Building your own EA G/08.

[Lukeye]

Re: Building your own EA G/08.

 

Hi,

 

here is my latest backtest.

My focus was on avoiding risky trades ( large drawdowns / long duration drawdowns).

 

It looks kind of good but is not finished yet.

 

h**p://depositfiles.com/de/files/zb49qqxg1

[asgard2]

Re: Building your own EA G/08.

 

Pretty Good Mate.. you are just starting with $80?

 

Anyone would be pleased to see that interest on their money.

[qwertyasdf999]

Re: Building your own EA G/08.

 

Hi,

 

can anyone translate this peace of code into plain english please ?

 

 // ïðîâåðêà HughLowLimit
  
  if(CountHighLowLimits > 0)
  {
     // óäàëÿåì âñå STOP îðäåðà, ðàñïîëîæåííûå â îïàñíûõ çîíàõ (BUY ñëèøêîì âûñîêî, SELL ñëèøêîì íèçêî)
     
     for(k = 0; k < CountHighLowLimits; k++)
     {
        switch(k)
        {
           case 0 : _iHL_LimitDistance = iHL_LimitDistance1;  break;
           case 1 : _iHL_LimitDistance = iHL_LimitDistance2;  break;
           default: continue;     
        }
        
        if((Counts[_OP_BUYSTOP] != 0) && (iHL_Low[k] > 0) && (MaxPrices[_OP_BUYSTOP] >= (iHL_Low[k] + _iHL_LimitDistance * _Point)))  
        {
           glDeleteAllDeferOrders(_OP_BUYSTOP, Symbol(), iHL_Low[k] + _iHL_LimitDistance * _Point, 0);
        }
        
        if((Counts[_OP_SELLSTOP] != 0) && (iHL_High[k] > 0) && (MinPrices[_OP_SELLSTOP] <= (iHL_High[k] - _iHL_LimitDistance * _Point))) 
        {
        if(PrintComments)  Print("CountHighlowLimit close.");
           glDeleteAllDeferOrders(_OP_SELLSTOP, Symbol(), 0, iHL_High[k] - _iHL_LimitDistance * _Point);
          
        }   
     }   
  }   

 

I ask because this one is the part which deletes stop orders the same minute they were opened.

 

 

This is translated comments(Google):

    // Check HughLowLimit 
  
   if (CountHighLowLimits> 0) 
   ( 
      // Remove all STOP warrants, located in hazardous areas (BUY too high, SELL too low) 

 

But, this piece of code is not the only part that deletes stop orders, you better check setting:MaxOrdersCount, if it is not 1, and if you already have an open buy order, any new buy stop orders will be deleted immediately, here is the code doing this:

   if((Counts[_OP_BUY] != 0) && (Counts[_OP_BUYSTOP] != 0))  
  {
     glDeleteAllDeferOrders(_OP_BUYSTOP, Symbol(), MinPrices[_OP_BUY], 0);
  } 

 

 

I think this is the reason, maybe I am worng because I just spend some hours to read the code, still have a lot to learn.

 

 

@asgard2

It is great to have you here to share your thoughts and experiences. :-bd

Very appreciated.

[Lukeye]

Re: Building your own EA G/08.

 

Cool qwerty,

 

how did you translate

óäàëÿåì âñå STOP îðäåðà, ðàñïîëîæåííûå â îïàñíûõ çîíàõ (BUY ñëèøêîì âûñîêî, SELL ñëèøêîì íèçêî

it with google ?

 

It does not work for me :o

[asgard2]

Re: Building your own EA G/08.

 

Who here knows how to get the time of the lowest lows for the last five days.

 

I am going to try to write this dynamic trade hour.. I am stumped with this problem but I am sure it is easy if you know how.

 

Edit

found the answer somewhere else.. for anyone that wants it..

 

int ix = ArrayMinimum(Low, (5*1440)/Period(), 0); // any chart

datetime minTime = Time[ix]; // date and time of min value
Reply With Quote 

[qwertyasdf999]

Re: Building your own EA G/08.

 

Cool qwerty,

 

how did you translate óäàëÿåì âñå STOP îðäåðà, ðàñïîëîæåííûå â îïàñíûõ çîíàõ (BUY ñëèøêîì âûñîêî, SELL ñëèøêîì íèçêî it with google ?

 

It does not work for me :o

 

 

change your .mq4 filename to *.txt, then you can use google to translate the file. :))

[Cyrillic]

Re: Building your own EA G/08.

