Bollinger Band Divergence-Indi Request (Profitably tested) | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

MT4 Breakout Alert Bollinger Band & Donchian Channel indi? 0 replies

ema bollinger band alerts 1 reply

Bollinger band difference indicator... 2 replies

Bollinger band settings? 1 reply

Bollinger Band Width for Metatrader 4 replies

Attachments: Bollinger Band Divergence-Indi Request (Profitably tested)
Exit Attachments

Bollinger Band Divergence-Indi Request (Profitably tested)

  • #1
  • First Post: May 22, 2012 3:40pm May 22, 2012 3:40pm
Hi,
is there any Coder that likes to code an Indi like the following.
I have tested nearly half a year with this kind of strategy and the Main-Indi would be like a "Bollinger-Band-Divergence"-Indi.
So I think of :
1. Compare X-Bars with each other.
2. All bars that have to be compared must be over the upper BB ( for a Sell-Trade ! ) / under BB ( for a Buy -Trade ) .
3. For a BUY:
If Price is higher then the the last bar (the Bar the indi is comparing with ! ) and the BB 's are not higher then there is a BB-Divergence and an Arrow (down ! ) should be painted over the bar.
For a SELL:
Vice Versa
4. Option for comparing the Middle Band (over /under ) of the BB should be included. (This feature is not tested very far but it's an option of a further strategy with this Indi ! )

I aprreciate if there are any ideas for this Indi and one of the fabulous Coder's out there in the FF-universe would like to code this Indi.
Greetings Zack
  • #2
  • May 23, 2012 9:27am May 23, 2012 9:27am
I there no Coder interested in this project.
Should be anyway worthy regarding my Backtesting .
Greeds zack
  • #3
  • May 25, 2012 2:05am May 25, 2012 2:05am
I get anything up to 10 messages every week asking me to code MT4 EAs. I (hopefully politely) decline virtually all of them.

No offense to zackery or anybody else, but it bugs me that many aspiring traders expect programmers to code their unproven ideas without offering any form of payment. If 90% of traders are unprofitable, then I'd suggest that more than 99% of (and I mean bug-free) EAs are likewise, simply because it's not possible for an EA to trade as intelligently as a human. If the EA turns out to be long-term unprofitable, then the harsh reality is that it's a total waste of time and effort for everybody involved.

Thanks largely to the library of routines that I've built up over the years, I can create most requested EAs in less than an hour. But past experience tells me that it frequently takes several e-mails for folk to explain exactly what they want, and then — not to mention those who modify their initial request as they continue to experiment with variation upon variation — anything up to several weeks to test the end result, as users allege that the EA has taken a trade that it shouldn't have, or missed a trade that it should have taken. In many cases, it's very difficult to troubleshoot price action that's been and gone, and by which point I'm usually immersed in another project; and having to down tools to revisit code that I wrote several weeks ago is a real nuisance. Given that whenever I start a job, I like to see it through to completion, it's simply not worth the hassle.

Maybe I'm a selfish, grumpy old man, but my time is precious to me. People expect to pay their accountant, plumber, or hooker, so why not their programmer? Just because some programmers are generous enough to code for nothing on forums like FF doesn't necessarily mean that we are unskilled, dime-a-dozen drones.

Rant over. LOL.

David
  • #4
  • May 25, 2012 6:15am May 25, 2012 6:15am
Quoting hanover
I get anything up to 10 messages every week asking me to code MT4 EAs. I (hopefully politely) decline virtually all of them.

No offense to zackery or anybody else, but it bugs me that many aspiring traders expect programmers to code their unproven ideas without offering any form of payment. If 90% of traders are unprofitable, then I'd suggest that more than 99% of (and I mean bug-free) EAs are likewise, simply because it's not possible for an EA to trade as intelligently as a human. If the EA turns out to be long-term unprofitable, then the harsh...
Well I really understand everything that is said here.
But I think if someone has a new idea ( don't know if my idea is published before! ) it may be a good opportunity for a Coder to get an additional Information about the market.
I understand that some Indis are not worth the effort to code it and it's a financial risk for a Coder to code something worthless but that's the way it goes sometimes . Sorry for that.

Anyway I will posting an example of what I have meant with my coding-request.

Maybe anyone of the Coders here have the time to do it and esspecially when it could be done mostly in 1 hour like "hanover" said.

