|
PureBytes Links
Trading Reference Links
|
Walter,
You've sure got the "adaptive" religion ;-) I think what you are looking for
is Cynthia Kase's PeakOscillator which uses the somewhat "unclear" Random
Walk Index. The RWI uses ATR as it's divisors.
The concept behind it is that by using a statistical measure of trend
instead of variations of the MA for the oscillator you can get historically
meaningful values for "oversold" and "overbought". The overcomes the basic
defect with all normalized momentum indicators (that they are useless beyond
the lookback period) and allows you to backtest against the historical price
series with valid results.
The devil is in the details though as the authors of both the formulas never
really lay them out. The RWI formula as is it presented in MS6.5 is just the
commonly accepted "best guess" of Poulous and Saitta's work, while Ms. Kase
never fully divulges how she works the RWI(high) and RWI(low) into her
oscillator since it seems to require only a single value of RWI.
Intuitively, you would just split the differerence between the RWI(high) and
RWI(low) but I sense this is "bad" math.
For more info, there was a long thread on this on the Omega list about a
year ago you could search or you may try contacting Ms. Kase through her web
site. Let me know if you solve the riddle!
hope this helps,
Rick
Tokyo, Japan
-----Original Message-----
From: Walter Lake <wlake@xxxxxxxxx>
To: metastock bulletin board <metastock@xxxxxxxxxxxxx>
Date: Friday, September 18, 1998 1:26 AM
Subject: ATR Volatility Indexing
>Thomas Stridsman's latest Futures Mag article discusses a standard
deviation
>based Dynamic Breakout System (DBS) and an Average True Range based DBS,
>with an RSI based "switch" to move from one system to the other system
>according to "up volatility" Vs "down volatility".
>
>Does anyone have any experience or information about ATR volatility
>indexing. Etzkorn briefly mentions it in his latest book also. Chande and
>Kroll instead refer to volatility indexing using standard deviation, CMO
and
>R-squared.
>
>Also, does anyone have any information on indexed ("adaptive") over-bought
>and over-sold lines used with "adaptive" Oscillators.
>
>Thanks
>
>Walter Lake
>
>
|