100% Mechanical PinBar system | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Open Range Break - 100% Mechanical 59 replies

Will this 100% mechanical system survive the long run? 0 replies

100% mechanical trading systems 69 replies

Profitable trading systems, which are 100% mechanical? 2 replies

100% mechanical systems 12 replies

Attachments: 100% Mechanical PinBar system
Exit Attachments

100% Mechanical PinBar system

  • #1
  • First Post: Edited Jan 25, 2011 4:10am Jan 18, 2011 4:58am | Edited Jan 25, 2011 4:10am
Purpose of the thread
To get forum members to assist in developing a long term profitable system that is 100% mechanical (i.e. can be traded with an EA) based on PinBar candles. I am not sure if this is achievable.

If you are not sure what a PinBar is please search the forum.

The attached EA is for members to test and share their results. Please post the results of your backtests using different settings and timeframes.

Suggestions to improve are welcome but let’s try to keep it a simple as possible. Constructive criticism is welcome. I want to keep the system 100% mechanical, i.e. no trader discretion, any filters should be programmable.

I’ll keep this first post updated with the latest versions of the EA, best strategies, current development etc.

Using the EA
The attached EA is to be used for backtests only, not for live trading. I’ve been programming for less than a month, i.e. I’m a total beginner programmer, the code does not cater for disconnects etc and will probably wipe out your account.

What the EA does?
The EA will work on any currency pair or timeframe.

The EA does not cater for different digit brokers; you may have to add a zero to your different pips settings (e.g. 300 for 30 pips).

The EA identifies a PinBar based on your settings.


Update V1.2
- I have sorted out the problem where stoporders are placed too close to current price. I am not using market orders as sometimes price may still be 1 pip from trigger level. The EA will change the standard entry level by 1 to 2 pips to still allow for a stoporder. It will print you a message confirming the change.
- Open_vs_Close: If set to TRUE, for a qualifying long entry the pinbar close has to be above the open, opposite for short entry.

Update V1.4
- Most important improvement, EA now uses market orders, order problems should be very limited, please let me know if you experience any
- Minimim size of pinbar can now be determined by % of ATR
- EA will draw arrows above/under qualifying pinbars
- You can use TPtoSLRatio to set your TP level, i.e. you can select a risk:reward ratio rather than a fixed number of pips

Update V1.5
- RSI, Stochastic, ADX filters added
- Descriptions added in EA setting to explain the different setting
Attached File(s)
File Type: ex4 Pasopvirpot-PinBar-EA-V1.5.ex4   16 KB | 2,932 downloads | Uploaded Jan 25, 2011 4:08am
  • #2
  • Jan 18, 2011 5:57am Jan 18, 2011 5:57am
without source code, its difficult to see what its doing.

can you make this work as indi only ? in that case , no trades are sent. AllowUserTrading=true/false

when you load the EA, can it scan the prev bars (LookBackBars=1000) and mark the pinbars ?

can it put a arrow up/down on a pinbar ?

can you allow user to specify the head color ? i mean in case of bullish pinbar, i want head as green. if close is below open, i dont trade it as a bullish pinbar. its an option

what should be the min head size as a percentage of bar length. this is different from where the head is located in the bar. this will eliminate doji type scenarios

finally, can it act as a universal scanner ? can it scan ALL pairs on chosen Timeframes and list the pin bars or place orders ? i dont want to put the EA on 50 different pairs and 6 timeframes. instead i want to scan every day morning and place orders. thats all

i know i am asking too much, but lets see...
  • #3
  • Jan 18, 2011 6:02am Jan 18, 2011 6:02am
Been tried to death, but good luck
  • #4
  • Jan 18, 2011 6:16am Jan 18, 2011 6:16am
Quoting Sim
Been tried to death, but good luck
Its like being told not to put your hand in the fire. You have to experience getting burned before the lesson really sinks in.

Sadly writing EA's based on nonsensical concepts is all part of the journey, you have to start somewhere and doing a bit of research of your own is a thousand times more beneficial than paying a bogus mentor 100 bucks a month to learn about pin bars.

Even if it only sharpen up his coding skills its worth doing.
  • #5
  • Jan 18, 2011 6:24am Jan 18, 2011 6:24am
If you can get a pin bar indicator to work successfully, I will give you a system you could code into an EA.

Good luck with that minor hurdle.

Juicyt
"Even if you fall down face first, you still went forward."
  • #6
  • Jan 18, 2011 6:32am Jan 18, 2011 6:32am
@ roundrock
This is just a simple EA for people to experiment with. It should be fairly easy to convert to an indicator to make arrows etc. Can it be used as a universal scanner? I'm sure it's possible and something I may try in the future but that is not the aim of this thread. Just running the EA in strategy tester visual mode will quickly give you an idea of the pinbars it identifies and the way it trades.

With regards to (1) minimum size of the head and (2) close vs open, these are filters that I won't mind programming if it will improve the profitability of the EA.

Can you clarify how you would suggest measuring the size of the head?

@Sim
Thanks for the comment, why would you say did most of these efforts not succeed?
  • #7
  • Jan 18, 2011 6:50am Jan 18, 2011 6:50am
Quoting Razzle
Its like being told not to put your hand in the fire. You have to experience getting burned before the lesson really sinks in.

Sadly writing EA's based on nonsensical concepts is all part of the journey, you have to start somewhere and doing a bit of research of your own is a thousand times more beneficial than paying a bogus mentor 100 bucks a month to learn about pin bars.

