Aurix Exchange

My EA`s Statement!

kerozin

Member
I have found a few EA`s over years of searching which are really profitable and safe!

Here is the statement of a few days with low Lot sizes which can be improved
even more if needed!
 

Attachments

  • DetailedStatement.jpg
    DetailedStatement.jpg
    16.1 KB · Views: 0

kerozin

Member
It`s not that good as i have tested many EA`s in this account, but with time
I remove those which don`t performs as expected and will add better ones!
 

kerozin

Member
Some of them are available in forums, like Wallstreet, and FxVoodo and some are costume made, so you can`t find them anyway!
 

kerozin

Member
Not for that, wanted to prove for those who don`t believe EA`s can be profitable that there are good EA which are still good and profitable, and anyone can find them if he looks for them deeply and don`t give up easily!
 

Capella

Active member
VIP Member
Not for that, wanted to prove for those who don`t believe EA`s can be profitable that there are good EA which are still good and profitable, and anyone can find them if he looks for them deeply and don`t give up easily!

Most EA's can be profitable during short periods of time, some also during long periods of time. A statement for 2 weeks doesn't prove anything, especially since you don't keep statistics on the different EA's that have made the trades. If you want to give helpful information, then run each robot on separate accounts, and state the name of each robot and what the settings are.
 

radicaltour

Moderator
VIP Member
I totally agree with Capella. At one point I had 5 EAs that made 1-2% everyday with very small drawdown. I thought I was in heaven, and this lasted for months...now they're all in the recycling bin. I think no EAs should be considered worthy without at least 1 year of profitability on a Live account.
 

kerozin

Member
I agree with you too, however those are carefully chosen, and they have about 1 year good forward test too, but I have started those ones just now to test!
And I think using 4-5 good EA`s in one account it`s much better, than only 1 because it`s more safer, and you get smaller DD also they trade more often, so your profits grow faster, also if you see one of them fails to deliver, just replace it with a new one, no big deal!
The problem is to have all of them profitable, if you have 3 winning ones and 2 which eats up the profits, you can`t make to much profit!
 

Capella

Active member
VIP Member
I agree with you too, however those are carefully chosen, and they have about 1 year good forward test too, but I have started those ones just now to test!
And I think using 4-5 good EA`s in one account it`s much better, than only 1 because it`s more safer, and you get smaller DD also they trade more often, so your profits grow faster, also if you see one of them fails to deliver, just replace it with a new one, no big deal!
The problem is to have all of them profitable, if you have 3 winning ones and 2 which eats up the profits, you can`t make to much profit!

I agree on that.

In order to use EA's in live accounts, they must prove that they are reliable during several years. Its' not realistic to demand several years of forward tests on live accounts, so running backtests is necessary. However, we must realize that backtests can never be 100% true. First, they are almost always never based on true tic-data. Second, they don't consider the actual broker prices (they comes from MetaQuotes). Nor do they consider the actual broker spread. However, it's the best we can do, besides running them on demo-accounts. A problem with demo-accounts though, is that the prices and spreads differs between the demo-servers and the live-servers for all brokers I've seen (of anyone know about any broker that has the same prices and spreads then let us know).

Most people just seem to stare at the profit curve, looking for the maximum profit, and hgh percentage of wins. However, something that is much, much more important is the drawdown, This must be kept as low as possible. I prefer to earn 5% per month if the drawdown is not higher than 2%, instead of earning 50% in a month with a drawdown of 75%. The first can be regarded as profitable trading, the latter is simple gambling. For me, any relative drawdown that is higher than 10% is simply unacceptable. I realize, that when trading on a demo account, or on a live account with $1000 or less, then it doesn't matter. But as soon as one start to trade with larger amount of money on a live account, large drawdowns are just not acceptable.

I trade live using several of my own robots on the same accounts. Running several robots on the same account is not a problem. Contrary, if they use different strategies and trade on different pairs, it's a way to hedge. I also run lots of different robots on demo-accounts, both my own and others (free and commercial ones).
 

pedro0410

New member
1st note

1st note

Hi
pretty new to all of that. just finishished my 1st ea which seems to be
kind of in profit. I backtested 2007 to may 2011 ( the quality of my test data seems to be not the best)

maybe somebody can help me getting better quality
it started in 2007 with usd 1000 end 2011 usd 4000

if somebody is interested


as i said my first robot

rgds
pedro
 

Capella

Active member
VIP Member
First of all, "mismatched data" and "modelling qiality: n/a" means that the historical data you've loaded from MetaQuotes have errors. Reload, and reload the historical data för EURUSD timeframe M1 and try again until there's no mismateched data.

