Forex AI

[Harre]

Just saw it while surfing at lunch time. A similar marketing/sales page to FAPT, showing live results that doesn't look too bad.
As far as I can tell they're trading eurusd only and it seems to be swing trades with a SL of around 90 pips.

Anyway, just stumbled over it, don't know anything about it, just throwing it in here since people like to compare and contrast EA's.

[Harre]

And the link to their results pages: http://forex-ai.com/LiveResults/tabid/64/Default.aspx

Following this but more interested in what the zenwonton EA can become :-)

[clolab]

interesting results....

[Harre]

Hmm, ok, while the Zenwonton super duper EA is being calibrated I decided to give this a try. Gonna set it up on a VPS starting trading Monday and will report results here. There's always money back guarantee :-) Not holding my breath but it's always fun to try it out.

[clolab]

yes please report your experience with this EA. Would be interested to.
What do you refering wiht "Zenwonton super duper EA" ?

[seabee40]

Besides FxPro, what other brokers can you use?

[Harre]

Zanwonton's EA that's under development here http://www.forexautopilotforum.com/showthread.php?t=529

I think any broker as long as they have charts for Gold and/or Crude oil as well as the eurusd. Haven't read the manual yet but will do so I can put it on a demo on Monday. Will use the Activtrades demo account I think. Will probably connect it to the mt4 stats page so I don't have to update the thread with every trade...

[thesecret]

Good Luck Harre. I'll be waiting for your results. Iam considering to purchase Forex AI also.:cool:

[reno77]

Hi all, I've purchased it too. It was $99 with 1 month money back guarantee. When you pay you can download a trial version for 1 month and when it expires you email them for the unlimited version.

It based on a Neural Network, ex4 file is pretty small only 10kb so I guess its a pretty simple AI :). Its not completely hand off as the neutral net need to be trained about once a week to adjust to the market. Hope its not the same logic as this free PNN neural network..

http://www.hernandezgomez.com/index.php/using-a-pnn-probabilistic-neural-network-and-metatrader-4-mql4-to-trade-forex/

I'll try on demo for this week and see how it goes.

[clolab]

Hi,
can you explain how you have to "traind" this EA? is it based on backtest or training is do on forward test?
I am interested to, but a bit skeptic, also dont like the 1 month "trial" story. I think if you purchase you should have full software...
:rolleyes:


thanks a lot for informtion.

[reno77]

Hi you have to backtest it on a time period of 3-6 months. I'm training mine from August 1 2008 to Today. The system works on H4 time frame and uses an alternate graph , either Oil prices or Gold to confirm the Buy/sell position to take.

There are a couple of variables that the system uses. When you run the Metatrader strategy tester it'll run for about 15 mins and give you the best values which give the lowest drawdown and highest profit. Then hopefully the values you get are good for your live trade...

You have to continuously do the strategy testing each week with updated data to get new variables that match market conditions.

[clolab]

Hi you have to backtest it on a time period of 3-6 months. I'm training mine from August 1 2008 to Today. The system works on H4 time frame and uses an alternate graph , either Oil prices or Gold to confirm the Buy/sell position to take.

There are a couple of variables that the system uses. When you run the Metatrader strategy tester it'll run for about 15 mins and give you the best values which give the lowest drawdown and highest profit. Then hopefully the values you get are good for your live trade...

You have to continuously do the strategy testing each week with updated data to get new variables that match market conditions.


ok i see. Juts my 2 cents. I would not include the data from dic15/2008 Gen08/09
So maybe only from sept/08 ->Dic15/08 Due christmas data would alterate results and not be good for live.

As i say, only my 2 cents.

[Harre]

...also dont like the 1 month "trial" story. I think if you purchase you should have full software....

You have the full version, it's just locked down to 1 month. That way if you don't like it it's very simple for them to just give you the refund. If you want to keep it you get an unlocked coipy. I certainly don't have a problem with that.

Will try to get it up and running on VPS tonight

[clolab]

what performance does backtests shown?


thanks

[Harre]

I haven't done any backtests and I don't think it's possible since you update the settings on a weekly or bi-weekly basis. If you look at the link on the previous page you can see their live results from a good while back.

[clolab]

yea i see.. Perhaps you can mate optimization for 1 month and the re-run the backtes to see how it performance with optimized setting on this periode.
Anyay, i will go purchase the EA.

thanks

[thesecret]

I haven't done any backtests and I don't think it's possible since you update the settings on a weekly or bi-weekly basis. If you look at the link on the previous page you can see their live results from a good while back.

So u must run optimization on a weekly or bi-weekly basis? :eek:

[seabee40]

That's what they recommend. They say you could do it monthly, but only if there isn't much turmoil in the market.

[Harre]

Ok, got the AI up and running on my Activtrading demo account now and the results will be shown here: http://harre.mt4stats.com/

If nothing else this EA has taught me how to optimize and backtest so already worth the investment :-D

[reno77]

Ok first trade forex-ai made last night looks like its going to be a winner :) Up $23 as of now.
http://farm4.static.flickr.com/3133/3192881518_5672d5bfc5_o.jpg

[clolab]

interesting... keep us up to date :)

[seabee40]

Reno, what are you using as alt_symbol in FXDD

[reno77]

Dang.. it closed with $ -22 loss after it hit a S/L at 1.3321. Back to the drawing board. I think i'll take this off live, retrain and try on demo account first. Good thing FAPT is doing good today so can cover the loss

[reno77]

Reno, what are you using as alt_symbol in FXDD

Hi I'm using XAUUSD since I cant find any Oil symbols.

