If you had $5000 to invest right now: Coins or Hardware? Whats your rationale?

[giszmo]

By my Calculations, even at a measly 5% difficulty increase per re-target, the hardware wouldn't have paid off even after a year.
And that's assuming you could get $5000 worth of Mini-rig Value hashing power delivered today.

From my calcs, you'd only have 821 BTC after a year, whereas you could buy 1000 BTC now and save yourself the electricity/heat/hassle of dealing with the hardware.

This calculation is assuming:

$5000 worth of "MiniRig Value" hardware
No Delivery Wait Time
No Electricity cost
Difficulty increase of 5% every 13 days
Block split in 210 days
1000 BTC bought today at $5 ea

Your calculations are flawed because they utilize a fixed value of Bitcoin based on today while all of your other numbers vary into the future. If you want a more realistic profit profile then you need to project the growth in Bitcoin value to coinside with your other projected growth variables. Drawing a line from the beginning of Bitcoin value to todays Bitcoion value and then extend into your desired future cashout date, say 12 months, will give your spread sheet a more realistic perspective of Bitcoin mining profitability. Of course, this assumes the projected growth of Bitcoin value is constant.

How do you do that? $0.4 per MHs sounds unrealistic.

mufa23 is only quoting the cost of the GPU's because he may have utilized existing MB/PSU/CPU/RAM. So for mufa23 he only had to spend $800 to become a miner. His next 2 GH/s will cost $.60 per MH/s.

For generating bitcoin, the exchange rate does not matter (a rise in difficulty is what reflects the rise in popularity and the technological progress in his model). For buying now it does not matter neither.

[1715036013]

1715036013

[1715036013]

1715036013

[1715036013]

1715036013

[1715036013]

1715036013

[mufa23]

How do you do that? $0.4 per MHs sounds unrealistic.

mufa23 is only quoting the cost of the GPU's because he may have utilized existing MB/PSU/CPU/RAM. So for mufa23 he only had to spend $800 to become a miner. His next 2 GH/s will cost $.60 per MH/s.

Yes, one rig I already had some spare parts, so I only spent $270 for the first rig.

Did this with the second, and will do with my next rigs.
$270 - Three 5830s
$20 - CPU
$5 - RAM
$45 - MoBo
$50 - PSU

$390 for a 1GH/s rig.

[Fuzzy]

By my Calculations, [...]

Your calculations are flawed because they utilize a fixed value of Bitcoin based on today while all of your other numbers vary into the future. If you want a more realistic profit profile then you need to project the growth in Bitcoin value to coinside with your other projected growth variables. Drawing a line from the beginning of Bitcoin value to todays Bitcoion value and then extend into your desired future cashout date, say 12 months, will give your spread sheet a more realistic perspective of Bitcoin mining profitability. Of course, this assumes the projected growth of Bitcoin value is constant.

That's because my question is based on a decision made today.
As stated previously, generating bitcoins does not depend directly on the value of bitcions.

ie, if you bought 1000 coins today ($5k worth), and a year later your hardware ($5k worth) would have generated 821 btc, you're still ahead if you had bought bitcoins outright, whether BTCs are worth $1 or $100
The value of the hardware however needs to be accounted for, as is the delivery time of said hardware.

How do you do that? $0.4 per MHs sounds unrealistic.

mufa23 is only quoting the cost of the GPU's because he may have utilized existing MB/PSU/CPU/RAM. So for mufa23 he only had to spend $800 to become a miner. His next 2 GH/s will cost $.60 per MH/s.

That's what I'm assuming. Unfortunately you can't plug a GPU into the wall and just mine away.

$270 - Three 5830s
$20 - CPU
$5 - RAM
$45 - MoBo
$50 - PSU

$390 for a 1GH/s rig.

Looking at those they seem like cherry picked numbers and cannot be found in bulk.

Also the avg hash rate for a 5830 is 300. Not bad for a $100 card tho.

And if you've ever run a rig for over a month, you know a $50 PSU just isn't gonna hold up.

[check_status]

For generating bitcoin, the exchange rate does not matter (a rise in difficulty is what reflects the rise in popularity and the technological progress in his model). For buying now it does not matter neither.

That's because my question is based on a decision made today.
As stated previously, generating bitcoins does not depend directly on the value of bitcions.

You are projecting into the future 12 months from Today the difficulty increases which shows a reduction in coin generation, highlighting a loss in production without determining what Bitcoins value is going to be 12 months from now. Your statistics are painting an inacurate picture of potential profit.

