Im not sure how bitcoin fees will be sustainable long-term.

[Decoded]

So i'm just wondering. I may be mistaken.

I've always thought that the bitcoin fee will increase linearly with each reward halving. So let's fast forward 100+ years to 2140, when there are no more coins being produced. Wouldn't the fee have to be somewhat near what it is today? Around 15 thousand dollars per block?

If we were to divide 1 megabyte (1000000 bytes) by a basic transaction size, maybe 300 bytes, we get 3333 transactions. Divide $15000 by 3333 and you get 4. Which is very high...

Am I mistaken?

[1714873280]

1714873280

[1714873280]

1714873280

[franky1]

yes and no

average transactions per block is more so 2500 at the moment
over the next few decades (and the last decade) technology changes and we should be able to expand transactions allowed per block that way cost per transaction is less. while also opening up more utility of bitcoin.

so far that has been delayed by alot of hostility and people thinking fee's need to go into panic mode rather then natural growth of transaction capacity.

say we did in 2013 move to 2mb..
say we will in 2017 move to 4mb..

thats obviously 2013=5000tx, 2017=10,000tx
then 2021=20,000tx, 2025=40,000tx, 2029=80,000tx
2032=160,000tx..

so in 16 years a tx would cost 10cents.

i know your thinking OMG 160,000tx bad bad bad..

but look just 16 years ago.
a hard drive was just 4gb MAX
internet was 56k MAX

today people can get harddrives 1000x bigger(4TB)
people can get internet 1000x bigger(56mb fibre)

so its safe to assume that over the next 16 years things will get bigger again.
imagine it complaining about the 2016 Call of Duty game needing 100gb of storage... but from the prospective of the year 2000
"WTF 100gb, its impossible, no one will ever play that game. its too big. it will only work on some data server"

yet millions of people are playing it.

so although blocksize needs to only double 6x.. and technology has been known to grow 1000 times.. there is no actual issue.
but hey. instead some cave dwellers want to hinder transaction utility/growth and cause a fee war now for payments that is not even an issue for DECADES

ask yourself. which is better
160,000 people paying 10cents each ($16,000)
4500 people paying $3.55 each ($16,000)

[AgentofCoin]

So i'm just wondering. I may be mistaken.

I've always thought that the bitcoin fee will increase linearly with each reward halving. So let's fast forward 100+ years to 2140, when there are no more coins being produced. Wouldn't the fee have to be somewhat near what it is today? Around 15 thousand dollars per block?

If we were to divide 1 megabyte (1000000 bytes) by a basic transaction size, maybe 300 bytes, we get 3333 transactions. Divide $15000 by 3333 and you get 4. Which is very high...

Am I mistaken?

1MB Block space limit will very likely not exist in 2140. SegWit alone will increase it to about 1.7MB in 2017.
As time goes on and technological advancements and efficiencies come about, the block limit will be raised to
higher levels that will be deemed safe, so that the network will be able to remain secure and decentralized as
well as scale for increased tx growth.

If by 2140 the block limit is still at 1MB, it will most likely be due to Bitcoin being dead and abandoned.
By 2140, it is highly likely a 1GB block will be considered a very small block.

[DannyHamilton]

- snip -
Around 15 thousand dollars per block?
- snip -

$15,000 per block?  Where did you get that number?

Right now the block subsidy is 12.5 BTC per block.
At the current exchange rate of about $715 per BTC, that's only $8937.50

The current transaction fees are about 0.64 BTC per block.
At the current exchange rate of about $715 per BTC, that's only $457.60

That's a total of block reward of only about $9395 (not $15,000).

You are also assuming that the total reward today is the minimum reward necessary.  It's quite possible that the reward today is MUCH more than it needs to be.  Perhaps in the future (as the block subsidy shrinks), we'll discover that the reward only needs to be about 20% of what it is today to support enough mining to keep bitcoin secure.  In that case, the fees would be only $1880 per block.

[QuestionAuthority]

Theoretically, if Bitcoin becomes widely used and survives that long, small percentages of a Bitcoin will be the commonly used denomination. mBTC and μBTC will be used like whole bitcoins are today. It will be very uncommon for anyone, except wealthy individuals, to hold a whole Bitcoin. The price of a single Bitcoin will be incredibly high. In all likelihood, businesses that rely on Bitcoin will take over mining. Most of these businesses will either mine Bitcoin to supplement their income or mine it simply to keep their primary business alive. This would still be a decentralized system where no single company would be trusted to keep the entire network alive. Chain pruning and other efficiencies developed over time will improve the size of the stored database.

