tmanbone
124 posts
msg #81280
- Ignore tmanbone |
10/16/2009 8:51:59 PM
I've been studying David Elliot at ThinkorSwim chat. He has a 2 line stochastic crossover indicator called the SOAP. The theory is you plot a full stochastic %K(8,3,5) and a full stochastic %K(5,3,3), all of the 3 next steps happen below the 50% line. #1 when the 5,3,3 crosses the 8,3,5 from below you've started wave 1, #2 directly after and without the 8,3,5 line crossing 50%, the 5,3,3 crosses the 8,3,5 from above back down through you've started wave 2, and #3 when the 5,3,3 crosses the 8,3,5 from below you've started your most profitable wave 3. I've been looking at this for several months on severals 1000's of charts and know the signal is profitable. The signal is fractal, I'm looking to use it for swing trades. The problem being I haven't been able to create a scan for stocks/etf's setting up on this signal. I've also been studying the SF forums for several months and know of your brillant minds. I appreciate any and all replies. Thanks
|
Cacher
121 posts
msg #81282
- Ignore Cacher
modified |
10/16/2009 10:29:37 PM
It sounds like what you are looking for is a process of screenings .....an interesting slant on screening... using those results of one screen for your next leg of the process the following day...the process would be endless, always chugging out results once each level gets symbols in the watch lists. The list would always need to be edited, adding new symbols, while moving some to the next watch list. I imagine the earliest a final result could show up would be in two or a few days or maybe even weeks...?
from the results, create or edit a watchlist called “WAVE1_results”
save these results from WAVE2, and create or edit another watchlist called “WAVE2_results”
Finally after a few days into this process, the rewards should start to be on your final and hopefully, rewarding results.
It would be an ongoing process...and the need to run each wave daily, and updating each of the watch lists daily, adding and moving symbols to the next watch list....it may be a bit tedious ... but then when the rewards are great, a little work could be warranted....and worth the effort.
|
tmanbone
124 posts
msg #81291
- Ignore tmanbone |
10/17/2009 10:38:53 AM
Yes Cacher you've got it. It's takes days to possibly a couple of weeks for the setup to occur. I'm thinking this could be applied to Kevin's idea of only screening IBD 100 stocks or of course any watchlist anyone prefers. Now from what I understand from David Elliot's SOAP indicator is that it is 2 stochastic %Ks one (8,3,5) the slower moving and one (5,3,3) the faster moving, the %Ds are deleted out. This screen on the final wave 3 should also play out with another indicator giving the probablity of an up move, one of my favorites being the MACD. An opposite screen or series of screens could be applied for moves in the opposite direction. Thanks for the input, great work.
|
chetron
2,817 posts
msg #81293
- Ignore chetron
modified |
10/17/2009 10:57:30 AM
PLEASE HELP ME....
tmanbone, AT THE BEGINNING YOU SAID, TO USE THE %K'S OF THE STOCASTIC,
Cacher, YOUR FILTERS ARE USING A %D
TMANBONE, DID YOU SAY TO DELETE THE %D?
TIA FOLKS
IF I AM CORRECT......
A LITTLE MORE ECONOMIC....
ONLY BELOW 50....
AS GOOD AS I CAN GET IT....
IS THAT BOTH BELOW 50????????????
MIINIMUM 3 DAY TURN AROUND.....
TYPOS.....
OOPS, DAY LATE...
MAYBE A LINE AT 50.....
WITHOUT ALL THE LINES.....
|
chetron
2,817 posts
msg #81295
- Ignore chetron
modified |
10/17/2009 12:42:51 PM
WHEN THIS RALLY FINALLY ENDS.....
OOPS, A DAY LATE.....
|
xxcheck
53 posts
msg #81297
- Ignore xxcheck |
10/17/2009 1:29:55 PM
Question Chet... This filter running with all 3 waves intact. Results from Wave 1, Wave 2 etc are already being filtered and finalized to Wave 3.
|
chetron
2,817 posts
msg #81299
- Ignore chetron |
10/17/2009 1:51:28 PM
THAT WAS MY HOPE.....
|
tmanbone
124 posts
msg #81301
- Ignore tmanbone |
10/17/2009 5:03:51 PM
Yes Chetron to omit the %D, looks like you're on it. The faster %K(5,3,3) can go above the 50 line, through the 3 wave process but the slower (8,3,5) shouldn't. I not very up to speed on the filter creating process so bear with me here. Is there much difference in a (5,3,3)%K line and a (5,1,3), the same for the (8,3,5) and the (8,1,5), curious. This post was sent without a lot of time to review the filters that you've posted; I will study futher. From what I can tell with a short look, we are on to something. Thank you for your input.
|
chetron
2,817 posts
msg #81321
- Ignore chetron |
10/18/2009 1:27:05 AM
the 2nd number of a full stochastic is the %D, by putting a 1 there you are saying that the %D = %K, which is a cheap way of deleting the %D.
|
trendscanner
265 posts
msg #81327
- Ignore trendscanner |
10/18/2009 9:21:11 AM
Chetron, I may not understand Fetcherese well enough to understand your coding of the filter so I may be wrong about this but looking at the first wave you've written, it appears to be looking for the event where days since Stochastic %K(5,1,3) crossed BELOW Stochastic %K(8,1,5).
The way Tmanbone describes the first event is when "#1 when the 5,3,3 crosses the 8,3,5 FROM below you've started wave 1". So should the first wave be looking for "Stochastic %K(5,1,3) crossed ABOVE Stochastic %K(8,1,5)" , i.e., it crosses %k(8,1,5) FROM below?
Interesting work, thanks for posting.
|