 

I have been following this post for some time now, i just had a great idea about trend identification of trend. If we were to use http://www.indo-investasi.com/viewtopic.php?f=52&t=5470 range bars from pacman scalper it would filter out all the noise and give the ea a greater advantage in terms of picking the direction.

[qwertyasdf999]

Re: Building your own EA G/08.

 

I have been following this post for some time now, i just had a great idea about trend identification of trend. If we were to use http://www.indo-investasi.com/viewtopic.php?f=52&t=5470 range bars from pacman scalper it would filter out all the noise and give the ea a greater advantage in terms of picking the direction.

 

i looked at the files, however the files are ex4 format, I am not sure if we can use it. if it is in mq4 format, i know how to.

[Cyrillic]

Re: Building your own EA G/08.

 

well all you have to do is decompile it i don't have windows so decompiler doesn't work on my system but you are free to tackle it

hxxp://www.4sh4red.c0m/file/127523099/d4bae370/ex4_to_mq4_224.html?err=no-sess I'm not allowed to post this link a apologize to mods.

Thx...

[Cyrillic]

Re: Building your own EA G/08.

 

Hello Asgard 2, by dynamic stop loss do you mean the trailing stop, or is there a special formula for it. Also could you explain more the dojie principle that you use in your code. Would it be if movement in this bar + previous bar is more than 80 don't trade or is it more complicated?

[asgard2]

Re: Building your own EA G/08.

 

Hello Asgard 2, by dynamic stop loss do you mean the trailing stop, or is there a special formula for it. Also could you explain more the dojie principle that you use in your code. Would it be if movement in this bar + previous bar is more than 80 don't trade or is it more complicated?

the Dojie rule says..

 

If the market is a 10 day high and there is a dojie.. do not buy until there is not a four hour dojie. So two consequetive dojies will give you 8 hours of no trade while the market makes up its mind.

 

Dynamic stop loss works on reverse logic.. mine if far from perfect and yours may be better.

 

If the reverse logic says sell.. it is give a value

If the drawdown has gone past a certain percent.. it is given a value

If the market has dropped at a fast speed it is given a higher value.

If there is a bearish engulfing patten for the day and the trade is a buy.. its lights out.. this is the highest value and will work by itself.

 

If all the values add up to 50..it closes all trades. This is not simple to write and it takes hours of back testing to cover most bases where you would normally want to quit.

 

The value of the back testing is you already know what happened in hindsight.. so you can check every time it made a mistake and should have quit.

 

The trick to writing any of the restrictive code.. is to go with it, rather than against it. It is not natural..

 

For example.. market is dropping.

 

You could write.. dont buy if market has dropped 20 pips in the last 15 minutes.

Or you could write.. dont buy if the market has no dropped 20 pips in the last 15 minutes.

 

sometimes I think so hard about this that my head literally hurts. My eyes have gridlines embedded into my pupils. :-S

 

When you get it right.. there is no much that can top it for satisfaction.

 

The dojie rules .. I also have bearish engulfing for the 4 hour are there to try to pick the market turn. If the market is high.. you will catch 50% of the turns this way.

 

The day rules can only pick up after the market turns. They are good for your stoploss.

[asgard2]

Re: Building your own EA G/08.

 

I did some work on averages ATTR (Average Time To Recover)

 

After a 4hr bearish engulfing pattern at a 5 day high.. on average over the past 12 months.. it takes 20 hours for the market to turn back up and mean it or an average of 39 hours when there is a downtrend.

 

Now.. the average for this varies a lot from the longest to shortest time, mostly in the past couple of months when the market has been particularly volatile.

 

It does give rise to the thought that for saftey you might have a timer on some of your rules... its a bit of pot luck but can certainly shift the odds in your favor for survival. It would not be too hard to have a trigger break the timer.. I will work on this idea.

 

Market is volatile = false //you need this or it will never come out of this mode.

If market is volatile (Last month highs and lows differ from the median beyond a certain pip value.)

Use mode x (safe)

use timer = true (if trade is buy, wait a number of hours after a dojie or bearish engulfing pattern)

inertia protection (tradetime breakeven) is on

take profit is 300

IHlL_limitDistance is 700 (do not buy high)

use rules x y z is on.

 

For my mind, most EAs that can survive are really low on trades and profit

 

I have found none that follow trade rules you would set for yourself if you were trading manually.

 

If an EA is going to be truly successful it has to change when the market conditions change.