You and I will never need to worry about this though because we'll be dead.

[mobnepal]

Transaction fee may not increase linerealy with each reward halving if price per bitcoin get increased significantly over time which is also quite likely to happen in coming years due to high demand. Cost of mining is calculated by miners on USD so most of miners try to choose recommend fee based on price of bitcoin as well as block reward. So over time when price will be high transaction fee in bitcoin may be lower in amount but fee when compared to USD may get increased to recover mining cost with decreasing reward.

[franky1]

$15,000 per block?  Where did you get that number?

not sure why dannyhamilton wants to knit pick an arbitrary number..

i think the OP was thinking of THE FUTURE.. not today and came up with that number maybe because in
2012: 50btc at $6 was only $300 a reward
2012: 25btc at $20 was only $500 a reward
2013: 25btc at $120 was only $3,000 a reward
2014: 25btc at $240 was only $6,000 a reward
2015: 25btc at $360 was only $9,000 a reward
2016: 25btc at $480 was only $12,000 a reward
2017: 12.5btc at $1200 <-maybe which will amount to $15,000 a reward
future: as danny hamilton meant to have said.. stagnates or doesnt rise as much (not sure why he thinks it will drop)

the reward due to deflationary would either be stagnant or rise. so $15k is an acceptable number.
i see no issue with using $15,000 as a base arbitrary number.
although i see it being higher. definitely not lower

i see more of a knit pick of why argue about an arbitrary number by countering it based on a single point of data.. but thats just me

anyway for the OP
try not to base a future of a couple decades->century on numbers of a single valuation today.. look to the past and then speculate the future based on multiple data points/possibilities

so as time goes bitcoins deflationary fiat valuation goes up and the btc amount goes down so that over time the total reward changes.
eg
when the reward is 6.25 the valuation maybe $2,000/btc = $12,500 reward - add on some fee's to get to $15k
when the reward is 3.125 the valuation maybe $4,000/btc = $12,500 reward - add on some fee's to get to $15k
when the reward is 1.5625 the valuation maybe $8,000/btc = $12,500 reward - add on some fee's to get to $15k
when the reward is 0.78125 the valuation maybe $16,000/btc = $12,500 reward - add on some fee's to get to $15k
then we may see the valuation slow down and/or stagnate
when the reward is 0.390625 the valuation maybe $25,600/btc = $10,000 reward - add on some more fee's to get to $15k
when the reward is 0.1953125 the valuation maybe $38,400/btc = $7,500 reward - add on some more fee's to get to $15k
at which point the reward:txfee are 1:1 with an arbitrary $15k the OP picked

which in 24 years time, its possible bitcoin can be worth near $40k each and possible for transaction capacity to cover the other $7,500 without issues due to only needing 80,000 tx (32mb blocks). which personally i can see possible in 16 years, so 24 years is no problem

where miners then concentrate on fee's as the main income after that point. where the reward is then and only then thought of as the subsidy(bonus)
until 2140 when the tx fee becomes the only income.

of course no one can truly estimate when the flip of income:subsidy goes from reward:txfee  to   txfee:reward. but 16-30 years is a good safe timeframe to concern ourselves with.
as oppose to now now now or 75-120 years

[Kakmakr]

Can someone possibly work out, when the break even point would be, when the miner reward & fees will not be enough to sustain the cost of mining. I have heard that +/- $200 per Bitcoin is about the breakeven point for mining to still be profitable at today's exchange rates.

I know there are a lot of variables, but it would be interesting to have a broad idea, even a guess at which point mining would not be profitable anymore. < Work on averages, if you really need to > 

[DannyHamilton]

Can someone possibly work out, when the break even point would be, when the miner reward & fees will not be enough to sustain the cost of mining. I have heard that +/- $200 per Bitcoin is about the breakeven point for mining to still be profitable at today's exchange rates.

I know there are a lot of variables, but it would be interesting to have a broad idea, even a guess at which point mining would not be profitable anymore. < Work on averages, if you really need to >  

There is no break even point.  Mining will always be profitable for someone unless everyone chooses to mine with unprofitable equipment in areas with unprofitable electricity rates.

