Replies to post #922 on A.I.M. Users Bulletin Board (AIMUSERS)

[extelecom]

Aptus, All, I think backtesting is a great way of testing. It would be nice to have a standardized test scenario. As was pointed out, there are many variations (and interpretations) of AIM here. Subtle changes can have a large impact (Vealies, delayed buying, minimum trade sizes, safe diddling) on the "end" results. I am convinced that there is no "Best" way to use AIM as there are too many variables. At least with a standard test set one can compare Apples to Apples. I think the easy way to backtest with Mark's software is a great step forward.

et

[Conrad]

Hi Mark,

I agree with everyting you say, except that back testing does not prove as much as it may appear to imply...with back testing you know what the stock is going to do and you can fiddle with the knobs to optimise for the dips and crests. This is precisely why I have come back on my idea of optimizing an AIM structure. Certainly it will work if the stock exhibits a recognizable pattern. So, my point is that if the stock behaviour is known we do not need any system, except Buyl Low Sell High.

So I still have that dilemm: How can we test a system if the stock behaviour is completely unknown?

Conrad