It cannot be pure indicator trading logic alone, it is going to need to think and the trading logic will need to be adjustable by the EA itself.

 

This is doable, it just takes a lot of work. You will never buy one, as good as the one you can create for yourself.

[Cyrillic]

Re: Building your own EA G/08.

 

This is very true, thank you very much asgard, I'm going to get to coding tonight I'll post a little surprise this week as a thank you for your support.

[asgard2]

Re: Building your own EA G/08.

 

A couple of days ago.. I decided to remove the greezly logic and export it into a brand new EA from scratch.

 

This is not easy by the way but I got sick of ploughing through code that I was unsure if it was affecting any of my changes.. anyway, these are my results so far. This is only over my base period so the results don't count long term but I need to work on some base period.

 

http://rapidshare.com/files/294941059/MyBetaModel_files.rar

 

Although the Maximum draw down is pretty good the relative draw down is to high..

 

I think, the first order should be 1% risk and the second order should be 2% risk. Given that the EA works on retracement, it is often the case the market entry is far better on the second order than the first.

 

The other thing I will do, is import the logic from another EA to catch the trend in the other direction. I already have one in mind which is Master Mind, I will let you know how it goes.

 

Edit:// managed to get the drawdown to a respectable level.

 

So for my two month test period here is the latest image with 112 wins.

 

http://img199.imageshack.us/img199/294/46356674.gif

[Cyrillic]

Re: Building your own EA G/08.

 

Interesting backtest, however i noticed that you only run it between may and july (in previous instances as well) you also use only buy no sell orders are placed. How does your ea favour on later or previous history tests if you don't mind me asking. Thank you.

[asgard2]

Re: Building your own EA G/08.

 

Interesting backtest, however i noticed that you only run it between may and july (in previous instances as well) you also use only buy no sell orders are placed. How does your ea favour on later or previous history tests if you don't mind me asking. Thank you.

 

OK..I cannot/have not been able to adjust the sell logic to work reliably.. the only time I got it going it was not as profitable as the buy only... "Yet!"

 

The reason I ran the backtest on that period is because it is a base period. It can get though it without any of my rules.

 

The idea behind it is that, if it can be extremely profitable at the base period (more is better), then I can introduce all my trading rules and not kill the profit to zero to get through the more difficult periods.

 

right now, the base is not the most profitable.. June July August is on the original ... but since my new EA, is only half working, I used the base period cause it will survive.

 

Actually.. the only reason (I was waiting for someone to pick it up) this EA looks so profitable is because.. it is opening way more orders than the two it is supposed to.

 

Obviously.. I missed something in the code.. I am fixing it now.

 

I actually posted it so people could see, that anything is possible.. and hopefully, they might have a go at learning some of this stuff.

 

How cool is it to be able to change the code at a whim, to suit no one else but yourself?

 

Anyway.. the main rule for myself is not to introduce any rule or code that is not genuine trading strategy if I was trading manually.

 

If anyone wants a copy of the extracted Greezly logic that I based my new EA on.. I am happy to post it. Sometimes I cannot work out, if anyone is interested in this topic so I don't want to fill it with code that people could care less about.

[Cyrillic]

Re: Building your own EA G/08.

 

I already have the old greezly :) If your logic is a bit different i would be happy to take a look, so far i have been checking with it to see if i can replicate and possibly improve profitable back tests. I find your code as well as this post interesting, the problem is most of the people here (or on any other forum) want a free ea that will make them a trillion dollars a week without putting any work into it. But i am sure that this topic also has a few silent readers that are improving their trading knowledge and possible they ea as we speak. So thank you for all your hard work and i hope you keep us posted.

[asgard2]

Re: Building your own EA G/08.

 

OK.. here is some code I wrote.. Not working as it is supposed to, so if anyone knows why.. I would be pleased to hear the solution.

 

This is a bearish engulfing pattern on a timer.. The timer works great and I was really happy with it..

 

What does not work great is the code that is supposed to kick in at a market high..

 

It kicks in anytime it feels fit and not necessarily at a market high..

 

In addition.. it sees each Monday zero bar as a bear bar when it may not be.. so I am stumped.