As the exchange rate drops, mining becomes unprofitable for those that have the highest equipment and electricity costs.  Those people with the highest costs stop mining (since they are losing money), and the difficulty drops. The lower difficulty increases the profitability of everyone that is still mining, so they continue to mine.

If the exchange rate drops more, then mining becomes unprofitable for those of the remaining miners that have the highest equipment and electricity costs. Those people with the highest costs  stop mining (since they are losing money), and the difficulty drops. The lower difficulty increases the profitability of everyone that is still mining, so they continue to mine.

If the exchange rate drops more, then mining becomes unprofitable for those of the remaining miners that have the highest equipment and electricity costs. Those people with the highest costs  stop mining (since they are losing money), and the difficulty drops. The lower difficulty increases the profitability of everyone that is still mining, so they continue to mine.

Taking this to an extreme, if the exchange rate drops enough then we will be left with only 1 person mining. They are getting EVERY block (144 block per day).  If the exchange rate is $1.00 per BTC and the block reward is 0.01 BTC per block (about 56 years from now), they will be earning about $43.20 per month.  If they are mining with a single Antminer U3, and they pay $0.11 per kWh they will be paying about $5.06 per month in electricity.  That's a profit of $38.14 per month.

That means that even at an exchange rate of $1, there is enough revenue for as many as 7 people to all mine profitably at $0.11 per kWh with an Antminer U3. They'd be paying a total of $35.42 in electricity, leaving 7.78 in profits to split between the 7 of them.  That's a monthly profit of about $1.11 per miner with 7 miners.

[Kevin77]

So i'm just wondering. I may be mistaken.

I've always thought that the bitcoin fee will increase linearly with each reward halving. So let's fast forward 100+ years to 2140, when there are no more coins being produced. Wouldn't the fee have to be somewhat near what it is today? Around 15 thousand dollars per block?

If we were to divide 1 megabyte (1000000 bytes) by a basic transaction size, maybe 300 bytes, we get 3333 transactions. Divide $15000 by 3333 and you get 4. Which is very high...

Am I mistaken?

No, you are right. I speculate when all the bitcoins will be mined out, miners will continue their process just for tx fees which will be still sum up more than 50 or 100 bitcoins or even more this will be possible even people pay 1 satoshi per kb as there would be huge number of tx and one block sizes up to 1 GB too.

[tertius993]

So i'm just wondering. I may be mistaken.

I've always thought that the bitcoin fee will increase linearly with each reward halving. So let's fast forward 100+ years to 2140, when there are no more coins being produced. Wouldn't the fee have to be somewhat near what it is today? Around 15 thousand dollars per block?

If we were to divide 1 megabyte (1000000 bytes) by a basic transaction size, maybe 300 bytes, we get 3333 transactions. Divide $15000 by 3333 and you get 4. Which is very high...

Am I mistaken?

No, you are right. When all the bitcoins will be mined out, miners will continue their process just for tx fees which will be still sum up more than 50 or 100 bitcoins or even more. It will be possible even people pay 1 satoshi per kb as there would be huge number of tx and one block sizes up to 1 GB too.

I'm not sure people will pay 1 satoshi per kb as 1 satoshi is the minimum you can possibly pay. It would be interesting to work out a viable ratio between fees and amount sent (1:100 on average maybe?) and then model the impact on fees and transactions as the price per bitcoin increases.

I suspect (but can't readily articulate) that it would show a practical price limit above which BTC couldn't rise, without significant developments.

[xdrpx]

There would be many factors to consider if the Bitcoin fees far ahead in 2140 would even be considered costly cause there's the monetary value that matters as well. A transaction fee of 4$ far ahead then would mean very very little. Just consider back in the 70's and 80's it was harder to earn 1$ than now, and thus far ahead then it would be much easier to earn $4 and this wouldn't be much of a fee to pay.

Also, applying Moores law, Hardware technological advancements keep happening and there would be new mining equipment capable of performing much faster computations than today, which in turn would also be at lower electrical consumption, costs and probably considering the higher value of Bitcoin then would also mean that miners would still be earning a lot more. Since Bitcoins would ultimately be scarce to produce, it's price might as well increase sharply over time.

[rapazev]

Can someone possibly work out, when the break even point would be, when the miner reward & fees will not be enough to sustain the cost of mining. I have heard that +/- $200 per Bitcoin is about the breakeven point for mining to still be profitable at today's exchange rates.

