Nobody knows what the difficulty is going to be, so lets make a table of possible values for a 1 Mh/s PMB at the current difficulty.
| Expected Increase | | Total Dividends |
| 30% | 0.00157759 |
| 20% | 0.00218436 |
| 10% | 0.00400466 |
| 7% | 0.00556492 |
| 2% (Moore's Law) | 0.01856706 |
| 0% | infinite |
EskimoBob and others are wrong. You can see that it is possible to make money from a PMB if you buy it at the right price. So, figure out what you think the difficulty will rise by every 2 weeks over the next several months, and don't pay more than the dividends you will receive.
Your analysis is incorrect.
You fail to include a time value of money for the present value of the bonds. Even at negative difficulty growth, the value of the income stream from mining is finite.
Given the risks the discount rate for the time value of money needs to be very high - at least 30% in my opinion.
Risks taken by buying a mining bond include
Failure of bitcoin protocol
Exchange rate risk
default risk by operator
Anyone with basic understanding of the situation will realize that the only way a mining bond would be profitable is if it is priced lower than the cost of equivalent mining hardware. Ultimately, sellers of mining bonds are exploiting people ignorant of the economics of bitcoin. The result is a guaranteed loss for the buyers.
Worse, many of these buyers will project their experience onto bitcoin itself, and paint the entire concept as a scam.