Bitcoin Forum

Miners running costs must pass trough exchanges and this will always be a downward pressure on the Bitcoin price.  Why must costs go trough the exchanges?  Because miners must pay their bills in currencies other than Bitcoins.

It's easy to estimate the combined running costs of all the miners, because they will always tend to the amount of total mining awards given.  This is due to the fact that mining is in a near perfect competition situation, and thus the profit margins will always be low.   

There are periods where profit margins could be temporary higher, like when the price of Bitcoin price jumps 30 times up, or when there are delays in delivering mining hardware, as was happening in 2013.   

So, what does this means for Bitcoin price?  It means that the higher the price of Bitcoin is, the higher the selling pressure becomes.  So, for Bitcoin to be a store of value, this must be offset with fresh buyers on the exchanges.

And this puts a limit on how high the price can go. 

Currently, this means that the total mining costs can go up to $3.6 million daily (price of $1000 per Bitcoin).  So, once we reach a somehow stable state, (where difficulty is not doubling each month), there will be a, perhaps $2 million of fresh cash needed to keep the price stable.

Also, when somebody says that Bitcoin can go up to $100,000 per Bitcoin (as Max Kaiser did), this means that there will be perhaps $200 million fresh money daily needed to prop up the price.   So, we can safely say that Bitcoin will never be that expensive.

Hi mestar.

According to BitPay blog (http://blog.bitpay.com/2013/09/bitpay-surpasses-10000-merchants.html):

According to my chart (http://divananalit.org/graphs/bitcoin-mining-cost-from-20130201-to-now.png), the current minings cost is about $100 per bitcoin. It's just an electricity cost.

And there are 3600 bitcoins generating every day.

So capacity to work for our crypto-currency costs $360,000 dollars every day, or $10,800,000 per month.

And we have just $6,400,000 of real operations per month.

Is that all a big ponzi-investment game or have I missed something?

Some people say that it is always cheaper to buy coin instead of mine. From a trader point of view it is true, but from a bitcoiner's point of view, it is very wrong

Every amount of hash power you put in the mining will strengthen the network and raise the cost of each newly generated coin. This is the fundamental support for bitcoin's market price. If everyone just bought coins since 2011 and no mining equipments, bitcoin would never rise to $1000, because every new comer can easily mine 1 coin per day with a $300 graphic card at 2011's difficulty, no one will buy coin at $1000 price

And miners don't need to cash out all the coins, if they hold mined coins, the coins supply on the exchange will dry up quickly (Speculator's sell were absorbed by the long term investors and mostly disappear from circulation), thus support the exchange rate. The operating cost will be so tiny compared to the bitcoin's value after a couple of years, so they most likely finance that through other means

The depreciation cost is also significant, it's not just electricity.  I did a quick calculation based on the numbers in the genesis block.  If 50% of the hash rate are 50GH/s singles and 50% are 400 GH/s Jupiters then it's about $145 / coin if you give the hardware a two year depreciation, and $250 for one year.

Though I'm not sure that makes it a ponzi scheme.  I don't understand why you're comparing the currency's value to its electricity cost.  No other currency is based on its electricity cost, so why should bitcoin?

BitPay transaction levels are irrelevant to the OP's analysis because commerce done with BTC-dollar conversion is neutral with respect to BTC and dollars quantities available on the market when people buy back their coins (which has become standard practice among the faithful for the explicitly goal of keeping valuation artificially high).

Say Bob starts with 1 BTC, Bob buys a Guitar from a merchant, the merchant uses BitPay which sells the coin on an exchange and gives the merchant $1000.  Then Bob goes onto the exchanges and spends $1000 to buy a single BTC.  The net effect is to move $1000 from Bob to the merchant and a Guitar from the merchant to Bob.  The value of BTC is irrelevant to the transaction and causes no net movement in BTC value because the same number of BTC and dollars are available on the market.

