Registered User
Joined: 6/30/2005
Posts: 30
|
How can I formulate my PCF's to scan for stocks from a weekly chart rather than daily? Any help would be appreciated.
OneDay
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
It depends on the scan. Post your daily versions here and I will try and convert them for you.
- Craig
Here to Help!
|
|
Registered User
Joined: 3/14/2006
Posts: 23
|
Craig, what did you do? I am also interested in that idea. I've been using some of the candlestick scans and would like to be able to apply them to weekly/monthly time frames as well.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 4,308
|
As Craig said, it depends on the scan. It is possible to convert some scans so they will look for patterns on a different time frame. But you'd need to post your daily version here so we can try to convert them for you.
- Doug
Teaching Online!
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
I utilize Stochastic 10,5,1 (simple)on a daily, two and three day basis. I'm interested in an upturn and not a crossover. Can a pcf be written for these multi day applications?
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Over how many bars should the Stochastics be "up"?
- Craig
Here to Help!
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
One bar.
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
To further clarify: I typically only scan (space bar) on the weekend or midweek, so the 3 day stochastic may be up from 1-3 bars. I just want the 2 day and daily to be up a minimum of one bar.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
To be specific...
You only want the stochastic value of the current bar to be greater than the value yesterday?
- Craig
Here to Help!
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
Since I may scan only once every three days, I have in the past written a daily stochastic that tells me if the value "turned up" for the current day, yesterday, or the day before. This was done so I don't have to scan everyday to catch the one day the value "upticked" from the previous day. Ideally that is what I would like to accomplish with the 3 day. I'm looking for the upturn in the fishing hook. All that said, having the upturn for just a single bar would be great, or if that's not possible or just too time consuming, having the current value today greater than yesterday would be light years ahead of what I am doing currently. Because my initial scan is looking for the 3 day to hook up, for the 2 day and daily I only want them to have a greater value today than yesterday.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
I'm not quite sure if this is what you are looking for or not, so I'm going to present a Daily formula along with 2-Day, 3-Day, 4-Day and 1-Week per Bar translations:
Daily:
STOC10.5 > STOC10.5.1 OR STOC10.5 > STOC10.5.2 OR STOC10.5 > STOC10.5.3
2-Days per Bar:
(STOC20 + STOC20.1.2 + STOC20.1.4 + STOC20.1.6 + STOC20.1.8) / 5 > (STOC20.1.2 + STOC20.1.4 + STOC20.1.6 + STOC20.1.8 + STOC20.1.10) / 5 OR (STOC20 + STOC20.1.2 + STOC20.1.4 + STOC20.1.6 + STOC20.1.8) / 5 > (STOC20.1.4 + STOC20.1.6 + STOC20.1.8 + STOC20.1.10 + STOC20.1.12) / 5 OR (STOC20 + STOC20.1.2 + STOC20.1.4 + STOC20.1.6 + STOC20.1.8) / 5 > (STOC20.1.6 + STOC20.1.8 + STOC20.1.10 + STOC20.1.12 + STOC20.1.14) / 5
3-Days per Bar:
(STOC30 + STOC30.1.3 + STOC30.1.6 + STOC30.1.9 + STOC30.1.12) / 5 > (STOC30.1.3 + STOC30.1.6 + STOC30.1.9 + STOC30.1.12 + STOC30.1.15) / 5 OR (STOC30 + STOC30.1.3 + STOC30.1.6 + STOC30.1.9 + STOC30.1.12) / 5 > (STOC30.1.6 + STOC30.1.9 + STOC30.1.12 + STOC30.1.15 + STOC30.1.18) / 5 OR (STOC30 + STOC30.1.3 + STOC30.1.6 + STOC30.1.9 + STOC30.1.12) / 5 > (STOC30.1.9 + STOC30.1.12 + STOC30.1.15 + STOC30.1.18 + STOC30.1.21) / 5
4-Days per Bar:
