|
PureBytes Links
Trading Reference Links
|
Benzie,
I don't know if you saw my building block posts on how I construct my
trend channels and set my targets and stops or not. I'll copy them below.
First of all you need a least a couple of weeks data to construct the
channel. For long positions, I will move the end point to the right
whenever there is a new high for the channel. If a new high isn't made
during the day, I don't change the channel. Therefore, the price moving up
will affect the channel, but the price moving down won't. However, I do
look at the channel every night and change the target and stop based on
where the next trading days data would plot. I only move the targets and
stops up in whole number increments with the target always being a whole
number and the stop always being a x 3/4 number. In theory, you should
never lower a target or stop, but I do cheat occasionally <G>.
JimG
-----------------------
I wanted to exchange some ideas on setting targets and stops based on
short term trend channels. First let me say that I always set my stops this
way. However, I often set my targets based on old highs or lows, but that
is fairly obvious and not the subject of this discussion. Also, for this
discussion I'll talk about up trend channels and long positions, but down
trend channels and short positions are just the mirror image.
I've been experimenting with using the MetaStock Standard Deviation
Channel for setting targets and stops. The main problem with most trend
channels is deciding where to start and end them. I don't like the idea of
just moving the end points around until the channel looks right, that's way
to subjective. I want any system I use to be very objective and not subject
to interpretation. After some experimenting, I came up with the following
methodology. First, I click on to the standard deviation channel on the left
tool bar. Then, I start the channel by placing the cursor immediately to the
left of the lowest low in the data being considered and hold the left mouse
button down. Then I stretch the channel to the right by moving the mouse to
the right while still holding the button down. Finally, I end it immediately
to the right of the highest high in the time frame by releasing the mouse
button. Then I set the deviation at 2 and extend the channel to the right.
Anytime a new high is hit, I'll drag the right end of the channel just to
the right of that new high. Using this methodology, I know that the trend
channels will be constructed the same way every time.
Now for setting the targets, I've been experimenting with setting the
deviation at 2 and setting the target at the whole number just under the top
of the channel. I like to use the whole numbers because specialists like to
take a stock to whole numbers to shake out the stops and limits. I then
move it up to the next whole number just as soon as the channel will let me.
I set the deviation at 2 because I like to let my profits run and not close
them because they hit an arbitrary target. A channel with the deviation set
at 2 will contain 95% of the stock moves. Therefore, the odds are that if a
stock moves outside the top of the channel it will correct soon. Of course
I'll miss an occasional rocket, but I'm willing to pay that price to have
most of my targets near the stock top.
For setting the stops, I set the deviation at 1 and set my stop at the
x 3/4 point just under the bottom of the channel. The reason for the 3/4
number under the channel is again that specialist like to drive a stock to
whole numbers, so a stop 1/4 under a whole number is less likely to be hit.
The reason for setting the deviation at 1 is that will contain 67% of the
stock moves. I want my stop to trigger quicker in the bottom half of the
channel than my target does in the to for two reasons. First, the long-term
market bias is up and I reinforce that by only taking long positions in
stocks that are trending up when the market is trending up. Therefore it
makes sense to let a stock run further on the upside than on the downside
under these conditions. Second and more import, the stop is there for
protection - to protect a profit or protect against a large loss. I want
that protection to kick in as soon as possible while still minimizing whip
saws. Picking a point that will contain 67% of the stock moves seems like a
reasonable compromise.
Another point I want to make is the amount of data necessary for
constructing the STUTC. I think the mistake I made on MSFT this week was
that I changed my channel starting point from the 10/08/98 low to the
11/16/98 low which I thought was the pivot point for the latest breakout.
The problem with this was that there was less than two weeks of data without
and significant reaction low. Therefore MSFT closed below my mental stop
Thursday. Luckily, Friday's action kept me in MSFT. If I hadn't moved my
starting point, MSFT would have remained well above the stop from that
channel. I'm going to make the general rule that I need at least 21 days of
data to construct a STUTC. I can fudge on that a little if there has been a
good reaction low in the data. I can ignore the rule entirely if the stop
is on a stock for which I haven't set a target and I want to tighten the
stop to protect a large profit. If I do that, I should recognize that the
cost could be a false stop and I should be willing to re-enter the stock
when it sets a new high.
That brings up two more points. Deviating from the system and not
setting any target. First anyone using a system for targets and stops must
thoroughly understand the system and be comfortable with it for it to work.
If you understand it you are more likely to stick with it most of the time.
When you do deviate, you only do so when it makes sense and with a full
understanding of the risks and potential corrective actions if wrong.
There are times when I feel so strongly about letting my profits run
that I won't set any target. This is usually when a stock has broken out to
a new all time high on good momentum. In that case I just set the stop as
described above and hope it won't get hit for a long time. Years is fine
<G>. Some would say that it never makes sense to set targets, but I
disagree with that. My experience is that reasonably thought out targets
pay off.
My closing point is that targets should be set as limit orders with
your broker but stops should always be mental stops. The reason is that if
you have a day job like I do, you can't watch the market all the time. You
want your position to be closed when the target is hit, so a limit order is
called for. Now to keep down the number of limit orders, I never submit
them until my stock gets within reasonable distance of the target. The
reason that stops should be mental is that I firmly believe that specialist
do gun stops. They like to take a stock down to shake out the stops just
before allowing it to run. Therefore my stop is mental and I require the
stock to close below the stop before closing the position the next morning.
