MARY4MONEY
806 posts
msg #51082
- Ignore MARY4MONEY |
4/14/2007 8:18:33 AM
all you have to do is use a high weekly rsi(2) with a low daily rsi(2) to get the rubber band effect- weekly rsi(2) above 90 last 10 days and rsi(2) below 10 will give you 67% winners for the whole market but if you change it to rsi(2) below 7 you get 100% winners thats really amazing
|
TheRumpledOne
6,411 posts
msg #51086
- Ignore TheRumpledOne
modified |
4/14/2007 12:09:55 PM
all you have to do is use a high weekly rsi(2) with a low daily rsi(2) to get the rubber band effect- weekly rsi(2) above 90 last 10 days and rsi(2) below 10 will give you 67% winners for the whole market but if you change it to rsi(2) below 7 you get 100% winners thats really amazing
Here's a filter that the guys who used HG's RSI(2) and NEVER GAVE HIM CREDIT FOR IT DEVELOPED:
|
maxreturn
745 posts
msg #51088
- Ignore maxreturn |
4/14/2007 1:16:22 PM
Ok TRO...I'm hoping you can answer this since I've had no luck getting responses from Mary in the past and you seem to know him quite well. How does Mary quantify winners vs. losers? I'm assuming it's a short term hold but what are the rules?
|
alf44
2,025 posts
msg #51091
- Ignore alf44
modified |
4/14/2007 2:04:08 PM
RSI(2) below 1...Close above the EMA(200)
----------------------------------
That gives a whole NEW meaning to the "KISS Principle" !
Regards,
alf44
PS. TRO...I get 0 hits on both of your filters ! fwiw
|
alf44
2,025 posts
msg #51092
- Ignore alf44
modified |
4/14/2007 2:29:26 PM
Now THESE work:
-----------------------------------
----------------------------------
Course you may want to put in some Price and Volume criteria to narrow down the list !
Regards,
alf44
|
alf44
2,025 posts
msg #51094
- Ignore alf44 |
4/14/2007 4:26:36 PM
hmmm...on second thought those may NOT be correct !
Can you mix weekly and daily criteria in the "count" statement ?
Little help here from the group plz ! tia
|
nikoschopen
2,824 posts
msg #51096
- Ignore nikoschopen
modified |
4/14/2007 5:49:19 PM
No you cannot throw the two in the same mix. You can see why by adding "draw weekly rsi(2)" and looking at the chart.
As you can see the weekly RSI(2) stays above the 90 for the duration of 10 weeks, not 2 as initially stated. The correct way to state the above premise is to treat days as weeks (5 days = 1 week). Hence it should be written
*Note: Not all fetched charts seem to conform to the above criteria.
[Edit] Oops, this should have been posted to the other thread.
|
ham1198
174 posts
msg #51183
- Ignore ham1198 |
4/18/2007 7:41:51 AM
TRO:
I took your HG filter and tweaked it with a couple of indicators I picked up elsewhere. I'm finally convinced after fighting the trend for years, the best stocks are those trending UP. Duh!! I love the "Draw Linear Regression Channel(60) line because it shows exactly what I'm looking for.
It's amazing how a simple little filter shows such great results. Yes Alf, K.I.S.S. Almost forgot, here's my revision:
/* daily rsi(2) and ma(200) filter */
date offset 2
rsi(2) below 5
close above ema(140)
close between 15 and 30
volume(90) > 100000
Draw Linear Regression Channel(60)
add column RSI(2)
add column Fast Stochastics(14,10)
add column Williams %R(10)
|
TheRumpledOne
6,411 posts
msg #51185
- Ignore TheRumpledOne
modified |
4/18/2007 11:38:34 AM
I added the TREND display.
HTH.
|
ham1198
174 posts
msg #51195
- Ignore ham1198 |
4/18/2007 1:52:29 PM
TRO, i hope this question doesn't sound too stupid.
What are the trend numbers telling me? I see numbers ranging from 1.0 to 111.00
Thanks for your help, it is greatly appreciated.
|