 

 

     extern int   EnglufSleepHours               =39;
     extern double EngulfingBarsize              =40; 

     bool   MarketHigh =        (lastOpen >= Highest5DayOpn);
     bool   Bar4Bear =          (iOpen(NULL,PERIOD_H4,1) >= iClose(NULL,PERIOD_H4,1));
     bool   Bar4Bull =          (iClose(NULL,PERIOD_H4,1)>= iOpen(NULL,PERIOD_H4,1));
     double Delta4HL_0 =        MathAbs(iClose(NULL,PERIOD_H4,1)-iOpen(NULL,PERIOD_H4,1));
     double Delta4HL_1 =        MathAbs(iOpen(NULL,PERIOD_H4,2)-iClose(NULL,PERIOD_H4,2));
     double Dif4ClsOpn =        ((iClose(NULL,PERIOD_H4,2)-0.0002)-iOpen(NULL,PERIOD_H4,1));



   if (Bar4Bear && iHigh(NULL,PERIOD_H4,1) > iHigh(NULL,PERIOD_H4,2)) int xz=EngulfingBarsize; 
     {
       static int EngulfTime; static bool Engulfed=false;
       if  ((MarketHigh) &&((Delta4HL_0 + (iHigh(NULL,PERIOD_H4,1)-iOpen(NULL,PERIOD_H4,1))) > 
       xz*Point) && (Delta4HL_0 >= Delta4HL_1) && (Dif4ClsOpn <= 0.0004)&& !Engulfed){             
       EngulfTime = iTime(NULL,PERIOD_H4,0); Engulfed = true;
       if (Engulfed) Print("Buy is Sleeping Util : " + TimeToStr(EngulfTime,TIME_DATE|TIME_MINUTES)); }   //Note the string needs fixing 
       if(((TimeCurrent()-EngulfTime)/60/60)<EnglufSleepHours && Engulfed) UseBuy=0; UseSell=1;
       if(((TimeCurrent()-EngulfTime)/60/60)>=EnglufSleepHours)Engulfed=false;
      }  
      
   if (!Engulfed) 
     {
      UseBuy=1; UseSell=0;
     }

 

Anyway.. if anyone feels like using it.. you will need to solve that problem and I would appreciate an update if you get it going. Else when I find out what I have done wrong.. I will post the solution.

[asgard2]

Re: Building your own EA G/08.

 

This is the New EA from May until today.. todays trades did not get a chance to close.

 

Yes, I know the draw down looks ugly.. since it is so new, I will just forgive it and fix it.

 

http://img12.imageshack.us/img12/8088/straster.gif

 

Initial investment of $5000 DD 53/68

[Cyrillic]

Re: Building your own EA G/08.

 

Looks interesting, going to to work a further look at it i am taking my time on this ea, i rushed my last 3 and now i get to kick back and relax a bit with this one :) I will look further into your code this week and try and help out.

[⭐ bbrain]

Re: Building your own EA G/08.

 

Hi,

 

Download MT5! You can debug the code in runtime.

 

Regards, bbr.

 

 

OK.. here is some code I wrote.. Not working as it is supposed to, so if anyone knows why.. I would be pleased to hear the solution.

 

This is a bearish engulfing pattern on a timer.. The timer works great and I was really happy with it..

 

What does not work great is the code that is supposed to kick in at a market high..

 

It kicks in anytime it feels fit and not necessarily at a market high..

 

In addition.. it sees each Monday zero bar as a bear bar when it may not be.. so I am stumped.

 

 

     extern int   EnglufSleepHours               =39;
     extern double EngulfingBarsize              =40; 

     bool   MarketHigh =        (lastOpen >= Highest5DayOpn);
     bool   Bar4Bear =          (iOpen(NULL,PERIOD_H4,1) >= iClose(NULL,PERIOD_H4,1));
     bool   Bar4Bull =          (iClose(NULL,PERIOD_H4,1)>= iOpen(NULL,PERIOD_H4,1));
     double Delta4HL_0 =        MathAbs(iClose(NULL,PERIOD_H4,1)-iOpen(NULL,PERIOD_H4,1));
     double Delta4HL_1 =        MathAbs(iOpen(NULL,PERIOD_H4,2)-iClose(NULL,PERIOD_H4,2));
     double Dif4ClsOpn =        ((iClose(NULL,PERIOD_H4,2)-0.0002)-iOpen(NULL,PERIOD_H4,1));