The cost of this is that you can get burned if a stock is dropping fast, but
I believe you save more in the long run by avoiding shakeouts. I like to
think I'm fairly rigid in this except I do, as in the case with MSFT, take
the next mornings market action into consideration. Again you need to
thoroughly understand what you are doing when/if you violate the system.
That's it, any comments pro or con? Does anyone disagree with setting
targets or stops? If so, why? What methods do you use and why?
JimG
------------------------
From: Jim Greening <JimGinVA@xxxxxxxxxxxxx>
To: Metastock <metastock@xxxxxxxxxxxxx>
Subject: Building Blocks - Trend Channels
Date: Sunday, December 13, 1998 12:52 PM
All,
Last week I talked about how I use trend channels to set targets and
stops to tell me when to exit a position. This week I'm going to go back to
the beginning to discuss how I use trend channels to tell me when to enter a
position. I'd appreciate any comments pro or con on this methodology.
I've come to depend more and more on trend channels as my primary
technical analysis tool. The theory of trend channel investing is simple.
First you have to construct the trend channel. After that is done the rest
is easy. For up trend channels you open a long position right after a stock
bounces off the bottom of a channel and close the position when it hits the
top of a channel. Vice versa for down trend channels. I treat horizontal
channels like up trend channels for trading purposes.
Of course, I expand on this simple methodology slightly <G>. First I
construct short, intermediate, and long term channels on all my charts. My
definition of short, intermediate and long term is flexible. In general
short term is days to up to a few months, intermediate term is a few months
to over a year, and long term is years. I like to open a position in the
direction of the long term trend channel when an intermediate term trend
channel is broken and the stock reverses in the direction of the long term
trend. Since this is a fairly rare occurence, my secondary method which
occurs much more frequently and is, therefore, the one I use the most, is to
enter a position when the short term trend channel is broken and the stock
reverses in direction of the intermediate term trend channel. In both
cases, the reversal is a better buy signal when it occurs in conjunction
with a bounce off the bottom of the longer term trend channel. Once a stock
has reversed its trend, I construct a new short term trend channel in the
direction of the longer term trend. I use this new Short Term Trend Channel
to set my targets and stops as I described last week if I entered a position
on the breakout of the old channel. If I missed entering the position at
the brekout, there are additional entry points every time the stock bounces
off the bottom of the new short term trend channel.
As you can see, using trend channels for signaling when to enter a
new position is straight forward and simple. The trick is how to
objectively and consistently construct the trend channels. For short term
up trend channels I use the Standard Deviation Channel built into MetaStock.
I start the channel immediately to the left of the lowest low in the data
being considered and end immediately to the right of the highest high in the
time frame. I set the deviation at 2 to determine the targets and at 1 to
determine the stops. I usually leave it at 1 for looking at the charts
since the stop is more important than the tartget. I extend the channel to
the right. Anytime a new high is hit, I'll drag the right end of the
channel immediately to the right of that high. After a few months with at
least two good reactions, I switch to Trader Vic type trend lines to
construct my short term channels. I only use Trader Vic type trend lines
for intermediate and long term channels. I like the Trader Vic methodology
much better than using a standard deviation channel, but it gives many more
false signals than standard deviation channels do for very short term
channels without any major reactions in the data.
For a Trader Vic up trend line, start with the lowest low in the time
frame being considered. Draw a line from that low to a low before the
highest high in the time frame such that the line doesn't pass through any
data. It's important that you don't draw the line through any data or to a
low that is past the highest high in the timeframe. A down trend line is
just the mirror image using the highs in the time frame. To construct the
up trend channel, extend the line to the right, then draw a parallel line
through the most extreme high in the time frame such that the line doesn't
intersect any other data points. This is easy with MetaStock. You can hold
the CTRL key down, then hold the left mouse key down once you are on the
original line and drag a parallel line to where ever you want it. You can
do the down trend channels in a similar manner. Horizontal channels are the
easiest. Just drop a horizontal line on the highest high and lowest low in
the time frame. My final touch is to change the line color and style: blue
and dashed for short term channels; red and dashed for intermediate term
channels; and magenta and dashed for long term channels.
That's all there is to it, what do you think? Does it make sense?
Jim
-----Original Message-----
From: bspark@xxxxxxxxxxxxxxxxxx <bspark@xxxxxxxxxxxxxxxxxx>
To: metastock@xxxxxxxxxxxxx <metastock@xxxxxxxxxxxxx>
Date: Tuesday, January 05, 1999 5:35 PM
Subject: Re: Jim Greening Trading Channels
Jim,
Could you please give me an idea of how you set targets using
Standard deviation channels. As the price starts to move Up (or
down) the channels upper and lower bands will change position, do
you dynamically shift your targets up as the band moves up or do
you fix the target price on the initial Standard deviation set at 2.
Once the channel has been in place for a while, and the lines are
extended, it is quite easy to place a target price, but as the
channel starts to form, it seems to move around quite a bit.
Regards
Benzie
************************
Brightspark
sales@xxxxxxxxxxxxx
Voice: +61 9 375-1178
Fax: +61 9 375-1668
Unit 7, 11-13 Marchant Way
Morley WA 6062 Australia
Visit our WEB Page
http://www.bspark.com.au
************************
|