Greeds Zack
  • #5
  • May 25, 2012 6:36am May 25, 2012 6:36am
BB-Divergence-Example attached.
Thanks for any response and any post that have been done here in this thread.
As you can see in the pic there many more examples for a very profitable Trade .
Along with a Strategy this is a superb additional Information.
But that's only my opinion. Thanks
Greeds Zack
Attached Image (click to enlarge)
Click to Enlarge Name: bb-divergence-example.jpg Size: 126 KB
  • #6
  • May 25, 2012 6:54am May 25, 2012 6:54am
Quoting zackery
BB-Divergence-Example attached.
Thanks for any response and any post that have been done here in this thread.
As you can see in the pic there many more examples for a very profitable Trade .
Along with a Strategy this is a superb additional Information.
But that's only my opinion. Thanks
Greeds Zack
As you can see in the picture the Price is making a new HIGH and the BB's are not making a new one.
Look for the red lines to underline this and the blue rectangle to show the Divergence with the BB's.
Z.
  • #7
  • May 25, 2012 8:03am May 25, 2012 8:03am
hey zackery..... have you by chance tried lookin around over at robert hill's yahoo group.... seems i've seen something like your request over there.....h
to trade and code, keep both simple... no call to impress....h
  • #8
  • May 25, 2012 8:17am May 25, 2012 8:17am
Quoting hayseed
hey zackery..... have you by chance tried lookin around over at robert hill's yahoo group.... seems i've seen something like your request over there.....h

Well thanks for that but I have tried to find this Indi everywhere and I think this is probably not done till yet.
Z.
  • #9
  • May 25, 2012 9:13am May 25, 2012 9:13am
Quoting zackery
I understand that some Indis are not worth the effort to code it and it's a financial risk for a Coder to code something worthless but that's the way it goes sometimes . Sorry for that.
If your idea is "tested profitably", why don´t you make money with it and then pay a coder to create the indicator ?

Don´t you think a coder should earn money ?
  • #10
  • May 25, 2012 9:25am May 25, 2012 9:25am
Quoting fxtr51
If your idea is "tested profitably", why don´t you make money with it and then pay a coder to create the indicator ?

Don´t you think a coder should earn money ?
As I have mentioned before it could be an additional Information for a Trade.
So I found it useful. Nothing more.

Sure , Coders could be earn money with that . So it's only a Question here in a Sharing-Forum.
If I have thought that this idea is not be valid for a Coder to "maybe" do it for free than I have not asked here in FF.

So if I have a Coding-Request that I think makes no sense for a Coder to code for free I surely ask what a probably fee he may take.
I have done this before. No doubt.

But I thought that the case is different.

Greeds Z.
  • #11
  • May 25, 2012 10:43am May 25, 2012 10:43am
Quoting zackery
As you can see in the picture the Price is making a new HIGH and the BB's are not making a new one.
Look for the red lines to underline this and the blue rectangle to show the Divergence with the BB's.
Z.
Not sure if this does exactly what you want since your verbiage seems to contradict the picture. But seeing you flounder around looking for help is difficult to watch. At the very least, it should give you a start so you can adapt it to your own needs. Good luck.
Attached File(s)
File Type: mq4 bb-div.mq4   4 KB | 2,242 downloads
  • #12
  • May 25, 2012 2:17pm May 25, 2012 2:17pm
Quoting zznbrm
Not sure if this does exactly what you want since your verbiage seems to contradict the picture. But seeing you flounder around looking for help is difficult to watch. At the very least, it should give you a start so you can adapt it to your own needs. Good luck.
@zznrbm:
first of all thanks a lot for that. Will have a look.
Is this Indi coded by yourself?

One more word about the idea .

The main thing is the divergence between two points:
At the first point you see in the picture a HIGH in the Price above the BB's . Please notice the position of the BB's.
The next HIGH of the Price above the BB's is lower then the price before but the BB's are higher then the last HIGH.

Please feel free to ask if there are any misunderstandings about the idea.
Thanks for that.
Greeds Zack
  • #13
  • May 25, 2012 2:27pm May 25, 2012 2:27pm
@zznrbm:
The attached Indi does not what I was looking for.
Have a look to the attached Screenshot.
Notice:
The white circles are from the indi.
The red,blue and aqua lines are drawn by myself.
Greeds Zack
Attached Image (click to enlarge)
Click to Enlarge Name: bb-divergence-example 2.jpg Size: 98 KB
  • #14
  • May 25, 2012 2:55pm May 25, 2012 2:55pm
Quoting zackery
@zznrbm:
The attached Indi does not what I was looking for.
Have a look to the attached Screenshot.
Notice:
The white circles are from the indi.
The red,blue and aqua lines are drawn by myself.
Greeds Zack
I'm sure it does not. You need to clearly specify how you determine the two price points for the comparision.

