Short version:[let me know if you think something else not addressed is biggie]
-Current collateral provides security for up to 2.25 price drop - that is enough to not need so smooth growth curve as well as to call the approach rather conservative.
-Interest of 5% annually is small compared to 130x price increase - so yes it is not included in my calculation. But even if I do I do not expect drastic changes. If you think one will need constantly offering 40% interest to place a short order this is a whole nuBit of a game.
-In my calculation I just used voluntarily re-shorting at predetermined increase [I used 33% or 50% - I will check exactly, if you want me to]
-On the soft collateral - we have discussed it once already. You thought it is BS, but in practice you need very insignificant amount for that purpose kept in BTS, if you place correct size-wise orders.
Certainly my calculations could be wrong; I am curious enough about this that I will likely do a more thorough analysis another day. But I am fairly confident that it is almost impossible to get 5x gain through just a shorting strategy over the period of time in which BTS rises from $0.02/BTS to $2.50/BTS.
My formula takes into account the minimum amount of soft collateral necessary to cover the short (unless you want to take your chances and let the DAC do it for you exactly 30 days after you opened the short position, but you are far more likely to end up covering at a bad time in the BTS price with that strategy).
The smooth growth curve is more about being able to cover and re-short in a short enough time period. You want to reduce the short interval to increase gains from compounding (that analysis ignores the cost of transaction fees but those are not very significant for reasonable lower bounds on the reshorting interval). If you have long periods of stagnation along with short bursts in the price, you will end up with less BTS than if the growth curve is smooth. As an example, imagine the growth curve looked like irregular steps where it alternated between being flat for some variable period of time and suddenly jumped up to twice its value. The 130x growth would be reached by ln(130)/ln(2) = 7 doubling jumps. If I had 2.5*X BTS, I could hold 0.5*X BTS as soft collateral and use the remaining 2*X BTS as hard collateral for a short position that held 3*X BTS locked as collateral and owed the network X*p BitUSD (where p was the price in BitUSD/BTS at the time of creating the short). If the price doubled some time later, I could use the 0.5*X BTS to buy X*p BitUSD at the new price of 2*p BitUSD/BTS. Then I could use the X*p BitUSD to cover the short and get back 3*X BTS. I would then have 3/2.5 = 1.2 times the amount of BTS I originally had. If I immediately used my new BTS to redo this process to compound gains, I could get a total multiplicative factor on my original investment of (3/2.5)^(ln(130)/ln(2)) = 1.2^7 = 3.6 (which is less than the 5x I showed was
theoretically possible in my previous post). More generally, if I reshort every time the price goes up by a certain fractional amount g (so g = 0.50 for the case of reshorting on 50% gains in price), then the multiplicative factor on the amount of BTS would be
f = (3/(2 + 1/(1+g)))^(ln(130)/ln(1 + g)). Notice the theoretical best case is right around 5x, but that is at a point where the reshort interval is so small that the transaction fees can really add up.
I am interested in incorporating interest into my formulas. Not sure how important of a role it will play. But if it is already hard enough to get 5x, it will be even less once the interest is added in. Also, don't be too sure that high interests are unlikely. If this investment strategy really is as good as you are claiming it to be and such a sure thing, everyone will want to get in on it. The market will thus force the interest rate high enough until few people remain who still think that the gains from shorting (with the effect of the high interest factored in) are a great deal.