[seabee40]

I was using that also on Fastbrokerfx and I couldn't get to optimize. I got optimization on Fxpro. FxPro hasn't had a trade yet

[Harre]

yea i see.. Perhaps you can mate optimization for 1 month and the re-run the backtes to see how it performance with optimized setting on this periode...

Yup, did this and that's how the optimisation actually works. Backtests look very good and I'll try posting them here tonight. I used data from October to the 19 December to avoid holiday trades.

[clolab]

Yup, did this and that's how the optimisation actually works. Backtests look very good and I'll try posting them here tonight. I used data from October to the 19 December to avoid holiday trades.

ok fine. I am curiouse to see results from that backtest (i know, it isn't reference for forward) but curiouse to see results from real peaple than the sellers web page. :)

[Harre]

First trade is in action :-)

Edit: Using 0.1 lots since that's what I am likely to start with, and have switched MM off /Edit

[thesecret]

Big dd some minutes ago...Good Luck boys...

[icg]

The backtest should look very good because you've optimized the parameters on that data. It goes through a huge number of iterations to see what would have worked best in your testing period.

Now, that's not necessarily a bad thing, but that's why it looks so good. The best way to test this kind of system is a walk-forward backtest. For example, "train" it using Jan-Jun. 2007 data then test it in July 2007. Keep doing that 1 month at a time. Unfortunately it's time-consuming and not fully automated, but that's the best way to test a system that relies so much on optimization.

Yup, did this and that's how the optimisation actually works. Backtests look very good and I'll try posting them here tonight. I used data from October to the 19 December to avoid holiday trades.

[Harre]

The backtest should look very good because you've optimized the parameters on that data. It goes through a huge number of iterations to see what would have worked best in your testing period....

Yea I know, which is why I said back testing this doesn't really make sense unless of course you go through the process of stepping forward as you say.

Anyway, forward testing is the way to go and in this case I don't think it really matters it being on a demo account. Starting with a SL just to make me happy I didn't put it live straight away :-P

[agathos23]

I took the plunge and bought the EA last Friday.

I must have gotten lucky and picked a good optimization parameter. So far it has done two trades. The first one last Friday I closed it manually with a 50 pip profit because I didn't want to carry an open position over the weekend. Then it opened another trade yesterday and it's still open right now and I'm up +170 pips.

This is with a live account with FXDD.

[Harre]

Nice Agathos! My first trade hit SL and the second one has been in the red all along, currently -43 pips. Not the best of starts but it's only demo and I was dubious to the optimisation I had done. Not easy to know what to include when we've had the holiday period. I just did October through to 19 Dec, which is a bit too short as well...

[agathos23]

Nice Agathos! My first trade hit SL and the second one has been in the red all along, currently -43 pips. Not the best of starts but it's only demo and I was dubious to the optimisation I had done. Not easy to know what to include when we've had the holiday period. I just did October through to 19 Dec, which is a bit too short as well...

I ran mine from 9/9/08 - 1/9/09 including the holiday period. Then I picked the setting that balanced profit with relatively lower drawdown percentage. So far so good. But then its still too early to tell with only two trades so far.

BTW, I have FapTurbo, Forex-AI, and FAP running on this live FXDD account that I opened during X-Mas with the free $100 balance. If I closed all my open trades that I have right now, my balance would be $390 or so. Not a bad haul so far starting from $100. :-)

[clolab]

Yea I know, which is why I said back testing this doesn't really make sense unless of course you go through the process of stepping forward as you say.

Anyway, forward testing is the way to go and in this case I don't think it really matters it being on a demo account. Starting with a SL just to make me happy I didn't put it live straight away :-P




"harre" i saw your Forex AI account stats. It' not that goos rigth now.
:(

[Harre]

I don't think it looks too bad. Lol, it's only been running for 2 days :-) Hit a SL with the first trade. The second one bounced between profit and loss and when it dropped back down again it stopped out on the trailing stop for -10. Sure it would have been nicer to have taken the 90 pip gain which it had but it's not a scalper, it's a day or swing trader this looking to make larger gains than the losses. It expects to have a couple of losses and make up for them by a larger profitable trade so let's see how it performs this week.

Like I said my optimisation was done between October and December 19th. I'll probably re-do it over the weekend and include all days until now to get fresher more recent data.

[agathos23]

BTW, the developer came out with a minor upgrade to the EA.

[xchaparro]

Ok, got the AI up and running on my Activtrading demo account now and the results will be shown here: http://harre.mt4stats.com/

If nothing else this EA has taught me how to optimize and backtest so already worth the investment :-D

nothing good so far ;)

[Harre]

DOH! The VPS stopped this from working. Rebooting the server now and think I'll re-do the optimization at the same time. CRAP!

EDIT: Ok, I'm also a £@$€@£$ idiot... Due to the bug in MT4 whereby you have to enter the value twice in certain fields (for optimization) I have used settings that are completely wrong :-(
Have now re-done the optimization from 6 October to today's date. Let's see how that'll work.
/EDIT

[clolab]

oh, so the results are non that good due the wrong optimization?
Please let us know when you have the optimized EA running on your account.


thanks

[Harre]

oh, so the results are non that good due the wrong optimization?
Please let us know when you have the optimized EA running on your account.


