Registered User
Joined: 10/27/2004
Posts: 36
|
Wow... Thank you... An early Halloween with Tricks and Treats!
Am I correct that using the easy scan "Price Percent Change 5-Day" trick to establish the previous trend (as in the non-boolean bullish engulfing easyscan) has been replaced by the ascending and descending sorts?
Can we load an easy scan such that it pulls up stocks with candle pattern A and candle pattern B? Probably not... forgive me... it's getting late.
Is it not possible to streamline, as opposed to all the 'ands?'
for instance 01<02<c1<03<c2<c3
Again, thank you, thank you.... I love to backtest these too.
Now if I could only profit in this market... How about one of your great videos on the subject of 'nerves of steel'
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
The way I would establish the trend is by sorting.
Sort the results of your EasyScans with either
Price Percent Change 5-day
or
Price Percent Change 26-weeks
depending on how long you want the trend to be.
Each PCF above tells you to sort in either ascending or descending order.
You cannot streamline with C>C1>C2. You have to use the "ands".
I am not sure what you mean by a stock having pattern A and B happening. I don't think two patterns can happen at the same time.
You could combine to look for Pattern A OR B by connecting two formulas with an or (be sure each formula gets put in its own set of parens.
Example
(formula A) OR (formula B)
so a PCF for Bullish Engulfing OR Tri Star would be:
((H1-L1)>(AVGH21.1-AVGL21.1) AND
C1<O1 AND
C>O1 AND
O<C1) OR
(ABS(C-O)<= (H-L)*0.05 AND
((C+O)/2)-L >= (H-L)*0.40 AND
((C+O)/2)-L <= (H-L)*0.60 AND
ABS(C1-O1)<= (H1-L1)*0.05 AND
((C1+O1)/2)-L1 >= (H1-L1)*0.40 AND
((C1+O1)/2)-L1 <= (H1-L1)*0.60 AND
ABS(C2-O2)<= (H2-L2)*0.05 AND
((C2+O2)/2)-L2 >= (H2-L2)*0.40 AND
((C2+O2)/2)-L2 <= (H2-L2)*0.60 AND
H1<L3 AND
H1<L1)
Be sure to watch the Visual Backtest video to see how to use these formulas to show spikes when they happened in th past.
QUOTE (islandrover) Now if I could only profit in this market... How about one of your great videos on the subject of 'nerves of steel'
I started a video titled "How to know with 100% accuracy which direction a stock will go tomorrow" but somehow lost my outline. I did not bother to look for it and scrapped the video. I didn't think anyone would be interested in it anyway. (Boy, I wish that was true!) 
- Craig
Here to Help!
|
|
Registered User
Joined: 10/27/2004
Posts: 36
|
Thanks, Craig. I did mean OR, not AND, the purpose being to collect stocks that show any or a select few of the bullish patterns into a common watchlist. For example, high reliability, morning star, morning doji star, three inside up (it may be that there is one algorithm that pulls up all three, but I have not advanced that far).
I import lists from multi-parameter fundamental screens from excel (via .txt), then I launch the technical analysis of daily data. Truth be known, all of this is pretty new to me so there is a whole lot of slogging going on - but the videos are incredibly helpful.
I did the visual backtest for both bullish engulfing and the 'calculated gamble' (different colors) from one of the evening Worden reports. I track the signal in the bottom window and track true/false price increases of 25%, 10% or 5% (different colors) in the middle window, my question being 'in my little stock world, would I be better off taking that first 20-25% profit or should I hold, and if I hold, just how far is it likely to drop after a break-out before it reaches that point again.' It is great training to run through time on a single stock or watchlists of particular types of stocks. If I could export the results I might attempt statistical analysis of what works (because hope looms eternal and isn't always projected accurately). Maybe user S2 has methods for measuring reliability and results?
My entries, exits, and lack of experience are my primary problems - not my tools (but I wish I had Telnet!).
I'm glad you didn't abandon us newbies after outlining your 100% method. What a pal!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
QUOTEmy question being 'in my little stock world, would I be better off taking that first 20-25% profit or should I hold, and if I hold, just how far is it likely to drop after a break-out before it reaches that point again.' It is great training to run through time on a single stock or watchlists of particular types of stocks. If I could export the results I might attempt statistical analysis of what works (because hope looms eternal and isn't always projected accurately)
We trainers cannot give this type of advice
QUOTEI'm glad you didn't abandon us newbies after outlining your 100% method. What a pal!
