Bob Gammon
Tom just mentioned your name, and it took a while for me
to connect you to a note from a friend of mine who
attended the LV conference
he said you gave a presentation talking about trading
aim on a daily basis
I really dont know exactly what parameters you had setup
except for the very low SAFE's. also I know that one of the
problems you mentioned was running out of cash easily
well, I think I have fixed the cash problem, which I also
ran into myself with these studies
the are two elements to ficxing the problem:
1) THE TREND
AIM is simply too fast recognizing trend changes on the daily charts, so you are always buying too high and selling to low. to counter that, I added an excel regression slope
column (13 days) to the spreadsheet, and then modified the
sell rule as follows:
-don't sell unless the slope is +.25 or better, and don't
sell if today's price is less than yesterday's. (ie don't sell into a downtrend and don't sell a downtick in an up trend)
buy rules were changed accordingly
2) THE PORTFOLIO CONTROL
for reasons I don't clearly understand, portfolio control
just grows too darn fast for daily trading - any value
from 10% to 25% works fine over a year and 1/2 or so, dont have test data for longer periods
anyway, these two things taken together for daily
trading end the cash crunch.
I suppose using margin does the same thing
btw, using either of these methods on a weekly AIM trade
will kill your results. also, I don't find that the results are scaleable (ie they do well for 50K and 10K accounts, but
performace relative to weekly trades begins to fail with a 100k account)
Bob Wallace |
