Lets pretend there are two mining bonds (each are 1MH/s per share, neither has default risk, both have same liquidity).
MINE.A is 1 BTC per share.
MINE.B is 0.01 BTC per share.
You are saying both would be a net loss?
Obviously not. Somewhere between 1.00 BTC per share and 0.01 BTC per share there is a price that for a given period of time and given future difficulty would break even. If the security is priced below that you would profit if you bought it. If the security is priced above that you lose if you bought it. "Investors" on GBLSE buying mining bonds have been "stupid" to date massively overpaying for bonds and creating a scenario where even under optimistic conditions they are unlikely to be profitable.
Quote
This constant pressure to compete against growing difficulty on existing equipment means BTC returns are always going to trend to zero (in which case the shares are worthless and any value was paid out with the reducing dividends), or fight a losing battle against the exponentially increasing hash rates.
And if before it aproaches zero the sum of dividends paid are > the the purchase price you would have a net gain. Take one of those "losing" bonds and instead imagine it was offered at 50% of the price it was offered at. Was it still a loser? The securities were overpriced. It is that simple. They were overpriced and investors who poorly understood the risks and economics created enough demand that they sold out at these over inflated prices. That led to other offerings at similar overinflated prices (I mean if you see a competitor sell out at 0.3 BTC per MH why would you offer your for less). It is called price discovery. Obviously (hindsight for some but foresight for many others) the offering prices were way way way too high and those investors were doomed to lose money.
Another way to look at it is NPV. The sum of any periodic payments over any period of time can be expressed as net present value. (i.e. $100 per month for the next 10 years is worth $x right now). Calculating NPV for 1 MH/s of mining power is challenging because determining future difficulty is tough but all mining bonds have an NPV (even if unknown until after they stop trading). If you buy it above NPV you lose. If you buy it below NPV you win.
TL/DR simplified version:
There is nothing wrong with the concept of a mining bond. However all mining bonds to date (and probably all mining stocks too) have been horribly horribly overpriced. Many are probably still overpriced despite falling 30% to 50% since launch. The mining bond market resembles the housing market. If you just took asking price for a new home in 2007 you probably have lost money. Does that mean the concept of owning a home is fundamentally flawed or does it simply mean you bought an asset at an overinflated price?
TL/DR even more simplifed version:
If you overpay for an investment you will lose money (or earn a lower return than you otherwise could have). It isn't more complicated then that.