First of all I can't say if the results are good or bad yet. This is a day/swing trader which expects more losses than profits but that the profits are substantially bigger. I.e. totally opposite from FAPT which lives on having loads of small wins to offset any stop losses that occur. In a month maybe it's time to say if it's working or not.

I have no idea if my settings caused the 2 first trades to end up in the red, The other chap who's using it didn't have the trades I had at least. I just know that I had very different settings to the ones I have now after optimizing the AI once more, this time with correct values.

It's up and running from yesterday with the new settings and the link in signature should update with the results. There was a short this morning so let's see where that goes.

[agathos23]

Ouch. I took a big hit with this EA this morning for a 96 pip loss. Three trades. Two winners and one loser. Overall I'm still up about 55 pips or so. Still too early to say if this EA is good or not. As harre said, this is a longer term evaluation. Being up 55 pips is not too bad actually and hoping to have more winners than losers.

[james_phuc28]

Hi,

Please try my settings :

+ Alpha = 56

+ Beta = 170

+ Mode = F (Fixed lot).

+ Alt_symbol : XAUUSD (or default, if your broker has crude oil).

+ Other settings : as default

With this setting, you will have maximum drawdown of around 14% for the last 6 months, and never dried out the account. This is very important, when you intend to use it live.

From back-testing of USD3,000, it doubled the account within 6 months. But according to the creator of this EA, in real trading (demo and live accounts), the performance is much better than back-testing results.

Same as FapturboEA that live trading results are much better.

To make sure that you will reduce the losing trades, when this EA is in profit of around 50 pips or more, please block in 20 pips profit by moving stop loss, and then, let it run until the end of the trend.

I figured this fact out when yesterday, it made a loss of 90 pips (buy trade), and then, when I adjust the Alpha and Beta as above, it made another Buy, and could win back more than 90 pips, and I closed this second trade manually.

Since this morning until now, it made 3 trades, because when the trade was in profit for 50 pips, I manually locked in 20 pips profit, but market retraced, and hit my trailing stop. That was way this EA made 3 trades today, and all these 3 trades were winners.

Because I forward-tested on live account, I have to do like this to prevent it from loss. But if you use it on demo account, please let it run until it jumps out to see how it performs in the next 30 days.

Have a good trading day to all !

James

[Harre]

Currently in a short of eurusd which is 117 in profit but the trailing stop is still only at break even :eek: Just have to wait and see but I would have thought a TS would have been maybe 50 pips or so...
Anyway, right this minute the EA is back to 0 but the second trade it made was also in profit before closing for a small loss.

[agathos23]

Dang, the EA lost another 70 pips. Now I'm down -20 pips overall. Don't know why it made two buys of EURUSD when it's been trending down all day today. Now there's a caveat to this. My first two winning trades were based on a different optmization settings. The latest two losses are based on settings I put in yesterday morning because it kinda looked better. I'll stick it out now though since drawdowns are expected. I just hope it doesn't blow up my account.

[clolab]

Dang, the EA lost another 70 pips. Now I'm down -20 pips overall. Don't know why it made two buys of EURUSD when it's been trending down all day today. Now there's a caveat to this. My first two winning trades were based on a different optmization settings. The latest two losses are based on settings I put in yesterday morning because it kinda looked better. I'll stick it out now though since drawdowns are expected. I just hope it doesn't blow up my account.

hope you are on Demo.....

[agathos23]

hope you are on Demo.....

It's live but not my money LOL. I'm using the free $100 I got by signing up with FXDD. Lost alot today but basically lost most of the profits I built up since around Christmas. Went from $100 up to $370. Now down to around $200 so overall I'm still up.

[Harre]

Not going to second guess this system but it had over 150 pips in the bag earlier so why hasn't the trailing stop kicked in? Now we're back to 70 pips and things can obviously go well from here too but wouldn't it have made sense to stop out for 100 pips profit on the rebound and look for a new entry??? That's obviously not how this one works but that's how I'd treat it if I had done this manually. Or to quote a Swedish saying: "Better one bird in the hand than ten in the woods" :)

[clolab]

maybe you should drop an e-mail to the seller and ask. Perhaps there is something wrong.

[Harre]

maybe you should drop an e-mail to the seller and ask. Perhaps there is something wrong.

Not really much point since the EA will only do what the settings tells it to do. The optimization will make all our settings different even though the software is the same. I will however see if I can get info on the trailing stop. For now I'll just let it run it's course since it's on demo.

Edit: Well it stopped out for 10 pips up or whatever, not so cool considering it had 150 earlier. I've emailed Gary at Forex-AI to see what he's got to say but as I said above it's basically using the settings you give it. I'd just like to know how it handles trailing stops and if you can set that yourself somewhere once you've backtested and checked it won't screw the results. /Edit

[Harre]

One thing that does impress me with the guy/team behind this is that they are quick to reply and are continuing to improve the EA with new features suggested by their customers. I also like that there are still only a limited number of people using it :-)

Psyched up about next week with FAPT during the night and this during the day :D

[thesecret]

Psyched up about next week with FAPT during the night and this during the day :D

Serious? Results are not good in your mt4stats.com :confused:
Be careful Harre...

