|
PureBytes Links
Trading Reference Links
|
Jim,
See also
http://groups.yahoo.com/group/amibroker/message/5432
from Oct 11, 2001 [well, 20000 messages in one year !!!]
for a moving AVRSI line, instead of the fixed 50. For many RSI curves
50 is not the realistic average for extended time periods and AVRSI
may be more expressive.
DT
--- In amibroker@xxxx, jvarney <jvarn359@xxxx> wrote:
> Dmitris Tsokakis' post (25590) on RSI
> gave me an idea. Instead of fitting RSI
> to a parabola, why not fit it to a
> standard deviation distribution? It
> seems obvious to apply a Bollinger Band
> concept to RSI. The bell curve should
> fit well, with the peak at 50 and the
> tails at 0 and 100.
>
> If you look at the attached chart,
> the dynamic RSI threshold has one
> characteristic I like. On vigorous
> breakouts the bands expand and allows
> the stock position to run. As RSI
> stalls and fades near the top, the bands
> contract and trigger later than a
> static threshold would. See the January
> to March trade.
>
> Good Trades,
>
> Jim
>
>
> // --- begin AFL ---
>
> // Dynamic RSI Thresholds -- J. Varney
> //
> //
> /*
> Instead of the traditional fixed RSI = 30 and 70 thresholds to
define
> oversold and overbought, this AFL uses standard deviation to
determine
> dynamic overbought and oversold theshold lines.
> */
>
> t = 10;
> myRSI = RSI(t);
> Adev = StDev(myRSI, 3*t);
> Hiline = 50 + 1.8*Adev;
> Loline = 50 - 1.8*Adev;
> Plot(myRSI,"",colorBlue,8);
> Plot(Hiline,"",colorRed,1);
> Plot(Loline,"",colorGreen,1);
>
> Title = " " +Name() +
> " Dynamic RSI Threshold, t = " +
> WriteVal(t, format=1.0) + " ";
>
> // --- end AFL ---
>
>
> __________________________________________________
> Do you Yahoo!?
> Faith Hill - Exclusive Performances, Videos & More
> http://faith.yahoo.com
|