|
PureBytes Links
Trading Reference Links
|
Keith, AW,
Thanks for the kind words. I'm glad that you both were able to gain
from it.
Preston
--- In equismetastock@xxxxxxxxxxxxxxx, "Keith Burris"
<Brokenrope@xxx> wrote:
>
> Preston;
> I think I learned more from your explanation about the
gobbledegook
> that we refer to as Metastock Formula Language than in hours of
> struggling through the books. Thanks.
>
> -- Keith
>
> --- In equismetastock@xxxxxxxxxxxxxxx, pumrysh <no_reply@> wrote:
> >
> > Gladly Paul,
> >
> > The theory:
> > In order to establish a normal value for something we need to
look
> > at its extreme low and high values over a period of time. It
really
> > doesn't matter what it is. In the case of rain we look at the
> Palmer
> > Index, which is what Mark had asked about. Its another way of
> saying
> > on average what value should we be seeing...are we above or
below
> > that. 50 is the norm.
> >
> > The principle:
> > Let's look at the indicator line by line using the following
> >
> > {1}(Fml( "Zero Lag MACD" )
> > {2}-LLV(Fml( "Zero Lag MACD" ),48))
> > {3}/(HHV(Fml( "Zero Lag MACD" ),48)
> > {4}-LLV(Fml( "Zero Lag MACD" ),48)
> > {5}+.0000001)*100
> >
> > Lets first look at line 3 and 4. Here we take the highest value
> over
> > 48 days and the lowest value over 48 days. This is now our range.
> >
> > Lines 1 and 2 establish where we are today. It will tell us how
far
> > we are from the bottom.
> >
> > Next we need to establish some type of value that we can relate
to
> > in a percent. So in line 3 you will note a division sign which
is
> > used to divide our range into our present position. The
resulting
> > output would be in decimals.
> >
> > In line 5 we do two very important things. We first add .0000001
to
> > our range value. This is done to eliminate a division by zero
> error.
> > We could have also used +prev-prev to do the same thing, again
for
> > the same reason. Finally we are going to multiply our result by
> 100.
> > This gives us a scale that is from 0 to 100.
> >
> > Normalizing or indexing in this fashion is a great way to
determine
> > what an indicator is doing. There are problems with the method
> > though that you should consider. One is the period of time. In
our
> > case we are only looking at 48 days worth of values. You really
> need
> > to consider whether this is going to be an appropriate amount of
> > time to establish a norm. A true statistician may feel that this
is
> > not enough time and may wish to increase the lookback period.
> >
> > Another problem with this form of normalizing is the range. We
are
> > basically range bound between 0 and 100. You will notice often
> times
> > that the indicator flatlines at either the top or bottom. This
is
> > when you are range bound. Its important to consider what is
> > happening at these points. If it happens often then it is
possible
> > that you are using a lookback that is too short.
> >
> > Another way of correcting the problem is to use a different
scale.
> > Instead of the 0 to 100 scale where 50 is your midpoint, you
could
> > use a bipolar scale. The midpoint of a bipolar scale is 0 with a
> > positive and a negative deflection. The beauty of using this
scale
> > is that you now are using a +/- scale of 100. It will allow for
a
> > much wider scale.
> >
> >
> > Hope this helps,
> >
> > Preston
>
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/equismetastock/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/equismetastock/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:equismetastock-digest@xxxxxxxxxxxxxxx
mailto:equismetastock-fullfeatured@xxxxxxxxxxxxxxx
<*> To unsubscribe from this group, send an email to:
equismetastock-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|