In the updated screenshot below, what criteria did you use to choose Point #1 and Point #2 for each divergence?

Do you use a set number of bars (which is how I interpretted Post #1 and how I coded it)?

Are you using new highs/lows as determined by a fractal-type calculation?

Without a detailed explanation of how you pick the data points, the indicator will never match your manual interpretations.
Attached Image (click to enlarge)
Click to Enlarge Name: bb-divergence-example.jpg Size: 164 KB
  • #15
  • May 25, 2012 4:45pm May 25, 2012 4:45pm
Quoting zznbrm
I'm sure it does not. You need to clearly specify how you determine the two price points for the comparision.

In the updated screenshot below, what criteria did you use to choose Point #1 and Point #2 for each divergence?

Do you use a set number of bars (which is how I interpretted Post #1 and how I coded it)?

Are you using new highs/lows as determined by a fractal-type calculation?

Without a detailed explanation of how you pick the data points, the indicator will never match your manual interpretations.
@zznbrm:
Okay, I try to find absolute Highs and Lows from the TF D1, H4 and H1 to compare the Highs and the Lows.
Hope this explains how I find them.
Hope that we can going forward with this.
Thanks anyway till now for your effort.
Greeds and nice weekend Zack
  • #16
  • Edited May 26, 2012 4:43pm May 25, 2012 8:59pm | Edited May 26, 2012 4:43pm
I've attached an image of Zack's chart, and highlighted two trades that I think meet his criteria, but he missed. BTW, both would likely have been profitable. (Sorry about the primitive drawing, but I'm typing this from my wife's computer, while mine is at the repair shop getting its power supply replaced).

Quoting zackery
But I think if someone has a new idea ( don't know if my idea is published before! ) it may be a good opportunity for a Coder to get an additional Information about the market.
Zack, now that you've explained your idea, I think it has potential (do you hear the chewing sound? That's me eating some humble pie, LOL). However, as zznbrm says, you need to clarify what you mean by an "absolute" high.

I will have a shot at it. You can explain if/how I'm wrong:

An absolute high is a candle whose high is above the upper BB, and has X lower candle highs to its immediate left, and X lower highs to its immediate right. X is a user-definable parameter. In other words, the absolute high can't be defined until X subsequent candles have completed. (Obviously the mirror image applies to define an absolute low).

OK, now we need to define the divergence. There must be two absolute highs, within Y candles of each other (Y is a user-defined parameter), where the second candle's high is lower than the first's, but the BB upper band value at the second candle is higher than the BB upper band value at the first candle. (Again, mirror image for a divergence of absolute lows).

If all of the above criteria are met, the indy plots a down/up arrow above/below the second candle, to highlight the bearish/bullish divergence. As I pointed out, the plotting would necessarily lag X candles after the second absolute high/low.

Anyway, hopefully that gets us one step closer to a strict mathematical definition that can be coded.
Attached Image (click to enlarge)
Click to Enlarge Name: zack1.JPG Size: 32 KB
  • #17
  • May 26, 2012 3:59am May 26, 2012 3:59am
Quoting hanover
I've attached an image of Zack's chart, and highlighted two trades that I think meet his criteria, but he missed. BTW, both would likely have been profitable. (Sorry about the primitive drawing, but I'm typing this from my wife's computer, while mine is at the repair shop getting its power supply replaced).

Zack, now that you've explained your idea, I think it has potential (do you hear the chewing sound? That's me eating some humble pie, LOL). However, as zzzbrm says, you need to clarify what you mean by an "absolute" high.

I will have a...
I like to hear those words of greed from every Coder

Hanover,first thanks that you jump on the Bandwagon.

Well I have not so much time today but in a short cut.
For example a Short -Trade:
An absolute HIGH is set above the BB.
Then there must be at least 1 Candle between that HIGH and a next HIGH that will set in the future abover the BB. The divergence is clear I think ,as HANOVER has described it exactly .
One criteria that is really important for my decision :
Between the two HIGHS which met the criteria there must be at least one Candle which have a CLOSE under or equal the BB.

I attach 2 Screenshots to hopefully clearing things up.

Please feel free to contact me if there are any questions about it.

