Price Action Education And Analysis - Page 4 @ Forex Factory









Forums
Calendar
News








































strangemartian










Log Out


New PMs: 0



New Posts: 154












Search Forums










Advanced Search











Visit Our Sponsor

























 
Forums
> The James16 Group


> James16 Trading Education





Price Action Education And Analysis
















User CP


FAQ
Members List



New Posts

Search vbmenu_register("navbar_search");


Quick Links vbmenu_register("usercptools");



















Search Forums













Show Threads
 
Show Posts



Advanced Search












Quick Links
Today's Posts
Mark Forums Read
Open Buddy List

User Control Panel
Edit Signature

Edit Profile
Edit Options

Miscellaneous
Private Messages
Subscribed Threads
My Profile
Who's Online












Go to Page...































Page 4 of 32

<
123 4
5678914
>
Last














 



Thread Tools
vbmenu_register("threadtools");



Search this Thread
vbmenu_register("threadsearch");




Rating:
vbmenu_register("threadrating");




Display Modes
vbmenu_register("displaymodes");



































 
#46  


 








01-07-06, 11:05pm
















james16



vbmenu_register("postmenu_37212", true);



wizard in training



 



Join Date: Feb 2005
Location: Dallas-Ft.Worth,Texas


Posts: 2,607




























Quote:





Originally Posted by WTB


Hello Jim,

I have a question about my favourite type of bar formation, the pin bar.

Today, I have been scrolling back through different charts scanning for
pin bars at fibonacci retracement in order to identify the end of the
pullback (and thus the resuming of the current major trend). Well,
after few hours worth of scrolling through charts, I sort of noticed
that pin bars are more present and significant the other way around,
meaning not at the fib retracement but rather at the swing's high/low,
pointing us out that we are due for a retracement inminently.

Have you also noticed this?

PS: I attach a screenshot so you have a visual picture of what I mean.





YES, they are most powerful at
these points as pointed out many times in the james16 chart thread. if
you or anyone as a trader would look for nothing but these bars at
extreme swing highs and lows to trade you could be successful. its not
always about confluence. i am dead serious when i tell you to go look
at any market you want. pull up a daily or wekly chart and look for
these bars with the long noses at swing highs and lows. you all know i dont make outlandish statements but here is one thats close but its true.
if you never trade anything else but these bars at these points on
daily and weekly charts you can bury any daytrader in the world over
time. this one post was worth your 99 bucks and much more.







__________________
james16group.com

"you cant rollerskate in a buffalo herd"
"you cant take a shower in a parakeet cage"
Roger Miller































james16


View Public Profile


Send a private message to james16


Send email to james16


Visit james16's homepage!


Find all posts by james16


Add james16 to Your Buddy List




















 
#47  


 








01-08-06, 11:34am
















firehorse



vbmenu_register("postmenu_37253", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532




















Backtesting Bub & Beb






Hi,

I've been doing some backtesting on the price action because I wanted
to see if I could determine a constant entry buffer, stoploss and
target.

So far I've done single bars, Bub and Beb on eurusd.

Definition for back testing of Bub

Close in top 30% of bar.
Open in bottom half of bar.
Enter 1 or more pips above high of BuB.
Stoploss at 1 or more pips below low of BuB

Additional rule I've added because it seemed to make sense at the time, high of Bub is higher than previous bar.

Opposite definition for Beb.

I've varied the entry buffer, stoploss and target to see the best
results for each year going back to 1999. Testing done in 1 hour
resolution. Data at GMT+2 so opening on Sunday 5pm EST is shifted to
midnight (5+5+2=12), so there is effectively no Sunday bar!
Results are currency, year, entry buffer, stoploss, target, pips
achieved, no of trades and recovery (Net profit divided by Max. system
drawdown)
//eurusd 05 - 6, 1, 125 -> +1352 in 41, rec 3.18
//eurusd 04 - 3, 1, 5 -> +97 in 71, rec 0.47
//eurusd 03 - 1, 1, 135 -> +1245 in 46, rec 1.88
//eurusd 02 - 5, 1, 165 -> +948 in 25, rec 1.59
//eurusd 01 - 3, 1, 5 -> +139 in 79, rec 0.87
//eurusd 00 - 1, 2, 170 -> +1395 in 34, rec 2.31
//eurusd 99 - 6, 1, 25 -> +198 in 49, rec 0.94

