Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

How serious are you about making money in Forex? 22 replies

How many real accounts have you blown before consistently making money in FX? 61 replies

Interbankfx - you making money with them yet? 16 replies

Are you making money 11 replies

Journal Factory posters - how much money are you making? 17 replies

Attachments: Which EA is making money for you?
Exit Attachments

Which EA is making money for you?

  • Post #1
  • Quote
  • First Post: Feb 19, 2008 3:06pm Feb 19, 2008 3:06pm
hi there. after reading hundreds of posts about (not) profitable Eas i still didnt get a straight conclusion: can a EA make constant profits ? tired of so many backtests, livetests, improvements, etc. I realize that this "dream" is never coming to an end. most of the time someone comes out and makes some modifications, (a.k.a.) improvements, saying he made one step closer to the perfection (I'm talking about already finished Eas). the thing is that a lot of people, including me, finding so many options (versions), doesnt know what to choose in the end. the possibilities are to many.

i want to ask you very straight: 1. Are you using a EA with your live account? Which EA is making constant profits for you ?

2. I ask you to post your results of your live tests/transactions only. Backtests are not accurate at all. I've made backtests on a couple of weeks in wich I also run the EAs live, and the results were not even close.

3. Please mention the Ea version and the settings you are using.

I'll start by posting my own live tests. I'm not using a EA in my live account yet. I didnt found a constant decent profitable one.

3000$ balance start, 10p3v0.02 EA, default settings:
GBP/USD m1
week 1: +119$
week 2: +77$
week 3: +38$
week 4: represents this week. still testing

GBP/USD m5
week 1: +62$
week 2: +22$
week 3: +1$
week 4: represents this week. still testing

The Eas were not stopped during any news. This EAs are the only one that showed to be capable of constant profits. I've tested other EAs also, but a negative result on a week, took them out of the game. Now, this 2 EAs are the candidates for my live account. I havent found a good one yet.

I hope many will benefit from this thread. Its up to you too!

P.S. Excuse my English.
Attached File(s)
File Type: mq4 10p3v0.02.mq4   11 KB | 1,489 downloads
Just not sure about long term results.

This will be an interesting thread.

Looking forward to it.

m
it looks ill have to manage myself as i did until now... anyway i'll post my best results, EAs and settings. bye
but I have a problem with my existing EA would appreciate if you could help meTicketOpen TimeTypeLotsItemPriceS / LT / PClose TimePriceCommissionTaxesSwapProfit794721872008.02.21 04:46balanceDeposit1 000.00794803262008.02.21 05:48sell0.01eurjpym159.24159.15158.892008.02.21 15:02158.890.000.000.000.32794807942008.02.21 05:56sell0.01eurchfm1.61961.61821.61612008.02.21 09:231.61820.000.000.000.13795216032008.02.21 09:23sell0.01eurchfm1.61791.61641.61442008.02.21 17:251.61440.000.000.000.32795486332008.02.21 11:00sell0.01eurusdm1.47381.47230.00002008.02.21 12:221.47230.000.000.000.15795487452008.02.21 11:00sell0.01eurchfm1.62081.62040.00002008.02.21 16:151.62040.000.000.000.04796002832008.02.21 15:02buy0.01eurjpym158.88158.89159.192008.02.21 15:29158.890.000.000.000.01796084082008.02.21 15:29buy0.01eurjpym158.93159.01159.242008.02.21 16:14159.240.000.000.000.29796350922008.02.21 16:14sell0.01eurjpym159.25159.10158.902008.02.21 17:31158.900.000.000.000.33796616922008.02.21 17:31buy0.01eurjpym158.91158.97159.222008.02.21 18:01158.970.000.000.000.06 0.000.000.001.65Closed P/L:1.65
Forex Price Action Hacked - Price Action Rockstar
duncanmc,

Thanks for starting this thread, please do continue to post your results and findings. I agree with you, there are so many ea's, and manual systems for that matter, being touted on here, but when it comes to putting them into actual practice they never seem to perform as well as hoped. Constant modifications, as you say, may work temporarily but usually quickly crash and burn. I am very interested in this topic, I am not using an EA on a live account but I have started using this 10points3 EA on a demo account with good results. Of course, it is a high risk because if one move goes against you it could be a huge loss, hopefully over time it would produce profit even with sustaining an occasional big loss.

Thanks again and don't be discouraged by the lack of feedback, I look forward to your continuing contribution.
Quoting duncanmc
hi there. after reading hundreds of posts about (not) profitable Eas i still didnt get a straight conclusion: can a EA make constant profits ? tired of so many backtests, livetests, improvements, etc. I realize that this "dream" is never coming to an end. most of the time someone comes out and makes some modifications, (a.k.a.) improvements, saying he made one step closer to the perfection (I'm talking about already finished Eas). the thing is that a lot of people, including me, finding so many options (versions), doesnt know what to choose in the end. the possibilities are to many.

i want to ask you very straight: 1. Are you using a EA with your live account? Which EA is making constant profits for you ?