Great weekend to all.
Greeds Zack
Attached Image(s) (click to enlarge)
Click to Enlarge Name: bb-divergence-example 3.jpg Size: 94 KB
Click to Enlarge Name: bb-divergence-example 4.jpg Size: 113 KB
  • #18
  • Edited 5:43am May 26, 2012 5:31am | Edited 5:43am
Quoting zackery
I like to hear those words of greed from every Coder
Greed?

I was merely trying to point you in the direction of explaining your idea in a mechanical way, so that an indicator could be coded from it.

From your last post, it seems that you are starting to think about rule specifics. That's good.
Then after you get an indicator written, and test your idea forward over maybe 500 trades, you'll be in a stronger position to refine the rules further: things like pairs to trade, timeframes, sessions, special exclusions. And hopefully you'll have devised some exit rules, too.
Finally, in maybe 6-12 months, you might be able to start a thread that's legitimately titled "profitably tested".

During 2009-2010 I coded more than 30 EAs for folk who claimed they had a good idea, and one that (like yours) that I thought was worth investing the time and effort needed to code and test it. Guess how many of these EAs are currently still being run as profitable? To the best of my knowledge, absolutely none of them. That's the lesson I've learned — there's a significant gulf between having an potentially excellent trading idea, and being a profitable trader.

To be fair, all you've asked for is an indicator. Indicators make great tools, because their output can be interpreted and filtered manually by an experienced discretionary trader. But that's a million miles away from having a fully automated (and profitable) trading system.

Any "greed" on my part has been more than vanquished by a large dose of realism.
  • #19
  • May 26, 2012 12:14pm May 26, 2012 12:14pm
zachery,

If there is one thing software developers dislike more than anything else, it is ambiguous/incomplete requirements.

Now, I know you are not a business analyst or technical writer. But let me make a suggestion:

Take an hour or so to jot down the criteria for a long signal and a short signal. You've been manually monitoring these signals for 6 months, by now you should be able to detail the criteria as easily as spelling your name. Pretend you are trying to explain the signal criteria to a 5 year old. Remove all ambiguity and give as much detail as possible. Why an hour? So you can really take the time to get all the details correct!

Once you have done that, post the indicator specs on this thread. I'm sure someone will code the indicator for you. Although some people think all the good programmers have migrated to the coding mecca of a certain piano teacher, I can assure you there are many excellent programmers on ForexFactory. If your requirements for your indicator are clear and concise, then someone will take on the project. But continue to post inconsistencies, and you will not get the help you want.

Don't make the developer try to guess what you want to do.
  • #20
  • Edited 3:45pm May 27, 2012 10:04am | Edited 3:45pm
Okay I try to describe it.

For example SHORT:

1. Indi should find the last X-candle-HIGHS of the 4 TF's ( M15 , M30, H1, H4 ). The HIGHS > upper BB . The position of the BB's at the time of the HIGHS should also being marked.

2. Then comparing all the X-HIGHS in every TF with the next upcoming/current Candle- HIGH > BB about if the HIGHS are higher and at the same time the upper BB is also higher or lower . If the Candle-HIGHS > upper BB and the upper BB is not making a new HIGH or if the Candle- HIGH is not higher then one of the last X- Candles but the BB are higher then there is a Divergence .

Means: Everytime a new HIGH > upper BB is set then the Indi compare it with all the last X-HIGHS.

Note: the comparing of each TF should be done separately and not being compared with the HIGHS of an other TF !

3. On condition that after a HIGH is set > upper BB then the next HIGH could first being set if 1 or more Candles will set a CLOSE or an OPEN < or equal then the upper BB .
The most conservative way is with the set:
If 1 or more Candles will set a CLOSE or an OPEN < or equal than the MIDDLE-Band of the BB. With that setting you will get the most profitable Trades. I think it would be fine to have this an OPTION in the INDI to choose whether the Candle-CLOSE/OPEN is set to "upper-BB" : TRUE/FALSE
or
"middle-BB" : TRUE/FALSE

This is just a thought.

4. Everytime a HIGH is set and compared with an other HIGH and there is a Divergence with the HIGH of the BB the Indi should plot an arrow over the Candle-HIGH. Every TF should have a different arrow-colour to separate the HIGHS from each other.

Vice Versa for a LONG .

Hope that everything is nearly being explained right.
Please be forbearing if there are any failure by myself regarding the explanation for the Indi. Additionally I am not a native English speaker.

Sunny Sunday to all.
Greeds Zack
0 traders viewing now