|
PureBytes Links
Trading Reference Links
|
Von Hef writes:
> >Myself, I try to be pretty mechanical; I use the various systems I'm
> >playing around with, and rarely do I use any fundamental knowledge of
> >the stock. As for an exit, it's a simple one, though I'd be glad to
> >be shown where its weak points are. It's a mix of SAR
> >(philosophically) and trailing low; I start by putting my stop at the
> >lowest low of the last 10 days. If the stock makes a new high, I
> >decrement 10 to 9. Every time a new high is made, I decrement the
> >number of days until I hit 5. So far, this has had me out of major
> >loss turnarounds w/o stopping me out overly quickly.
>
> My goal is to be mostly mechanical also. I like your idea on your stops,
> have you written code for your exit method?
Not really. I have a custom indicator which is just:
n := input("Number of periods:", 2,100,10);
llv(L, n);
which I plot on the charts in which I have a position. (Similar for
high on the short side)
That way, when I check my charts nightly and see that my stock has made
a new high (since I opened the position), I can just adjust the
properties and decrement N. It then replots, and I cancel/change my
stop at the broker if needed, which is simply a GTC stop-loss.
Recently I've been looking a lot at the TTT talk that's been going
around, and Widners support/resistance. Both of those are based on
breaking new highs/lows over the last X days, so it has a Donchian
appeal to it. I have not done any testing into these yet, but using the
"eyeball" method, they seem sound, if a bit shorter term than I am used
to.
--
========================================================================
Men's rules for women: 16. Ask for what you want. Subtle hints don't
work.
http://www.unixgeek.com/cgi-bin/motd.pl - PGP email preferred
|