:: The Dream EA What's yours? ::

The reason I ask is simple. I've used a variety of EA's on demo accounts over the last three months. ALL - without exception - have lost money. I am still using them. All of them in this forum and others I've found elsewhere.

IMO the discussion here reminds me of a childhood story about a king with no clothes. Many talk about consistent profits. But I just haven't seen them in my demo accounts.

The amount of work that people go to here to develop EA's and then share them is commendable. Ladies and gentlemen, do not for one minute believe I am criticising your work. My experience is what it is.

If you had unlimited time or funds to pay someone to develop an EA for you what indicators would you use? What settings? Hours? Restrictions? Limits? Checks? Confirmation? Strategy? Money management?

All comments - negative and positive - appreciated.

i would just want somethin that closes 2 trades when they are in profit by a certain number of pips or dollar amount.

My Dream EA is firebird but I am still testing.

Is it in profit so far?

Well check the firebird threads. Its in profit overall but I believe it got roughed up this past week.

Some EA are good for range bound, some are good for trends. I have yet to see an EA that works for both.

Currently, a simple 2 MA crosses + RSI + MACD could have made you 300 pips within 10 days when you placed them onto GBPUSD since AUG 30...

Over at InterbankFX website, they have a contest for manual and EA executed trades...
http://www.interbankfx.com/mini_chal...st_results.php
You'd find some people doing no less than 4000 trades in a month suggesting more than 100 trades in a day? Obviously a scalping EA.

Perhaps that could be a winner.

My dream EA is one that wins me the Automated Trading Championship and banks me 40 large! I'm trying to code her right now, but I don't know if she'll be ready by the deadline.

andersenws

In my first few weeks of doodling with forward testing of various EAs i can sure say (Like Yardie and BK) that each has a personality that must be suited to it's market type. In a ranging market EAs such as FB can easily yeild over a thousand pips in a week. And in a trending market there are some i've tested that can also yeild over 1k in pips a week.

Here are my 2pips worth;

Somehow, there needs to be an EA-Manager created. The purpose of this particular EA will be nothing more than determining the current market condition; Up, Down, or Sideways. THEN the EAM would turn on the respective EA for that type of market. The user could designate which EA s/he felt was appropriate for the type U/D/orS, after all there will always be arguments as to which is the best in any given market, so we would not have to force say a FireBird on someone for a ranging market, and your favorite EA for trending markets. . . I know this is e<DO do conditions.< market current the for tool proper off on turn to work its EAM let then and choose let?em Just market. trending a in similar or interested] if links follow google p>asier said than done. But in the end i believe an EA Portfolio of sorts will eventualy win the day. That is the direction i am heading and researching.
Cheers,
Thom

I think anything is possible, I think depending on who is using the EA and how, you can make successfull EA's on anytime period but my year of coding led me to a new trading system I love, that I will be trading by hand.

Everyone loves the scalping systems, and the quick exact entries on price movement and the quick exits. It's trading to a machines strength's something that would be impossible to do 24/5 as a human. The only problem I had with these sort of EA's is live market conditions can make short work of them.

1. Try to find a broker who actually lets you trade 1000+ times a month might be a bit tough

2. EA's have to play by the brokers rules. They have no choice, they have limited ability to deal with platform freezes, pip spreads, news, you can obviously code around this stuff, however it does tend to hurt the integrity of the system you have created in the long run.

3. EA's are built on solely Technical Analysis, widely regarded to fail in the long run. I tend to agree with that statement for scalping systems and short term trades but dis-agree for longer trades, position trades that can take days-weeks.

I do think that on 4hr time+ charts and systems that use a few bars up to 20+ bars to create and confirm signals are reacting to fundamental changes being priced into the market. I do think you can code around live broker issues of platform freezes, pip spreads, etc. much easier. When your entry is well confirmed, aggressive risk management isn't as scary.(by aggressive mean maybe bumping up to 5% risk...not 20%!)

The best EA I wrote traded just over 120 times in 3 years back test...made crazy amounts of money. Because the trades were so general taking hours to confirm, it traded at 80% good trades. Didn't matter what type of data you fed it, 90% or 10% modeling quality, it hit the same solid trades, and it hit them in the now 2 months of forward testing. The big plus is it "shouldn't" have much trouble with pip spreads(targets are always 100pip+ away), platform freezes during during news or volatility actually help keep me out of trades that may be wrong or unconfirmed.

In the end my favorite brokers don't have metatrader...so I will be running it on a DEMO and sending myself emails about signals, and entering trades manually on my preferred broker.... I figure I can handle 50 a year or so! I generally have a few hours to reach a computer and place a trade, it's also easy to see when a trade is developing and when a potential entry is coming within a day or 2.

Anyhow my opinion, I think there are people who are smart enough to create very profitable scalping EA's, I just haven't been one of them yet. If my current strategy does make money, I will be re-visting fast EA's as my next goal.

Also as a tip for beginners, focus on creating an EA that run's on 1 currency. There is more than enough money to be made off 1 currency, and it's very difficult to focus on a bunch of currencies. My system that trades 50 times a year, is only profitable on EUR/USD. It even enters buy and sell positions completely differently. Price action is just to different across currencies for the bulk of EA's to run well on. Also once you have studied your current EA's build on the price data for the 2000+ time, you begin to just feel that pair's price movement alot better.

Again just opinions, you really need to find your own way in this...just because it's automated doesn't mean you are emotionally un-attached to the trades. You still fund the account, you still apply the EA, and you still watch it play with your money. You have to be comfortable with it!!!