[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Capturing more of MFE (was: efficient exits)



PureBytes Links

Trading Reference Links

I got more ideas from that one post than all the posts in the past 6 months
combined.

-----Original Message-----
From: Gray, Gabriel [mailto:gabriel@xxxxxxxxxxxxxxx]
Sent: Tuesday, October 05, 2004 7:14 PM
To: Mark Johnson; omega-list@xxxxxxxxxx
Subject: RE: Capturing more of MFE (was: efficient exits)


Is that all you got Mark? Stop holding out.

Gabriel

-----Original Message-----
From: Mark Johnson [mailto:janitor@xxxxxxxxxxxx]
Sent: Tuesday, October 05, 2004 6:39 PM
To: omega-list@xxxxxxxxxx
Subject: Capturing more of MFE (was: efficient exits)


One generic approach that seems to work well
is the following.

1.  Add two new parameters to your system,
     "Thresh" and "tighter".

2.  Add a rule that when a trade moves in
     you favor by "Thresh" amount, you'll
     move your stop tighter (protecting more
     of your profits) by an amount "tighter".


You can make "Thresh" be (some number of ATR's),
or some constant times (the Linear Regression Slope),
or some constant times (the 20 day range (HH20 - LL20)),
or some number of standard deviations of price,
or some fraction of the move from the Pivot Low to your entry, or some
number of days on which the trade was net profitable, or some multiple
times your initial stop, or some fraction of the average MFE for your
system, or any other darned profit target you can dream of.


You can make "tighter" be (reduced #days in a moving avg exit) or
reducing the chandelier height in the chandelier stop, or reducing the
#days in a CBO exit stop, or adding a new exit, very tight CBO, after
you hit the Thresh, or reducing the get-me-out threshold of the Yo-Yo
exit, or boosting the acceleration parameter in the Parabolic stop, or
modifying the indicators that get you out (Stochastic, RSI, DMI)
     such that they'll get you out quicker,
or any other darned method of becoming quicker to exit

I realize that it's controversial to add parameters
to a system.  But if you try it I think you may be
pleased with the results.

  -- Mark Johnson