[Harre]

The results aren't good??? I think I'll remove that link cause there's no point having it there. If people are going to evaluate an EA by looking at 3 trades I would feel guilty if it makes a profit next week and half the world mortgage their house and but all the money on this EA...
I appreciate your concern thesecret but first of all my 500 buck accounts are for evaluation purposes and I really don't care if they get drained. Secondly I am still not putting this live as I am waiting for the broker to get back to me about my application(don't want to use the FxPro account for more than 1 EA to be able to keep track on how FAPT is doing).

[Harre]

Answer from the developers reg trailing stop>

It is programmed to only move the trailing stop at the end of the 4H block.
The reason it is programmed this way, because in fast moving market the trailing stop would need to move with every tick (when it reached new high or low) and the broker would detect that we are using automatic trader.

Unfortunately some brokers have the ability to put your account in manual execution mode when they see it too profitable and/or detect that you are using a profitable EA.
I can make the S/L to move like a trailing stop, but will make it as an option as I believe it may not going to work on live accounts with every broker that way.

I understand your frustration because a potential 100 pip became only 13, but I look at the big picture and see that the EA has performed pretty well over the past 3 month even though it didn't always exit well.

>Please advise, and is it possible to see how much the trailing stop is in the settings?
The trailing stop is set at 90 pips and called 'StopLoss" in the parameters and it is moves at the beginning of each bar only based on the previous bar closing price.

As mentioned above, I will have an option to move the StopLoss with every tick in the new version, due to be released in 2-3 weeks.

Gary

[grandaevus]

I think, I've found the magic numbers for the week.

143, 20, 144, 173

191, 173, 124, 58

Optimized Maximal Drawdown: %5

fixed lot: 0.10, alt_symbol:GOLD

Test range

09/11/2008 - 01/18/2009 (profit:4472.55, max. drawdown: %2.64)

01/01/2009 - 01/18/2009 (profit:391.46, max. drawdown: %1.52)

01/11/2009 - 01/18/2009 (profit:138.66, max. drawdown: %0.97)

[lukx]

What I should put in alt_symbol if I don;t have oil or other things that are in manual?

http://img91.imageshack.us/img91/4371/screenshot001kg9.jpg (http://imageshack.us)

[Harre]

You've got Gold, use that. It does say that in the manual ;-)

[Harre]

I think, I've found the magic numbers for the week.

143, 20, 144, 173

191, 173, 124, 58


Superb, I'll enter those into my AI as well and wishing best of luck. Still on demo here and waiting for broker to get back to me.

EDIT: Although will need to check these cause I am using GOLD and not CRUDE OIL as my alt indicator... /EDIT

[grandaevus]

Superb, I'll enter those into my AI as well and wishing best of luck. Still on demo here and waiting for broker to get back to me.

EDIT: Although will need to check these cause I am using GOLD and not CRUDE OIL as my alt indicator... /EDIT

I also use "GOLD" as my alt_indicator.

[Harre]

Cool. I did a backtest of your settings and they looked super so I guessed you were on gold as well. Will be interesting to see what happens to it this week.

[Harre]

Pissed off with the FAPT melt down last night I've put this live on FXPro for the time being using my settings from 14 Jan.

[agathos23]

Frigging Forex AI opens a buy of EURUSD when the thing is tanking. The parameters I'm using are not looking too good right now. I've had three consecutive stop losses of 96 pips now.

[Harre]

Too bad we both missed the short they did yesterday. Came in mighty nicely that one. I closed the long position earlier when it dropped down to 0 but it opened another while I was out shopping... Not doing swell either but there we go.

[reno77]

My Forex AI bought in at 1.2935 last nite too. It was actually up when Obama was giving his speech but went down soon after to close at the 1.2844 stop loss.

Its opened another buy at 1.2918 which looks good as EURUSD is going up now. I've manually adjusted the stop loss to 1.2918 so as not to lose anything.

[kostolany]

My Forex AI bought in at 1.2935 last nite too. It was actually up when Obama was giving his speech but went down soon after to close at the 1.2844 stop loss.

Its opened another buy at 1.2918 which looks good as EURUSD is going up now. I've manually adjusted the stop loss to 1.2918 so as not to lose anything.

any news about the Forex AI?

[clolab]

I am curiouse to... any news?

thanks

[Harre]

I've been offline a while and so has my trading. Back on Monday. I know the developers are working on new features regarding Trailing stop and take profit that you will have the opportunity to set yourself. Still think this might be a good EA but can't really say yet. I like the idea behind it and the response we get from the devs.

[Harre]

Forex-ai has made just over 40 pips the last 2 days. Looks like it makes decent entries and I quite like it so far even though it hasn't had a big win yet. Keeps the losses managable and feels like a well thought out product without marketing ploys.

[JillyB]

Forex-ai has made just over 40 pips the last 2 days. Looks like it makes decent entries and I quite like it so far even though it hasn't had a big win yet. Keeps the losses managable and feels like a well thought out product without marketing ploys.

Hi Harre

I'm reading this with interest. Please keep posting and let us know how it goes.

[Harre]

Got another 65 pips on a eurusd long this morning and since I put this live it's been profitable.
This EA is still improving. At the moment I think it makes sensible entries and once the trailing stop functionality and take profit option is added I think it will be even better at taking home the pips by exiting at the right time.
I'm currently risking more than they do on their live accounts. Gary and co only risks 3% max even though higher risk would have given them amazing results for marketing purposes. For me this just makes the EA more appealing. I'm not looking for holy grail make me millionaire in 3 months from a 500 buck account stuff. Sustainable profit in the long run and since this EA changes every time you run the optimisation I think it's got a reasonable chance to stay profitable.