I then optimized 99-04 and ran the results on 05
//eurusd 99-04 - 4, 1, 140 -> 1031 in 212. Rec 0.49
//eurusd 05 - 4, 1, 140 -> 342 in 33. Rec 0.33

I then tried changing where the bar needed to close for 99-04
//eurusd 99-04 - 4, 1, 140 -> 1524 in 198 trades. Rec 0.82 (0.25)
//eurusd 99-04 - 4, 2, 140 -> 1284 in 185 trades. Rec 0.71 (0.20)
//eurusd 99-04 - 4, 1, 185 -> 2435 in 139 trades. Rec 1.18 (0.15)
//eurusd 99-04 - 3, 5, 5 -> 521 in 167 trades. Rec 2.47 (0.10)
//eurusd 99-04 - 1, 5, 10 -> 418 in 74 trades. Rec 1.24 (0.05)

I then ran the best pip and best recovery factor for 05
//eurusd 05 - 4, 1, 185 -> -1020 in 23 trades. Rec -0.86 (0.15)
//eurusd 05 - 3, 5, 5 -> -179 in 44 trades. Rec -0.54 (0.10)

Please note that they may still be errors in the code.
The timing of where the day starts will of course make a difference.
Note that the above only tests single bars without regard for anything else, like s/r, trendlines, retracements, etc.

I will try other ways to see if there is a consistent way to make money from singles bars ie finding an edge.
If anyone else has any ideas on how to consistently "mechanically" make
money from single bars then please let me know, as long as it can be
programmed!

Best regards
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#48  


 








01-08-06, 12:27pm
















bluemeasure



vbmenu_register("postmenu_37262", true);



Trader



 



Join Date: Aug 2005
Location: Japan


Posts: 125



























Hi Alan,

Nice work!

A few questions:
1) I'm wondering what kind of results you would get if you were to use higher timeframes, say 4H or Daily data.
2) What was your exit strategy (target) for the results you posted?
3) What platform are you doing your backtesting on?

Thanks,
BlueMeasure

































bluemeasure


View Public Profile


Send a private message to bluemeasure




Find all posts by bluemeasure


Add bluemeasure to Your Buddy List




















 
#49  


 








01-08-06, 1:05pm
















firehorse



vbmenu_register("postmenu_37264", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi,
Quote:





Originally Posted by bluemeasure


Hi Alan,

Nice work!

A few questions:
1) I'm wondering what kind of results you would get if you were to use higher timeframes, say 4H or Daily data.
2) What was your exit strategy (target) for the results you posted?
3) What platform are you doing your backtesting on?

Thanks,
BlueMeasure




1) I was backtesting 'daily' bars at 1 hour
resolution/periodicity, because you can get confusing results when
testing daily bars if the following bar is engulfing (higher high and
lower low). You don't know if your profit or stoploss was hit first. I
went down to 1 hour to calculate the profit and stoploss points as I
thought that should give enough accuracy for a guideline as to the
profit or loss. It just takes longer to calculate the more lower
timeframe used to backtest.

2) For the above testing, limit exit only. So the trade either hit stoploss or limit exit.
I tested entry buffer from 1 point to 8 points step 1 point.
I tested stoploss from 1 point to 6 points step 1 point.
I tested limit exit from 5 points to 200 points step 5 points.
I have not yet tested any case of moving to breakeven or trailing stoploss.

3) Amibroker 4.70.5. Data from forexite.

Note that I may yet find an error in the program that would totally invalidate all the above results so don't take it 'too' seriously but it's the best model I've got at the moment.

Best regards
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#50  


 








01-08-06, 1:41pm
















WTB



vbmenu_register("postmenu_37266", true);



Member



 



Join Date: Sep 2005



Posts: 399



























Sorry Alan, I am too stupid to understand your results. So, to sum it up, were they possitive or negative pips-wise?

