Second, it's not possible to give you good feedback as long as you don't post the actual orders. So save you StaretegyTester.htm and corresponding gif file as a zip-file and upload for us to review instead of taking a snapshot of low quality that is hard to read.

Third, try running it on smaller timeframes to see which works best. Also try to run 2 to 3 optimized settings at the same time for stoploss, takeprofit, traling stop, etc. Don't adjust them all at once, only 1, 2 or maximum 3.You can step in 5 pips each first and then later on step on 1 pip to find the best optimized settings. Run the test from 1 January 2008 till todays date. If the last days ends up with large losses, the change the last date to a date where the last trades does not have any big losses or big profits, but normal ones (compared to the recent ones). When you have optimized the EA for this larger period, then run it again for the last year, and last 3 months and optimize the settings again, so that you can confirm if there's any need to adjust the settings for this recent period compared to the last 3 years.

There's much more to say.

Good luck wth your EA.
 

kerozin

Member
Most people just seem to stare at the profit curve, looking for the maximum profit, and hgh percentage of wins. However, something that is much, much more important is the drawdown, This must be kept as low as possible. I prefer to earn 5% per month if the drawdown is not higher than 2%, instead of earning 50% in a month with a drawdown of 75%. The first can be regarded as profitable trading, the latter is simple gambling. For me, any relative drawdown that is higher than 10% is simply unacceptable. I realize, that when trading on a demo account, or on a live account with $1000 or less, then it doesn't matter. But as soon as one start to trade with larger amount of money on a live account, large drawdowns are just not acceptable.

Yes, I know the advantages and disadvantages of every system, and it`s not easy to find EA`s with low DD and High Profit Factor and high winning ratio and good Risk:Revard ratio which can survive every market conditions
for long therm that`s for sure!
But I don`t think my Ea`s has big DD, as I remember it`s under 1%, and they make arround 200 pips per/day, arround 3000/pips/month so depending on lot size that will be more than 60-80%/month for sure,
Try to make that every day manually and you will know about what I`m talking about!
 

Capella

Active member
VIP Member
<snip>...
But I don`t think my Ea`s has big DD, as I remember it`s under 1%...<snip>

According to the picture you uploaded from the StrategyTester, the relative DD was about 11%, not 1%. But 11% is ok if your EA can hold that for years. So run some tests from 1 Jan 2008 til now to confirm this. Keeping the DD low is crucial.

Unfortunately the StrategyTester doesn't calculate Standard Deviation, Z-Score, HPR, Sharpe ratio, Balance of growth, MFE/MAE, GHPR, LR correlation, and other statistical values which are important to look at in order to correctly evaluate trading systems. But for all who developing their own EA's, I strongly suggest that you read more about those statistical values, and how to calculate and read them.
 

kerozin

Member
According to the picture you uploaded from the StrategyTester, the relative DD was about 11%, not 1%. But 11% is ok if your EA can hold that for years. So run some tests from 1 Jan 2008 til now to confirm this. Keeping the DD low is crucial.


According to the forward test I uploaded the DD was 1.35% exactly!
I don`t need to run test as I`m full of them, they don`t tell you
much anyway!
I see You don`t give too much attention to forward test I have attached,
I know it`s not for long term, but I can upload new ones anytime!
You see them as Strategy Tests, which isn`t the same thing!
Don`t worry, I know for what and where too look for!
 

Capella

Active member
VIP Member
According to the forward test I uploaded the DD was 1.35% exactly!
I don`t need to run test as I`m full of them, they don`t tell you
much anyway!
I see You don`t give too much attention to forward test I have attached,
I know it`s not for long term, but I can upload new ones anytime!
You see them as Strategy Tests, which isn`t the same thing!
Don`t worry, I know for what and where too look for!

Forward tests are of course better than backtest. But 2 weeks forward test cannot be compared to 3+ years of backtests. If the backtests shows a drawdown of 11% then you should also expect about the same on forward tests.
 

Aurix Exchange
Top