Cheese
1,374 posts
msg #154448
- Ignore Cheese |
10/27/2020 1:19:37 AM
Graf,
How does this list of picks compare to yours?
|
graftonian
1,089 posts
msg #154452
- Ignore graftonian |
10/27/2020 9:05:13 AM
|
nibor100
1,031 posts
msg #154453
- Ignore nibor100 |
10/27/2020 10:52:47 AM
@Graftonian,
I'm still working with the original filter you posted in this thread and have come up with an approximate 80% solution, shown in the filter down below.
1. To reduce complexity I've changed your trigger from a 3 part trigger to a 2 part trigger without the STC component.
Then I added the STC component check back in by using this line:
add column stccrosscount equals 1
Then to show the new stocks that meet your 3 trigger criteria I added the following 2 lines:
add column trigger 1 day ago
sort on column 7 ascending
So the only thing missing from my modified filter is the check of which stocks met the STC component today that did not meet it yesterday so they also would be included at the top of the list. I'm guessing that means about 20% of the new stocks are co-mingled in the sort column in those stocks that have the number 2.
2. Bottom line is my modified filter finds all of the same stocks and stacks many of the first timers at the top of the list but is co-mingling some of the first timers in the lower part of the list, but it should help you some. and I plan to work a little more on it before I give up entirely.
Ed S.
|
nibor100
1,031 posts
msg #154455
- Ignore nibor100 |
10/27/2020 12:14:08 PM
@ MAC and CHEESE, (I couldn't resist :)
Here is the link to the PDF from the Schaff Trend Cycle originator:
https://www.omnitrader.com/currentclients/proforum/get-attachment.asp?attachmentid=2503
There is a double use of Stochastics on the MACD in his original implementation.
Ed S.
|
Cheese
1,374 posts
msg #154457
- Ignore Cheese |
10/27/2020 1:13:12 PM
haha
Thanks, Ed
Yes, I know that the STC that you and graftonian developed using MACD
is the correct one, and it does reconcile with the STC on other platforms.
The STC that other sF members developed using only EMAs is not good
enough for STC purists. But for basic subs who don't need to reconcile
the STC values, then the "STC using EMAs" can give us a rough idea.
|
Mactheriverrat
3,153 posts
msg #154467
- Ignore Mactheriverrat |
10/27/2020 10:59:11 PM
Well I just threw my code out there. I'm not really that crazy about it.
|
graftonian
1,089 posts
msg #154483
- Ignore graftonian |
10/28/2020 5:50:45 PM
Both the original and modified STC codes picked ZDGE, a bright spot in a sea of red.
|
nibor100
1,031 posts
msg #154486
- Ignore nibor100 |
10/29/2020 2:46:25 AM
@ Cheese, Graftonian,
It appears to me your comparisons are not quite fair as each filter is using different parameters than the other such as different length moving averages.
I've modified Graf's comparison filter below so basic parameters are equal, and it appears in many cases there is only a slight difference, yet still significant differences in many others.
Ed S.
|
graftonian
1,089 posts
msg #154490
- Ignore graftonian |
10/29/2020 9:30:45 AM
Perhaps IFT(5,9) is a better substitute for STC, and it is a canned function.
|
Cheese
1,374 posts
msg #154493
- Ignore Cheese |
10/29/2020 10:48:01 AM
@ nibor100 and graftonian
Thank you for coding your MACD-based STC and your IFT(5,9) ideas.
Your MACD-based STC is true to Doug Schaff's intentions.compared to the EMA-based "STC".
Your MACD-based STC correctly reflects the values and threshold levels.
Although the EMA-based STC is mainly a workaround for basic subs,
its smoother, rounder gradual move may have appeal to some,
compared to the sharper edges of the MACD-based STC.
|