My suggestion is to mail them and ask for a trial. I know they don't advertise it but considering their customer base is very small so far I reckon they might see it as an opportunity to get some good word of mouth going.

[reno77]

I am curiouse to... any news?

thanks

Sorry I took the EA off live last week after 2 losses. am waiting for conditions to stabilise in US ..hopefully next month.
Currently I'm manual trading now on GBP/JPY by taking tips from forexfactory's Auslanco GBP/JPY thread.

[Harre]

The AI entered a short but got caught out in a retrace and hit SL for -91 pips this morning.

[JillyB]

The AI entered a short but got caught out in a retrace and hit SL for -91 pips this morning.

How is it doing overall Harre?

Is it making profit yet?

[Harre]

Since I put it live I'm about break even (couple bucks profit). I'll keep posting here since I've had requests from a few peeps.

[Harre]

The AI entered a beautiful short and made 200 pips this afternoon. I've closed it manually to have margin enough for FAPT trades tonight. So update on the performance is that it's made just over 200 pips since I put it live. Will be interesting to see what the result had been if I had not closed it, but with the small account I have I needed the margin.

[clolab]

interesting... i am currenlty looking to purchase another AI advisor form Itrepidfs.com with some incredible results:
https://intrepidfs.com/DetailedStatement.htm

I have currently using 4.1 but i do not suggest using that versione as it need manuall optimizaton a few times a day. So i suggest to go with ver. 4.71 that has automatic otpimization.

just my 2 cents

[Harre]

Why a statement from February 2008?

[JillyB]

interesting... i am currenlty looking to purchase another AI advisor form Itrepidfs.com with some incredible results:
https://intrepidfs.com/DetailedStatement.htm

I have currently using 4.1 but i do not suggest using that versione as it need manuall optimizaton a few times a day. So i suggest to go with ver. 4.71 that has automatic otpimization.

just my 2 cents

Let us know how you find it Clolab. You could set up a new thread dedicated to it if you wanted, then we can all see how things are going.

Do they offer any money back guarantee with the product?

[clolab]

about statement.. i just found this one.. (didnt note it was from last year)
There should be a liv statement from a smaller account for begin of jan/09
Will post it when i find it.

about moneyback, i think yes i pay by paypal so moneyback should not be a problem.
I don know much about this EA i just discovered last week by navigating over the net.
Maybe we can open a new thread I have to upgrade to vers.4.71 as i don like to optimize 1-2 times a day manually.

[Harre]

Forex-ai went long on eurusd at 22.00 GMT last night and just got stopped out -90 so profit down to 110.

[Harre]

Another long that was entered on Friday got stopped out this morning -90 so profit down to 20.

[kostolany]

about statement.. i just found this one.. (didnt note it was from last year)
There should be a liv statement from a smaller account for begin of jan/09
Will post it when i find it.

about moneyback, i think yes i pay by paypal so moneyback should not be a problem.
I don know much about this EA i just discovered last week by navigating over the net.
Maybe we can open a new thread I have to upgrade to vers.4.71 as i don like to optimize 1-2 times a day manually.

hello clolab,
any news from ver. 4.71? Thanks.

[clolab]

hello clolab,
any news from ver. 4.71? Thanks.


hi, i ask the developer if he can send me the latest week statement of vers. 4.71. Hope he will send soon. I would be intersted to update to latest versione but like to see some more results.
Will let you all know as soon as i have news.

[agathos23]

Another long that was entered on Friday got stopped out this morning -90 so profit down to 20.

Same here. I've been having a string of losses. I checked the forum at the developer's website and found out that he uses 6 months for his optimization and will go down a month if get doesn't get something good. The last couple of weeks I've been using about 3 to 3 1/2 months. I've redone the optimization for a full 6 months today and I went by his recommendation of over 50k profit and drawdown of 16% or less. I entered the new settings to the EA today replacing the ones I used over the weekend. Let's see how it goes.

[clolab]

hello clolab,
any news from ver. 4.71? Thanks.


Hi, this is what i got from developer of neuronet about this weeks stats.
BTW. perhaps we should opena another thread as i do not make confusion:
----------
I don't have any live statements from last week.

There was a market anomaly where it was reporting bearish all week but was continuing to plummet due to world economic conditions and a few news announcements in the United States. This is a very rare scenario but it make it very difficult to predict any market movement last week so I chose not to trade live during that period--and will continue to do so until the final vote on an 819 billion dollar economic stimulus passes in the U.S.
-----------

[kostolany]

Hi, this is what i got from developer of neuronet about this weeks stats.
BTW. perhaps we should opena another thread as i do not make confusion:
----------
I don't have any live statements from last week.

There was a market anomaly where it was reporting bearish all week but was continuing to plummet due to world economic conditions and a few news announcements in the United States. This is a very rare scenario but it make it very difficult to predict any market movement last week so I chose not to trade live during that period--and will continue to do so until the final vote on an 819 billion dollar economic stimulus passes in the U.S.
-----------

i have read their support forum at https://intrepidfs.com/drupal/?q=node/6
i think, they are searching for better settings for this EA. only my 2 cents.

[Harre]

i have read their support forum at https://intrepidfs.com/drupal/?q=node/6
i think, they are searching for better settings for this EA. only my 2 cents.

