Take 10 strategies, do a manual ranking of them based on your gutfeel / preferences / etc, then look for formulas that will rank them accordingly. Include CL always-in / 1 contract in the mix if you want.
Thanks. Fortunately I figured out what was nagging at me so no need for gutfeel empiricism. Incidentally the CL data you included is insufficient for calculation of the Kelly fraction. Weird but true. SAS == 6*k*E*(1+|E|/sigma)*sgn[ E ]*min[ 1, N/1000 ] , where k is the Kelly fraction, E is the expected value of the trade returns (%/100), sigma is the standard deviation of the trade returns (%/100), sgn[ E ] is -1 when E is negative, is +1 when E is positive, N is the number of trades in the performance evaluation SAS[ +2, -1] = 6*.25*.5*(1 + .5/1.5)*1 = +1.0000
Why would you rank profitable systems. Put them all to work in real time and enjoy the money they make for you. Use your math skills to add the profits.
For the same reasons that you rank anything else. So you can focus on the best and discard the ones that aren't worth the time and effort. What does that mean?
Trade-list attached - I don't use the Kelly fraction, so I wouldn't be able to compute the SAS for this system, but if you don't mind doing it and sharing the result that would be great. I trade this system using a 50,000 account for 1 contract.
Sorry, I don't have Excel. But it's not hard for you to get your Kelly fraction. Just use Excel's Solver routine to solve the Kelly equation in the first post.
I've decided that I really don't care how negative-expectation systems stack up against one another. If you share that sentiment, here's the version of the SAS for you. SAS == 6*k*max[ 0, E ]*(1 + E/sigma)*min[ 1, N/1000 ]
Here's what I hope is the final upgrade to the System Achievement Score : SAS == 4*k*max[ 0, E ]*PF*min[ 1, N/1000 ] , where k is the solution to the Kelly equation (see the OP), E is the expectation (%/100), PF is the profit factor (see below), N is the number of trades in the performance evaluation. PF is the ratio of the gain total to the absolute value of the loss total. PF == sum[ max[ 0, Ri ] ]_i=1toN / sum[ max[ 0, -Ri ] ]_i=1toN SAS[ +2, -1 ] = 4*.25*.5*2 = +1.0000