Mining though is a continual creation process that puts a downward pressure on prices IF the newly mined coins are allowed to reach the exchanges.  Currently most newly mined coins are being hoarded by miners out of the same kind of collective market manipulation that the Bob example demonstrates.  This can't be maintained forever, eventually miners will over-invest (they may already be) and difficulty will squeeze out nearly all profit forcing them to start selling.  That will begin a long bear market trend.


komer:  Your numbers look right to me (but I use 4k as a rough mining production per day during a period of exponential hash-growth, difficulty lags a bit and this means mining averages ~9 minutes per block), the dirty secret of BTC is that (on average) it dose not circulate.  Generally new coins get used at most once in a real transaction and then disappear into hoards.  This is all by design of the BTC crowd that actively promote the disappearance of coins into hoards so that the available pool of coins shrinks causing exchange rates to balloon on artificial scarcity.  Also BTC is best described as a Pryamid scheme mixed with a stock-bubble, it is not a Ponzi scheme.

I would say the total coins coming into the market = Mining supply + Divestment of existing holders.

Divestment of existing holders / per year - I would say is around 20%. (can be higher or lower that this).
Lets say there are 12 million coins = 
20% of this = 2.4 Million coins

Mined coins per year = 1.3 Million

Total Supply = 3.7 Million

at 1000$ price -> you need 3.7 Billion/year just to hold the price.


 

What difficulty increase are you using for those calculations? Assuming current difficulty increases (which are likely to hold true for at least a couple of months with current pre-orders) a 50GH/s single will generate less income than electricity cost by may 2014. Linky (http://thegenesisblock.com/mining/a/6a4f466ea8)

While i consider it unlikely that BTC will ever be that expensive that prediction highly depends on when you expect that price. If you fast foward to the year 2028 the block reward will drop to just above 1.5BTC. In which case the pressure from miners selling freshly minted coins at that price is reduced to a "mere" $22.5M per day.
Well, you have to put that figure in some kind of perspective so i´ll pull some random numbers out of my a** which you may or may not consider reasonable assumptions.
Assumption one, as a payment network bitcoin may not be more expensive than current existing options. PayPal fees exceed 2% of the transfered value, so lets take 2% as an upper limit (and conveniently ignore current additional costs like exchanges for now).
While transaction costs seem low for btc you basicly got to add the cost of freshly minted coins towards that number, since they are basicly the running overhead for network maintenance.
Since i got to choose some figures i will go for a 1% fixed fee and assume that the actual transaction fees are at least as large as the hidden overhead from freshly minted coins (that is, total network cost does not exceed the aforementioned 2%), so basicly the transaction fees are at least $22.5M as well, and since i already said that is 1% of the actual transaction value, i will assume a total daily transaction value of at least $2.25B.
For comparision, Paypal (using 2010 numbers) processed around $300M worth of transaction every day. That equated to 18% of the total online B2C commerce.
To achieve a value of $100k by 2029 a combination of the following has to happen:
-Online retail will surge
-Value of the dollar will decline more rapidly than currently
-BTC will be a/the dominant player in online commerce and/or achieve a strong foothold in offline commerce as well

So basicly, i dont consider that scenario very likely, ymmw, but i´m not saying its impossible.

Very true.  However, it is the electricity cost that has this unique property that it has to hit the exchanges if miners want to pay their bills.

Miners' hardware investments don't interfere with the exchanges, and thus are not influencing Bitcoin price at all. 

Bitcoins used in transactions, where a person buys them on an exchange, uses them to buy something, and merchant then quickly sells them, also don't influence the price much. (depending on the velocity of the whole process).

So, we are left with two other major factors in Bitcoin price formation:  speculators and miners.  Most of Bitcoins bought are a pure speculation, hoarded in a hope that the price will go up.  I would say this is the biggest one.  The other would be miners selling them to cover the running costs.  This one can easily go up 10 times, and start to put some real pressure on the price.

So, this is what Bitcoin is all about.  Speculation and mining.  Also, an interesting fact is that once Bitcoins are bought and "hoarded", they don't influence their price on the exchange (at least not directly.)  So, in a sense, speculators are free-loading on the whole network, making money if the price goes up with no costs at all.  This situation can be sustained for a little while, while fresh cash is coming in, or when the total network costs are not huge.  But in the long term, the only possible stable state is where the costs of the network are comparable to the value created by the Bitcoin network. 

From the thread https://bitcointalk.org/index.php?topic=404164.0 (https://bitcointalk.org/index.php?topic=404164.0) we have two estimates of current mining cost of one Bitcoin:

10% - $100 , Avalon #2 miner
2% - $20, the most efficient hardware (?)

