This is the situation AFTER the IPO have been paid back. How much revenue will be used to R&D is anyone's guess. But if BTC keeps rising, that is getting smaller and smaller.
No, the 0.25 BTC return that I estimated does average out the period of time
after and before shareholders have recouped their IPO investment. In other words, ASICMINER will represent 10% of the global hashrate on average, from the moment they launched until the end of 2013 (IOW right now they are at 15% and it will go down to 5%).
In the real world (dunno, haven't been there a while) I could have sworn that 10% profit a year was a good year.
No, that would not be a 10% profit. What I explained is that if you buy a share at 0.6 BTC today, and recoup 0.25 BTC by the end of 2013, and if after that the share's worth is close to 0 (because by that time ASICMINER will generate negligible profits), then you have not made any profit. You have lost 0.35 BTC per share.
When you take the electricity prices into account you cant only take BTC to it, but instead the price a BTC has in exchange for USD for example. And i think the powercosts wont be something that will have such a heavy weight.
Even if you make power free or ignore all operating costs, ASICMINER won't generate significant profits: if the difficulty increases by 30x by early 2014, which is a reasonable estimate, then one share (assuming they are at 50 Thash/s) will generate merely 0.014 BTC per month ignoring operating costs. Assuming you buy a share at 0.6 BTC today, and assuming you recoup about half the investment by early 2014 (by my estimate in my previous post), you have 0.3 BTC remaining to be recouped, and at 0.014 BTC per month, at a constantly increasing difficulty, you will basically never do it...
The only shareholders who will make money are those who bought low (0.1-0.3 BTC, eg. those who went in the IPO), or those who manage to buy & sell at the right time (the share may be volatile enough to have lows & highs despite being overpriced). That, or you may make money if ASICMINER's future unknown plans about scaling beyond 50 Thash/s with more efficient chips turn out to be at least moderately successful. Maybe they will but that is not certain at all.