Truth be told, if I were to ever come up with something like this... you would find me on an island somewhere not making videos.
- Craig
Here to Help!
|
|
Registered User
Joined: 10/27/2004
Posts: 36
|
I use the visual indicator to track the results of that question - trigger in the bottom window, follow thru results in the middle window - both as true/false. My question was not intended for you, but Telechart lets me see the results easily (but not export them back to spreadsheets - I guess I made a too-thickly veiled request).
See you on the beach, someday soon - right?
islandrover....
|
|
Registered User
Joined: 10/27/2004
Posts: 36
|
can you annotate the original post with the 'traditional' reliability info, similar to that which S2 put on the original PCF/candlestick thread? (feel free to remove our banter, too)
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I don't feel comfortable tagging something on to a post...
I am sure you can understand why.
- Craig
Here to Help!
|
|
Registered User
Joined: 4/6/2005
Posts: 239
|
Craig,
Thanks for including this thread for boolean approaches to candle patterns.
The reliablity factors I posted earlier in the other sticky candle pattern thread came from one of the web sites mentioned. I can not verify those reliability factors, but am trying to. So far my analyis is not conclusive. The Nisson and Bigalow books do not provide any reliability factors. Bigalow does have two sections, "Major Signals" and "Secondary Signals". Some of the patterns that Bigalow classifies as secondary Nisson spends more time addressing in his book.
My finding so far indicate that the reversal patterns are between 50% and 100% accurate, providing the market goes strongly in that direction; and less than 50% accurate when going against that direction. So pick your pattern and which way the market will go and you'll have pretty good success. 
|
|
 Gold Customer
Joined: 8/15/2005
Posts: 71
|
when I cut and paste "falling 3 methods it returns an error in syntax message. Please check. Thanks
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
annja45,
I've think it's fixed. Did you see my answer to your question here?
rising three methods
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 5/20/2005
Posts: 178
|
Thanks for the candlestick patterns.
I am wondering if it is possible to find M and W shape chart patterns?
thanks
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
If by M and W shaped chart patterns you mean Double Top and Double Bottom, pretty much no, but you may be interested in reviewing the following topics to see both why and to get an idea of how to at least narrow the search somewhat:
Scanning for patterns
Scanning for Stocks Printing a Double or Triple Bottom
How to create PCF to scan double top / double bottom price pattern?
Stochastics Double top and bottom
double tops or double bottoms
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
Hi. There are syntax errors in the Meeting Lines & Three Line Strike pcf's as originally posted. They're probably typos on the AVG function. Corrections are below:
Meeting Lines
C1<O1 AND
(H1-L1)> (AVGH21.1-AVGL21.1) AND
O1<MINL3.3 AND
C>O AND
C<C1*1.01 AND
C>C1*0.99
*Sort in ascending order*
Three Line Strike
C2>C3 AND
C1>C2 AND
(H3-L3)> (AVGH21.3-AVGL21.3) AND
(H2-L2)> (AVGH21.2-AVGL21.2) AND
(H1-L1)> (AVGH21.1-AVGL21.1) AND
O>O3 AND
C<O3
*Sort in descending order*
Cheers!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I corrected them. Thanks!
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
Hi Craig -
Could you possibly post boolean pcf formulas for High Wave & Bull/Bear Long Legged Doji? Thanks.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Sure... define them for me and I will write the PCFs.
- Craig
Here to Help!
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
Craig I don't usually wonder around this areas, in fact end up here by wrong click.
Anyways, I have to say you did a superb jog on tacking down those candle patterns.
Keep up the good work.
good luck.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Thanks!
- Craig
Here to Help!
|
|
Registered User
Joined: 4/18/2005
Posts: 2
|
Craig S.
I am not really computer literate...tried to create some
PCFs to be able to sort Optionable Stocks by Candlestick
indicators such as Three White Soldiers & Three Black
Crows...I think I followed Michael's instructions o.k.,but
I keep getting "Syntax error in Formula" when I try to
CALCULATE them....the formulas were "copy/pasted" from
your Boolean PCFs for Candlesticks posting...
What might I be doing wrong????
"kip&" Webster
(email removed by Moderator)
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Copy and paste here what you are copying and pasting into your system so I can see what the problem is.