And if you've ever run a rig for over a month, you know a $50 PSU just isn't gonna hold up.

He's probably using 2 Dynex 500W PSU's.

[Fuzzy]

For generating bitcoin, the exchange rate does not matter (a rise in difficulty is what reflects the rise in popularity and the technological progress in his model). For buying now it does not matter neither.

That's because my question is based on a decision made today.
As stated previously, generating bitcoins does not depend directly on the value of bitcions.

You are projecting into the future 12 months from Today the difficulty increases which shows a reduction in coin generation, highlighting a loss in production without determining what Bitcoins value is going to be 12 months from now. Your statistics are painting an inacurate picture of potential profit.

That's because it doesn't matter what they're worth in $ if what you're interested in is how many bitcoins either option nets you in 12 months.

And if it weren't for the 12 week delivery (during which you miss out on ~300 BTC), I'd gladly forego the 200 BTC difference and electricity costs if it meant I got to have a piece of kickass Bitcoin Mining hardware which may one day become a Bitcoin Artifact of sorts.

[sunnankar]

That's because it doesn't matter what they're worth in $ if what you're interested in is how many bitcoins either option nets you in 12 months.

And if it weren't for the 12 week delivery (during which you miss out on ~300 BTC), I'd gladly forego the 200 BTC difference and electricity costs if it meant I got to have a piece of kickass Bitcoin Mining hardware which may one day become a Bitcoin Artifact of sorts.

Yes, using BTC as the numeraire does complicate things because expenses are in $$ or Euros for electricity. But no one has even discussed the discount rate for determining the NPV. That should then be applied to alternatives like BITBOND or GIGAMINING and then corrected for downtime of hardware and labor value.

For example, there are currently 3,924 GIGAMINING available at 1.51 BTC each. $5,000 would buy 662 of them or 3,310 MHs of production in perpetuity with no downtime or labor. That is the opportunity cost which you could then project for a discount rate and difficulty changes to model for 1,000 BTC NPV.

[check_status]

By my Calculations, even at a measly 5% difficulty increase per re-target, the hardware wouldn't have paid off even after a year.

In your argument, Bitcoin will continually lose value over time because of difficulty increases. This conclusion is invalid because you don't take projected Bitcoin value into consideration. You assume a static, unchanging Bitcoin value over time while difficulty does change over time. You are intentionally weighting your chosen outcome.

[Fuzzy]

By my Calculations, even at a measly 5% difficulty increase per re-target, the hardware wouldn't have paid off even after a year.

In your argument, Bitcoin will continually lose value over time because of difficulty increases. This conclusion is invalid because you don't take projected Bitcoin value into consideration. You assume a static, unchanging Bitcoin value over time while difficulty does change over time. You are intentionally weighting your chosen outcome.

Actually, if you look carefully I NEVER take the value of bitcoin into consideration, beside what I can get Today in terms of Coins or Hardware. It's as if you won a contest and got to choose between 1000 BTC or a 8.2 GH/s FPGA machine.

[...]

[...]

For example, there are currently 3,924 GIGAMINING available at 1.51 BTC each. $5,000 would buy 662 of them or 3,310 MHs of production in perpetuity with no downtime or labor. That is the opportunity cost which you could then project for a discount rate and difficulty changes to model for 1,000 BTC NPV.

Looking at my previous model, given a steady difficulty increase, the total number of bitcoins mined looks asymptotic, which is really weird, to see the reward of 10 GHs reach 0.01 BTC per 14 days  

But that's what the calcs show, and plugging in 3,300 MHs, at 5% diff re-targeted at 14 day periods, yields 348 BTC in a year, and only 417 four years later.

[sunnankar]

Looking at my previous model, given a steady difficulty increase, the total number of bitcoins mined looks asymptotic, which is really weird, to see the reward of 10 GHs reach 0.01 BTC per 14 days  

But that's what the calcs show, and plugging in 3,300 MHs, at 5% diff re-targeted at 14 day periods, yields 348 BTC in a year, and only 417 four years later.

Not sure that is the best premise to use. Perhaps you should review your calculus and differential equations for your next financial modeling.

[sunnankar]

Actually, if you look carefully I NEVER take the value of bitcoin into consideration, beside what I can get Today in terms of Coins or Hardware. It's as if you won a contest and got to choose between 1000 BTC or a 8.2 GH/s FPGA machine.