Can you please go and hi-jack somebody elses thread? Or even better, create a new one.

[Harre]

Ok, have changed broker but now back online with Forex-ai and new settings from optimisation yesterday. It took a short position during last night but was stopped out for -90. That means -70 live so far but still confident this will turn out to be a good little investment.

There's a beta version out now with added settings for Trailing Stop, Take Profit level etc and I like the way this team listens to their customers. Mind you, at the end of the day it's dollars that counts in this business so time for some profitable trades :-)

Their live accounts, one using OIL and one using GOLD as alternative indicator did reasonably well in January. 1 only managed break even but the other up 65%, and this with a risk of 3%.

[JillyB]

Ok, have changed broker but now back online with Forex-ai and new settings from optimisation yesterday. It took a short position during last night but was stopped out for -90. That means -70 live so far but still confident this will turn out to be a good little investment.

There's a beta version out now with added settings for Trailing Stop, Take Profit level etc and I like the way this team listens to their customers. Mind you, at the end of the day it's dollars that counts in this business so time for some profitable trades :-)

Their live accounts, one using OIL and one using GOLD as alternative indicator did reasonably well in January. 1 only managed break even but the other up 65%, and this with a risk of 3%.

Hope it works out Harre. The EA sounds really interesting as it has to be optimised to the markets, so will pick up variations in the market environment.

Keep posting, please. :)

[Harre]

AI entered a short earlier today and currently floating at 95 pips profit. Tempted to take it of course but will let it run this time.

[Harre]

Exit on the eurusd short a while ago for 70 pips profit so back to 0 in total.

[Harre]

The AI entered a long on eurusd last night 22.00 GMT. Don't remember the exact entry but think it's up about 50 pips at the moment. Will update when I get home.

Edit: Locked in 150 pips profit on it cause I don't want risking it over the weekend. It's still in action tho.../Edit

Edit2: Trade exit for 150 pip profit so approx 150 dollars up on this EA live. /Edit2

[Harre]

Have had one SL and one break even trade since the last one so EA currently up 60 since live start. Currently in a long on eurusd which is 95 pips up and has SL at break even. They've released a new version which has option for trailing stop which I'll be installing tonight.

[clolab]

ok interesting.

[Harre]

The previous long came back to break even so still 60 up in total. I've now updated to v3 which has trailing stop available and currently in a long at 1.2858 on eurusd, at the moment 41 pips up with SL at b/e. Another decent entry to test the support.

[Harre]

Another long entered at 1.2813 and currently locked in 70 pip profit, still open.

I think it makes sensible entries and now with the trailing stop updating every tick instead of every 4 hours it will require less manual intervention. I've personally gone in and changed SL levels to break even a couple of times before but should now be able to leave it alone.

EDIT: Closed at 70 pips profit. In total 130 pips up since going live. /EDIT

[Harre]

Quick update. I didn't have the AI running earlier in the week which was a pity since it would have entered a short which would have come in nicely.
It made 60 pips earlier today and has now entered a long on eurusd at 1.2531, apparently expecting support and rebound.
In total 190 pips plus since going live.

[Harre]

Forex-ai exited the buy for a profit of 189 pips and have now reversed the position and shorted eurusd at 1.2720. Total profit since I went live now 380 pips. Enjoying this more and more :-P

[JillyB]

Forex-ai exited the buy for a profit of 189 pips and have now reversed the position and shorted eurusd at 1.2720. Total profit since I went live now 380 pips. Enjoying this more and more :-P

Since you got version 3 things seem to have picked up with this EA, Harre. The trailing stop really appears to be locking in profit now instead of letting the price reverse on you. Are you happier with the latest revision?

[Harre]

I am happier yes, even though backtests show that not using trailing stop makes more money. I'm cool with smaller but more consistant profitable trades and you can select which setting you feel comfortable with.
The improvement in results I think are mainly due to using 6 months data in the optimisation rather than 3-4 months like I did previously. The entry points have been superb lately and today's trades spot on. It rode the retracement right up to the end where it exited and then entered a short exactly at the right time (66 pips plus as I type).

Still way too soon to give a final verdict of course but I really like the attitude these guys maintain, actively seeking input from the traders and refrain from "A MILLION IN A WEEK" crap. I'm not going to recommend it to anybody but if the results keep up I wouldn't mind supporting these guys with a little affiliate blog somewhere :-P

[Harre]

Just an update. The short is currently 143 pips in profit with 100 locked in so an outstanding week for the AI.

[JillyB]

Hi Harre,

On the Forex AI homepage I notice it says $99 now "and If I don’t make at least $5,000 within two months after trading on my live account, I will not pay the balance of $200. In fact, I will owe Gary White nothing and I can keep my copy of the Forex-AI AutoTrader."

Have you had your Forex AI for more than two months yet? Have you been asked for $200?

[Harre]

I haven't had it for 2 months and still only trading 0.1 lots so doubt the 5k barrier will be broken (although you never know). If I did make 5k I certainly wouldn't have a problem paying 200 of them to Gary for his efforts :-)

The previous trade did exit for 100 pips profit making a grand total of 480 so far and it's entered a new short which is currently 40 pips in profit.

[JillyB]

I haven't had it for 2 months and still only trading 0.1 lots so doubt the 5k barrier will be broken (although you never know). If I did make 5k I certainly wouldn't have a problem paying 200 of them to Gary for his efforts :-)

The previous trade did exit for 100 pips profit making a grand total of 480 so far and it's entered a new short which is currently 40 pips in profit.

