|
PureBytes Links
Trading Reference Links
|
One clarification here, forgive me if this is too obvious. Cycle length
changes when trend changes. As the trend period lengthens then the cycle
length increases and vice versa. So if the trend has been up or down and is
about to change, then you want to look for the hint of the cycle length to
shorten. That is how I would use the Hilbert period, along with some
divergence stuff.
BobR
----- Original Message -----
From: "BobR" <bobrabcd@xxxxxxxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Tuesday, February 29, 2000 6:04 AM
Subject: [RT] Re: Gen:Adv Dcl FFt
> Using the John Ehler's Hilbert Period ela in a recent TASC magazine the
> period of the nyse advances minus declines makes an interesting picture.
> The current period as of Monday close is 15 days. It cycles from a low of
> bout 8 to a high of about 17.5. Instead of using (H+L)/2 in John's code
use
> the C of data2 - C of data3 where data2 is adv and data3 is decl. Then
> making another plot of the Hilbert period using up volume minus down
volume
> reveals that the volume period can be out of sync with the issue period,
i.e
> start getting shorter before the issue period, like before corrections,
etc.
>
> Enjoy,
> BobR
>
> ----- Original Message -----
> From: "Ronald McEwan" <rmac@xxxxxxxx>
> To: <realtraders@xxxxxxxxxxxxxxx>
> Sent: Monday, February 28, 2000 8:49 PM
> Subject: [RT] Gen:Adv Dcl FFt
>
>
> > I ran a Fourier Analysis of the daily NYSE advancing issues minus
> > declining Issues to see if there were any useful component frequencies.
I
> > came up with a frequency of 78 days. If the holds out the market (S&P
> > 500) will have a change of direction on or around May 30, 2000
> >
> > Ron McEwan
>
|