Another potential flaw in your logic.

Like a gold mine which produces a commodity but has input costs, like fuel and trucks, denominated in a different currency; the FPGA machine has operational costs that are fiat currency denominated and must be born by someone and therefore has opportunity cost even if you can have those costs born by someone besides you. The FPGA machine also likely has an expected salvage value just like the used dumptrucks and gold mine would after the asset has been depleted. You fail to take into account these calculations with the discounted future cash flows.

But unlike a gold mine the FPGA machine has alternate uses that could possibly generate more revenue than the value of generated bitcoins and use that revenue to purchase bitcoins (this ties in with my earlier post about difficulty; miners who can generate more value from their processing power from an alternate activity will reallocate processing power from bitcoin mining to activity X).

You have also failed to take into consideration any tax implications; For example, 179 deductions for hardware, deductions for electricity, capital gains, etc.

Therefore, the decision between bitcoins or hardware is more akin to bar of gold or gold mining company whose miners/equipment/mine can also perform excavations and foundations for skyscrapers than between lump sum or annuity.

[Fuzzy]

Looking at my previous model, given a steady difficulty increase, the total number of bitcoins mined looks asymptotic, which is really weird, to see the reward of 10 GHs reach 0.01 BTC per 14 days  

But that's what the calcs show, and plugging in 3,300 MHs, at 5% diff re-targeted at 14 day periods, yields 348 BTC in a year, and only 417 four years later.

Not sure that is the best premise to use. Perhaps you should review your calculus and differential equations for your next financial modeling.

I'd be interested in which premise you could come up with for this situation. As you can see, I'm doing a manual step calculation here, not an integral function.

And looking at those charts, since the crash, the difficulty has been climbing at reasonably steady rate of about 4.895% per retarget, so my assumption is pretty close.

Actually, if you look carefully I NEVER take the value of bitcoin into consideration, beside what I can get Today in terms of Coins or Hardware. It's as if you won a contest and got to choose between 1000 BTC or a 8.2 GH/s FPGA machine.

Another potential flaw in your logic.

Like a gold mine which produces a commodity but has input costs, like fuel and trucks, denominated in a different currency; the FPGA machine has operational costs that are fiat currency denominated and must be born by someone and therefore has opportunity cost even if you can have those costs born by someone besides you.

The cost of electricity in a year for that scenario (using $0.10 / KWhr), is $356.
But that assumption is covered here:

This calculation is assuming:

• [...]
• No Electricity cost
• [...]

The FPGA machine also likely has an expected salvage value just like the used dumptrucks and gold mine would after the asset has been depleted. You fail to take into account these calculations with the discounted future cash flows.

That is true, but since the value of bitcion mining hardware is tied to how many bitcoins it can mine, that value is very low at the end of a year, and it's simply impossible to predict it's value retention. If you have a suggestion, I would love to hear it and take it into account here.

But unlike a gold mine the FPGA machine has alternate uses that could possibly generate more revenue than the value of generated bitcoins and use that revenue to purchase bitcoins (this ties in with my earlier post about difficulty; miners who can generate more value from their processing power from an alternate activity will reallocate processing power from bitcoin mining to activity X).

Please name one use of the FPGA that would yield more bitcoins than mining. If you are just predictiong that there may be a use some time in the future, then that is an even harder guess to make than the original question I posed.

You have also failed to take into consideration any tax implications; For example, 179 deductions for hardware, deductions for electricity, capital gains, etc.

I would like to point out that there are no set Tax rules for Bitcoin yet, but if we use them as investment options, as most people will probably do, then you bring up a VERY good point relating to hardware/electricity deductions.

Providing that Bitcoin succeeds (if it doesn't, then you're no better off if you'd bought bitcoins outright), then you make back the value of the hardware when you realize your investment by converting to fiat and subtracting the cost of hardware from the taxes you owe on your capital gains.

Brilliant  

[check_status]

Actually, if you look carefully I NEVER take the value of bitcoin into consideration, beside what I can get Today in terms of Coins or Hardware. It's as if you won a contest and got to choose between 1000 BTC or a 8.2 GH/s FPGA machine.

