|
PureBytes Links
Trading Reference Links
|
>From: Michael Crain <MichaelC@xxxxxxxxxxxxx>
>To: "'Robyn Greene'" <greene@xxxxxxxxxxxxxxx>, Jim Allen
><jallen7@xxxxxxxxxxxxx>
>CC: Mark Brown <markbrown@xxxxxxxxxxxxx>, omega-list@xxxxxxxxxx
>Subject: RE: Aberration & Bet Size
>Date: Wed, 26 May 1999 13:32:39 +0200
>
>I have to agree with Robyn "a relatively small percentage of the trades
>accounted for a relatively large percentage of the gains....... I think
>it's
>prudent to throw out a handful of best and worst trades to eliminate the
>possibility that the system results aren't simply the result of luck (good
>luck or bad luck).". The upper and lower "extreme" results can skew the
>results of a study. You can naturally use various methods statistical or
>otherwise to do it. I personally remove the upper and lower 5 percentile
>(just my rule of thumb).
>
As a trend trader I "assume" that markets go sideways then trend up or
down. So any trading logic I write tries to bring me through the side-
ways period with some kind of money still in my account (very careful
money management), the higher expectation of said logic though is to
make sure a position is in place for the once in awhile major trend,
if the system even hints of missing one of these moves it is immediately
discarded. Abberation will do this okay, but in my view
needs more emphasis on money management. I trade weekly data.
|