Analysis of Buying a Rig for Mining

[Gameover]

Here is an analysis of a 3x5850 mining rig over the next year:

assumptions:
15%   bitcoin value increase per 10 days
50%   difficulty rate increase per 10 days
1050   hash rate
$2.02   power cost/day (700w @ 0.12/KWh)

As you can see even if the value of bitcoin increases to over $1200 in a year, you are not even going to make your money back on a dedicated rig, at 90 days you are going to start losing money!  This is using a 50% difficulty bump when it is currently around 70%.  If you use 70% you start losing money in 60 days.  Even at difficulty rate increase of 50% and bitcoins increase 30% per 10 days (putting them at a value of over $100,000 per coin in 1 year!) you will still lose money at day 160 with a maximum income of $976.  Subtracting your rig would be a profit of $300 or so, I think its definitive, the ship has sailed on buying rigs for mining.

So what if difficulty slows considerably to say 30% per 10 days (less than half of the current rate)?  Assuming a bitcoin increase of 15%/10 days (value of $1225 in 1 year), you will start losing money at day 190 with a total profit of $1124, hardly enough to be worth the risk.

If you're not doing a simple calculation such as this before buying a rig you should seriously reconsider buying one at all.

So what about more efficient rigs, free power, ASIC, and FFPGAs?  They don't matter, even assuming a power cost of $0 your returns on coins are so minimal after a few months due to the difficulty rating it won't be worth it.  Your cost to simply replace hardware will outweigh the returns.

And the nail in the coffin, "Every 210,000 blocks, the Bitcoin reward per block is cut in half. Right now, the payout is 50 Bitcoins per block. Sometime soon, the payout will halve to 25 Bitcoins per block."  None of this analysis takes this into consideration.

So in conclusion something has to give:
- people will go bankrupt with new rigs, dropping the difficulty rating
- bitcoin will fail altogether and crash
- the value of bitcoin will skyrocket

Hard to say which, maybe all of them a little, maybe one of the a lot.

[1714855588]

1714855588

[1714855588]

1714855588

[nazgulnarsil]

why is the first difficulty increase at 81%?  is that the current prediction?

[Gameover]

why is the first difficulty increase at 81%?  is that the current prediction?

I assume you mean the 440,000?  That is happening in 2 hours, already a known fact.  The estimate is actually 434734 right now.

[nazgulnarsil]

I think extrapolating the difficulty curve that far is specious.  the difficulty will taper off HEAVILY if profitability goes down that much.

[JJG]

I think extrapolating the difficulty curve that far is specious.  the difficulty will taper off HEAVILY if profitability goes down that much.

No one is going to argue that difficulty will continue increasing once mining nears the point of being unprofitable.

But it doesn't matter, because at that point it will be unprofitable or minimally profitable.

[Horkabork]

Awesome work, but you're also assuming that the miner cashes their bitcoins into dollars the instant they get them. If bitcoins truly are increasing at 15%/10 days, then it would behoove the miner, if he's fine with the additional risk, to keep bitcoins for the duration.

In other words, compound interest needs to be in there, if it isn't already. That first week's money of $172 will become $528 in 80 days according to your own assumptions.

Yeah, your running total will still peak at the same time, but it'll be a lot higher.

[Freakin]

if you believe BTC will increase at 15%/10 days then doing anything besides buying BTC directly is an awful idea

[Gameover]

I think extrapolating the difficulty curve that far is specious.  the difficulty will taper off HEAVILY if profitability goes down that much.

It doesn't matter, even at 30% difficulty increases it is still a bad investment.

Awesome work, but you're also assuming that the miner cashes their bitcoins into dollars the instant they get them. If bitcoins truly are increasing at 15%/10 days, then it would behoove the miner, if he's fine with the additional risk, to keep bitcoins for the duration.

In other words, compound interest needs to be in there, if it isn't already. That first week's money of $172 will become $528 in 80 days according to your own assumptions.

Yeah, your running total will still peak at the same time, but it'll be a lot higher.

If you aren't going to cash in your bitcoins then why mine at all, just simply buy bitcoins.  Also I assume cashing in all bitcoins because I think that is what most would do, cash in until at least your original rig was paid for.

if you believe BTC will increase at 15%/10 days then doing anything besides buying BTC directly is an awful idea

Doing anything besides buying BTC directly is an awful idea no matter what the scenario: if the rate of increase for BTC is lower you are never going to recoup your rig cost, if its higher you should have bought BTC directly even more so.

