1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Why EA's don't work - except maybe...

Discussion in 'General Forex Talk' started by WaveRider, Oct 13, 2011.

  1. WaveRider

    WaveRider Sergeant

    Joined:
    Mar 7, 2011
    Messages:
    304
    Likes Received:
    15
    The ... is the real attention grabber. EA's can work, can't they? We are greedy people. I believe everything I say but I would like anyone to challenge my reasoning.

    Theoretically, I might make a billion dollars this week by doing nothing but letting software trade for me. In theory that is true. I would gladly pay this guy $99 to find out if that is possible. Because I am an intelligent consumer I will read reviews (that are on his site) and choose the best one. Well his didn't work but I will try this other guy. Think of it, isn't it worth paying $100 for a billion dollar pay off? That's a good investment.

    The lapse in logic about EA's is if you or I made an EA that worked, neither you nor I would sell it. End of story. We could stop there since that is an insurmountable fact.

    Selling an EA is a profit making venture that requires a team of programmers, web developers, a marketing strategy and customer service. That's like owning, then selling the golden goose but being happy to be the goose groomer for a few bucks a month. That's bad business. Unless the goose doesn't lay golden eggs after all. Then selling a goose that theoretically might lay a golden egg is a good deal because I can be the goose groomer for a few bucks a month. Now that's good business. Sell a dream. Hey, you never know.

    What about strategies that trade near certainties like the Sunday gap or that trade the weekly pivot point? Sive Morten says that's a 70-80% trade certainty. Cachinga! But if the gap is 60 pips, the market can easily go 200 pips the wrong way before it comes back to close the gap. When do you bail out for the inevitable 20-30% that won't go your way? That math will destroy your account. I mean that math does in fact destroy people's accounts. And ther are EA's that trade that way destroying peoples' accounts as we speak.

    Can an EA work? We might think of an EA as a tool, like a hammer or a shovel. It works when a person makes it work. It is an advisor. It is not a decision maker. From what I have read and experienced, there are no leave it alone EA's. Sorry to burst that bubble, for all the would be suckers. All EA's have to be tinkered with, directed, turned on and off and a real person with a real brain is the actual decision maker. Not optimized. Tinkered with and directed. Because of that, it is insane to use an EA that you don't understand the logic of very intimately. That's really any commercial EA. Spend the $99 on lottery tickets. At least you'll know a lot faster that you didn't win.

    EA's that are most likely to work? One's you develop yourself or have intimate knowledge of the strategy. It would have to be based on a strategy that you know works. If you knew that, you would already be a pretty good trader and not need an EA anyway. Anyone who uses an EA should know enough source code to understand what it's doing and how to tinker with it. Still it will never be a leave alone EA.

    Here's an example. Because bankers and other people are on vacation in December, the market is really sleepy, then everyone comes to work in January and wants to trade their brains out. January usually is a great trending month. Your December strategy would get killed in January and your January strategy would get killed in December. An EA won't know this. A well trending EA on in January will likely do very well, but only in January. And not every January trends as well as others. Again a person would have to make that call.

    Okay. Pick me apart. Tell me why I'm wrong. Or agree.
     
  2. Pharaoh

    Pharaoh Colonel

    Joined:
    Oct 3, 2007
    Messages:
    19,685
    Likes Received:
    2,230
    Your December/January market comparison shows both the biggest flaw and the biggest opportunity in EA development.

    I'm not a programmer, but I can't figure out why EA developers don't build in a simple "market state detector" based on one or two indicators. If a strategy is designed for trending markets, there are indicators that show the strength of a trend. How hard could it be to not open trades when a trend indicator falls below a pre-specified level?

    The optimal solution would be to have at least 2 strategies built into the EA. One for trending markets and one for ranging markets. If the trend indicator shows a solid trend, use trending strategy. If it shows ranging conditions, use the ranging strategy. If it's somewhere in-between, open no new trades.

    I'd also like to see some sort of volatility indicator. This would be used to prevent new trades from opening if volatility is too low or too high.

    My suspicion is that most EA purchasers are looking to see the EA get started trading right away. A cautiously written EA might skip trading for quite some time. That wouldn't be good for sales.
     
  3. Johnfa

    Johnfa Private, 1st Class

    Joined:
    Jun 16, 2008
    Messages:
    60
    Likes Received:
    0
    EA NEVER WORK! That's the fact!

    WHY?

    If it works, the winners of the Automated Trading Championship would be on the top 10 richest people in the world!:)

    If it works, every traders would be a billionaire!;)

    Hope that help!
     
  4. ondrej

    ondrej Recruit

    Joined:
    Oct 11, 2011
    Messages:
    1
    Likes Received:
    0
    Automated systems can work, just not the once you can buy on the web!
    Anyone who buys an ea online and hopes to make a living is delusional and should wake up before they lose all their capital.

    These 2 books might help you with the matter:
    Amazon.com: Evidence-Based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading Signals (9780470008744): David Aronson: Books
    Amazon.com: Quantitative Trading: How to Build Your Own Algorithmic Trading Business (Wiley Trading) (9780470284889): Ernie Chan: Books
     
  5. PipStar

    PipStar Sergeant

    Joined:
    Oct 14, 2007
    Messages:
    223
    Likes Received:
    5
    EA's don't work. Don't waste any more time on it.
     
  6. RahmanSL

    RahmanSL Major

    Joined:
    Jan 16, 2010
    Messages:
    2,526
    Likes Received:
    218
    Legend has it that there is a genuine honest-to-goodness out-of-this-world (and smacking licking good) EA left behind by some Space Aliens.