I know there are a lot of variables, but it would be interesting to have a broad idea, even a guess at which point mining would not be profitable anymore. < Work on averages, if you really need to > 

that's impossible to predict... if the price keeps raising it will always be profitable in some places(electricity price) using the right equipment.
even when there's no more reward per block and only fee's, if the price is high enough it will be profitable.

another thing, we are talking about bitcoin how it is today. there are development going on and in 10~20 years everything will be possibly different.

[franky1]

Can someone possibly work out, when the break even point would be, when the miner reward & fees will not be enough to sustain the cost of mining. I have heard that +/- $200 per Bitcoin is about the breakeven point for mining to still be profitable at today's exchange rates.

I know there are a lot of variables, but it would be interesting to have a broad idea, even a guess at which point mining would not be profitable anymore. < Work on averages, if you really need to >  

there is a part of speculation called a resistance point. this is the lowest point the price can get to before the majority of people refuse to sell any cheaper.
though people in general buy and sell at different prices (speculative area) there is always a bottom line where the majority resist selling.
after looking at the charts and doing some other maths based on known manufacturing costs(not retail) of asics, electric etc, and putting those numbers against the hashrate. and also doing this for the historic data.

right now the resistance point is over $480
EG
in 2011, speculation went up to over $32 but the resistance point was about $2
in 2013/14, speculation went up to over $1000 but the resistance point was about $200

this year speculation is at over $700, but resistance is about $480.
prior to the halving, it was around $600 speculative high, and $300 resistance point.

so now you know the area of speculation vs resistance (profitable area). you then have to look at the mining specifically.

some farms like antpool have no upfront costs. yes thats right zero.
this is because the manufacturing costs of an asic(bitmain) is far lower then the retail price they sell to competition.

its been worked out bitmain for every asic they sell retail, can actually produce 4-5 rigs for that single unit retail price.
they can either
hand one to the customer and keep 3-4 to themselves to run.. and then use the reward for paying bills
or
hand one to the customer produce 2-3 for themselves to run and then put ~$400 towards electric (2 rigs electric over 6 months ~$400)

basically it boils down to electric cost.


which i done the maths ages ago in regards to the bitmain S9 (13thash, 1.3kw/h) unit

one asic
at 20cents/kwh for 6 months electric = $900
at 10cents/kwh for 6 months electric =$450
at 5cents/kwh for 6 months electric = $225
and the $1600 to buy an asic

anyway, im now going to use round numbers just to get the point across ASAP. (someone else can be more refined if they want)

so knowing asia is about the 5c/kwh and antpool costs $0 for the asic. makes the 'cost' of mining by bitmain/antpool $225 per asic required ATMOST over an expected 6 month period. (i say atmost due them able to repay the electric at no cost thanks to retail income. but lets stick with $225 costs)

at 125,000 asics antpool have. and a ~18%(over 6 months) of blocks solved by antpool = 4700 blocks =~59000btc
which makes 0.471744btc per asic
which at current price value is an ROI of over $330 per rig right now.(based on a $700/btc valuation)
which is profitable compared to their $225 'cost'(electric explained earlier)

but say an american miner buying a $1600(retail price) and then paying average american electric for 6 months $450(10c/kwh) totals $2050 upfront.
but as i have just worked out the average btc earned over 6 months per rig is  0.471744btc ($330 per rig right now, based on a $700/btc)

so americans are not profiting. infact they are at a 83% loss right now.

summary.
throwing all the numbers together..
for antpool (lowest cost) to remain in profit btc has to be 1btc=$476 (where the mining income of 0.471744=$225 cost)
but as i say the costs can be lower if they prepay the electric via the competition buying the rigs

americans(retail purchase priced rigs) will only break even if bitcoin was $4345.

so asia can sell right down to $476 and break even
america wont sell and wait for speculation to jump to well over $4000 before selling. or be dumb and sell at a loss.

hope this helps

[arransiv]

I don't think it's high because after every halving the miners have to earn more in some way so I think that's the perfect thing, after 100 years miners will have something good to earn.

[clickerz]

I don't think it's high because after every halving the miners have to earn more in some way so I think that's the perfect thing, after 100 years miners will have something good to earn.

It is ok for me as we need also miners to process transaction. We need them in order for the  system to run efficiently. Compared to other money transfer, so far bitcoin is the cheapest.