um... ok. Mathematically it helps if you don't think in naive terms as stock, company, team, risk, price etc. First of all, there are no stocks. A stock is just a call at zero strike with infinite maturity. So you short a stock, you're just going short an instrument at some weight. The weight is the number of stocks you short. So you short 5 shares of google, the spot right now is 497 and the weight is minus five. So you do this with a bunch of equities ( google, apple, linkedin etc ). Then you obviously get a vector of spots and a vector of negative weights. The money you make is simply the inverse of the dot product of both vectors. What do you do with that money ? Obviously you don't sit on it. You buy protection on upside and speculate on downside simultaneously. ie. You buy X OTM calls on goog, say at 550 strike the OTM call is 70 cents so you buy that. Then you speculate on the downside ie buy say Y 485 weekly put at 70 cents.

So if your bet is right, the Y puts make money, the X calls lose money, and overall you come out winner. You wait until google is say 487 and then buy back your shares making 10 buck per share plus the money off your Y puts minus the money from the X calls. So thats just another dot product ie. C = 5 times 10 + Y times a - X times b.

Now say the trade goes south. Then you lose on the put, lose on the short, but make money on the X calls, so the dot product looks like above but with different weights. In either case, you can only lose a fixed sum worstcase ( so you statement " there's no limit to your potential losses" is definitely false ). So maximizing the money is then a constrained linear optimization problem. In a polynomial vector space, you can find X & Y quite easily using Dantzig. ( http://en.wikipedia.org/wiki/Simplex_algorithm )

HTH.

Embarrassingly enough, I'm not smart enough to parse your comment for correctness. If I were, I would be building a product to model that exact understanding of the market if it is, indeed, accurate.

EDIT: grammar