WTB


View Public Profile


Send a private message to WTB


Send email to WTB



Find all posts by WTB


Add WTB to Your Buddy List




















 
#51  


 








01-08-06, 2:29pm
















firehorse



vbmenu_register("postmenu_37270", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi,
Quote:





Originally Posted by WTB


Sorry Alan, I am too stupid to understand your results. So, to sum it up, were they possitive or negative pips-wise?




'Bub & Beb, by themselves, without any other information,
using the above test conditions', my thoughts 'so far' is that it's a
bit of a hit and miss.

Using the same entry buffer, stoploss and limit exit for every year is 'not' the best way to trade BuB & Beb.

Under the conditions tested:
Using a stoploss of 1 pip seems like a good idea.
The best entry buffer varies between 1 and 6 from year to year.
The best limit exit varies between 5 and 170 between year to year.

I was posting my results in the hope of someone either proving me wrong or being able to improve my results.

I'm still waiting for James to answer some of my questions to see if I can backtest more correctly.

Best regards
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#52  


 








01-08-06, 8:19pm
















pipmaker



vbmenu_register("postmenu_37299", true);



Member



 



Join Date: Nov 2004



Posts: 127




















Amibroker






@firehorse

Thanks for your results on backtesting. I use amibroker too, but I´m far
away from beeing able to do back testing with this software. Would it be
possible for you to explain how you do this? Eventually in a seperate
Thread? That would be a great help for me and other amibroker users.
Thanks, Chris.

































pipmaker


View Public Profile


Send a private message to pipmaker




Find all posts by pipmaker


Add pipmaker to Your Buddy List




















 
#53  


 








01-08-06, 8:38pm
















james16



vbmenu_register("postmenu_37307", true);



wizard in training



 



Join Date: Feb 2005
Location: Dallas-Ft.Worth,Texas


Posts: 2,607




























Quote:





Originally Posted by firehorse


Hi,'Bub
& Beb, by themselves, without any other information, using the
above test conditions', my thoughts 'so far' is that it's a bit of a
hit and miss.

Using the same entry buffer, stoploss and limit exit for every year is 'not' the best way to trade BuB & Beb.

Under the conditions tested:
Using a stoploss of 1 pip seems like a good idea.
The best entry buffer varies between 1 and 6 from year to year.
The best limit exit varies between 5 and 170 between year to year.

I was posting my results in the hope of someone either proving me wrong or being able to improve my results.

I'm still waiting for James to answer some of my questions to see if I can backtest more correctly.

Best regards
Alan




here you go alan. the truth is price action bars will probably be
tough to backtest because there are so many variables involved. please
keep in mind that i am not advocating trading every bar you see. far
from it. there are good areas to trade them and not so good. the idea
is to incorporate them into your trading thus giving you an edge on
where you think the market is likely to go in the near term. trying to
backtest every bar takes the deescretion of being a smart trader and
attempts to try to trade every bar which should not be done anyway. i
am still interested in your work however and enjoy your skills. try to
refine your selection process to better reflect real descretionary
trading as opposed to every bar. good stuff your doing. jim

1. Is there any requirement of where the open needs to be, like bottom
half, third, quarter, or it only needs to be below the close?

2. Also any requirement for it to be of a particular size compared to some sort of average range?

3. Any requirement for the high of the bar to be higher than the previous bar or low to be higher than previous bar?


1. in a BUB example the best bars tend to be those where the open is in
the bottom quarter and the close is in the top quarter . my experience
is the higher on the bar the close the better.

2. markets vary of course but you will notice when scouring for them that a relatively long bar is best.

3. most good BUB's or BEB's tend to be breakaway bars especially out of channels or off of support or resistence.[/quote]







__________________
james16group.com

"you cant rollerskate in a buffalo herd"
"you cant take a shower in a parakeet cage"
Roger Miller








Last edited by james16 : 01-08-06 at 8:45pm.































james16


View Public Profile


Send a private message to james16


Send email to james16


Visit james16's homepage!