[ensign_lee]

Do you mind uploading this spreadsheet to googledocs or something similar? I'd love to use it and plug in my own numbers

[Gameover]

is there somewhere i can post it to a public google doc?  or if you create an account for it ill post it there.

its pretty simple, here are the formulas for row of day 10, use the image to put in the starting values and your good to go

10   =B49*(1+$B$41)   =$B$42/B50*1000*$B$43   =D49*(1+$B$40)   =D50*C50   =E50-$B$44   =F50*10   =G50+H49

starting values
0.15   bitcoin value increase
0.3   difficulty rate increase
1.01   coins/difficulty
1050   hash rate
2.02   power cost/day

[Freakin]

Doing anything besides buying BTC directly is an awful idea no matter what the scenario: if the rate of increase for BTC is lower you are never going to recoup your rig cost, if its higher you should have bought BTC directly even more so.

Sorry, I was referring to people who are justifying their unprofitable mining rigs by saying "but bitcoin is going to go up so high!"

[Gameover]

Sorry, I was referring to people who are justifying their unprofitable mining rigs by saying "but bitcoin is going to go up so high!"

Yeah with current difficulty increases even if bitcoin increases to $100,000 per coin you still are only going to make a few hundred dollars at most.  Along the lines of what JJG said, even if difficulty does taper off or even decrease, your profits are going to be so minimal its not going to be worth it.

The reason for the huge boom in the last month is GPU mining and news outlets, after that levels off there will be no other similar event, even if ASIC and FFPGA had 3 orders of magnitude better performance like the jump from CPU to GPU, because of 2 reasons: 1) the difficulty will be so high it won't matter, and 2) no one has a ASIC or FFPGA like they already have a CPU and GPU, using one of those is a 100% direct speculation on mining, whereas now everyone and their monkeys brother has a CPU or GPU to dunk into the mining pool.

[w128]

Doing anything besides buying BTC directly is an awful idea no matter what the scenario: if the rate of increase for BTC is lower you are never going to recoup your rig cost, if its higher you should have bought BTC directly even more so.

I think there are two things that keep people from doing this:

1.) People are simply enamored with the idea that mining is creating money out of thin air.

2.) Mining is easy relative to trading. If everyone had the capacity to intelligently buy and sell BTC for profit they'd already be getting rich off traditional markets.

[bcpokey]

This again... I'm going to pick on a few things to start with:

"sometime soon the bitcoin reward will halve", yeah if you consider 2013 soon.
"$2.02 per day (700w @ .12/kWh" -- 3 5850s use absolutely nowhere near 700w. 3OCd 5870s use about 600W from the wall, 3 5850s will use closer 500W or less.
"You will start losing money on day xx" -- You stop mining when it is no longer profitable, no one put a gun to your head to keep going. You can also cash out of your rig if you really don't think you'll ever mine again and have no use for it, recouping even 60% of your initial investment (a VERY conservative estimate if you drop out 90/160 days after building) puts your profit back up > $1000. Not everyone is trying to replace their job with mining, that is a hefty ROI. Find a traditional investment schema that can offer 70 - 100% returns?

As has been mentioned no one needs to cash out immediately, power bills are delayed by 30 days, the minimum time for a forced cash out (or I suppose when the CC bill comes in, however if you purchase on the cusp of your billing date you have 30 days before the account closes, then 30 days to pay, so a 60 day window before any interest accrues on that bill).

Again, I'm not encouraging people to build mining rigs, I agree that the profit is really not there anymore, but your numbers are as flawed in this thread as they were in the previous one I responded to, skewed only to support your assertion that other people stay away, which I feel is somewhat disingenuous at worst, sloppy at best. A better analysis would be why investment and encouraging further investment will yield longer term gains, but that is a lot trickier I suppose.

[w128]

This again... I'm going to pick on a few things to start with:

"sometime soon the bitcoin reward will halve", yeah if you consider 2013 soon.
"$2.02 per day (700w @ .12/kWh" -- 3 5850s use absolutely nowhere near 700w. 3OCd 5870s use about 600W from the wall, 3 5850s will use closer 500W or less.
"You will start losing money on day xx" -- You stop mining when it is no longer profitable, no one put a gun to your head to keep going. You can also cash out of your rig if you really don't think you'll ever mine again and have no use for it, recouping even 60% of your initial investment (a VERY conservative estimate if you drop out 90/160 days after building) puts your profit back up > $1000. Not everyone is trying to replace their job with mining, that is a hefty ROI.

As has been mentioned no one needs to cash out immediately, power bills are delayed by 30 days, the minimum time for a forced cash out (or I suppose when the CC bill comes in, however if you purchase on the cusp of your billing date you have 30 days before the account closes, then 30 days to pay, so a 60 day window before any interest accrues on that bill).

Again, I'm not encouraging people to build mining rigs, I agree that the profit is really not there anymore, but your numbers are as flawed in this thread as they were in the previous one I responded to, skewed only to support your assertion that other people stay away, which I feel is somewhat disingenuous at worst, sloppy at best. A better analysis would be why investment and encouraging further investment will yield longer term gains, but that is a lot trickier I suppose.

The precision of this analysis isn't really important.  evidences the most important point which is basically "No, miners don't guarantee a static return. They are not money printers."

