Re: 'switches'
Having done testing with the AI program's backtesting function I've observed that sometimes using nothing, MACRO on/off, volumizer on/off (or various combinations of such) plus tweaking the amount of initial cash reserve can cause the returns to vary considerably, sometimes to the good, sometimes not so. The problem, of course, is that there's no way to know if the benefits of a particular set of "switches" being thrown one way or another will persist into the future, or that the run you've made, whilst impressive, is as good as it's going to get. Similarly, there's no way to know in advance if any particular combination makes sense either. So perhaps simplicity is the best of all.
At some point I'd think even AIM will run into the law of diminishing returns - for all our efforts to wring out another 1% of performance - is it worth the time and effort expended?
I keep leaning toward the idea of some self-directed system that would automatically adapt to the market but if such a thing had been developed - well I doubt if we'd hear about it. Even if we did we couldn't likely afford it. Besides, if it was too widely used would the markets grind to a halt? We've sort of touched on this before so i don't want to restart the whole thread again, per se. Still it's one of those ideas that sounds so simple in theory but to make it go, well, to quote The Brain, "Therein lies the difficulty, therein lies the genius!" Narf.
Is 1% worth the time or effort? Yes if it's 1% of 1 million! ($10,000). That ain't chicken feed to me.
As for self directed systems, I think Mark Hing (Automatic Investor) has mentioned Neural Network approaches to this. I believe there may be other products out there that purport to use NN. I don't know of any specifics, because the concept is beyond me other than I think that the software 'learns' and 'adapts'.
Best Regards, Steve (The Grabber)
