Profitable EA? | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

New profitable System - Sidus 1,849 replies

Profitable strategies in range trading conditions 22 replies

How long the learning time to breakeven/be profitable? 8 replies

Neural Programming for Profitable trading 22 replies

PivotMagic v4 EA: Something new, could prove profitable! 0 replies

Attachments: Profitable EA?
Exit Attachments

Profitable EA?

  • #1
  • First Post: Jun 19, 2015 5:12am Jun 19, 2015 5:12am
Hi guys,

I need your help. I wrote simple (is important) EA and I wonder is it worth to develope it more.
I have tested this EA for period 2011.01.01-now. EA has constant SL and positions opened by buystop/sellstop. Only for H4, EURUSD (but it works also with other currencies).
Just ~500 transactions.
I attached charts and data from tester. First two attachements present EA without any Money Management.
Another two pics presents this EA with MM added.

Plz give me advice, what number of positive transactions (period) you think is essentiall to think about going live ?
What other factors I should test/check to confirm it could be profitable on-live.

without MM (only trailing SL)
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-17 10.23.16 am.png Size: 39 KB

Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-17 10.25.17 am.png Size: 45 KB


with MM and trailing SL
Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-17 10.21.16 am.png Size: 46 KB

Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-17 10.29.40 am.png Size: 49 KB


thx
PP
  • #2
  • Jun 19, 2015 9:55pm Jun 19, 2015 9:55pm
Numerous things to consider before going live my friend here are a few:

Slippage, if it is a scalper or does not have a news filter it can be affected by Slippage easily. All EA's should use a News Filter imo, unless it's an EA based around trading the News.
Delay, Delays in execution might be an issue
Spread, try Increasing the Spread significantly and see the impact on the EA

Did you test it with reliable Tick Data?
Test for a Longer period of time. How long is the longest period of stagnation, how about period of losing trades?
Can you expect it to lose a few months in a row and then come out on top eventually? Are you willing to wait/trust your EA's logic enough to wait during such a time and not turn off the EA?

Ask yourself the following: What logic makes your EA superior to other methods? What's it's Edge, and what makes you believe it's Edge must exist in the future?
A Positive Mind, and Positive Spirit lead to True Happiness
  • #3
  • Jun 19, 2015 11:23pm Jun 19, 2015 11:23pm
Yes this is just tiny tiny picture of trading. You not really giving us a lot to work with here, is it a momentum based system, breakout, trend, counter trend, price action, night trader? ? IF the TP and SL are less than 40 pips your live results may differ quite a bit due to slippage and spread factor. What does the test look like when you double the spread tested ?
All posts are my personal opinion
  • #4
  • Jun 20, 2015 7:15am Jun 20, 2015 7:15am
Thanks a lot Rparm, I have to re-think all of the tips u delivered.
EA has TP=170 and SL=150 for H4 period so I think I can eliminate only spread role.
Slippage and Delays play a big role i think.
Also, I don't know I have reliable Tick Data.
Thanks, I'm going to make tests for slippage and delays.

Rtm,
Thx for your response, presented EA is a mix a momentum and breakout system. It opens buystop/sellstop on specific moment on candle and wait for momentum on that bar.
Than it trails the stop.
If momentum doeasn't appear it delete sell/buy stops and waits for another bar.
I told u it is a very simple ea, so i wonder it could be good start to develop.
  • #5
  • Jun 20, 2015 4:37pm Jun 20, 2015 4:37pm
Quoting pthread
Thanks a lot Rparm, I have to re-think all of the tips u delivered. EA has TP=170 and SL=150 for H4 period so I think I can eliminate only spread role. Slippage and Delays play a big role i think. Also, I don't know I have reliable Tick Data. Thanks, I'm going to make tests for slippage and delays. Rtm, Thx for your response, presented EA is a mix a momentum and breakout system. It opens buystop/sellstop on specific moment on candle and wait for momentum on that bar. Than it trails the stop. If momentum doeasn't appear it delete sell/buy stops and...
Glad to see some people writing EAs that use decent time frames and TP/SL levels. A 10 pip TP on a 5 minute chart is no way to trade imo.
  • #6
  • Jun 20, 2015 6:22pm Jun 20, 2015 6:22pm
pthread,

Can you try this EA on USDJPY and let us know.
  • #7
  • Jun 20, 2015 6:58pm Jun 20, 2015 6:58pm
Most people test their EAs with historical data. That won't work; you're running it right now, not five years ago.

Perform an optimization of parameters using data from one to three months old, and make adjustments every couple of weeks.
This will help you to match your EA to the actual market conditions.

Hope this help to change your approach.
Kind regards,
MF.
Trade, don't bet.
  • #8
  • Jun 21, 2015 2:08am Jun 21, 2015 2:08am
Quoting MarketFox
Perform an optimization of parameters using data from one to three months old, and make adjustments every couple of weeks.
Spot on, again.
I go back around 6 months, not more.
  • #9
  • Edited 3:28pm Jun 21, 2015 11:34am | Edited 3:28pm
Hi,
Thanks for all responses, I made some test with USD/JPY.
Without any changes EA works fine only for GU.
UJ doesn't work becouse of the night hours are not tradeable (with EU settings). So, I changed it a little by including night hours and optimizing params.
I present only the finest resuts, I used whole history from my MT4 candles (have only H4 data from 2010).

Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-21 06.11.13 pm.png Size: 39 KB

Attached Image (click to enlarge)
Click to Enlarge Name: 2015-06-21 06.11.30 pm.png Size: 50 KB


I chose this results becouse of the number of transactions in test: 1555.
Other settings have better R/R (event 90/10) but it has low number of transactions.
I've tried to optimize it every month, 3m, 6m but for now it doesn't make sense.
Maybe I can filter more, if u have any clues for that plz give an answer.
Thx
P

ps. second screen has baloon help with canadian dollar but it is only my mistake during taking screen, chart is for USD/JPY
  • #10
  • Jun 22, 2015 3:39am Jun 22, 2015 3:39am
Quoting pthread
Hi, Thanks for all responses, I made some test with USD/JPY. Without any changes EA works fine only for GU. UJ doesn't work becouse of the night hours are not tradeable (with EU settings). So, I changed it a little by including night hours and optimizing params. I present only the finest resuts, I used whole history from my MT4 candles (have only H4 data from 2010). {image} {image} I chose this results becouse of the number of transactions in test: 1555. Other settings have better R/R (event 90/10) but it has low number of transactions. I've tried...
I don't mean to discourage you. But your EA is not consistent. I don't deny the profitability of EAs, I have coded dozens and so far, there is only one EA which I find profitable, reliable and consistent. However, manual trading is better (than the profitable EA which I created).

If an EA requires too many optimisation, then it probably will not work. If it only works for a certain period of time, let's say one year (from 10 years of data), then it probably won't work the year after.

Your EA should not be based on curve fitting, it should be based on an edge which you have discovered, not based on curve fitting of indicators.
  • #11
  • Jul 18, 2018 1:43am Jul 18, 2018 1:43am
Try this on M15

Minimum balance: $50 to $100
Attached File(s)
File Type: ex4 LionForce EA.ex4   56 KB | 691 downloads
  • #12
  • Jul 23, 2018 2:29am Jul 23, 2018 2:29am
Hi @pthread, would you share your EA? i want to test it on real account. Thanks.
  • #14
  • Oct 18, 2018 3:17am Oct 18, 2018 3:17am
Quoting rod777
Hidden

this one don't open any trade
  • #15
  • Oct 18, 2018 3:23am Oct 18, 2018 3:23am
you have to download the history data and activate dll imports
  • #16
  • Oct 18, 2018 3:33am Oct 18, 2018 3:33am
Quoting rod777
you have to download the history data and activate dll imports
on the tester work but when added to the chart for live trading don't work.
  • #17
  • Oct 18, 2018 6:05am Oct 18, 2018 6:05am
GBPUSD Only
Attached File(s)
File Type: ex4 Guufuj EA.ex4   13 KB | 698 downloads
  • #18
  • Nov 27, 2018 9:59am Nov 27, 2018 9:59am
Quoting Guufuj
GBPUSD Only {file}
Nice1
Attached File(s)
File Type: ex4 ReCharged_SMG EA.ex4   56 KB | 456 downloads
Gig-B
  • #19
  • Feb 13, 2019 1:17pm Feb 13, 2019 1:17pm
Hi guys,

I am back after more then four years
I just open the MT, set the strategy and pressed 'Start'
Parameters are exactly the same as 4 years before, without optimalization. H4 150/170.
This time I extended the test period to 2011-2019 (four new working without optimization).

Attached Image (click to enlarge)
Click to Enlarge Name: Zrzut ekranu 2019-02-13 o 20.08.37.png Size: 52 KB


Attached Image (click to enlarge)
Click to Enlarge Name: Zrzut ekranu 2019-02-13 o 20.08.05.png Size: 54 KB


One thing I noticed that many error showed in the report, I've been using other trading software so the data maybe corrupted since I left MT4. Have no idea.
I have no comments.. yet, couldn't believe that this strategy would not lose all money.
It was optimiized 4 years ago for the period 2011-2015.

Publishing it rather for fun then to have any suggestion. Cheers.
Peter
  • #20
  • Feb 13, 2019 3:36pm Feb 13, 2019 3:36pm
If your trading analisys doesnt have a TRUE advatange coded in the expert then you will never going to have consistent results (every trade will have a random outcome) period..

It doesnt matter if you are testing 3 million years of data to find the best settings for the indicators you chose to use, you are just curve fitting the results to the past.. and thats the definition of "hunting" for the magic combo of indicator/setting that "could" end profitable..

See how you, right from the start, dont have a solid/consistent advantage that can give you a high probability trade consistently over time ??


BTW im not an asshole or something like that.. Im just giving you a dose of reality
0 traders viewing now