- Craig
Here to Help!
|
|
Registered User
Joined: 4/18/2005
Posts: 2
|
CC>C1 AND
C1>C2 AND
C>O AND
C1>O1 AND
C2>O2 AND
ABS(C2-O2)>(H2-L2)*0.50 AND
ABS(C1-O1)>(H1-L1)*0.50 AND
(H-L)>(AVGH21-AVGL21) AND
O>O1 AND
O<C1 AND
O1>O2 AND
O1<C2 AND
O2>O3 AND
O2<C3 AND
C>H*0.85 AND
C1>H1*0.85 AND
C2>H2*0.85
raig S....here is one of the PCF's giving me grief.....
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I think the double CC at the beginning of the above is a typo. I copied and pasted the above:
C>C1 AND
C1>C2 AND
C>O AND
C1>O1 AND
C2>O2 AND
ABS(C2-O2)>(H2-L2)*0.50 AND
ABS(C1-O1)>(H1-L1)*0.50 AND
(H-L)>(AVGH21-AVGL21) AND
O>O1 AND
O<C1 AND
O1>O2 AND
O1<C2 AND
O2>O3 AND
O2<C3 AND
C>H*0.85 AND
C1>H1*0.85 AND
C2>H2*0.85
And had not trouble. There must be another PCF in your system causing the problem. When you copy and paste this one and TEST it do you get the error?
- Craig
Here to Help!
|
|
Registered User
Joined: 3/1/2006
Posts: 11
|
How, I find the information in the videos and discussions like this one very helpful. I'm working on a couple of PCFs. One for a matching high and one for a matching low. What changes would I need to make to the PCF's for matching high and matching low so that the following conditons occur:
Matching low:
1) A long blackday occurs
2) The second day is also black with its close equal to the close of the first day.
3) Both days have little or no lower shadow.
Matching high:
1) the first day is a long white day.
2)The second day has the same closing price as the first day
3) Both days have little or no upper shadow
Here is what I think the PCFs should be:
Matching low:
(C1<O1) AND
ABS(C1-O1)>(H1-L1)*0.50 AND
C<O AND
C<C1*1.01 AND
C>C1*0.99 AND (C1=C) AND ABS(C - O) / (H - L) * 100
Matching high:
(C1>O1) AND
ABS(C1-O1)>(H1-L1)*0.50 AND
C>O AND
C>C1*1.01 AND
C>C1*0.99 AND (C1=C) AND ABS(C - O) / (H - L) * 100
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Here is how I would do these:
"Matching Low"
C1<O1 AND ABS(C1-O1)>(H1-L1)*.8 AND (H1-L1)>(AVGH40-AVGL40)*1.2 AND C<O AND C=C1 AND C1<=L1*1.01 AND C<=L*1.01
"Matching High"
C1>O1 AND ABS(C1-O1)>(H1-L1)*.8 AND (H1-L1)>(AVGH40-AVGL40)*1.2 AND C=C1 AND C1>=H*0.99 AND C>=H*0.99
- Craig
Here to Help!
|
|
Registered User
Joined: 2/14/2006
Posts: 8
|
I have been with TC NET only about 1-2 mo's.
Have traded maybe 3 yrs, since retiring.
I tried to do what is advised in FAQ, re:"hammers and hanging man". the "add conditions for pcf's" are missing...
Was the video changed? thanks.bb
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
No, you have to create the PCFs.
Watch these short videos, read the instructions again in the original post and let me know if you have any more questions.
Welcome to the TeleChart community!
How to create a Personal Criteria Forumula (PCF)
Using EasyScan to find stocks that meet your own criteria
- Craig
Here to Help!
|
|
Registered User
Joined: 9/24/2005
Posts: 15
|
the shooting star formula isnt working can u give me one and a formula that says stocastics 14 3 3over 80
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
str520.
Please try the shooting star again, it does work and has nothing to do with stochastics.
I provide a formula and method in your post in ASK A TRAINER for finding stochastics over 80. Please look for it there.
- Craig
Here to Help!
|
|
Registered User
Joined: 6/27/2006
Posts: 286
|
Hello,
Dark Cloud Cover
ABS(C1-O1)>(H1-L1)*0.50 AND
C1>O1 AND
O>H1 AND
C<((C1+O1)/2) AND
C>O1
I have use this formula posted above to look for dark cloud cover candles. It found 106 patterns. If I use this formula:
C1 > O1 AND O > C1 AND C < (C1 - (C1 - O1) / 2) AND C > O1 it finds 198 patterns.
Can you explain what the differance is and why?
Thanks dave
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
They are asking for different things.
