A.I.M. Users Bulletin Board

[Conrad]

Clive,
I have been searching for some information on your 50/50 suggestions but found none that give 50-50 as an optimum starting split for investing in arbitrary stocks. In one example one "advisor" came up with 3 different ratios for 3 different investors buying the same equity. For a conservative beginner he suggested 60% or even more in cash and he himself said "I like more risky approach, so I go for 70 stock as in the long run it earns me more on the average. . .as cash does not earn enough". . . this reflects more or less my point of view, apart from considering the particular price level of the stock in the trading range at the time of stepping in..

Since we have discussed the issue without having started out from well defined statement as to what the goal is I think we are each beating a different horse without knowing if the horses are supposed to run or to dance. You talk about a “50/50 adjusted loss/reward”. . . which might mean that the in the end the chance of losing a buck is the same as the change to win one. . it is easy to interpret it that way, so that in the end on the average nothing is gained because the loses balance the gains. . .this sounds rather silly to me as an investment objective and you probably do not mean it that way, so we are probably talking about "apples" on the one hand and "3-inch ss bolts" on the other hand. Besides that I do not limit the investment method to AIM but for the sake of argument let's consider only AIM-like methods.

I look at it this way to get an answer to the question . . . I do not understand your answer so I try to make sure that I understand at least what I mean

Suppose we do 100 10-year "optimisation" runs of arbitrary real stocks that existed 10 years ago and maximise the yield on Portfolio Value(PV)and let the optimisation determine all the parameters that one has available in an AIM System that one uses. For my Testing I use 6 Vortex AIM parameters including the Min Trade Amount Which can be modelled as a fixed amount or as a percentage of the equity value. For a standard AIM there are also the same 6 variables so in that respect the only difference between the methods is the algorithms for the trades and the PC Update. During the optimisation all these variables are optimised. In principle the optimisation is exhaustive so that the absolute maximised PV is found. In the end I end up with 100 CER values ranging possibly from 100/0 to 0/100. Some stocks that were selected would have gone to zero value manybe. In reality one would allow his judgement to intervene when stock dive deep so for the optimisation that could be used for Bailing out, or at least some other way to deal with it. . .this could simply be normal investing till the cash runs out. . .this would appear the best option for the experiment I think.

I would the average all the CERs and see what comes out of it. Then I note the sum of the 100 PV's.

Next I would redo the 100 optimisations but start each run with a 50-50 CER and at the end I would calculate the sum of the 100 PV's and compare. Would you predict that the optimised PV-sum for the 50-50 CER Start Option would be generally higher than the PV-sum of the optimised CER Start Option? This would in my opining be a good way to test our statements on the question what the Optimum Star CER would have to be. If you suggest that 100 optimisations is not enough we do a 1000 of them. I woukld just do it as often as you think is enough or as far till we get the result of optimisation of all the stocks in the world that existed 10 years ago. That should settle the issue for anyone that can not completely follow your answer to the question on the optimum value for the initial SER for any stock.

The same type of experiment could be done with Standard AIM or any of its derivatives.