   if (Bar4Bear && iHigh(NULL,PERIOD_H4,1) > iHigh(NULL,PERIOD_H4,2)) int xz=EngulfingBarsize; 
     {
       static int EngulfTime; static bool Engulfed=false;
       if  ((MarketHigh) &&((Delta4HL_0 + (iHigh(NULL,PERIOD_H4,1)-iOpen(NULL,PERIOD_H4,1))) > 
       xz*Point) && (Delta4HL_0 >= Delta4HL_1) && (Dif4ClsOpn <= 0.0004)&& !Engulfed){             
       EngulfTime = iTime(NULL,PERIOD_H4,0); Engulfed = true;
       if (Engulfed) Print("Buy is Sleeping Util : " + TimeToStr(EngulfTime,TIME_DATE|TIME_MINUTES)); }   //Note the string needs fixing 
       if(((TimeCurrent()-EngulfTime)/60/60)<EnglufSleepHours && Engulfed) UseBuy=0; UseSell=1;
       if(((TimeCurrent()-EngulfTime)/60/60)>=EnglufSleepHours)Engulfed=false;
      }  
      
   if (!Engulfed) 
     {
      UseBuy=1; UseSell=0;
     }

 

Anyway.. if anyone feels like using it.. you will need to solve that problem and I would appreciate an update if you get it going. Else when I find out what I have done wrong.. I will post the solution.

[asgard2]

Re: Building your own EA G/08.

 

Ok.. this is where I ended up about 3 or 4 days ago with the greezly logic in a new EA with only the first three rules posted here.

 

I threw out everything in the logic except the high/low limits.

 

Since that time I have introduced my rules to lower drawdown with fair sucess. I have one rule to go and I will post the results.

 

The addition of the rules lowers the profit level to around 600,000 but also lowers the draw down to around 18% so far. I am trying to improve it and I have some ideas and I still need to tweak the rules after I have finished adding them.

 

I will post the results as soon as I get the drawdown to under 10% and the profit over 700. Hopefully that will be soon.

 

http://img16.imageshack.us/img16/1781/1000000d.gif

 

There are only six rules so far and 3 of them I have posted already.

 

Here are two more.. you might need to tweek them a bit cause I will. They are somewhat rough but very effective (and costly). If you have not managed to get though the last few months.. these will do it for you.

 

           if (BearDAY && MarketHigh) // Rule 4. If last day is a bearish Engulfing pattern do not buy for this day. Set Time of The pattern
           static int SetEventTime;   
           static bool EnglufEvent=false;  
           {
           if ((HiLowDelta_0 + (iHigh(NULL,PERIOD_D1,1)-iOpen(NULL,PERIOD_D1,1)) > 200*0.0001) && (HiLowDelta_0 >= HiLowDelta_1) && (DifBullClsBearOpn <= 0.002))
                {
                if (type == _OP_BUY) MayOpenDeferOrder = false; 
                 if (!EnglufEvent)   EnglufEvent=true; SetEventTime=Time[0]; 
                }
           }
           

          if (EnglufEvent)  //Rule 4a While the time since the last Bear Engulf is recent. Set trade hour to Midday. 
           {                 //Set buy price lower than normal. While each day low is lower than the previous by more than 30 points, don't buy.
           if(((TimeCurrent()-SetEventTime)/60/60)>= hours)EnglufEvent=false;
           if (EnglufEvent ==true) BuyDistance = 700; //if (EnglufEvent ==true && (Ask + BuyStopLevel) > iLow(NULL,PERIOD_D1,1)) MayOpenDeferOrder = false;
           if ((EnglufEvent ==true)&&(DayOfWeek()==1) && (Hour() <= 12) && MathAbs(iOpen(NULL,PERIOD_D1,0) - iClose(NULL,PERIOD_D1,1)) > 0) MayOpenDeferOrder = false;
           if ((EnglufEvent ==true) && (iHigh(NULL,PERIOD_D1,2)- iClose(NULL,PERIOD_D1,1) > 0.003)) MayOpenDeferOrder = false; 
           if (EnglufEvent ==false)BuyDistance = 1500; }
           
