|
PureBytes Links
Trading Reference Links
|
Perry Kaufman says the same when he states for trend following approach :
"Stop losses are counterproductive. Say you read somewhere that you should
only risk 1% of your trading capital on any given trade and you have $25,000
in trading capital. You've calculated that you can only risk ù250 on a
trade. You're arbitrarily putting in a number that's convenient to you but
has nothing to do with the market. When you look at the price fluctuations
of a particular market, you'll notice that the fluctuations are close to the
random distribution that you would associate with noise. With that
information, you can predict that a $250 stop-loss will be hit twice as
often as a $500 stop-loss, which in turn, will be hit twice as often as a
$1,000 stop-loss. So it comes down to the closer you set your stop-loss, the
more often it's going to be hit. This has very little to do with your
trading system. Setting small stop-loss almost guarantees that you're going
to be stopped out.
The safest thing to do is to first test your system without any stop-loss
method superimposed on it. Basically, look at the natural risk in your
system. Your system should go long when market goes up and go short when
market declines. Accept the risk of the system and plan for it, or plan for
twice as much risk or three times as much risk and then trade it that way.
The system itself does some sort of natural adjusting to market events. If
you try to put a stop loss on top of a system, you end up getting stopped
out when the system itself still says the trend is up."
So then you're faced with a dilemma. Of either trusting the signal or trying
to get back in. That's not a normal trading signal because the system has
already given you a trend signal, and if you decide to get back in, you then
have to decide how to do so. If you stay out, you may miss an opportunity.
If you get back in, then there was no reason to get out. Either way you're
no longer performing according to your system profile, and you can no longer
expect the results generated during testing."
"Q : Is there any situation where a stop-loss method might make a difference
?
Perry Kauffman : My experience indicates that when you use a stop losses you
don't help the system; The only stop-losses that ever seem to work in
testing are the ones that get you out of a catastrophic move.
"
So it seems that Perry Kaufman agree with Gallagher or vice-versa on this
question, that a stop-loss is better to be not a simple $ or % money
management stop-loss but rather, if needed, a one that is pertinent to both
the market breath/noise/ direction and the system logic.
Now the psychological- emotional impact on the trader's choices to
risk/reward, losses and drawdown remain the main question.
It seems that there is no general answer, nor general path to a unique
personal goal and constraints; only personal one.
Philippe
> -----Message d'origine-----
> De : David Colin [mailto:davidcolin@xxxxxxxxx]
> Envoyé : lundi 27 septembre 2004 19:07
> À : jack zaner; Omega-list@xxxxxxxxxx
> Objet : Re: Risk
>
> Well, Gallacher's point is that in the long run, you'll be taking _many
> more_ losses by placing stops this way (adding up to the same amount as if
> you had placed stops "correctly"). For instance, in your example, what
if
> the market retraces $1500 from your entry 3 times and then goes on to a
> huge gain? You've lost just as much (more actually). In the grand scheme
> of things, the initial trade was actually a winner that didn't have enough
> room to breathe. As I mentioned this concept seems borne out in my own
> testing.
>
> -David
>
>
>
> At 09:54 AM 9/27/2004, jack zaner wrote:
>
> >Hi Dave: Let's say I am trading a simple SAR system and I buy a long
> >breakout with a $4000 risk. If I wait until the short reverse signal, I
> >lose $4000. If I put in a $1500 stop, haven't I saved $2500- - all other
> >things being equal. I agree that arbitrary stops have no relation to
market
> >action- - they are more of a psychological crutch- - but they can, and
do,
> >prevent catastrophes. Much more complicated than Gallacher makes it
seem.
> >Regards, Jack.
|