Analysis of $/BTC price correlated to the number of ASICs sold

[codro]

The current price of a BFL Single SC is roughly 108 BTC. (at $12/BTC)

With the reward drop to 25 BTC per block coming soon, there are only going to be 3600 BTCs mined every day.

Let's assume miners expect to make their initial investment back in 10 months. In these 10 months, there will be roughly 300(days)*3600(BTC)= 1,080,000 BTCs mined.

Since it will only be feasible to mine on ASICs, it's safe to assume that all the mining power will be comprised of ASICs.

This means that it will take 10000 ASIC units of equal hashing power 10 months to each mine 108 BTC and break even before they start making a profit.

Note that you can consider a Mini Rig SC to be roughly 25 BFL Single SCs, so only 400 Mini Rigs need to be sold before the price of $12/BTC will start becoming unattractive for the miner.

This scales linearly, if there are going to be 20000 Single ASICs sold (or 800 mini rigs, or any combination of the two), it would take the miner 20 months to break even. I suspect that the BTC price would then have to at least double so that the break even point can get back to 10 months.

So, if ASIC prices stay the same, for every 10000 ASIC units sold*, the $/BTC price needs to go up by +$12.
 
Thoughts?

* 1 ASIC unit = 1 Single SC, 25 ASIC units = 1 Mini Rig SC, etc

[codro]

I'm probably being too optimistic about miners only expecting to break even after 10 months, if we make that 5 months instead, it scales linearly as well, with just 5000 ASIC Units per $12 $/BTC increase.

2000 ASIC Units if break even is expected in 2 months (and price stays at $12/BTC). Consequently the price would need to be $48/BTC for 10000 ASIC units to be able to break even in 2 months, for example.

[Fjordbit]

Thoughts?

Economies are iterative cycles that need to be modeled with differential equations. A future miner could buy 110 coins and then get a 1299 ASIC from BFL, who then sells the 110 coins on the market to another future miner buying 110 coins. I don't think any of these calculations hold water.

[DeathAndTaxes]

The price doesn't need to go anywhere.  It is very possible that some miners have naive expectations and so much hashing power will be sold that it pushes prices out to a 33% annual ROI (3 year break even).   As the market becomes saturated some miners may become disillusions and simply sell their rigs for a loss.

Price drives difficulty but difficult doesn't drive price.

[codro]

The price doesn't need to go anywhere.  It is very possible that some miners have naive expectations and so much hashing power will be sold that it pushes prices out to a 33% annual ROI (3 year break even).   As the market becomes saturated some miners may become disillusions and simply sell their rigs for a loss.

Price drives difficulty but difficult doesn't drive price.

That is true, my point was that the only way anyone will make a profit from ASIC mining once everyone embraces it is if the price goes up.

If it doesn't, and stays at the current level, once enough people get into it, no one will break even at current market prices.

I mainly just calculated the number of ASIC systems needed to reach that point. Which - doesn't seem like a lot... A few months of sales?

[DeathAndTaxes]

Yeah I see the wording was just kinda weird.  Also I would point out that ASICs have an interesting cost structure.  A very high upfront cost (NRE) and a very low unit cost.  The non-NRE per unit of the ASICs offered likely is 30% or less of the current retail price.  This means that when sales slow (because market becomes saturated and ROIC% plummets) BFL and other ASIC providers can slash prices.   If difficulty skyrockets to say 100 mil a Single at 108 BTC doesn't look so attractive but a Single selling for 50 BTC suddenly looks much better.   It can be only 50 BTC because the price of USD:BTC has doubled but it also can be only 50 BTC because BFL cut the price in half.