    Many believe it's hidden somewhere deep in a top secret Indiana Jones location inside one of the great Pyramids of Giza, where even the Son-of-Ra'a have no knowledge of (and hence that accounts for you-know-who becoming a Forex Trader in search of lost knowledge).

    All other EAs are based on technological fragments from that original honest-to-goodness Alien EA and hence just doesn't work.

    Hope this helps to shed some light and impart enlightenment on this EA thing.

    Peace, and may the force of the market be with you all....and have a happy stress free weekend.

    P/S: By the way, the real honest-to-goodness meaning of "EA" stands for "Electronic Alien"....this lend further credence to the truth of that legend.
     
  7. Forexwatchman

    Forexwatchman Sergeant

    Joined:
    Sep 17, 2009
    Messages:
    198
    Likes Received:
    0
    Well put WaveRider, there is no such thing as a Golden Goose being sold for $99.

    It really is an obvious fact, but greed and desperation overrule most newbies judgement. What cracks me up the most is how once you purchase the EA, you also receive access to some exclusive forum area where everyone gathers around and discusses all the ways to play around with the EA and actually make it work right. How these people don't realize it's far more rewarding and efficient to learn simple technical analysis than to play with a program is beyond me.
     
  8. f-man

    f-man 4Xangels Representative

    Joined:
    Dec 5, 2010
    Messages:
    762
    Likes Received:
    14
    I agree that commercial EAs are not profitable on the long run, but I do not agree that EAs do not work at all.

    If a good trader interferes and corrects the trading of a good EA, it can be profitable. It just saves time for the trader and reacts without the psychological obstacle of the human mind.


    I agree 100% with your title "EA's don't work, except maybe..."
     
  9. WaveRider

    WaveRider Sergeant

    Joined:
    Mar 7, 2011
    Messages:
    304
    Likes Received:
    15
    EA's

    "Seriously, ours really does work, mooooahahah!!! Thanks for the $99."

    I probably should say that commercial ea's don't ever work, but one you make yourself probably won't. "Probably doesn't work" is still a step better than "can't work." Big institutions use automatic trading all the time. They also have a team of PhD's in biological sciences who can program. And they aren't selling their scripts for $99. Never have I heard of a script that a big firm uses that they can ignore, though. Are those guys telling their supervisors that the account is being whittled away and they don't know why, or they are searching for the magic settings? No way. Those EA's are like a blood hound in the hands of an intelligent police officer, and both together find the bad guy. The "indicator" that weeds out bad trades is the human mind. You can't script wisdom, sorry iPhone 4.

    Both of the Giants that contribute to this forum (I'd love for them to weigh in on this) would never trust their company's funds to any script, other than a stop order script (meaning they let price come to them). And my money is worth more to me than that of a firm.

    Please correct my thinking, have any of the EA's tested here been profitable? Any? Hello? Not for a month or 3 months but a real reliable track record? How long do you try out for a job as a trader before the company trusts you with their money? Probably at least a very well documented year of solid analysis and risk management in a LIVE account. EA makers, I'm tempted to open a scam investigation on all of you! Not really. But I would like to take this moment to thank the EA guys for selling hope and for having a great 3 months before you explode, change company name, issue a new EA and enjoy your vacations in Antigua and the Dominican Republic. Puerta Plata is really nice. Glad you enjoyed it. I'm enjoying my hope sandwich, though I can't afford any cheese or mustard.

    To Pharaoh, respectfully, I haven't seen anything that makes me believe an indicator can tell you anything your own market analysis can't and all really good indicators seem to be the last 5% of confirmation about what you already suspected. All indicators are lagging and price reactive, never price predictive. Never ever regardless of their name. If a guy says to me that he wrote a January specific EA, I'd believe it could theoretically work, but I'd have to see it for myself and have to see what the code is and what it's doing. He's assuming January will behave like most Januarys. I cannot believe that an EA has an indicator or bunch of them that tells them markets are trending or not. Those indicators are right until they are wrong. Then it's whip, whip, whip, busted. The years of searching for a new indicator, instead of just learning how the markets work, is a snare that puts people forever in the 95% category who make the 5% richer.


    Thanks for the book recommendations. Every successful trader has read a few books by other successful traders.

    Thanks to the heavy weights who dignified this thread by giving their 2 cents.

    If I find an EA that proves me wrong, I'll PM everyone on this thread with the script. Keep the $99. I'll be making all the free money I want. Maybe I'll send the scrip free and send everyone $99 paypal since I can afford it.
     
  10. Pharaoh

    Pharaoh Colonel

    Joined:
    Oct 3, 2007
    Messages:
    19,685
    Likes Received:
    2,230
    Didn't you read the ads. Aren't there at least 3 or 4 for sale for $99 or less that were "stolen" directly from the big banks?

    Of course, we know that the big banks write all of their trading software in MQL4. :p


    I'll go out on a limb and say it's impossible to write a perfect one that never needs some sort of settings updates. Some of the allegedly intelligent ones self-optimize, but I wonder if any of them can really do that. My suspicion is that none of the ones for sale can do it.

    I do see a small number of commercial ones running on live accounts in Performance Testing that are beginning to look at least a little tempting. Of course, if I did shell out my $99, I wouldn't get rich instantly. I'd start on a demo, then transfer to a nano account for months before even thinking of letting a robot play with a significant amount money.
     

Share This Page