This formula
ABS(C1-O1)>(H1-L1)*0.50 AND
C1>O1 AND
O>H1 AND
C<((C1+O1)/2) AND
C>O1
Requires the following to be true:
a) The body of yesterday's candle must be more than half the size of the overall candle
b) The close yesterday must be above the open yesterday
c) The open today must be greater than the high yesterday
d) The close today must be less than the mid-point of yesterday's candle body
e) The close today must be above yesterday's open
The second formula you list: C1 > O1 AND O > C1 AND C < (C1 - (C1 - O1) / 2) AND C > O1
Requires the following to be true:
a) The close yesterday must be aboev the open yesterday (same as b above)
b) The open today is above the close yesterday
c) The close today must be less than the mid-point of yesterday's candle body (same as d above)
d) The close today is above yesterday's open (same as e above)
I am not sure which of the two fit you better. Look at the stocks they find and see which you like better.
- Craig
Here to Help!
|
|
Registered User
Joined: 6/27/2006
Posts: 286
|
Thanks for that info Craig.
|
|
Registered User
Joined: 6/27/2006
Posts: 286
|
Another question for you when I do somthing like this:
ABS(C1 > O1 AND O > H1 AND C < L1)
and make an easy scan it keeps asking me for a value but on the formulas you made it does not. why is that?
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
The formula I made uses the ABS differently. It does not ABS a Boolean, it ABSs an equation.
What are you trying to do with this formula?
- Craig
Here to Help!
|
|
Registered User
Joined: 6/27/2006
Posts: 286
|
Ah I see, I'm just playing with the formulas to see what gives me the best results Then I keep track of the stocks it gives to see what happens a few days to a week or two later. The formula I posted above is for a bearish engulfing that includes the shadows of the day before. I took out the ABS and its working fine. I will track about 20 or 30 stock for a week or two so I can find out how this pattern works and what percent of the time it works.
Thanks for your help and HAPPY 4th
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
You may find this interesting as well: Visually Backtesting Specific Symbols
- Craig
Here to Help!
|
|
Registered User
Joined: 6/27/2006
Posts: 286
|
Yes I use that with some other PCF I made. It wont work to track if the pattern failed and how much it gained if it worked. It is good to look for confirmation of the patterns. I like it when I use it to confirm the pattern.
Example:
Here are six stocks I found to confirm the bullish engulfing pattern on 6/29.
If you bought on 6/29 and still hold it you would be up or down the amount in the profit colum.
Bullish Engulfing Candle Tests
100---6/29/06-7/4/06----P/L------Cost-------Sell------Profit
VLTR--$14.50--$15.64--$1.14---$1,450.00--$1,564.00--$114.00
PENN--$38.79--$39.06--$0.27---$3,879.00--$3,906.00--$27.00
GMXR--$29.83--$31.40--$1.57---$2,983.00--$3,140.00--$157.00
SPPI---$3.83----$3.79---($0.04)--$383.00----$379.00----($4.00)
SURW--$18.93--$20.48--$1.55---$1,893.00--$2,048.00--$155.00
USAP--$28.23--$29.48--$1.25---$2,823.00--$2,948.00--$125.00
|
|
Registered User
Joined: 2/24/2006
Posts: 1
|
Thank you for the Candlestick Boolean Equations; I have been using above and modifying them to suit my viewpoint. I would like to use Linear Regression to determine if the basic trend is Up, Down or sideways before using equations to filter for various Candelstick Pattterns. How can I incorporate Linear Regression into PCF? I use Linear Regression on my charts and this is very helpful.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
apalmieri00,
You may wish to review the following:
Using Linear Regression vs Classical Peaks/Valleys for Divergence Analysis
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 11/28/2005
Posts: 38
|
Craig,
I have a few scans set up on my Telechart that I got from the (removed by Moderator) website. However, the scans look different to the ones you have listed, can you tell me which are correct?
For example, for a bullish engulfing pattern I have: ((O1 > C1) AND (C > O) AND (C >= O1) AND (C1 >= O) AND ((C - O) > (O1 - C1)))
Thanks,
Kardaya
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Here is their version (cleaned up):
O1 > C1 AND
C > O AND
C >= O1 AND
C1 >= O AND
C - O > O1 - C1
Here is mine
ABS(C1-O1)>(H1-L1)*0.50 AND
C1<O1 AND
C>O1 AND
O<C1
differences:
I require that yesterday's body takes up more than half of the candle.
Mine is a tad more efficient.
If C1<O1 AND
If C>O1 AND
If O<C1
then C has to be greater than O so there is no reason to put it in there.
- Craig
Here to Help!
|
|
|
Guest-1 |