Find all posts by james16


Add james16 to Your Buddy List




















 
#54  


 








01-08-06, 11:27pm
















james16



vbmenu_register("postmenu_37384", true);



wizard in training



 



Join Date: Feb 2005
Location: Dallas-Ft.Worth,Texas


Posts: 2,607




















Update 1-8-06






example chart










Attached Images

 

















__________________
james16group.com

"you cant rollerskate in a buffalo herd"
"you cant take a shower in a parakeet cage"
Roger Miller































james16


View Public Profile


Send a private message to james16


Send email to james16


Visit james16's homepage!


Find all posts by james16


Add james16 to Your Buddy List




















 
#55  


 








01-09-06, 1:50am
















firehorse



vbmenu_register("postmenu_37416", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi,
Quote:





Originally Posted by pipmaker


@firehorse

Thanks for your results on backtesting. I use amibroker too, but I´m far
away from beeing able to do back testing with this software. Would it be
possible for you to explain how you do this? Eventually in a seperate
Thread? That would be a great help for me and other amibroker users.
Thanks, Chris.




Would love to help but that would be a full-time job! It's taken
me 18 month's of solid hard slog to understand a tiny bit of how
Amibroker works. Thinking in Amibroker arrays still does my head in.
Each backtesting program is still being built from first principles
rather hashing together old routines.

The best advice I can give is to join the Amibroker mailing list, read,
read and read and then try for yourself. The benefit of Amibroker is
that 'if' your program & data is correct, the results are accurate.
Also for optimizations, it's damn fast!

The numbers I've given out is what I believe to be the minimum you need to have an idea of what went on during the backtesting.

Each year has been tested separately to show that one set of numbers
don't work across different years. The number of trades has been given
because you still need to take the spread off per trade. Recovery is
given so you know how much of a relative drawdown there was during the
year.

Best regards
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#56  


 








01-09-06, 2:05am
















firehorse



vbmenu_register("postmenu_37417", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi Jim,

I'm backtesting to see if there was anyway 'for me' to define a good
Bub & Beb. I understand it will work 'better' with confluence but
I'm looking to see if there are any conditions by which a Bub or Beb
would give me an edge in general

By prodding you with questions, I'm hoping to bring out the little nuggets that you do automatically without thought


Quote:





Originally Posted by james16


1.
Is there any requirement of where the open needs to be, like bottom
half, third, quarter, or it only needs to be below the close?

2. Also any requirement for it to be of a particular size compared to some sort of average range?

3. Any requirement for the high of the bar to be higher than the previous bar or low to be higher than previous bar?


1. in a BUB example the best bars tend to be those where the open is in
the bottom quarter and the close is in the top quarter . my experience
is the higher on the bar the close the better.

2. markets vary of course but you will notice when scouring for them that a relatively long bar is best.

3. most good BUB's or BEB's tend to be breakaway bars especially out of channels or off of support or resistence.




There you go, a few more nuggets

1. Easy to program in.
2. I'll try an ATR of some sort.
3. Pass! I don't how to program detection of channels, s/r

Big thing missing is moving to BE. Let me see if I can solve that.

Best regards
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#57  


 








01-09-06, 2:37am
















firehorse



vbmenu_register("postmenu_37422", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi
Quote:





Originally Posted by james16


example chart




Thank you for the example.

It's interesting to note the Bubs have higher highs and higher low, and
BeBs have lower low and lower high! Would this be a general rule?
(Looking for another nugget )

What is 'more' interesting is what is 'not' labelled.

Bar A - BeB - Not taken because still in trade from Bar 1
Bar B - Bub - Not taken because still in trade from Bar 4, also
following bar was inside bar so didn't activate 5 pip entry buffer.
Bar C - Bub - Perfect Bub but I think this wasn't taken because it was up against a resistance.
Bar D - Beb - Closed just above 0.25, equal to 0.3 so no trade.
Bar E - Beb - No idea why this was not taken
Bar F - Beb - Good Beb but I think this wasn't taken because it was down against a support.

Are my points correct and can you explain bar E?