The people who understand depreciation, interest, borrowing and resale which your argument assumes aren't likely to be the same people trying to figure out how build their first miner.

Also, "You don't have to pay your power bill right away!" as an argument is laughable.

[bcpokey]

This again... I'm going to pick on a few things to start with:

"sometime soon the bitcoin reward will halve", yeah if you consider 2013 soon.
"$2.02 per day (700w @ .12/kWh" -- 3 5850s use absolutely nowhere near 700w. 3OCd 5870s use about 600W from the wall, 3 5850s will use closer 500W or less.
"You will start losing money on day xx" -- You stop mining when it is no longer profitable, no one put a gun to your head to keep going. You can also cash out of your rig if you really don't think you'll ever mine again and have no use for it, recouping even 60% of your initial investment (a VERY conservative estimate if you drop out 90/160 days after building) puts your profit back up > $1000. Not everyone is trying to replace their job with mining, that is a hefty ROI.

As has been mentioned no one needs to cash out immediately, power bills are delayed by 30 days, the minimum time for a forced cash out (or I suppose when the CC bill comes in, however if you purchase on the cusp of your billing date you have 30 days before the account closes, then 30 days to pay, so a 60 day window before any interest accrues on that bill).

Again, I'm not encouraging people to build mining rigs, I agree that the profit is really not there anymore, but your numbers are as flawed in this thread as they were in the previous one I responded to, skewed only to support your assertion that other people stay away, which I feel is somewhat disingenuous at worst, sloppy at best. A better analysis would be why investment and encouraging further investment will yield longer term gains, but that is a lot trickier I suppose.

The precision of this analysis isn't really important.  evidences the most important point which is basically "No, miners don't guarantee a static return. They are not money printers."

The people who understand depreciation, interest, borrowing and resale which your argument assumes aren't likely to be the same people trying to figure out how build their first miner.

Also, "You don't have to pay your power bill right away!" as an argument is laughable.

Your response would make sense if this was merely a post saying "You should be careful as mining is not a money machine" (as you put it) "here is why it is a risk". It does not, it clearly says you "will never pay off your rig" and "you will lose money" in no uncertain terms. Here is the very first line of this post after stating the theoretical rig specifications:

As you can see even if the value of bitcoin increases to over $1200 in a year, you are not even going to make your money back on a dedicated rig, at 90 days you are going to start losing money!

The rest is cloaked in numerical analysis which is the exact opposite of what you are trying to claim here. Seem to have contradicted yourself. I also stated that I was merely going to be picking on those points at the start, not that it was the utter defeat of the argument. Sloppy assumptions lead to faulty conclusions, rigor isn't an optional thing in an "analysis".

The concept of delayed payments vs accrued interest is laughable? Good luck to you in the real world then. If there is a significant gain to be made from sitting on a commodity rather than instantly liquidating it then you need to take that into account. To think any other way would be as you put it, laughable.

[mechatronic]

I generated some btc a while ago, then I ignored it for awhile. Credit to the early adopters for sticking with mining. That said, I feel like the current "wealth" in the btc economy is divided amongst those few early adopters, and in the blink of an eye mining became unfeasible. I'd like to get more involved by creating a client/exchange, along with a mining mini-farm, but at the same time I feel like I've missed the boat. This strawman analysis if taken at face-value confirms it - I think. In my pool I am generating about 1 btc / 40h with two 5870's on a rapid downward trend.

The main reason I am interested in btc is due to the mining tech/ distributed clients etc. Now that mining is moving out of reach of the individual, I wonder if the community will shift from tech-centric to business / investor centric.

[nanotube]

you forget that difficulty only increases while there's expectation of making a profit on the miner hardware+ongoing costs - hardwareresale at the end.

difficulty increases will simply not happen, unless bitcoin price increases sufficiently in order to encourage new mining power to come in.

there's a feedback loop in operation here, things will level out eventually.

[Gameover]

As has been mentioned no one needs to cash out immediately

If you do not cash out immediately you are only speculating on bitcoin price, and if you are doing this simply buy bitcoins rather than mine them.

but your numbers are as flawed in this thread as they were in the previous one I responded to, skewed only to support your assertion that other people stay away, which I feel is somewhat disingenuous at worst, sloppy at best.

The numbers are all 100% real, and very conservative if you ask me, with difficulty increases from 30% to 50%, and bitcoin increases from 15%-50% per 10 days, how can you call that not conservative?  If anything I am try too hard to make it look like you will have a good return on a rig, not the other way around.

A better analysis would be why investment and encouraging further investment will yield longer term gains, but that is a lot trickier I suppose.

It is a lot trickier because it is impossible, there are simply no longer term gains to be had, and after about the 30-50day period out, no gains at all, only losses.

[xf2_org]

Judging from the forum traffic, people are already starting to sell their rigs.