I really don't see why the argument.
Estimates that dividends could repay cryptx's loan by X date are woefully short, because of the huge reinvestment % . Dividends don't resume when sum(dividend+reinvestment)=loan. Dividends resume when sum(dividend)=loan.
Make a copy of cryptx's spreadsheet at
https://docs.google.com/spreadsheet/ccc?key=0AjI5bgsiFJAidHgwa0UyTHNEVG1lbDdvN2FMTExvOHc&usp=drive_web#gid=32 and change Col B to add 10PH per interval. If the first IPO batch sells out, shareholders will only see dividends in Sep. If the IPO does not sell at all, dividends will be seen in Dec. All while the mine is steadily growing in hashrate, but absolutely not keeping up with the network.
Using cryptx's own spreadsheet so that nobody argues about the hosting fees, reinvestment calculations, BTC exchange rate, etc etc.. All I'm changing is the rate of growth of network hashrate to 10PH per interval, from cryptx's 6PH projection. Yes, it's linear, so the rate of increase may easily be far, far greater. At least this choice should appease people who point out that hashrate cannot rise exponentially forever.
- If the IPO doesn't sell at all, the
loan amount could potentially be 3760 - 150 - 300 = 3310 BTC.
- if a 1000BTC "loan" is taken from cryptx, it will only be fully repaid from dividends in Aug '14, and shareholders finally start getting lean and fast-declining dividends in Sep '14. I don't know how cryptx calculates it, or how the loan amount is merely 1009 BTC if 10000 shares (950BTC) are sold at IPO.
- I don't see why the IPO will be bought up if the dividends quickly trend towards zero before reaching ROI, and since shares on the market are cheaper.
- Note how % network still drops during the first few periods of heavy reinvestment. This shows that the reinvestment strategy is still futile at this rate of network growth.
We should now agree on how horrible the outcome is even with a linear 10PH increase. Perhaps certain anonymous people will claim that 10PH is way too much. ASICMINER has indicated
expected delivery of 109PH of chips shortly. That's just one manufacturer. How about the rest?
So the question really is what you think the network hashrate / difficulty growth will be.
I don't think we should argue about other possible factors such as BTC exchange rates because these are the same figures as in cryptx's projections.
I also don't know whether other cloud mining operations can be sustainably profitable, but that isn't the point we're discussing.