Sharing my free Study EA | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Sharing our Best EAs and Indicators Free of Cost 13 replies

Emacsen Study, Study, Study 0 replies

Sharing hurt your method? 29 replies

Free study materials for Series 3? 4 replies

Attachments: Sharing my free Study EA
Exit Attachments

Sharing my free Study EA

  • #1
  • First Post: May 21, 2020 9:51am May 21, 2020 9:51am
Hello guys,

I've been trying to do some robots in the past weeks. I tried to create robots using generators on ProQuant application and ForexRobotAcademy, but I have noticed that those robots tend to start good and then give less performance over time. I now I enter a new phase where I try to create robot without indicators based only on Market behavior and resistance level.

But the problem is that I am not a coder, and I don't know MQl4, I am using the interface of EAbuilder.com to create my robots. I have tested several ideas. This robot I am sharing below is one of them. If anyone can improve the performance of this robot or can use it as a template to create a more complicated and sophisticated robots, please share back any improvements.

Thanks guys!

The robort has been optimized with a LiteForex ECN demo account for the period January 1st 2019 - April 3rd 2020, but the test period is until May 20 2020. Max spread is fixed to 2. I don't know but maybe the profit factor, the win ratio can be improved a little more.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Capture2.JPG Size: 96 KB
Click to Enlarge Name: Capture3.JPG Size: 74 KB
Click to Enlarge Name: Capture4.JPG Size: 88 KB
Click to Enlarge Name: Capture5.JPG Size: 67 KB
Click to Enlarge Name: Capture6.JPG Size: 73 KB
Click to Enlarge Name: Capture1.JPG Size: 88 KB
Attached File(s)
File Type: mq4 Sixth Study EA.mq4   13 KB | 1,117 downloads
File Type: ex4 Sixth Study EA.ex4   29 KB | 724 downloads
  • #2
  • May 24, 2020 2:50am May 24, 2020 2:50am
Quoting descaboy
Hello guys, I've been trying to do some robots in the past weeks. I tried to create robots using generators on ProQuant application and ForexRobotAcademy, but I have noticed that those robots tend to start good and then give less performance over time. I now I enter a new phase where I try to create robot without indicators based only on Market behavior and resistance level. But the problem is that I am not a coder, and I don't know MQl4, I am using the interface of EAbuilder.com to create my robots. I have tested several ideas. This robot I am...
Can you explain the logic behind your ea?
  • #3
  • May 24, 2020 9:46am May 24, 2020 9:46am
Quoting ayush9342
{quote} Can you explain the logic behind your ea?
The EA is based on Price Action on two Time Frame. First considering the price on H1, when price is above the SMA 10 (H1_3), and H1_1 (shit0) high has moved a certain amount of pips (Plus_H1) from the open of the previous candle H1_2, then entry point is on M15 when the high of current candle crosses above the high of previous candle (Candle_Index_P). This is for Buy signal. The reverse is for Sell signal.

I created the logic, then optimized it to find best values. But I do not have 99.9% data quality yet, only 90%. I am looking to find a way to test my EA's with 99.9% data quality to be sure of their performances. Maybe someone with MQl4 skills can add some custom codes to make it a better EA.

Best Regards
  • #4
  • May 24, 2020 11:33am May 24, 2020 11:33am
Can You share a picture of perfect setup on H1 and M15? Maybe You could number the candles involved in setup...
  • #5
  • May 24, 2020 12:22pm May 24, 2020 12:22pm
Quoting descaboy
{quote} The EA is based on Price Action on two Time Frame. First considering the price on H1, when price is above the SMA 10 (H1_3), and H1_1 (shit0) high has moved a certain amount of pips (Plus_H1) from the open of the previous candle H1_2, then entry point is on M15 when the high of current candle crosses above the high of previous candle (Candle_Index_P). This is for Buy signal. The reverse is for Sell signal. I created the logic, then optimized it to find best values. But I do not have 99.9% data quality yet, only 90%. I am looking to find...

Hello, first you need to know that this is not a system that I have traded manully, i't s just a logic created from the robot then optimized to find best values. It's one of my robot, I have other robots based on other market behavior like the Second Study which based only on Time. The backtest of the robot is on M15. I provide below a few pics of visual backtest maybe they will help.

