Results 1 to 8 of 8
 5 Attachment(s)    

Thread: DragonFly EA

  1. #1
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default DragonFly EA

    Dear Mr. Funyoo,
    nice to meet you and thanks for your great help to people of this forum.

    Just few words to introduce myself: i'm not a coder, but I had an idea that seems working. I've studied Mql hardly, but is not my field . Sorry...

    But I've developed my idea and I've created this little mq4, named DragonFly, and, unbelivable, it works. And well, I think.

    Best setting I've tried in MT4 Backtesting?

    Gbp/Usd - 130 trades from July - 83.70% win. Timeframe 5min, but is useless, because I've included timeframe in the code.

    This EA, I think, is very flexible and could be used in each pair, almost, with right settings. Is not a scalping system, is not a trend system. Is a intraday system.

    I can work again on trading rules. I think it could be strongly improved.

    The problem is that I cannot develope a good exit strategy. I'm not a coder. I've tried, but is not easy assembling codes from various EA samples. I will study, I promise, but I will not be ready before next year...

    In the end, the system took 24 pips per trade, but can lose 350 pips in just 1 trade. I have to win 15 trades to cover the loss... (but this system don't need a large stoploss. If the entry is right, it goes immediately in the right direction. I think...)

    If You have time to help me to develope this little and absolutely free EA, let me know, please. I'll be gratefull.

    Best regards

    In the attachment, the strategy report.
    Attached Images Attached Images DragonFly EA-strategytester6-gif 
    Attached Files Attached Files

  2. #2
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Hi R3d and welcome,

    Where is this mq4 file ?

    Have you tried a hidden stop loss with a reverse option ?

  3. #3
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default

    Thanks for the reply, funyoo.

    I've just followed the rules

    First step, statement. Now, if you have time, the idea.

    I've tested it with some more strong parameter. It wins 77.5% and close in positive the backtest, but the dd is sometimes very hard.

    Philosophy is simple: chart is ever retracing the MA. With some filter, we can choose to buy when price is below the MA and sell when the price is above the MA. But the hard part is when to buy or sell. I've tried to find this moment, and my work is not finished.

    Best regards, and thank you for your help.
    Attached Images Attached Images DragonFly EA-gpbusd-gif 
    Attached Files Attached Files

  4. #4
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default

    PS: i've included every explanation in the mq4 file

  5. #5
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by R3d View Post
    PS: i've included every explanation in the mq4 file
    I suggest you to do a demo detailed statement. And to specify then the rules as I have not the time to review your code and the explanations included in it.

  6. #6
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default

    Ok, mr funyoo, right. Sorry, I know you've not much time. My fault.

    I've used the following indicators:

    Stochastic 5, 15, 30 min
    DeMarker 5, 15, 30 min
    CCI 5, 15, 30 min

    (I will try with MFI and LaGuerre too)

    All the indicators I want to use (and try) have levels (20/80 or -100/100 etc...)

    I think we are very closed to a change of direction when indicators are very high or low, but not at the same level in timeframe (CCI 5min > 100; CCI 15 min >80; CCI 30 min> 60, i.e.)

    I set a MA that cut the chart (34 to 55);
    I set 2 MA's Levels (+/-12)

    (Next time I want to try with envelopes...)

    Trading rules consider just bars under MA down level for a bullish signal and over for a bearish signal.

    When we have a bar enough high or low, above or below the MA level and a coincidence between all the indicators in each timeframe, I wait for a little change in Stochastic 5 min (Sto[0]>or<Sto[1]) and an Open in current bar > or < low of previous bar to buy/sell.

    The exit strategy is based only on Trailing Profit and opposite signal. For me, that part of code is not easy to understand. I've tried to add a stoploss, but I've received a lot of errors. Don't know why.

    In a sideway market it works fine. More problems with a hard trend, because the system receive some false. I think I will filter this problem maybe with a second level of MA, putting a new rule: do not trade below/above this level. Or maybe with a fast MA. Don't know, I have to try.

    Most of lost money depend from this (24 oct. 2008, 6.30 - 12 GTM, i.e.).

    I'm trying to assembley the right amount of indicators, I'm testing in demo this system (no trade, in two days. Rules are maybe a little restrictive), I'm working on trading rules, but I cannot write complex code, only copy and modify some parts from samples, in the hope it works.

    By default, the Dragonfly EA have good settings:

    TP: 24 (but maybe is useless with those exit rules. I've tried with 100 and works fine)

    TS: 20

    MA period: I suggest 34, 46 or 55, but you could try with 16 or 21, following the chart fastly.

    MALevel1: is the up/down level of MA for signal.

    I don't know if is a good or bad system. For now, backtesting on MT4 provided by ODL give nice results (70 - 83.7% win) that suggest me to work on exit strategy and to change some entry rule. A lot of errors in backtest on MT4 provided by Alpari. Maybe the code in not so good

    I will test again and again and return to upload statements.

    Thanks for your time.

  7. #7
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default

    Problems: backtest is drastically changed. I've included MM and StopLoss. I've to work a lot around this EA. I'll be back soon. Good profit win, but a bad DD.

    Thanks again, Mr Funyoo. I'll be back soon.

  8. #8
    Junior Member R3d's Avatar
    Join Date
    Jan 2009
    Posts
    6

    Default Update

    Hi, traders. Hi Mr. Funyoo.

    Just an update.

    I've found a little error in code.

    Correct the error (in DeMarker trading rules);
    Added Money Management;
    Added StopLoss;
    Modified the exit strategy;

    I have some good result, with a very large win ratio (83 - 87%) on GpbUsd and USDJPY.

    Entry strategy is enough good.

    I've to work on exit strategy. Is very poor. The drawdown is very bad on GpbUsd.

    Please: DO NOT USE THIS EA, FOR SURE NOT ON EUR. THERE IS A WORK IN PROGRESS TO IMPROVE IT.

    Attachment: strategy tests (English this time, not Italian)
    Attached Files Attached Files

Tags for this Thread

100, alpari, backtest, backtesting, change, color, copy, demo, eur, free, free ea, funyoo, high, indicators, intraday, intraday system, mq4, mql, mt4, pips, profit, report, review, scalping, signal, simple, stop, stoploss, strategy, study, system, thank you, time, trading, trailing, trend, usd

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •