Semi automatic EA (Driven Version) | Page 26 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

EA tool & strategy for semi-automatic interactive trading 82 replies

semi-automatic fibs in MT4? 0 replies

Testing a semi-automatic EA with a strategy tester 1 reply

Geoff's Manual and Semi Automatic Trading Setup 988 replies

Trading a (semi) automatic approach...but want more 11 replies

Attachments: Semi automatic EA (Driven Version)
Exit Attachments

Semi automatic EA (Driven Version)

  • #501
  • Mar 7, 2009 3:48pm Mar 7, 2009 3:48pm
Interesting Divyendu perhaps Xchaparro would be so kind as to help us with that? I would love to start testing on a Demo. Thanks for the input.

J
  • #502
  • Edited 4:42pm Mar 7, 2009 4:04pm | Edited 4:42pm
For the guys trying to automate this Ea. I wanted to try and contribute something that i personally use in my trading also to determine reversal of daily trend that seems to be very exact for me on trending pairs. It is based on price itself instead of cross overs,and I personally believe it has something to do with the fibonaci mathematical phemona. One of the main numbers in fibs is .5 or 5.. the pros generally like a .5 retracemnt. There seems to be something very profound about that number in the universe.

My obervation doesnt really have to do with fibs but i feel it is tied in somehow. i am pretty new to to forex so im no pro, just making observations.

Instead of a MA crossover or at least in combination with an ea crossover, if you plot a 5 period ema on any chart, doesnt matter the tf. If price CLOSES above the 5 period ema then trend is up. If price CLOSES below the 5 period ema then trend is down and usually continues to do so. It works on any time frame any pair and is great sign to exit a trade or keep an eye on opening a new one. Now it is not 100% perfect as nothing is but on daily charts is much more reliable and could help identify trend reversal if ti could be coded in with the ma crosses and such.

sure hope it may help
bulls

ps real live use would be to look @ eu right now on daily. plot the 5 period ema. you would see that price HAS closeD above 5 period ema and we could be looking @ a temp revesal. we got price action AND close above 5 ema and i like trix,,,is also showing green, bounce off lower end of channel. all good signs, but eur is weak so how much up we'll see.
Attached Image (click to enlarge)
Click to Enlarge Name: 5 ema close.gif Size: 23 KB
Whats happening in EUR/USD should be happening to you!
  • #503
  • Mar 7, 2009 9:46pm Mar 7, 2009 9:46pm
Quoting BullsAndBear
I read every page on this thread and ................................................. I got confusing to the middle of this thread. so much tal;k of an .................. I was very tempted to organize ALL of it on a pdf to help out but quite honestly i am so tired from reading this whole thing. ............bulls
In generall there is no siginificant update here on this thread, in the middle of the thread I tried to filter the trade by only trading in a high posibility only (when the HA and Arrow is in line), but still need a tweak. Some of our friends here still trying to automated the EA but I havent found some significant improvement.
Quoting BullsAndBear
....... i will run this live with small lots next couple months and be back to post results.bulls
Make sure U get the feel first spescially in difficult times, and deposit enough balance.

Regards,
Samurai Power
  • #504
  • Mar 7, 2009 10:02pm Mar 7, 2009 10:02pm
The only major problem is to recognize the daily trend, when its done then the case is solve.
I think the key is Price action, fibo and trend line on 4H chart, coz I think we can see where the daily chart goes by looking at 4hour chart behaviour, and we would get an early alert when the price tries to change the trend. But any suggestion is highly appreciated. Thanks

Regards,
Samurai Power
  • #505
  • Mar 7, 2009 10:22pm Mar 7, 2009 10:22pm
Hi again guys,

I found a great tools at some other thread here, something that can really helps us to select the indis that might work with this system, i encourage u get some experience with this.

Quoting Quantem
Here is a great tool to help you get experince faster, by trading on a simulated market. You can enter and exit trades like you would in a real market, so it gives you a more realistic feeling.

You can speed it up, and slow it down as well. I have been able to practice trade a month in about 1 hour using this tool.

The system runs in MetaTrader 4

When learning to trade, there's no substitute for experience. A trader who has spent many months trading a live market will have a feel for the market that is almost impossible to explain.

They...
http://www.forexfactory.com/showthre...69#post1856969

Regards,
Samurai Power
Attached File(s)
File Type: zip Trade Simulator.zip   491 KB | 2,149 downloads
  • #506
  • Mar 8, 2009 4:03am Mar 8, 2009 4:03am
Yes, this really is a great tool, I aplied it yesterday and you can run virtual market in the past, over the weekend when market is closed.
  • #507
  • Mar 8, 2009 4:49am Mar 8, 2009 4:49am
Quoting gasgas
Yes, this really is a great tool, I aplied it yesterday and you can run virtual market in the past, over the weekend when market is closed.
yes i knew this tools about a year ago, but not sure what is it for, i thought its just the same as strategy tester, now we can test indis that might work with this system in just a minute. Any other indis suggestion will be great.

Regards,
Samurai Power
  • #508
  • Mar 8, 2009 10:32am Mar 8, 2009 10:32am
Hi xchaparro, Fredella, and others

I did some chart studies (see post 500), and here are some filter settings that would cut down the risk for bad trades quite a lot. Of course some good trades would be affected, but the overall winning statistics I feel could be improved if something like this could be implemented:

Stoch + ADX filter: apply on [Firebird v63G trader seven]

Indicators on 4hour time base
Stoch 15, 9, 9
ADX 28

1) ADX falling and Stoch within 30/70 = all trades accepted

2) ADX falling and Stoch over 70 or below 30= no positions entered in the direction of oversold / overbought

3) ADX rising and Stoch within 20/80= no position entered against Stoch direction of travel

4) ADX rising and Stoch beyond 20/80 = no trade

The specific settings of what would constitute overbought/oversold could be made external to allow calibration.

I will try to think of the code myself, but I'm afraid I'm terribly amateurish, most likely to introduce bugs as big as your fist...

D
  • #509
  • Mar 8, 2009 10:47am Mar 8, 2009 10:47am
Gasgas fren,

This is what u might need and u might find it very usefull.
Just install it on your PDA, control your chart on your VPS, and now your trading is in mobile whereever u are.

Regards,
Samurai Power
Attached File(s)
File Type: zip RDC Mobile.zip   298 KB | 618 downloads
  • #510
  • Mar 8, 2009 10:53am Mar 8, 2009 10:53am
Quoting makron
Can anybody add to code of this EA price stoploss (for example 1.4060). I think its very useful
I think the EA has the stoploss in it, just check it out once again.

Regards,
Samurai Power
  • #511
  • Mar 8, 2009 11:00am Mar 8, 2009 11:00am
Divyendu,

any ideas are most welcome here, hope xchapparo or Fridella sees that, and try to modified that in ea. I will set it up, and run some tests, I'm playing with other indis.,but most of them repaints or lag....

Samurai thanks, I find that most useful, seems like you, I to need constant observation whenever I am

Regards
  • #512
  • Mar 8, 2009 11:02am Mar 8, 2009 11:02am
Quoting Divyendu
Hi xchaparro, Fredella, and others

I did some chart studies (see post 500), and here are some filter settings that would cut down the risk for bad trades quite a lot. Of course some good trades would be affected, but the overall winning statistics I feel could be improved if something like this could be implemented:

Stoch + ADX filter: apply on [Firebird v63G trader seven]

Indicators on 4hour time base
Stoch 15, 9, 9
ADX 28
.....
D
Nice analisys, i'll look in to it
  • #513
  • Mar 8, 2009 11:07am Mar 8, 2009 11:07am
-Semi auto-

No trading for tomorrow since its monday and no good signal on my chart.


Regards,
Samurai Power
Attached Image (click to enlarge)
Click to Enlarge Name: chart.gif Size: 19 KB
  • #514
  • Mar 8, 2009 7:12pm Mar 8, 2009 7:12pm
Hi guys

Changed tradelogic in line with my earlier post to include filtering with the help of Stoch and ADX. Backtesting over a three week period tend to increase a 10000 account by 500 or so.

took away some unused functions

Feel free to test, tweak and comment

D
Attached File(s)
File Type: mq4 Firebird_v63SF_TraderSeven.mq4   19 KB | 528 downloads
  • #515
  • Mar 9, 2009 6:52am Mar 9, 2009 6:52am
Quoting Divyendu
Hi guys

Changed tradelogic in line with my earlier post to include filtering with the help of Stoch and ADX. Backtesting over a three week period tend to increase a 10000 account by 500 or so.

took away some unused functions

Feel free to test, tweak and comment

D
thanks for helping out man, will test it today.
Is this ment to be fully automatic, or is still needed control for set up Short or Long??

I still use original trader seven EA, but today I notice, price went over 80pips from first short, and it didn't open that trades (pip step =40). It was on Usd/Jpy. Someone notice same problems in testing?? Some bug in EA mabey?

Regards
  • #516
  • Mar 9, 2009 8:15am Mar 9, 2009 8:15am
Hi Gasgas

My thought is that the filter will filter out bad trades, so there should be no need to manually block long or short positions.
It must be stressed that this is not yet tested and is certain not to work yet in many cases - so no live trading with this one yet!!

The target for developing a fully automated EA should be to put the odds more and more in ones favor - for the filters to kill off more bad trades than good ones (in pip value), and to lessen the likelihood of big drawdowns (larger positions can then be entered without too much risking blowing ones account).

I have one more idea for putting the odds in our favor:
Bad trades need to travel all the way to the stop loss which is further away than the take profit level. Therefore bad trades are normally longer lasting than good ones. Lets say for the sake of argument that most of the good trades in a particular currency cross last no more than 100 bars and most of the bad trades last more than 150 bars. Then it would be possible to put a trade-kill function at 125 bars.
It would kill a long position as soon as the Stoch cycle reaches a peak and a short position as the Stoch cycle reaches a valley.
In this way a loosing trade could be less costly by many many pips.

This type of function is beyond my comfort level to code, so please, anybody with the skills.. We need your help!

D
  • #517
  • Mar 9, 2009 10:34am Mar 9, 2009 10:34am
Hello Divyendu,

Your EA works fine but on some pairs it doesn't trade at all for the last three months. Maybe you reduce your filtering a bit more :-)
  • #518
  • Mar 9, 2009 11:06am Mar 9, 2009 11:06am
Hi ssvl

Which are the pairs that haven't given trades? I'd like to check it out
Thanks
D
  • #519
  • Mar 9, 2009 1:30pm Mar 9, 2009 1:30pm
Hey Divy

would you be able to attach the indicators by which you expanded the EA


your idea looks great, but i want to see the indies on the chart

and many thanks Samurai for the main idea

thx
  • #520
  • Mar 9, 2009 1:46pm Mar 9, 2009 1:46pm
Divy

I made a backtest onto the last 3 weeks /EURUSD 5 min chart/

the results looks great

u did it , i hope


here u are
Attached Image (click to enlarge)
Click to Enlarge Name: StrategyTester.gif Size: 8 KB
0 traders viewing now