Thanks
Alan










Attached Images

 











































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#58  


 








01-09-06, 3:57am
















james16



vbmenu_register("postmenu_37444", true);



wizard in training



 



Join Date: Feb 2005
Location: Dallas-Ft.Worth,Texas


Posts: 2,607




















Reply






REPLY
Quote:





Originally Posted by firehorse


HiThank you for the example.

It's interesting to note the Bubs have higher highs and higher low, and
BeBs have lower low and lower high! Would this be a general rule?
(Looking for another nugget )

What is 'more' interesting is what is 'not' labelled.

Bar A - BeB - Not taken because still in trade from Bar 1
YES


Bar B - Bub - Not taken because still in trade from Bar 4, also
following bar was inside bar so didn't activate 5 pip entry buffer.
YES


Bar C - Bub - Perfect Bub but I think this wasn't taken because it was up against a resistance.
NO, NEXT BAR DID NOT BREAK BAR C OR IT WOULD HAVE BEEN TAKEN


Bar D - Beb - Closed just above 0.25, equal to 0.3 so no trade.
THE BEST BUB OR BEB CLOSE IN THE LAST QUARTER OF THE BAR AND THE
FURTHER THE BETTER. IN THESE EXAMPLESTHIS ONE DID NOT MEET THE BEST OR
OBVIOUS CRITERIA I WAS LOOKING FOR


Bar E - Beb - No idea why this was not taken
ONCE AGAIN I WAS JUST POINTING OUT THE VERY BEST ONES. THIS ONE WAS CLOSE AND BY THE WAY IT WAS A GOOD ONE

Bar F - Beb - Good Beb but I think this wasn't taken because it was down against a support.
NEXT BAR NEVER BROKE PAST BAR F TO THE DOWNSIDE

Are my points correct and can you explain bar E?

Thanks
Alan












__________________
james16group.com

"you cant rollerskate in a buffalo herd"
"you cant take a shower in a parakeet cage"
Roger Miller































james16


View Public Profile


Send a private message to james16


Send email to james16


Visit james16's homepage!


Find all posts by james16


Add james16 to Your Buddy List




















 
#59  


 








01-09-06, 10:46am
















firehorse



vbmenu_register("postmenu_37497", true);



Member



 



Join Date: Mar 2005
Location: London


Posts: 532



























Hi,

Thanks for taking the time to answer.

I feel like a right plonker for Bars C & F

Interesting point for bar E.

Many thanks
Alan

































firehorse


View Public Profile


Send a private message to firehorse


Send email to firehorse



Find all posts by firehorse


Add firehorse to Your Buddy List




















 
#60  


 








01-09-06, 11:04pm
















dnbae



vbmenu_register("postmenu_37591", true);



-



 



Join Date: Dec 2005



Posts: 23




















Pin bar?






Hi Jim,

Attached is a trade I took on the 4hr USD/CAD. It looked like a good
pin bar except that the right eye was higher than the left eye. I
thought it was a good trade also because this bar occurred at the top
of a declining trend channel. Would you have taken this trade even
though the eyes are a bit off? Isn't it your criterion that in an
upward protruding pin nose, the right eye be slightly lower than the
left eye? Or is that not so important? Thanks so much for your help,
Jim!

http://www.forexfactory.com/forexfor...tid=2912&stc=1










Attached Images

 











































dnbae


View Public Profile


Send a private message to dnbae


Send email to dnbae



Find all posts by dnbae


Add dnbae to Your Buddy List

























Page 4 of 32

<
123 4
5678914
>
Last



























Currently Active Users Viewing This Thread: 1 (1 members and 0 guests)




strangemartian















Thread Tools


Show Printable Version


Email this Page





Subscribe to this Thread

















Display Modes



Linear Mode




Switch to Hybrid Mode




Switch to Threaded Mode














Search this Thread










Advanced Search















Rate This Thread




Excellent
Good
Average
Bad
Terrible






















































































© Forex Factory
Contact
Disclaimer


Preferred Brokers
Preferred Services
Forums
Calendar
News





Top