Replies to post #9475 on A.I.M. Users Bulletin Board (AIMUSERS)

[Conrad]

Steve,

When I gave the two options for the relation between the PC and the Number of Shares

1 PC=C1*(1-2S)*N (S is the fractional SAFE)

2 N=PC/(1-2S)/C1


I did not presume that these were the only options. They are simply a starting points for discovering how to solve the problem. What is done could for example be strongly dependent on the expected share price development for a real stock. I have tried out a few options and concluded that the requirement that the price run-away and the share quantity erosion be altered will force you to enter a condition with a virgin set of relationships with multiple variables, which cannot be solved unless a new set of relations are set up from an external motivation. This means that as long as you have no specification for what you want or need then there is no solution. This is the same like not knowing the travel time to cover an unknown distance to your work. With a given speed input you would never know when to leave for work!

So, the criteria for selling and buying are completely free, and you can arbitrarily create a specification in the same way as you did before with the Lichello algorithms. For a given price structure it is possible to specify the optimum buyin and selling prices but then the buy and sell amount remain free. The decision to Buy/Sell small parcels or large parcels cannot be made unless you provide an Optimization Goal: Profit maximization? Share growth maximization? Or what?

For a given price structure in a Test Run there is no risk involved and you could find the optimum strategy by an optimization routine, taking into account trading costs and interest and still you have a degree of freedom to go for Value optimization of Share Quantity growth.
Without setting criteria for these options there is no solution, and setting the criteria is not a mathematical procedure but one of defining exact objectives.

If you want trading symmetry then you can achieve this with the Ratio System and the PC-Update as I have previously shown:

Buy/Sell = (PC-SV)*M
PC Update: PC=SV after each trade

With M selected initially arbitrarily and then having it optimized for a given price history. The SAFE becomes redundant. You can apply the brakes or accelerate the trading as you wish. All you need is to set the Trade Resistances and optimize these on the basis of the price history.

If you are willing to specify Buy/Sell relationships for the Buy amount and the Sell amount and the PC-Update etc. then price structure for the remaining trade triggers is rigorously determined.

ZZzzz.