[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Correlation Coefficient



PureBytes Links

Trading Reference Links

Thank you for this well extensive explanation and too for active attending
the List again. We, as your users, rely gladly on your contributions in general
and in this at the List, in particular.

Perhaps in the near future you can inform the List about the new Bulletin Board,
announced, but not yet made available at your excellent website.

Regards,
Ton Maas
  
-----Oorspronkelijk bericht-----
Van: Equis Support <support@xxxxxxxxx>
Aan: 'metastock@xxxxxxxxxxxxx' <metastock@xxxxxxxxxxxxx>
Datum: maandag 24 augustus 1998 20:26
Onderwerp: RE: Correlation Coefficient


>You may use the Explorer to find stocks that meet your criteria in
>Correlation analysis.  You must decide what the criteria is.  Please
>understand I do not use Correlation analysis in my trading so I am not
>recommending any parameters.  But, for example:
>
>Place this in your filter:  Correl(P ,C ,x ,x ) >.5 and only those
>issues whose correlation is greater than .5 will be found.  The x's
>indicate the time and shift parameters you must select.  P is the
>independent variable and C is the dependant variable.
>
>You may also want to simply place the Correl(P ,C ,x ,x ) in a column
>which will show you the exact correlation of each issue you explore.
>
>To make this work, you must open a chart of the S&P, or whatever index
>you wish to use and then select it.  When you select it it will have
>"handles" on the ends of the plot.  Run the exploration and the list of
>issues will be those that have a greater than 50% correlation.
>
>Again, select your own parameters.
>
>Equis Support
>
>
>-----Original Message-----
>From: A.J. Maas [mailto:anthmaas@xxxxxx]
>Sent: Sunday, August 23, 1998 4:59 AM
>To: metastock@xxxxxxxxxxxxx
>Subject: Re: Correlation Coefficient
>
>
>As Equis(Support guys and gals) are watching the List hope they can jump
>in this accurate question and provide its users with the missing manual
>pages
>on this subject as an Exploration.
>
>As an Indicator by itselve it is possible. But to my knowledge, this is
>NOT possible
>in an Exploration.
>Though the manual(MS65-p.431) says that it is, I asume it meant this as
>in a
>standard indicator using 2 price plots or using the indicator within a
>Composites
>Chart(MS65-p.151).
>
>A spinn-off workaround can be made by using Tip 5(Exploration
>Tips(MS65-p.364))
>and then when finished exploring, in the report, to RANK(descending way
>down)
>using the Sort button or to click on the horizontal bar right above
>results first listed sec.
>
>Another work-around can be made by OLE linking the Correlation Indicator
>form a chart to Excel. Very time and handling and safety procotions
>consuming though.
>This will give you instant view and results of ALL of your securities,
>if required daily
>updated as you go along in future/time as well. 
>
>A futher solution(but also quite time and handling consuming) is to
>"manualy"
>COPY and PASTE Exploration results to Excel or when looking at a chart
>to
>right click the Corr-Indicator and do a COPY and PASTE to Excel.
>These will give you the results ONLY PER ONE single security, hate to
>think you have
>to do this on 500 secs.
>
>Regards,
>Ton Maas
>Ms-IRB@xxxxxxxxx
>    
>  
>-----Oorspronkelijk bericht-----
>Van: CMA <cma6@xxxxxxxxx>
>Aan: metastock@xxxxxxxxxxxxx <metastock@xxxxxxxxxxxxx>
>Datum: zondag 23 augustus 1998 0:47
>Onderwerp: Correlation Coefficient
>
>
>>I would like to calculate the correlation coefficient (vs. S&P 500)
>>for a number of stocks in an Exploration.
>>Is that possible in ver. 4.5 or 6.0?
>>              CMA
>>
>>
>>
>>
>>_________________________________________________________
>>DO YOU YAHOO!?
>>Get your free @yahoo.com address at http://mail.yahoo.com
>>
>>
>