2. I ask you to post your results of your live tests/transactions only. Backtests are not accurate at all. I've made backtests on a couple of weeks in wich I also run the EAs live, and the results were not even close.

3. Please mention the Ea version and the settings you are using.

I'll start by posting my own live tests. I'm not using a EA in my live account yet. I didnt found a constant decent profitable one.

3000$ balance start, 10p3v0.02 EA, default settings:
GBP/USD m1
week 1: +119$
week 2: +77$
week 3: +38$
week 4: represents this week. still testing

GBP/USD m5
week 1: +62$
week 2: +22$
week 3: +1$
week 4: represents this week. still testing

The Eas were not stopped during any news. This EAs are the only one that showed to be capable of constant profits. I've tested other EAs also, but a negative result on a week, took them out of the game. Now, this 2 EAs are the candidates for my live account. I havent found a good one yet.

I hope many will benefit from this thread. Its up to you too!

P.S. Excuse my English.
10p3 based EA are martingal EA so they will blow the account (any account) one day or the other but such EA can still be used if they are baby sitted (closing all position when the market does not show any sign of reversal of trend resuming )
this is very good post, but ...

my thing is, i'm ready to test on a "live" account,

after approx. 2 months of pretty good ( profitable )

trading on a demo account.

but, i read that the results will not, or may not,

be the same as the demo account results.

is this true ? if so, then i wonder why ?

also, so that i don't risk too much for my first live account test,

is it possible to open and account and trade

with just a couple hundred dollars ( $200 ) ??


thanks a bunch
I'm testing out an EA on the GBP and EUR 5MIN charts with a live account. Trading very small amounts.

The EUR has done decently.

I'm going to tinker with the SL and TP this weekend to see if I can generate more profit. It obviously recognizes break outs and enters the trade, but telling it when to exit is the key to it's long term success.
Capital Preservation is key to long term wealth accumulation
  •  AdamN
  • | Joined May 2009 | Status: Trading&Programing as Hobby | 209 Posts
This is an interesting thread, looking forward to see the result of 10p3 in the long run

Anyone else have a good EA used on real account?
Quoting MuddBuddha
I'm testing out an EA on the GBP and EUR 5MIN charts with a live account. Trading very small amounts.

The EUR has done decently.

I'm going to tinker with the SL and TP this weekend to see if I can generate more profit. It obviously recognizes break outs and enters the trade, but telling it when to exit is the key to it's long term success.
Are you thinking of posting it so that others can also play around with this ea and come up with some suggestions?
Will you be able to find or develop a retail EA that might make a few quick $ in the short term? Possibly. But what is more likely is that it will eventually fail, and you will enter the cycle of EA-jumping until you wake up and realise that trading profitably is about much, much more than finding the right "system" or EA. Contained in this below post is one quick story about EA's and good full time traders which may interest you.

http://www.forexfactory.com/showpost...0&postcount=16

My suggestion? If you want to have a bit of fun and have a bit of a "punt" on the markets for a short while, then by all means play around with EA's. But if you want to get serious about trading and have ambitions of long term, consistent, sustainable profitability then forget EAs and start the long journey of learning how to trade properly.

Sorry to put a dampner on this thread but sometimes where there is vision and hope, there also needs to be a dose of reality. But I don't expect you need to take my word for it though - that reality will hit home if you play around with retail EA's for long enough.
I love these type of threads.

Why do you think banks and other financial institutions employ traders rather than spending millions on developing fully automated systems.

The reason is that it cannot currently be done to any significant percentage.

They do however have semi-automated systems which after millions of calculations will give BUY, SELL or CLOSE signals but the decision is left to the human trader to accept or ignore the advice being given.

A purely technical driven EA or system alone cannot be profitable long term without taking into account economic and fundemental analysis which currently cannot be coded.

Scoobs.
Quoting scooby-doo

A purely technical driven EA or system alone cannot be profitable long term without taking into account economic and fundemental analysis which currently cannot be coded.

Scoobs.
I disagree. I think a profitable automated system can be developed, that is purely technical in nature. What I am certain of, is that it would have to be a hell of a lot more complex and comprehensive than anything you will find on a public forum.
Quoting moneymikie
this is very good post, but ...

my thing is, i'm ready to test on a "live" account,

after approx. 2 months of pretty good ( profitable )

trading on a demo account.

but, i read that the results will not, or may not,

be the same as the demo account results.

is this true ? if so, then i wonder why ?

also, so that i don't risk too much for my first live account test,

is it possible to open and account and trade

with just a couple hundred dollars ( $200 ) ??


thanks a bunch
Hi MoneyMikie

Here is a good example of why demo results are not accurate.
The spike in this demo account produced a trade closure that resulted in a substantial gain to the account yet in a live account you will not see this spike.

All the best to you in your future trading.