Regards
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Capture7.JPG Size: 211 KB
Click to Enlarge Name: Capture8.JPG Size: 216 KB
Click to Enlarge Name: Capture9.JPG Size: 218 KB
  • #6
  • May 24, 2020 1:07pm May 24, 2020 1:07pm
The results from strategy tester look good. Your explanation of the entry I can follow, just, but you haven't defined the exits.
I have looked at the code, however, in my experience looking at someone else's or generated code is always more difficult than your own?
I have a modular EA that I can plug rules into very quickly if I know what they are.
I am happy to be involved and will run tester on the ea later.
Thanks for sharing.
  • #7
  • May 25, 2020 5:55am May 25, 2020 5:55am
Quoting Spikehunter
The results from strategy tester look good. Your explanation of the entry I can follow, just, but you haven't defined the exits. I have looked at the code, however, in my experience looking at someone else's or generated code is always more difficult than your own? I have a modular EA that I can plug rules into very quickly if I know what they are. I am happy to be involved and will run tester on the ea later. Thanks for sharing.

hello, the only exit rule for this EA is either Take Profit or Stop Loss. But if you can find a setting that works, it could be an adding feature.
  • #8
  • May 25, 2020 7:18am May 25, 2020 7:18am
Hi, it is the second attempt that I make to answer this discussion, we hope it is successful, I wanted to tell you to be careful of the false test results in demo, but especially when you use TRAIL, CANDEL with INDEX 0, and Price Bid and 3 years that I use that site to make my EAs, I would like to exchange some opinions with you, but in no way I can contact you, probably because I am new in this Forum and I have never written any posts.
Congratulations on your studies.
Forgive my English
  • #9
  • May 25, 2020 7:24am May 25, 2020 7:24am
Quoting descaboy
{quote} hello, the only exit rule for this EA is either Take Profit or Stop Loss. But if you can find a setting that works, it could be an adding feature.
You must have settings otherwise you wouldn't be able to get the results posted..
  • #10
  • May 25, 2020 7:32am May 25, 2020 7:32am
Quoting Spikehunter
{quote} You must have settings otherwise you wouldn't be able to get the results posted..
Each setting you add or change in the EA it will behave differently. Having only TP and SL the EA has a certain behavior, adding an exit rule, the EA change, need to re-optimize to find best parametters again. Some strategies may win only with TP and SL some others may be better with exit rule. The possibilities are limitless.
  • #11
  • May 25, 2020 9:00am May 25, 2020 9:00am
First of all you cannot backtest two timeframes at the same time with mt4. So you should made an ea with only one timeframe if you want to backtest, otherwise you should do forward test on demo.
  • #12
  • May 25, 2020 3:31pm May 25, 2020 3:31pm
Quoting jo24
First of all you cannot backtest two timeframes at the same time with mt4. So you should made an ea with only one timeframe if you want to backtest, otherwise you should do forward test on demo.
Nope, the logic consider two TF but if you look at the backtest image, you'll see it has been done on M15. the logic doesn't mean the robot will trade on two different TF, Price Action on H1 is the main condition but M15 is the TF to trade when the condition on H1 it met.
  • #13
  • May 25, 2020 3:44pm May 25, 2020 3:44pm
Quoting leleitaly
Hi, it is the second attempt that I make to answer this discussion, we hope it is successful, I wanted to tell you to be careful of the false test results in demo, but especially when you use TRAIL, CANDEL with INDEX 0, and Price Bid and 3 years that I use that site to make my EAs, I would like to exchange some opinions with you, but in no way I can contact you, probably because I am new in this Forum and I have never written any posts. Congratulations on your studies. Forgive my English

Hello I know that backtest is only a potential of what the robot can do. Now I have purchased Tick Data Suite, I'll optimize again all my EA with 99.9% data quality. Even then, the performance will be only a figure, but it will give a better idea of the robot's performance.

I don't really see interest in backtesting a robot for long period of time, since market is evolving fast and I think it's best to optimize the robot from time to time to have best parametters to face actual market conditions. The rest in only a matter of the robustness of the strategy involved in the EA. And I think a trader should not fall in love with a robot, it's just a piece of code. The robot is working it's fine, if the strategy can no longer produce profit, then get a new EA. This is the best strategy to keep getting profits in this changing market.
  • #14
  • May 25, 2020 4:05pm May 25, 2020 4:05pm
Quoting descaboy
{quote} Hello I know that backtest is only a potential of what the robot can do. Now I have purchased Tick Data Suite, I'll optimize again all my EA with 99.9% data quality. Even then, the performance will be only a figure, but it will give a better idea of the robot's performance. I don't really see interest in backtesting a robot for long period of time, since market is evolving fast and I think it's best to optimize the robot from time to time to have best parametters to face actual market conditions. The rest in only a matter of the robustness...
Hi, forgive me if I allowed myself to give you my opinion, but even if I am a novice, I know the site you use very well, and I was pleasantly surprised to find another user who uses it, sincerely you are the first in a forum, returning to the strategy you use and which I am clearly testing too, with Alpari demo data more or less the results are similar, I also backtested with EURUSD, a couple that I use frequently but the results are very different, you have by chance of the settings also for EURUSD?
  • #15
  • May 26, 2020 6:53am May 26, 2020 6:53am
Quoting leleitaly
{quote} Hi, forgive me if I allowed myself to give you my opinion, but even if I am a novice, I know the site you use very well, and I was pleasantly surprised to find another user who uses it, sincerely you are the first in a forum, returning to the strategy you use and which I am clearly testing too, with Alpari demo data more or less the results are similar, I also backtested with EURUSD, a couple that I use frequently but the results are very different, you have by chance of the settings also for EURUSD?

