I feel that mining bonds are not the means to the end, aka, just buy the bond and hold it. Although, if you bought at the original IPO, you would still be ahead of the game by the 4%.
This thread claims that mining bonds are the turds. Could it be that "investors" strategies in owning mining bonds are the real problem?
Bitcoin mining should be considered to be a depleting asset and as such, one must understand depleting assets in order to understand how to invest in them. It is becoming common place knowledge now that reinvestment is an essential part of any mining strategy, whether you are buying mining equipment or you are buying mining bonds. Otherwise, you lose on both ends because of technology advances eroding value in your mining equipment and difficulty increases eroding your dividend payments.
If you are not willing to reinvest, mining and mining bonds are probably not for you.
This is basically false.
To a first order approximation your investment decisions are linearly superposed together. In month 1 you invest X1 in mining equipment/bonds and that investment will generate a NPV of Y1 over its lifetime. In month 2 you invest X2 in mining equipment/bonds and that investment will generate a NPV of Y2 over its lifetime. Y2 is independent of X1 and Y1 is independent of X2; any one of the investment decisions will either be profitable or not, and your total profit is the sum of the individual profits. If E[Y1]>X1 then investing in month 1 was a good decision, no matter what you do next. If E[Y2]>X2 then investing in month 2 was a good decision, no matter what you did before.
This is truer for buying bonds than for equipment, since economies of scale with the latter may make it more lucrative to keep a consistent stake (which means reinvesting in more MH/s). I don't think this effect is large though, and for bonds it doesn't apply at all.
To a second order approximation you need to correct for your risks, and this supports having a consistent stake; but in the same way that martingale strategies can't gain you expected profit when gambling, "investment strategies" can't get you expected profit unless there is expected profit in the individual decisions. If there is ever expected profit to be made, it can be obtained by buying and holding forever.
Otherwise, you lose on both ends because of technology advances eroding value in your mining equipment and difficulty increases eroding your dividend payments.
You're double-counting the loss. Either you compare against selling now, in which case you only care about the current traded price; or you compare against keeping forever, in which case you only care about the future dividends. In a reasonably efficient market, the two are commensurable; pick one to focus on.
If the dividends you have obtained in a unit of time are greater than the depreciation of your holdings, the original investment was good whether you reinvest or not. If the dividends you have obtained in a unit of time are less than the depreciation of your holdings, the original investment was bad whether you reinvest or not.
Basically, the argument is that what miners do is use real money (bitcoins) at a present day to purchase equipment with a fixed MH/s, thus a bond holder is really performing the same action that miner is now. However, just look at that statement. If you wanted to get into mining, would you buy equipment now? With ASIC difficulties coming along with the reward halving, I certainly wouldn't buy a GPU. I wouldn't even consider a non-BFL FPGA. So my option is a BFL FPGA.
You can pre-order BFL ASICs. And if you issue new bonds you can price them according to ASIC prices. Currently signal-to-noise ratio is too low so issuing new deterministic bonds doesn't make much sense. But any known future development and upgrade path can be trivially priced in, giving a price which is fair for both issuer and investor.
The point with deterministic bonds is that the concept of "issuer" is fundamentally meaningless. A MH/s is a MH/s, it will generate Y BTC over its lifetime and traders should be able to place bids and asks according to their estimation of Y (or its correlation with their underlying operations), with leverage of course. The traded price for a deterministic perpetual MH/s will converge to an equilibrium reflecting the opinions and needs of everyone on the market. It just happens that the best way to sell MH/s currently is as an issuer on GLBSE, leveraging personal trust rather than collateral.
Because of the way difficulty retargeting works, a large discrepancy between the total cost of physically obtaining MH/s and its lifetime earnings (hence its traded price) can't be sustained for long periods of time.