Even if it only sharpen up his coding skills its worth doing.
Great comment, exactly what I'm trying to achieve, programming and testing my own EA's have taught me a lot in a very short time, even though results have been very disappointing. And my coding skills improve with every attempt...
  • #8
  • Jan 18, 2011 6:54am Jan 18, 2011 6:54am
Quoting juicyt
If you can get a pin bar indicator to work successfully, I will give you a system you could code into an EA.

Good luck with that minor hurdle.

Juicyt
The EA already correctly identifies pinbars based on my criteria, and any additional criteria should be easy to add, why do you say this is a hurdle?

I would love to try out your system.
  • #9
  • Jan 18, 2011 7:49am Jan 18, 2011 7:49am
2011.01.18 19:14:52 2008.09.12 00:00 Pasopvirpot-PinBar-EA-V1.1 EURUSD,Daily: Ordersend failed with error #130 50


pl add error handling code, its difficult for users to know what went wrong. pl print the order price, sl, tp so that users can see the logs.
  • #10
  • Edited 8:54am Jan 18, 2011 8:35am | Edited 8:54am
Quoting roundrock
2011.01.18 19:14:52 2008.09.12 00:00 Pasopvirpot-PinBar-EA-V1.1 EURUSD,Daily: Ordersend failed with error #130 50

pl add error handling code, its difficult for users to know what went wrong. pl print the order price, sl, tp so that users can see the logs.
Will update error handling code in next version.

One problem I need to eliminate: the EA places stop orders at the end of the Pinbar + buffer. In a few instances price has already moved too close to the entry level (e.g. if it breaks the pinbar right at the open of the next bar), the stoporder is then not accepted (error 130).
  • #11
  • Jan 18, 2011 9:02am Jan 18, 2011 9:02am
Quoting pasopvirpot
Will update error handling code in next version.

One problem I need to eliminate: the EA places stop orders at the end of the Pinbar + buffer. In a few instances price has already moved too close to the entry level (e.g. if it breaks the pinbar right at the open of the next bar), the stoporder is then not accepted (error 130).
In that case, if the current price is beyond your intended entry price, i.e. you want to be in the market, then you could code to place a market order. Need to code for handling context busy msgs as well if the EA is placed on more than one chart and you run into multiple orders trying to be placed at the same time.
  • #12
  • Jan 18, 2011 9:06am Jan 18, 2011 9:06am
Quoting fxtradera
In that case, if the current price is beyond your intended entry price, i.e. you want to be in the market, then you could code to place a market order. Need to code for handling context busy msgs as well if the EA is placed on more than one chart and you run into multiple orders trying to be placed at the same time.
Thank you, as you say I will have to to use market orders in these instances.
  • #13
  • Jan 18, 2011 9:32am Jan 18, 2011 9:32am
Quoting pasopvirpot
Thank you, as you say I will have to to use market orders in these instances.
some brokers wont accept market orders with SL, TP. You need to first place market order and then modify it with SL, TP
  • #14
  • Edited 9:57am Jan 18, 2011 9:46am | Edited 9:57am
Thanks for sharing,

A strategy based only on pinbars will need one heavy money management strategy to work. The exit strategy is what is important here.

I will check out your EA and see what I can contribute

Iemand hier is van Suid Afrika hehe
  • #15
  • Jan 18, 2011 3:52pm Jan 18, 2011 3:52pm
Quoting Saidar
Iemand hier is van Suid Afrika hehe
Ja hier is bloubulle op hierdie forum!
  • #16
  • Jan 18, 2011 4:28pm Jan 18, 2011 4:28pm
Quoting pasopvirpot
Great comment, exactly what I'm trying to achieve, programming and testing my own EA's have taught me a lot in a very short time, even though results have been very disappointing. And my coding skills improve with every attempt...
Pasopvirpot,
Last year I worked with alouisos to come up with a pinbar EA. I am attaching the EA here. It may save you some work or give you some ideas:
http://www.forexfactory.com/showthread.php?t=222791

Kostas

P.S. In that thread you can find V2.0 as an mq4 file
Attached File(s)
File Type: ex4 Pin Trader EA V3.0.ex4   15 KB | 1,818 downloads
  • #17
  • Jan 19, 2011 12:52am Jan 19, 2011 12:52am
Quoting pasopvirpot
Ja hier is bloubulle op hierdie forum!
Hehe goed om te sien hier is normale mense
Ek ontwikkel nou al vir so 'n jaar lank EA's, so ons moet bietjie praat een of ander tyd.

About the robot, like I said it is the exit that is important right now work on your order and money management.
  • #18
  • Jan 19, 2011 3:21am Jan 19, 2011 3:21am
Updated the EA to V1.2.
  • #19
  • Jan 19, 2011 5:00am Jan 19, 2011 5:00am
You need to first have an indicator that will find all the circled long wicks, then you will be able to construct a system.

In the mean time, I'll see what else I can dig up.
Attached Image (click to enlarge)
Click to Enlarge Name: longwick.jpg Size: 139 KB
"Even if you fall down face first, you still went forward."
  • #20
  • Jan 21, 2011 5:14am Jan 21, 2011 5:14am
Quoting Razzle
Its like being told not to put your hand in the fire. You have to experience getting burned before the lesson really sinks in.

Sadly writing EA's based on nonsensical concepts is all part of the journey, you have to start somewhere and doing a bit of research of your own is a thousand times more beneficial than paying a bogus mentor 100 bucks a month to learn about pin bars.

Even if it only sharpen up his coding skills its worth doing.
It's not a nonsensical concept at all, it's just you'll find out this way also that the amount of pinbars an script will find is far more than the naked eye, because with discretion we naturally see the best pins, which stick out on swing points, have good OHLC, etc

But a script will just humble us every time, finding pins which fall in the same rules but are just in chop, and really have zero predictability.
0 traders viewing now