So this means that the difficulty must increase 10 to 50 times before this current bubbly situation starts to sort itself out.  If it doubles each month, this means 3 to 7 months.

There is asymmetry of miners' costs lag in up market and down market.  In the up-market, it takes time for miners to get new hardware, and we have a situation where current miners are hugely profitable.  In an down-trend market, the least efficient miner will always be at zero profitability, and there will always be full mining saturation. 

In other words, when the price goes up, breaks on the price have a lag.  This makes Bitcoin very prone to bubbles, and this is exactly what we have seen so far.  We are in a third major bubble in less then 3 years.  Is this the never-ending feature of Bitcoin, bubble after bubble?

You are right, all this "costs will catch up with price" business and cost numbers are for the next 3 years.  Then half the amounts for the next 4 years.

Bitcoin the transaction vehicle can function, no matter what the price is.  You buy them today, the other party sell them tomorrow, it can work if the price is $1, or if the price is $100k per BTC. 

But if you want to "invest" in Bitcoin, a non yielding instrument (so by the future cash flow calculation its value should be 0), and with a hidden costs (that also increase together with the Bitcoin price, a truly unique property),  if you want to invest, you should be very concerned in what happens in next 3 or 7 years.  I predict speculators will be hit hard.



Currently, Bitcoin will not win against Paypal on price of transactions, since you can have two exchange transaction fees, costs of getting money into exchange, cost of getting the money out of exchange, and an actual Bitcoin transaction fee, (which can be zero).

However, Bitcoin will win in the markets that Paypal will not touch, like gambling, poker, sending money to Iran and others.  Those areas are auto-win for Bitcoin. 


I agree.

Only until reward per block will halve! ;)

That's just looking at today's cost.  It'll of course vary over the coming months, according to $/GH/s and W/GH/s of new hardware.

By reading all this, it can be concluded that Bitcoin as a way of paying, has little impact on the price. Would that ever be possible? I mean, would the Bitcoin price change if it would be used to pay employees and they would use it again to pay for groceries as the dollar or euro would?

Sure, the total amount of BTC circulating has to have a big enough value to pay for all trades done within a time frame. Simple example, if all coins together were worth the equivalent of $20M you couldnt buy a yacht worth $50M even if you had all the coins. The trade is simply impossible. If town had e.g. 100k inhabitants and every one get an average salary of $3k/month and spends all coins every month the total supply of coins must exceed a value of $300M. A somewhat more complicated calculation was done by the Bank of America. Basicly they calculated the value of a bitcoin under the assumption that one day 10% of the online trade is done in bitcoins, that would justify a value of $1300/BTC. However bitcoin is very, very far away from that goal. Current prices are imho far too inflated by speculation.

The idea this thread brings about has been on my mind alot lately. For other alt coins as well.

OP assumes every miner needs to offset his costs. I can only speak for myself, but mining was simply my way of purchasing Bitcoins. I had a certain amount of fiat which I wanted to convert into Bitcoin and I purchased mining gear (GPUs at the time) and then spent the rest on electricity. When the fiat ran out, I sold the hardware directly for Bitcoins.

So... I mined and never put any downward pressure on the exchanges. Quite the opposite, I removed (through evil hoarding) coins from circulation.

To grasp the total "downward pressure" that miners put on the exchanges, one would have to know the financial situation and intentions of every Bitcoin miner, which is quite impossible.

It's very hard to know how much BTC is used for payments, right? I know we can see how much BTC is send in total, but we don't know what the change is. If someone sends 10BTC, it could mean he just bought something for 9.99BTC. In that case the total BTC in circulation could be very small, and for now has minimal influence onto the BTC price.

No. You didnt apply any preasure in either direction. Upward pressure is buying. Hoarding is pressure neutral.

The situation is different nowadays. Simple put there are 4 possibilities.
Miner buys with fiat and does not sell or spend coins. "Old case", since e.g. Ati didnt sell graphics cards for BTC. No coins are bought or sold on exchanges, pressure neutral.
Miner buys with BTC and does not sell or spend coins. Initial downward pressure on purchase since the manufacturers needs to sell BTC for fiat as the supply chain (e.g. chip manufactures) and staff get (at least mostly) paid in fiat.
Miner buys with fiat and sells coins on exchanges (or buys goods via services such a Bitpay). Delayed downward pressure since the BTC can only be mined after the hardware has been delivered.
Miner buys with BTC and sells coins on exchanges (or buys goods via services such a Bitpay). Both initial and delayed downward pressure.
As you can see there is not a single upward pressure case.
The "best" case for the BTC price from mining is neutral.
I dont know the percentages for each of those case, your guess is as good as mine. But mining obviously generates downwards pressure.

