| Welcome Guest, please sign in to participate in a discussion. | Search | Active Topics | |
Registered User
Joined: 2/5/2005
Posts: 55
|
I would like to creat a scan that looks for stochastics levels that are below say 11 on weekly charts. Is there a way to do this? Thanks!!
Dave
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Check out this video, I think you will see your solution here:
Sorting price and indicators by their Actual Value
- Craig
Here to Help!
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
Can also be done with a PCF. Post the parameters you want to use for the Stochastic if you are interested, and I'll gen up a solution for you.
Jim Dean
|
|
Registered User
Joined: 2/5/2005
Posts: 55
|
Thanks Craig!
Tanstaafl
i want to set up the scan with standard settings and look for values lower then 11 on the weekly charts. I hope this is what you need!!!!
Thanks!
Dave
|
|
Registered User
Joined: 2/5/2005
Posts: 55
|
Whoa, that video showed me everything that I needed..... Thanks!!!!
Dave
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
Hi, Dave:
I need to know the first two numbers that you are using for your *plotted* Stochastics line ... the first is called the "Period", and the second is called the "%K". Note that %K is "typically" either 1 or 3 (fast or slow), but there is really no "standard" setting for the Period value ... you need to tell me what you are using before I can create the formula for you. I will assume that you are using "Simple" for the average type.
Jim Dean
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 4,308
|
Glad you saw the value in the video. If you set your chart to a weekly time frame, if your sort by actual value, you can quickly scroll down the list to find the stocks where the actual value of Stochastics is less than 11.
- Doug
Teaching Online!
|
|
Registered User
Joined: 10/7/2004
Posts: 14
|
Dear Jim,
I would appreciate if you could post the pcf formula for a weekly stochastic with period setting 21 and %k 6.
Many thanks ,
Stephen
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
Hi, Stephen:
I've not been ignoring your request.
I've been using a certain expansion-rule for a while to create N-day Stoc PCF's, which seems to work well versus the databox values of TC's internal Stoc plot, when the chart is set for the same N-days.
However, before posting anything here, I have been rechecking the "reference" formula for Lane's stochastic carefully against TC's internal plots, just for Daily bars, and ran into a contradiction. As long as you treat TC's internal-plot daily Stoc results as "correct", then my expansion formula works just fine. But if you were to compare TC's values for daily or other N-day bars to an Excel calc done the "right" way, or presumably to other charting packages that do it the "right" way, then you will see some discrepancies ... I feel that is important to point this out to avoid "challenges" to the veracity of my PCF.
The errors in the daily-bar calc seem to normally be only about 0.5-2%, with more error appearing when the Stoc value gets relatively high or relatively low. But it is a *disturbing* observation, since the Stoc calc is pretty well documented and can be modelled in a PCF readily.
Here is what I noticed - you can see it just by creating two PCF's, and assigning them to WL tab columns. The first PCF uses TC's standard internal Stochastic with periods=5 and %K(slowing)=3:
Stoc5.3
The second PCF is simply a longhand calc (per George Lane's formula) of the same value, in PCF-eze:
(((C-MINL5)+(C1-MINL5.1)+(C2-MINL5.2))
/((MAXH5-MINL5)+(MAXH5.1-MINL5.1)+(MAXH5.2-MINL5.2)))*100
These two PCF's SHOULD give exactly the same answers, but they don't. The first PCF does match the plotted Stoc's values (top pane databox). If you plot the second PCF as a Custom Indicator, the databox shows the same values as the Stoc5.3 PCF column ... ie still about 1% error.
So, what's the big deal? Well, users (like me) sorta expect TC to do the calc right. But unless a W programmer chimes in here to explain why the PCF#2 does not match their internal calc, I guess we will never know.
OK ... so ... having said all that and qualified the answer that you were looking for in the first place with the proviso that the N-day Stoc PCF-expansion will be off from the "correct Lane-based Excel calc's" by that same general amount of error, here is the PCF you wanted for a 5-day model of a period-setting =5 and a %K-smoothing = 6 (%D set =1).
To check it out, please PLOT the Stoc21,6,1 on a 5-day chart on the top pane. Then create and calc the PCF below, and assign its results in a WL tab column. Turn on the pointer databox, position the pointer to the most recent day that has been downloaded (turn off TCNet's realtime if you have it), and Spacebar your way down the WL, comparing the values between the databox and the PCF. You will see they agree precisely.
PCF for weekly 21-bar Stoch with 6-bar %K-smoothing:
(Stoc105.1.0+Stoc105.1.5+Stoc105.1.10+Stoc105.1.15+Stoc105.1.20+Stoc105.1.25) /6
I'm sure that you (or somebody) might be curious how to solve this problem for other cases of N-days, given a Stoc period of P bars and a %K-slowing period of S bars (these all assume that the signal line %D=1).
So, the *generic formula* for an N-days/bar *PCF*
model of the TC-world graph of the native Stoc(P,K,1)=
(STOCX.1.a + STOCX.1.b + STOCX.1.c + ...) / K
where X = P*N, and the series has K terms, with the final a,b,c... parameter of each of the terms equal to an integer multiple of K, starting with 0*K for the first term, 1*K for the second, and so on.
That should explain the whole thing pretty thoroughly for you, and provide you (and others) with a way to model other N-day Stoc cases. Thanks for your patience with my initial qualifiers to the calc's.
Jim Dean
|
|
Registered User
Joined: 10/7/2004
Posts: 14
|
Dear Jim,
Thanks for the explanation and formula.
Just to clarify, PCF for weekly 21 bar stoch with 6-bar %K smoothing for previous week would read as
(Stoc105.1.5+Stoc105.1.10+Stoc105.1.15+Stoc105.1.20+Stoc105.1.25+Stoc105.1.30) /6
I think this is correct.Please confirm.
Stephen
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
You got it!
Enjoy!
|
|
|
Guest-1 |