|
PureBytes Links
Trading Reference Links
|
Tom,
I input the MESA formulas in Metastock and found the results to be less than
hoped for as well. I think he did carefully choose his data for the TASC
article. As a sidenote, I recently developed a S&P 500 futures neural net
that seems to be doing excellantly at predicting the start of major/minor
uptrends +/- a few days, many times acting as a leading indicator (go to my
website www.mindspring.com/~blee7/ and view "High Probabability Long Entry
Setup.gif"). It was last updated as of the close last Friday indicating a
highly probable near term uptrend. Most movements of the Signal off
the -1.0 base tend to indicate the beginning of a new uptrend, again +/- a
few days. I plan to update the chart maybe 2-3 times a week just to see how
well it continues to track. 333.gif shows the training data period, 222.gif
shows the testing/validating period and 111.gif shows the true one year
out-of-sample period performance on the the day I developed the model. Not
an infallable model, but it tends to give decent entry signals. Like MESA
claims, it tends to work in both trending and nontrending modes. I have yet
to develop its counterpart, a "Go Short" model. - Regards, Brian
-----Original Message-----
From: Tom Strickland <tstrickland@xxxxxxxxxxxxx>
To: metastock@xxxxxxxxxxxxx <metastock@xxxxxxxxxxxxx>
Date: Tuesday, April 18, 2000 10:26 AM
Subject: RE: === MESA
>I had the same experience at first. I copied the code from the Equis web
>site and got the "bad opcode" error message several times. However, it
>finally plotted correctly and I haven't had an error message since. It
>worked the first time when I plotted the instantaneous trendline prior
>to plotting the sinewave indicator. Don't know if that's important but I
>kept doing it that way. Even when it works, it's slow to plot - must be a
>lot of computation involved.
>
>Once I got it to work, I put it in a template and applied it to several
>commodities. I didn't find the results to be nearly as clear-cut as those
>carefully chosen examples presented in TASC.
>
>I'd be interested in the experience of others.
>
>Tom Strickland
>
>
>
>
>At 11:20 PM 4/17/2000 -0600, you wrote:
>>I tried to input the MESA code. The instantaneous trendline worked just
>>fine, but I keep
>>getting the error message (bad opcode) on the Sinewave Indicator. Has
>>anyone else had this
>>error? Or does anyone know why I am getting this error?
>>
>>
>> > ** Original Subject: RE: === MESA
>> > ** Original Sender: Kevin243@xxxxxxx
>> > ** Original Date: Mon, 17 Apr 2000 19:08:47 -0500 (CDT)
>>
>> > ** Original Message follows...
>>
>> >
>> > The latest MESA code as shown in Stock & Commodities Mag can be copied
>> from
>> > the Stock & Commodities web site.
>> >
>> > Kevin Campbell
>> >
>> >
>> > In a message dated 4/17/00 8:34:17 AM Central Daylight Time,
>> > Michel.Amelinckx@xxxxxxxxxx writes:
>> >
>> > > There are some interesting articles here on the site you mentioned
>> > > http://www.mesasoftware.com/pub/index.html I believe he explains
many
>> > > indicators included in MESA plus others and some have ONLY the
>> EASYLanguage
>> > > code however. Did you tried to put this in Metastock already Ton ?
>> > >
>> > > Greetings
>> > >
>> > > Mickey
>> > >
>>
>> >** --------- End Original Message ----------- **
>>
>> >
>>
>>
>
>
|