As opposed to a user of the currency. Probably not too many of those yet.
User buys coins and buys goods via e.g. Bitpay. Initial upwards pressure from buying, later downwards pressure from purchasing. Sum is pressure neutral.

And the last case, the speculator.
Speculator buys coins and does not spend them any time soon. Creates a direct upwards pressure on the price.

As price for BTC rises the downwards pressure from the miners increases (assuming the general composition does not change, that is the same percentage of miners still sell coins) and the upwards pressure from speculators decreases since the same amount of fiat invested will now buy less coins. And vice versa if the price for BTC drops.

And well, since your guess on the intentions of miners is probably as good as mine (10% of the miners selling? 30%? 50%? Maybe even 80% of total freshly mined coins?) you can directly calculate the required upward pressure required to keep the price merely stable. And dont forgot that a large amount of the mining pressure is delayed pressure, mining equipment worth three digit million dollars still waiting to be shipped ...

Blockchain.info has a chart (https://blockchain.info/charts/estimated-transaction-volume) that tries to guess which part of the transaction is the change and which part is the actual transfer. I have no idea how good their guess is though.

If a miner wants to cover his running costs of mining, he can do two things: 
-sell part of his mined coins
-pay the bills himself

In both cases, the same amount of fresh cash is needed.  In your case, you decided to pay the bills yourself.



We would also have to know the price of electricity for each miner. 

However, sometimes total numbers are easier to estimate.  For example, we know that the total electricity used can't be more than 3600 BTC daily.  We know this because people are not stupid.

We also know that it can't be 360, at least not for long, because, then you have a magic machine that takes 1 dollar of electricity and turns it into 10 dollars, with no other inputs.  Everyone would want this money printing machine, and soon your 1 dollar of electricity will turn into 1.1 dollar.  Mining can never be profitable for long, because, free money.  And soon, trough the magic of ebay, all the hardware will move to the regions where electricity is cheaper.

You're expecting everyone to be reasonable, which is a bad assumption.

I had a poll before, you can see that more than half of the miners hold their coin at least for a year
https://bitcointalk.org/index.php?topic=296264

And for traders, similar distribution
https://bitcointalk.org/index.php?topic=295753.0

Would be interesting to see if that distribution has changed due to price rising, when the price was 100ish i only sold half the mined coins, now i sell all.

And the first answer there indicates that price of BTC heavily influences that decision.

But anyway, those are figures that can be used for calculation. So, assuming 30% of the freshly minted coins are sold that equates to roughly 32k coins being sold per month, or, at current prices about $30M. Gox is responsible for 20% of the BTC trades currently, assuming the fresh coins get sold evenly that would mean 6k BTC being sold on Gox, which would cause the price to slip from 952 to 848 within a month.
That is assuming no new speculators ofc.

Actually, I can pay bills with bitcoin, and now I can even pay taxes with bitcoin, thanks to snapcard.  Your leading assumption requires reassessment in the present environment.  You should also remember the halving in 2016 and 2020.

I'd be interested to see a blockchain analysis of the turnover rate of new coins.  That would provide an evidence-based assessment of exchange stability.

Of course price stability does not strictly depend on marginal demand at exchange.  PQ=MV.  Even if no fiat inflows, as long as PQ/V grows proportionately to issuance, prices will be approximately stable.

I'm inclined to believe the rate of liquidation is even lower then 30% given the way the market moves and how tight it has become, Gox generally has a coin depth equal to ~3-6 days worth of mined coins now a days.  A poll on the forums may be missing large 'institutional' miners that I'd suspect to be even more holding oriented then the forum visiting and poll taking membership.

Ultimately a block-chain analysis would be the ideal way to answer the question as it's possible to know when coins move through an exchange you could find out the average period that coins wait and what fraction move and even if they move in response to market swings.  I would not be surprised if smart market watchers already do this but so far I've seen nothing in the more academic analysis papers about answering this rather vital question.