Well I've taken the plunge and bought the EA. I paid via Paypal, but how do I get to download it? I've not had any email from Gary or his team. Does it take a while for them to get in touch?

I have to admit that my internet explorer did shut down right after I'd paid, so if there was a link to a webpage, then I didn't get it. Do you think I should email him?

[kostolany]

Well I've taken the plunge and bought the EA. I paid via Paypal, but how do I get to download it? I've not had any email from Gary or his team. Does it take a while for them to get in touch?

I have to admit that my internet explorer did shut down right after I'd paid, so if there was a link to a webpage, then I didn't get it. Do you think I should email him?

Hi JillyB,

you can register a new membership account on their homepage with your paypal paymentID. After some hours you will receive notification of your access and you can download the EA from the membership area.

[JillyB]

Hi JillyB,

you can register a new membership account on their homepage with your paypal paymentID. After some hours you will receive notification of your access and you can download the EA from the membership area.

Thanks for that. I'll do that now and hopefully I can get the EA set up for use on Monday.

[chrisandrews12]

Well I've taken the plunge and bought the EA. I paid via Paypal, but how do I get to download it? I've not had any email from Gary or his team. Does it take a while for them to get in touch?

I have to admit that my internet explorer did shut down right after I'd paid, so if there was a link to a webpage, then I didn't get it. Do you think I should email him?


Hi JillyB,

I am also interested in this ea but have been in contact with Alpari UK, my broker, and they said they don't have info on oil or gold or deal in those comodities, so does that mean I cannot use the ea.... or is there a way around this.

I would be grateful if you could advise me of your broker so that I can move over to them.

Also, a big thank you to Harre for introducing this ea to the forum

Chris

[Harre]

It is possible to use other indicators but oil and gold are the preferred ones. Sorry I don't recommend brokers anymore after having been stung by Fxpro. Just have a look around, most brokers do at least gold.

Don't thank me :-S At least not until you've made a truck load :-) I don't endorse or recommend anybody to buy this EA, I'm just trying it out like you and so far like what I see.

Last trade stopped out for -41 pips btw so total is about +440 since going live.

[chrisandrews12]

It is possible to use other indicators but oil and gold are the preferred ones. Sorry I don't recommend brokers anymore after having been stung by Fxpro. Just have a look around, most brokers do at least gold.

Don't thank me :-S At least not until you've made a truck load :-) I don't endorse or recommend anybody to buy this EA, I'm just trying it out like you and so far like what I see.

Last trade stopped out for -41 pips btw so total is about +440 since going live.

ok - so if I wanted to try something other than gold or oil then do you know what i should go for....

[Harre]

Well if you look at their live results page they've started a new account which uses AUDUSD as the secondary indicator. It's only been going for a week I think so no telling if it's as effective as oil or gold.

[JillyB]

Hi Harre,

How long does it usually take for you to do the optimization? I've set it going today and it tells me it will take 223 hours and 5 minutes. This is a week and a half!!! Does it normally take this long?

[seabee40]

Mine takes about 3-5 min.

[JillyB]

Mine takes about 3-5 min.

Are you using 'open prices only'?

I was using 'every tick' as it says

"If you are using “T” Trailing stop loss instead of “P” Periodical, you should
change the model to “Every Tick” otherwise the training will not be accurate.
It will take the training much longer than a few minutes. We usually train using
the Periodic Stop Loss to speed things up."

[seabee40]

I'm using everything default, download, optimize and to the chart. The only thing I am doing is trying different alt pairs. XAU/USD AUD/USD USD/CHF I have 3 charts running at once. I have a buy in USD/CHF and sells in the other two.

[Harre]

Are you using 'open prices only'?

I was using 'every tick' as it says

"If you are using “T” Trailing stop loss instead of “P” Periodical, you should
change the model to “Every Tick” otherwise the training will not be accurate.
It will take the training much longer than a few minutes. We usually train using
the Periodic Stop Loss to speed things up."

Oh... I hadn't seen that and I'm still just doing the open bar one. Hmm, will have to look into that.

Btw, it opened a sell this morning 1.2968 currently up 150 pips :-)

[Harre]

....and its weekly/mothly "training" is probably nothing more than curve fittng

Another intelligent comment Miksa... Curve fitting is applied to make back testing show brilliant results. Forex-ai don't show back test results, only live forward testing. The optimisation is done on 6 months data and updated weekly or bi-weekly to AVOID curve fitting.

Since you "think" and have your opinions without any idea what you're talking about allow me to have a little ponder:
I think you're out of work, in a trailer somwhere wondering where it all went wrong. Where did you lose everything? You've got f*ck all money to even use on your own magnificent EA's which are nothing but half finished stuff you found on various developing forums, filled with rip off trojans and viruses. I feel sorry for you.

[JillyB]

Oh... I hadn't seen that and I'm still just doing the open bar one. Hmm, will have to look into that.

Btw, it opened a sell this morning 1.2968 currently up 150 pips :-)

Hi Harre,

I decided that I would do the optimising using 'open price' only - I couldn't stand the wait for it to complete the testing using 'every tick' mode.

I started the test parameters from 1 Sept 2008 - today. I've run them several times and get different results each time. Have you had the same experience?