Dan

Attached Image (click to enlarge)
Click to Enlarge Name: 1.jpg Size: 174 KB
Couple things here...testing the 10p3 EA is a waste of your time, although I understand the fascination with it. You're months, if not weeks, away from demolishing your account (assuming you're letting it trade in auto both ways).

Secondly, completely automated systems can never work on their own and generate consistent profits. If you disagree with that statement, you simply have not been around long enough, or your brain can't grasp the concept that any mathematical formula will have boundaries in which it operates. These boundaries are constant. The market knows no such boundaries.

Take it for what it's worth.
MM
Quoting Aldente
I disagree. I think a profitable automated system can be developed, that is purely technical in nature. What I am certain of, is that it would have to be a hell of a lot more complex and comprehensive than anything you will find on a public forum.
An EA must be simple, otherwise it is not better than discretionary trading. Too many key variables or "if-the-else" in your EA and you run the risk of over-optimization which maybe worked in the past but will fail for sure in the future.
You don't find good EA's here because they require programming skills and finding some inefficiency of the market. The only thing is, one doesn't disclose it on a public forum.
Quoting mr.marketz
Secondly, completely automated systems can never work on their own and generate consistent profits. If you disagree with that statement, you simply have not been around long enough, or your brain can't grasp the concept that any mathematical formula will have boundaries in which it operates. These boundaries are constant. The market knows no such boundaries.
I've coded a variant of DIBS which is profitable. It is not extremely profitable on a daily basis but you can double your account every 2 years using the pairs which trend the most.
Not trying to call you out, or anything. But if you're telling me that you've been running this thing "hands off" for two full years and it doubled your account... I simply don't buy it. If it's based on back testing, then great. Otherwise, your claim is as believable as defying the laws of gravity... an extreme rarity in the physical domain.
The biggest problem IMO is that the most 'robust' technical concepts are 'visual', abstract, discretionary, etc, which are not easy to define mathematically. How do you program an EA to rank the quality of support and resistance levels, or of a chart or candle pattern, for example? Yet, for an experienced analyst, these levels and patterns are 'obvious': his eyes are subconsciously and immediately drawn to them.

I read everywhere that profit is made in the 'exits', by letting profit run, and cutting losses. If entries are unimportant, then let's suppose we enter on a MA crossover, with an initial SL just beyond the recent swing point, and then trail the SL. That's very easy to code. So why aren't more of these EAs profitable? (more here)

In general, indys (which are easy to code) work profitably when their calibration happens to coincide with market cycles, and fail when they don't. For example, sometimes a MA(20)/MA(30) cross will align itself nicely with price swings; sometimes a MA(50)/MA(100) cross will work better. In my pre-forex days I spent some time writing indy-optimized EAs, and, after testing them across suitable sample sizes (several hundred trades across different markets), ended up chasing my tail. Hence I'm now very wary of the dangers involved in curve fitting. I reached the conclusion that the underlying market drivers are not indy-based, hence indy-based EAs will always be hit or miss.

The psychology that drives the market might remain constant (fear and greed are near-irresistible emotions, and humans are invariably creatures of habit), but the problem remains that it's impossible for an EA to know who the participants are at any given point, and what the nature of their sentiment is. Forex is more complex than other markets because every pair is affected by two different currencies, and the market incorporates a mix of both speculative and non-speculative players. No doubt there are other reasons, also.

Another point is that (if I may generalize) is that, in order to remain profitable indefinitely, an EA must be smart enough to adapt to any changes in market conditions (behavior, patterns, etc), and also smart enough to know how conditions have changed, and hence what the nature of the required adaptation is.

If I understand correctly, there are quants who have attempted to solve the problem using complex math and physics based algos, but these algos (wave theory, chaos theory, high powered stats, etc) fall well beyond my rather limited understanding, LOL.

Bottom line is that (as has already been pointed out in this thread) if somebody is smart enough to devise an EA that exploits a real inefficiency, in such a way that it's guaranteed to deliver profit consistently and indefinitely, they're certainly not going to sell it, let alone give it away for free on a public forum.
In it's simplest form, trading, whether with an EA or manually, is a flat out matter of probabilities.

Given the market conditions that exist and the values demonstrated by whatever indicators you've placed on your chart, is this market condition likely to continue on long enough to make your position profitable given your stop loss and profit target?

Unfortunately, currency pairs rarely continue in the same behavior for extended periods of time. Their movement patterns come in waves and cycles. I don't mean ranging, trending, etc., but increases and decreases in daily volatility.

This is where most EA's, and honestly, most traders fail. Learning to adapt to changing market conditions is not something that comes naturally to most human's. So why should we expect any less from an EA?

Ronald Raygun has been working on a learning EA, and he has come a long way with it. It's very good. It still depends on the trader to set in the parameters, so there is always that weak point, but on it's own, it's excellent.
Capital Preservation is key to long term wealth accumulation
0 traders viewing now