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

Re: cycles



PureBytes Links

Trading Reference Links


In response to Jean-Roland, I would say first take a look at
http://www.equis.com/free/taaz/fourier.html  
I think the explanation there is pretty clear and it would be
difficult for me to do much better, but I'll try...

Here is an example... apply the FFT indicator to a security
of your choice.  You will get a few "predominate cycle
lengths".  Let's say (for sake of this example) that the
strongest cycle length is 205 days (which is the length
used for U.S. Steel on the web page above).  A very very
rough way to use this is to find the low price over the
past year, then, advance 102 days (half of the cycle length
of 205) and this is where you could roughly expect the price
maximum, then advance another 103 days (to fill the cycle)
and you could roughly expect a (local) minimum.
[if you're good with your computer, you could try overlaying
a sin-wave with the same cycle-length to see where the
maximum and minumum prices "should" be according to the
Fourier transform].

I personally have not used this indicator or tried to analyze
or forecast security prices using cycles, because I'm
on other types of wild goose chases :)



Jean-Roland Quastana wrote:
>Could you tell us mor precisely how to work with FFT indicator, maybe =
>with a practical example.
>I've been trying to understand this indicator for a (too long) while...
>
>Thanks,
>
>Jean-Roland
>  ----- Original Message -----=20
>  From: Jeff Haferman=20
>  To: metastock@xxxxxxxxxxxxx ; metastock@xxxxxxxxxxxxx=20
>  Sent: Tuesday, December 12, 2000 12:34 AM
>  Subject: Re: cycles
>
>
>
>  Chuck,
>  Take a look at the FFT stuff built into Metastock=20
>  (online look at http://www.equis.com/free/taaz/fourier.html)
>
>  This will tell you the predominate cycle lengths for the
>  time series under investigation.
>
>  You can look at the cycles fractally also (see Edgar Peters "Chaos and
>  Order in Capital Markets").
>
>  Good luck,
>  Jeff
>
>
>  Charles Warren wrote:
>  >
>  >Hi,
>  >
>  >I am looking to develop an algorithm for an exploration of stocks =
>that =3D
>  >will determine the frequency (number of time periods) from valley to =
>=3D
>  >valley dips, and where the stock is at relative to a projected next =
>=3D
>  >valley (or peak).  If you manually use the cycle lines feature on a =
>=3D
>  >single stock chart you can place a starting point (date) and change =
>the =3D
>  >frequency of the time intervals among a series of vertical lines, =3D
>  >attempting to match up the peaks or valleys in a cycling stock over a =
>=3D
>  >time frame.  The goal of the algoithm is to determine if the stock is =
>=3D
>  >near a projected valley (or peak) based on its past tendencies to =
>form =3D
>  >valleys or peaks at certain intervals.
>  >
>  >Chuck
>  >
>
>------=_NextPart_000_00B5_01C0643F.1CC406E0
>Content-Type: text/html;
>	charset="iso-8859-1"
>Content-Transfer-Encoding: quoted-printable
>
><!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
><HTML><HEAD>
><META content=3D"text/html; charset=3Diso-8859-1" =
>http-equiv=3DContent-Type>
><META content=3D"MSHTML 5.00.2919.6307" name=3DGENERATOR>
><STYLE></STYLE>
></HEAD>
><BODY bgColor=3D#ffffff>
><DIV><FONT face=3DTahoma size=3D2>Could you tell us mor precisely how to =
>work with=20
>FFT indicator, maybe with a practical example.</FONT></DIV>
><DIV><FONT face=3DTahoma size=3D2>I've been trying to understand this =
>indicator for=20
>a (too long) while...</FONT></DIV>
><DIV>&nbsp;</DIV>
><DIV><FONT face=3DTahoma size=3D2>Thanks,</FONT></DIV>
><DIV>&nbsp;</DIV>
><DIV><FONT face=3DTahoma size=3D2>Jean-Roland</FONT></DIV>
><BLOCKQUOTE=20
>style=3D"BORDER-LEFT: #000000 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: =
>0px; PADDING-LEFT: 5px; PADDING-RIGHT: 0px">
>  <DIV style=3D"FONT: 10pt arial">----- Original Message ----- </DIV>
>  <DIV=20
>  style=3D"BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: =
>black"><B>From:</B>=20
>  <A href=3D"mailto:haferman@xxxxxxxxxxxxxxxxxxxxxxx"=20
>  title=3Dhaferman@xxxxxxxxxxxxxxxxxxxxxxx>Jeff Haferman</A> </DIV>
>  <DIV style=3D"FONT: 10pt arial"><B>To:</B> <A=20
>  href=3D"mailto:metastock@xxxxxxxxxxxxx"=20
>  title=3Dmetastock@xxxxxxxxxxxxx>metastock@xxxxxxxxxxxxx</A> ; <A=20
>  href=3D"mailto:metastock@xxxxxxxxxxxxx"=20
>  title=3Dmetastock@xxxxxxxxxxxxx>metastock@xxxxxxxxxxxxx</A> </DIV>
>  <DIV style=3D"FONT: 10pt arial"><B>Sent:</B> Tuesday, December 12, =
>2000 12:34=20
>  AM</DIV>
>  <DIV style=3D"FONT: 10pt arial"><B>Subject:</B> Re: cycles</DIV>
>  <DIV><BR></DIV><BR>Chuck,<BR>Take a look at the FFT stuff built into =
>Metastock=20
>  <BR>(online look at <A=20
>  =
>href=3D"http://www.equis.com/free/taaz/fourier.html";>http://www.equis.com=
>/free/taaz/fourier.html</A>)<BR><BR>This=20
>  will tell you the predominate cycle lengths for the<BR>time series =
>under=20
>  investigation.<BR><BR>You can look at the cycles fractally also (see =
>Edgar=20
>  Peters "Chaos and<BR>Order in Capital Markets").<BR><BR>Good=20
>  luck,<BR>Jeff<BR><BR><BR>Charles Warren=20
>  wrote:<BR>&gt;<BR>&gt;Hi,<BR>&gt;<BR>&gt;I am looking to develop an =
>algorithm=20
>  for an exploration of stocks that =3D<BR>&gt;will determine the =
>frequency=20
>  (number of time periods) from valley to =3D<BR>&gt;valley dips, and =
>where the=20
>  stock is at relative to a projected next =3D<BR>&gt;valley (or =
>peak).&nbsp; If=20
>  you manually use the cycle lines feature on a =3D<BR>&gt;single stock =
>chart you=20
>  can place a starting point (date) and change the =3D<BR>&gt;frequency =
>of the=20
>  time intervals among a series of vertical lines, =3D<BR>&gt;attempting =
>to match=20
>  up the peaks or valleys in a cycling stock over a =3D<BR>&gt;time =
>frame.&nbsp;=20
>  The goal of the algoithm is to determine if the stock is =
>=3D<BR>&gt;near a=20
>  projected valley (or peak) based on its past tendencies to form=20
>  =3D<BR>&gt;valleys or peaks at certain=20
>intervals.<BR>&gt;<BR>&gt;Chuck<BR>&gt;</BLOCKQUOTE></BODY></HTML>
>
>------=_NextPart_000_00B5_01C0643F.1CC406E0--
>





  • References: