, so what I mean more accurately is the supply of miners will be reduced - and the lowest cost producers would be the only which remain. Bitcoin also has the feature of difficulty adjustment - so again assume the demand and price both drop causing miners to exit. The network hashpower will drop and it will take longer than 10 min. to find blocks on average, resulting in a lower difficulty. The lower difficulty has the same effect as raising the price in terms of $ x BTC/day. You can either raise the price or raise the # of BTC expected per day to induce miners to return.

The reason somebody wouldn't produce gold with a cost of $1M per oz. is they would quickly go out of business providing there are many other gold producers who operate at a marginal profit.

As for demand spurring investment this is almost certainly true - the issue is that it can lead to 'mal-investment'.  A good example is happening now with shale oil producers.