Hello, no problem to give your opinion my friend.

I do have a couple of robots to trade different pairs like Euro, Audusd, Nzdusd... But recently I was more interested to find a good robot that can trade the Cable, a pair that has more range and that is moving more than the Euro. That's why all my Study EA's are for the cable, but you can try to optimize it for the Euro to see what you can get.

Volatily in the Euro has dropped significantly so I am not too focus on this pair for now.
  • #16
  • May 26, 2020 7:22am May 26, 2020 7:22am
This is a 99.9% tick by tick backtest on M15 with this EA on ICM from 2011 Jan - 2020 May.
Past experience with these plug and play models tells me that your chance at long term positive expectancy is higher if you use it to create divergent strategies as opposed to convergent ones. There are some good threads on this forum adopting this approach, but you need to be able to live with the constant closing at small losses in search for that major trend explosion that provides you with most of your profits.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 1.gif Size: 97 KB
Click to Enlarge Name: 2.JPG Size: 240 KB
  • #17
  • May 26, 2020 11:22am May 26, 2020 11:22am
[quote = descaboy; 12963428] {quote} Ciao, nessun problema a dare la tua opinione amico mio. Ho un paio di robot per scambiare diverse coppie come Euro, Audusd, Nzdusd ... Ma recentemente sono stato più interessato a trovare un buon robot in grado di scambiare il cavo, una coppia che ha più portata e che si sta muovendo più del Euro. Ecco perché tutti i miei EA di studio sono per il cavo, ma puoi provare a ottimizzarlo per l'euro per vedere cosa puoi ottenere. Volatilmente l'euro è sceso in modo significativo, quindi per ora non sono troppo concentrato su questa coppia. [/ Quote]


too bad, both only in Demo, but today excellent operation of $ 280 with lot 0.40 ..
  • #18
  • May 26, 2020 11:24am May 26, 2020 11:24am
Quoting carddard
This is a 99.9% tick by tick backtest on M15 with this EA on ICM from 2011 Jan - 2020 May. Past experience with these plug and play models tells me that your chance at long term positive expectancy is higher if you use it to create divergent strategies as opposed to convergent ones. There are some good threads on this forum adopting this approach, but you need to be able to live with the constant closing at small losses in search for that major trend explosion that provides you with most of your profits. {image} {image}
Hi, where did you get the data service, and how much it costs you, it would help me too, I'm also testing various EAs compiled with this platform.
  • #19
  • May 26, 2020 3:46pm May 26, 2020 3:46pm
Quoting carddard
This is a 99.9% tick by tick backtest on M15 with this EA on ICM from 2011 Jan - 2020 May. Past experience with these plug and play models tells me that your chance at long term positive expectancy is higher if you use it to create divergent strategies as opposed to convergent ones. There are some good threads on this forum adopting this approach, but you need to be able to live with the constant closing at small losses in search for that major trend explosion that provides you with most of your profits. {image} {image}
Hello, thank you for your input.

When I uploaded the robot it has been optimized with 90% data quality only. I didn't have access to 99.9% data quality. Only recently I have bought the Tick Data Suite, and I am in the process of re-optimize all my robots with 99.9% data quality. But I do test the robots on Demo and keep records of their performances. Even with 90% data quality optimization I have some robot reaching 500% profits. So improvement is constant. I am trying to build a real portfolio of robots which can successfully increase an account balance month after month.
  • #20
  • May 26, 2020 3:48pm May 26, 2020 3:48pm
Quoting leleitaly
[quote = descaboy; 12963428] {quote} Ciao, nessun problema a dare la tua opinione amico mio. Ho un paio di robot per scambiare diverse coppie come Euro, Audusd, Nzdusd ... Ma recentemente sono stato più interessato a trovare un buon robot in grado di scambiare il cavo, una coppia che ha più portata e che si sta muovendo più del Euro. Ecco perché tutti i miei EA di studio sono per il cavo, ma puoi provare a ottimizzarlo per l'euro per vedere cosa puoi ottenere. Volatilmente l'euro è sceso in modo significativo, quindi per ora non sono troppo concentrato...

Yeah, I have 3 robots on my live account, they all have made money today . Market is finally moving after days of stagnation.
0 traders viewing now