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

RE: [RT] Long (very) Term Fourier Concepts



PureBytes Links

Trading Reference Links



Hello Clyde,

 

As you may recall, years ago I acquired your separate Fourline_MEM and Hurst Channel indicators as a set, then I worked to combine them into a semi-automated design. The intent was to utilize a “nearest neighbor” approach using the top N strongest cycles and combine them on a real-time basis to create a set of prediction bands. The resultant combined code works well in non-realtime situations, but it’s still a bit buggy on live data (the projection lines randomly go asymptotic every X bars) and I haven’t had the time or willpower to debug it. At any rate, I’ve always felt that this approach had merit, particularly when looking for a turning point within an error band. What I did notice is the tendency, as the market cycle analysis changes, for the projection bands to be “pushed” as new bars are added. In other words, the projection would show a turning point as imminent, and keep showing that condition until it simply changed it’s mind, making it difficult to act on it. Similar to how Bollinger Bands can keep stretching out as the trend continues without adding to the knowledge of when a reversal might occur.

 

I have noted over the years that you’ve worked at least some aspect of these ideas into your code. I believe that there could be predictive value when the cycles reverse themselves on real-time short duration bars (or other types of synthetic bars). I would be very interested in any update you might have to document your progress and testing over this development time.

 

With best regards,

Gene

 

From: realtraders@xxxxxxxxxxxxxxx [mailto:realtraders@xxxxxxxxxxxxxxx] On Behalf Of Clyde Lee
Sent: Sunday, April 19, 2009 10:48 PM
To: swingmachine
Subject: [RT] Long (very) Term Fourier Concepts

 




Hurst indicated that a CENTERED MOVING AVERAGE was a proper

method of examining cyclicality in a price series -- I agree.

 

The attached shows two relative long term (186 weeks--3.6 years and 386 weeks

--7.4 years).  These lengths are taken from of MEM evaluation of cycles in the

DOW weekly data.

 

The attached pictures show just how far this approach can project into the

future and whether it is correct or not remains to be seen.

 

In evaluating studies of this length we can only examine if the behavior of

prices lie within a series of "bands" which we calculate around the Centered

Moving Average of prices and the Fourier projection of the selected period

of cyclicality.

 

LOOK HARD AT THESE RESULTS.

 

This approach is not yet a "commercial" product.  Any feedback as to whether

this approach has merit would be appreciated.

 

Clyde

 

= = = = = = = = = = = = = = = = = = = = = = = = = = =
Clyde Lee                                SYTECH Corporation
7910 Westglen                       Phone:  713.783.9540
Suite 105                                    Fax:  713.783.1092
Houston, TX  77063     www.theswingmachine.com
= = = = = = = = = = = = = = = = = = = = = = = = = = =
This e-mail and any attachments may contain confidential and
privileged information. If you are not the intended recipient,
please notify the sender immediately by return e-mail, delete this
e-mail and destroy any copies. Any dissemination or use of this
information by a person other than the intended recipient is
unauthorized and may be illegal.

 

 

 

 



__._,_.___


Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe

__,_._,___