Yeah, we're there. Lets be practical. If the S3 sells for more BTC than it can mine @ 18% diff increases, then lets not buy it.
There has to be some incentive in it for the miner.
OK ?
Unfortunately that's not how real world economics work, ASIC manufacture's have figured this out but the people haven't.
Simply put: In a market with imperfect data, price is determined by the most optimistic (STUPID) investors.
BTC is the poster child of imperfect data: Electrical costs vary greatly, difficulty rate increase estimations vary, did I break even in terms of BTC or fiat, reliability of vendor and shipping time, extremely short term skew of supply vs demand.
The Asic manufacture's understand that with sooooo much guessing going into ROI calculations that they simply need to find a number that enough stupid people will buy up everything they produce. Anything less than this is leaving money on the table and not smart business.
Yeah, although I have to say, at 2 TH/S that I've bought since March/April, I ROI'd post power costs a few days ago, and its all bitmain equipment, so I mean, they ARE doing something right. You just can't be one of the dumb sheep spending an extra .1 or $400 when ROI is doubtful. In an industry like this, mining a small amount of coin makes sense if you already hold a decent amount of btc, but as an investment in and of itself its pretty nonsensical. Why invest thousands of dollars in equipment that you can't sell in an instant sell on an exchange if something happens that you think will ruin btc. You will be left with a paperweight. At least with bitcoin, you could still get out of the trade and perhaps jump back in if/when the dust settles.
Spending any money on bitcoin mining is way riskier, and honestly miners should demand, and buy based on a 60-120 day ROI schedule. 60 being 5-10% diff increases and 120 being 20% every cycle. Anything more than this is absolute bullshit and miners should not have to take that risk.