That is why I used the terms 'invalid conclusion' and 'flawed', as your goal is to extract value. Today value is measured in Dollars, it is the world standard in which all other valuations are made. You have, in theory, $5000; You want to buy hardware that costs $5000 or Bitcoins that cost $5000. You are seeking either break even or profit, measured in what? Dollars. For that, you will need to know what the value of a Bitcoin will be in the future. And then, you can determine what is the most profitable direction, hardware or outright purchasing of Bitcoins.

The trend graph starts in July 2010, the next data point, $1.70/BTC 4/24/11, then $2.20/BTC 11/19/11, then $4.40/BTC 2/19/2012, and finally $5.00/BTC 5/12/12.
This shows the value of 1BTC may be $6.50 just prior to the split, if the trend continues. This translates to approximately a $.25 increase per month per BTC.
The halving is artificial scarcity and like all scarcity, value increases. While there may be some wild speculation when the halving happens, the value will continue it's upward trend.

[MarketNeutral]

What is the approximate dollar threshold at which it would obviously be more profitable to invest in mining instead of purchasing bitcoins outright?

I know people disagree wildly, but I'm curious.


EDIT: Also, how does this figure differ depending on the expected time frame, such as if I want to break even at 3 months versus 3 years?
The mining calculators have been of some help, but I still must read between the lines.

Also, I'm surprised that $5000 is so disputed. Is 5k that close to the mark?

[Edited for grammar.]

[Fuzzy]

Actually, if you look carefully I NEVER take the value of bitcoin into consideration, beside what I can get Today in terms of Coins or Hardware. It's as if you won a contest and got to choose between 1000 BTC or a 8.2 GH/s FPGA machine.

That is why I used the terms 'invalid conclusion' and 'flawed', as your goal is to extract value. Today value is measured in Dollars, it is the world standard in which all other valuations are made. You have, in theory, $5000; You want to buy hardware that costs $5000 or Bitcoins that cost $5000. You are seeking either break even or profit, measured in what? Dollars. For that, you will need to know what the value of a Bitcoin will be in the future. And then, you can determine what is the most profitable direction, hardware or outright purchasing of Bitcoins.

Exactly, and at the end of the day year you will either have X bitcoins or Y bitcoins. My question is, will X be larger than Y. the vallue per bitcoin will be the same, whether its $1 or $100, what matters is how many you have.

Also, I'm surprised that $5000 is so disputed. Is 5k that close to the mark?

The $5000 is an arbitrary number. It's just an investment, $1 or $1million, that's up to the investor to decide. The variable here is what you are buying with the funds available.

From my perspective, no FPGA bought today will be able to beat a direct bitcoin buyout.
Unless you can write the cost of hardware off of your "taxes paid on capital gains", that's where things get murky, as I've presented a rather ideal scenario for hardware.

[check_status]

This Trend Graph is based on the market prices for each month, projected out until march 2013.

[Fuzzy]

This Trend Graph is based on the market prices for each month, projected out until march 2013.

I find it odd that most of the values on that chart are below the projection line.

[check_status]

I chose to base the trend graph on the actual market prices instead of correcting for the Tsunami spike of June 2011. The second spike is nowhere near the magnitude of the first. Whether the values are above or below the line aren't important if what your interested in is will the value increase based on previous history. To graph future profitability you would utilize the pattern you noticed, say 10% below the trend line, to create a more conservative valuation.

[arklan]

very interesting thread.

of course, i think 1 year is short sighted. afterall, the hardware isn't ALL going to melt down into piles of goo in exactly 12 months. it'll keep earning. also, as we've just seen, the difficulty is far more fungible then a steady 5% increase.

i say hardware - and the lower the electricity it uses, the better!

[giszmo]

very interesting thread.

of course, i think 1 year is short sighted. afterall, the hardware isn't ALL going to melt down into piles of goo in exactly 12 months. it'll keep earning. also, as we've just seen, the difficulty is far more fungible then a steady 5% increase.

i say hardware - and the lower the electricity it uses, the better!

oh, i guess in one year you will not burn all your hardware but with my 8 cards the trend was going in that direction.

[arklan]

very interesting thread.

of course, i think 1 year is short sighted. afterall, the hardware isn't ALL going to melt down into piles of goo in exactly 12 months. it'll keep earning. also, as we've just seen, the difficulty is far more fungible then a steady 5% increase.

i say hardware - and the lower the electricity it uses, the better!

oh, i guess in one year you will not burn all your hardware but with my 8 cards the trend was going in that direction.

...ouch. sympathies. maybe i've been lucky, i don't know. been running full tilt for 8 months or so now without a mining related death.