I haven't seen this mentioned so I'll just throw it out for consideration:

It seems likely that the income from large scale miners, the fees from the biggest exchanges, and the charges by the large mining pools, will not be sold on any exchange.  The most likely assumption is that they are selling to big buyers over the counter.  These 'investors' don't intend to sell in the short term.

Net effect (given demand):  upward pressure on prices.

Mining is buying bitcoins.

And your Snapcard providers sells those bitcoins on Coinbase.  So, tell me again, which assumption do you think I should change?

You seem to be lost in unimportant details for this feedback mechanism work.  On exchange, off exchange, doesn't really matter, if miners want to sell, that amount of fresh cash is still needed.

Those polls are price-irrelevant, it is a general strategy of miners, the higher the price, the higher the difficulty, the average mining income actually dropped. But old coins indeed worth more when price rises, so when a price spike happens, many old coins will get sold. Once they are sold, they will spread into more people's hands, thus further reduce the future sell pressure

FED is printing 2.8 Billion USD every day, there are enough amount of fresh cash, the rise of bitcoin price is unstoppable before they stop QE

No, its creating bitcoins. Which these days means buying or renting mining equipment. The mere act of mining does not change the price in any way.
Or is mining gold the same as buying gold on the market?

The 2016 block difficulty adjustment attempts to keep the supply of newly found bitcoins steady.

If I want to obtain Bitcoins I have basically two options. I can purchase them from someone who already has a supply (previously mined bitcoins), or I can purchase hashing power and pay for electricity in an attempt to find block rewards. Either way, I'm trying to obtain bitcoins and either way it has an effect on the exchange rate.

The rotation of the earth causes periods of darkness known commonly as "night".
Factually correct and just as irrelavent.

No, buying increases price. Buying mining equipment increases difficulty.
If buying mining equipment had a direct effect on the exchange rate (i assume you imply going up there as well) then even cpu mining would still be profitable, since buying mining equipment just increases the price of bitcoins. While that would be really cool for miners i dont really think thats how it works.

The only coins that affect the exchange rate are those put up for sale.

The block reward, which tries to happen every 10 minutes, are coins that are up for sale.

This is a common misconception.  Those coins are only "up for sale" when they're put on an exchange and their sale is advertised or immediately executed.  Think about it.  Coins going into a hoard have no effect on the market price.  Coins not up for sale are not part of the economy.  Off-exchange sales have an indirect effect on market value, but I don't think you're arguing that.  I'll say it again: As far as the current market value is concerned, coins going into a hoard might as well not even exist.

If a coin is being hoarded, it isn't being put up for sale. I don't see how you can argue that this doesn't affect the exchange rate.

Holliday/traderCJ :  I think your disagreement is coming from what your assumption of 'baseline' is.  CJ is operating on a 'present market supply of BTC and USD' as his baseline so an activity which doesn't add or subtract from the market is neutral to him.  Holliday seems to me assuming a baseline of 'all coins' liquidated or on the market somehow, so hoarding anything is a reduction of available coins from that baseline.

Thing is that the history of BTC has ALWAYS been one in which most coins are hoarded very quickly.  The distinction were talking about is if they miners are the chief hoarders or if they are just operating a narrow margin business that supplies coins to a market from which hoarders make purchases.  In the GPU era I think the former kind of miner was dominant, the barrier to entry was low, the ongoing power cost was the primary expense and the miner liquidated his mining output immediately to lock in profits.  Now you have big barriers to entry and infrastructure dominates so the miner today is more likely to be a self-hoarder or at the least to be in a much longer-term repayment horizon so they hold coins longer because deflationary gain is necessary to make the mining process profitable. 

This is a much more tenuous basis for a market then the old GPU era, that market badly overbought equipment and the new dynamic is even MORE conducive to over-investment, without the Chinese bubble starting in November we would already be deep into a collapse in which miners would be seeing no ability to ever break even on their machines.  China pushed back that data a few months but it would take a never ending bubble to hold it off indefinitely.

That really sounds a lot better than miners got screwed on pre orders because they didnt expect nearly as many people ordering mining rigs.
However cutting losses is just as valid a tactic as hoping for the best case (further price increases).

Thanks Impaler, good info.  I am warming up to your line of thinking on Bitcoin.