(STOC40 + STOC40.1.4 + STOC40.1.8 + STOC40.1.12 + STOC40.1.16) / 5 > (STOC40.1.4 + STOC40.1.8 + STOC40.1.12 + STOC40.1.16 + STOC40.1.20) / 5 OR (STOC40 + STOC40.1.4 + STOC40.1.8 + STOC40.1.12 + STOC40.1.16) / 5 > (STOC40.1.8 + STOC40.1.12 + STOC40.1.16 + STOC40.1.20 + STOC40.1.24) / 5 OR (STOC40 + STOC40.1.4 + STOC40.1.8 + STOC40.1.12 + STOC40.1.16) / 5 > (STOC40.1.12 + STOC40.1.16 + STOC40.1.20 + STOC40.1.24 + STOC40.1.28) / 5
Weekly:
(STOC50 + STOC50.1.5 + STOC50.1.10 + STOC50.1.15 + STOC50.1.20) / 5 > (STOC50.1.5 + STOC50.1.10 + STOC50.1.15 + STOC50.1.20 + STOC50.1.25) / 5 OR (STOC50 + STOC50.1.5 + STOC50.1.10 + STOC50.1.15 + STOC50.1.20) / 5 > (STOC50.1.10 + STOC50.1.15 + STOC50.1.20 + STOC50.1.25 + STOC50.1.30) / 5 OR (STOC50 + STOC50.1.5 + STOC50.1.10 + STOC50.1.15 + STOC50.1.20) / 5 > (STOC50.1.15 + STOC50.1.20 + STOC50.1.25 + STOC50.1.30 + STOC50.1.35) / 5
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
Thanks Bruce, this will give me something to play with this weekend. Can't wait! Gives me a little understanding how the software thinks on the multi-day applications. Very helpful.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
You're welcome. Even if the Daily formula isn't exactly what you are after, it will hopefully be possible to re-write it and then re-arrange the multi-day per bar formula components to match.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
Still having a problem getting the results I'm looking for regarding Stochastic: Period 10, SK 15, SD1 (simple). I ran the 3 day scan you graciously provided, and it selected several stocks that according to the indicator should not have been included. Look at TEX for example. The three day stochastic has been in a down trend for 6 periods, and yet the 3-days per bar pcf selected it as meeting the criteria. Any suggestions? I didn't try the 2, 4, and weekly. I assumed they had the same problem as the 3 day.
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
Point of correction in my original post of 3/15/05. Should have read Stochastic 10,15,1 and NOT 10,5,1. Sorry for my confusion. My post of 4/2/06 correctly identifies the Stochastic I utilize. What change do I need to make? I hadn't caught my error until after I had sent my last post.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 4,308
|
Are your PCF's updated for All Items in System?
- Doug
Teaching Online!
|
|
Registered User
Joined: 3/15/2005
Posts: 10
|
Think I figured it out. To change the 3 day Stochastic 10,5,1 to Stochastic 10,15,1 I took the pcf and included fifteen values between the parenthesis and then divided by 15. Makes for a pretty long pcf, but the results appear favorable. If you could confirm this is correct, I can do the grunt work to rework the forumlas you provided earler. Thanks for your patience.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
steadyprofit,
That would be the correct way to convert the formulas, but we may want to test some different versions of the daily formula before going through the grunt work (maybe not if it is working for you as is). The current version just tests the current value to the previous three periods to see if there was an upturn over any of those intervals:
STOC10.15 > STOC10.15.1 OR STOC10.15 > STOC10.15.2 OR STOC10.15 > STOC10.15.3
Maybe something that tests for a downtrend before the uptrend would be better:
STOC10.15 > STOC10.15.1 AND STOC10.15.3 < STOC10.15.4 AND STOC10.15.4 < STOC10.15.5
Both of these formulas take shortcuts to some extent in that they only look at the final result and don't really don't explicitly test for what's happening during each of three periods of interest as mentioned in your Thursday, March 30, 2006 10:57:45 PM post. It wouldn't be difficult to do in the daily version, but would make for long formulas when converted to multi-day per bar versions. I'd be tempted to use a MIN() function for the daily version, but that would also be problematic when translated to multi-day per bar versions of the formula.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
|
Guest-1 |