|
PureBytes Links
Trading Reference Links
|
The original RSI version, compliant with Wilder's book is the RSI ( and NOT the
so called RSI Classic)
Due to various misunderstading over time, some developer used the classic
version that removed the pseudo exponential smoothing, used by Wilder in most of
his work ( Dmi + - used the same).
Sincerely,
Pierre Orphelin
www.sirtrade.com
Tradestation 2000i, TradeStation 7 sales and support
Safir-X, neurofuzzy logic trading system builder
To subscribe to our FREE trading system newsletter:
http://www.sirtrade.com/newsletter.htm
-----Message d'origine-----
De : unicorn@xxxxxxxxx [mailto:unicorn@xxxxxxxxx]De la part de Alex
Matulich
Envoye : mardi 5 aout 2003 04:00
A : omega-list@xxxxxxxxxx
Objet : Something's wrong: RSI vs RSIClassic
Has anybody ever compared the output of the RSI to RSIClassic functions?
For example (using lookback of 6 bars):
Price RSI RSIClassic
984.50 32.46 18.18
984.25 28.74 20.00
984.50 37.35 37.50
984.75 45.29 28.57
984.75 45.29 28.57
985.00 53.73 42.86
985.00 53.73 75.00
984.75 43.96 75.00
Not even close to being the same!
So the question becomes: which is the "correct" RSI? And why?
It looks like TS tried to take some short cuts with the RSI
function, updating a couple of running averages without
recalculating them afresh on each bar. The result looks like it
uses exponential averages instead of just sums. Anybody have any
thoughts?
This has bedeviled me all day (why my own code, which has its own
RSI calculations, isn't working how I expect), until I just now
plotted them out together.
--
,|___ Alex Matulich -- alex@xxxxxxxxxxxxxx
// +__> Director of Research and Development
// \
// __) Unicorn Research Corporation -- http://unicorn.us.com
|