          // if (BearDAY && HiLowDelta_1 > 200 *Point) { 
          // if (HiLowDelta_0 < 35*Point && (DifBullClsBearOpn <= 0.002) && (iLow(NULL,PERIOD_D1,1)-iClose(NULL,PERIOD_D1,1) > HiLowDelta_0 *4)) BuyTrailingStop = 20 * Point} 
           
           if (type == _OP_BUY)           
           static int SetDojieTime;   // provide a "Declare Event Time"
           static bool DojieEvent=false; 
           double DayAngle= (iMA(NULL,0,1440,0,MODE_EMA,PRICE_CLOSE,1)-iMA(NULL,0,1440,0,MODE_EMA,PRICE_CLOSE,5));
             {
              if ((MarketHigh && (iHigh(NULL,PERIOD_D1,2) < iHigh(NULL,PERIOD_D1,1) && HiLowDelta_0 < 20 *0.0001))
             ||((DayAngle < 0 && BullDAY) && (iHigh(NULL,PERIOD_D1,2) < iHigh(NULL,PERIOD_D1,1) && HiLowDelta_0 < 20 *0.0001)))
              {
                MayOpenDeferOrder = false; //Prevent any buy orders for the day
                 if (!DojieEvent)   DojieEvent=true; SetDojieTime=Time[0]; }}
                 
          if (DojieEvent)  //Rule 5// If it is not false, 
           { 
           if(((TimeCurrent()-SetDojieTime)/60/60)>= hours)DojieEvent=false;
           if (DojieEvent ==true) BuyDistance = 700; //if (EnglufEvent ==true && (Ask + BuyStopLevel) > iLow(NULL,PERIOD_D1,1)) MayOpenDeferOrder = false;
           if ((DojieEvent ==true)&&(DayOfWeek()==1) && (Hour() <= 12) && MathAbs(iOpen(NULL,PERIOD_D1,0) - iClose(NULL,PERIOD_D1,1)) > 0) MayOpenDeferOrder = false;
           if ((DojieEvent ==true) && (iHigh(NULL,PERIOD_D1,2)- iClose(NULL,PERIOD_D1,1) > 0.003)) MayOpenDeferOrder = false;     
           if (DojieEvent ==false)BuyDistance = 1500; }// Print ("EnglufEvent is true");
           

 

If you have any questions on how these work.. just ask and I will be happy to answer.

 

A summary is that.. first rule is a "day bear engulfing pattern" if such an event is true, dont trade for that day and set a timer.

After that day and while the timer is active. Don't buy as high and change the tradehour to midday. If the market is still droping at more than 300 points per day don't buy.

When the timer expires set things back to normal

Second rule.. says the same thing except it applies for a dojie pattern.

 

Hope this is helpful.

 

Hopefully.. a dojie arrives before the bear... very often the case. If I was trading manually and I saw these patterns, I would wait... my EA should do the same.

 

  
     bool   BullDAY =           (iClose(NULL,PERIOD_D1,2)>= iOpen(NULL,PERIOD_D1,2));
     bool   BearDAY=            (iOpen(NULL,PERIOD_D1,1) >= iClose(NULL,PERIOD_D1,1));
     double Highest5DayOpn=     iOpen(NULL, 1440, iHighest(NULL, 1440, MODE_OPEN,5,1)); 
     double HiLowDelta_1=       MathAbs(iOpen(NULL,PERIOD_D1,2)-iClose(NULL,PERIOD_D1,2));
     double HiLowDelta_0=       MathAbs(iClose(NULL,PERIOD_D1,1)-iOpen(NULL,PERIOD_D1,1));
     bool   MarketHigh =        (lastOpen >= Highest5DayOpn);
  

[asgard2]

Re: Building your own EA G/08.

 

Ok this is where I ended up when I put in all the rules.. actually it has a loss of 80c in there somewhere on a spike, so only 190 wins but not too bad overall.

 

[center:1plw25c8]http://img14.imageshack.us/img14/5356/finalfj.gif[/center:1plw25c8]

 

Full backtest is here http://rapidshare.com/files/298104702/final.rar

 

Last post on this subject unless someone wants to know something..

 

Food for thought though.. those rules would be good for any EA that is halfway profitable to begin with.

[Cyrillic]

Re: Building your own EA G/08.

 

Wow very impressive, you have eliminated all the bad draw downs, this just shows that hard work can bring wonders. I have been very busy all last week so i didn't have any time what so ever to work on my ea's, however this is still all very interesting. Thank you very much for all your help asgard2 and i hope everything works out for you! Good luck!

[Rio]

Re: Building your own EA G/08.

 

I haven't been closely following this thread for a while, but it just goes to show that the only halfway decent EAs out there are ones you develop yourself.

 

Actually I've been working on an EA of my own for a while. It's based off the Steinitz Fractal Breakout indicator (except it doesn't rely on the indicators). It's hugely profitable... on any timeframe on any currency, except that I can't quite get the rules working right. I could use some help. Maybe I will borrow somew of the ideas introduced here.