[Harre]

Yeah, the results differ slightly. There's a thread about it on the Forex-AI members forum where Gary explains why. If you can't get to it I'll do a little digging when I get home tonight.

183 pips up on the short now with just under 100 locked in :-D I sincerely hope you guys get a good start with it cause the last week or so has been ridiculously good.

[JillyB]

Yeah, the results differ slightly. There's a thread about it on the Forex-AI members forum where Gary explains why. If you can't get to it I'll do a little digging when I get home tonight.

183 pips up on the short now with just under 100 locked in :-D I sincerely hope you guys get a good start with it cause the last week or so has been ridiculously good.

Hi Harre

I'll go and have a look for the thread on the members forum.

I think I must have missed that trade as I've entered the settings, but it hasn't opened a trade. I will just have to be patient. :) So far it seems really good and I like the way it keeps moving the sl to reduce the amount you're exposed to. Once you're in profit it keeps locking in more and more as it goes.

At least with your open trade Harre, you know you will take some profit on it. There is no way that it can give back all those profits and turn negative, because the EA has kept moving the sl down to protect your profit.

I have to say that I was very impressed with the manual. It's very detailed and professional. I've seen so many of these that are a just reguritated, repeats of the same thing and have about 2 pages of information that you actually need. But this is different - even down to the handy hints.

The question and answer section at the end is good too and answers some very particular problems - one of which I happened to be having trouble with too.

[Harre]

Yeah I prefer the trailing stop even though it will cause me to miss a few of the big winners.
Agree tha manual was good. In general I think they're doing a good job and have been very professional so far. Having spent such a small amount on the EA my expectations were low but I have been very impressed with how they treat their customers. Of course it's easier to do that when you're starting up and there's not 1000 emails with questions every day...
The EA itself is just plodding along nicely and most entries have been good but obviously way too early to tell if it's going to be a consistent winner. So far so good is all I can say and at least here the results they get and I get are very similar :-)

[Harre]

Ok, closed the trade manually for 240 pips. That's 680 up since going live. Why did I close it manually? Because I am over the moon with it and there's no need to get greedy. It's found a bit of support and it's a coin flip which way it goes from there :-) If it breaks through the support there's another 190 pips or so to be made but there will be no tears from me if that happens :-P

[JillyB]

Ok, closed the trade manually for 240 pips. That's 680 up since going live. Why did I close it manually? Because I am over the moon with it and there's no need to get greedy. It's found a bit of support and it's a coin flip which way it goes from there :-) If it breaks through the support there's another 190 pips or so to be made but there will be no tears from me if that happens :-P

Well done Harre,

Glad to see it's working for you. I'm still waiting for my first trade to kick in. Hopefully it'll happen today or tomorrow.

[Harre]

Got a short this morning at 1.2740. Not optimal perhaps but it may well run down somewhere in the 1.25 area. Can afford a few SL hits now if needs be :-)

Gl Jilly

EDIT: AI stopped out for -57 and has entered a new short at 1.2789 . 620 up since live /EDIT

[clolab]

Hi Harren,

has this versione, you are using now, a trailing stop featuere?
If yes, can you set trailing stop yourselve?

Thanks

[JillyB]

(Sorry, I typed this at 8.30am this morning, but forgot to press the submit button)

Hi Harre,

First trade opened at 4.00am. Sell ????? It's way under water at the moment, which I would expect from looking at the charts - good doji/pin-bar at the bottom of the trend and start of a new upswing? Looking at the charts it was the perfect buy setup not a sell.

Has it opened any trades for you?

[JillyB]

Hi Harre,

Yes I had the short at 1.27027 and got stopped out for -91 pips. Not a very good start to the system, but it's early days.

It's now in another short trade entered at 12 noon at 1.28280 which is in profit and the stop loss has been moved down to lock in 21 pips - so at least I know it's not going to be another losing trade.

[Harre]

Hi Harren,

has this versione, you are using now, a trailing stop featuere?
If yes, can you set trailing stop yourselve?
Thanks

Yea it has trailing stop which you can switch on if you want to but there's no "settings" for it. It will simply follow the trade from the start with a 90 pip gap.
I am using it and I've been pleased with it so far.

[JillyB]

Harre,

I can't help but wonder if the system took the sell trade this morning because neither of us were in a trade, where as in actual fact it would still have been in an earlier sell trade and this one wouldn't have triggered (if you understand what I mean).

I didn't have a trade because I had just set up the system and you had got out early on the existing sell trade. What do you think?

[JillyB]

Last trade closed for +21 pips - so totally the system is -70 pips for me.

It's now opened another sell trade at 1.28455 - however the sentiment does seem to be definitely against it at the moment, so I am expecting another losing trade here.

[Harre]

Proper crap day for the AI which has been caught in the see-saw. I've got -57, -4 and -83 for a total of -144 pips. Approx 530 profit since going live.

[agathos23]

I took forex AI off of my live account because it was losing a bit more than winning. I left it running on my demo account for now (and that one is currently running at a slight loss but close to getting profitable).

Edit: hey guys/gals: As Jilly pointed out, the trailing stop training takes a long time and if I let it run, it'd run for a week. Do you train the EA with the Periodical stop but run with the trailing stop in the live settings?

[JillyB]

Hi Harre,

If you're around could you help me with some questions please.

I'm still a bit perplexed with the weekly training.

1. Still unsure if I set the optimization going using 'every tick mode' or 'open price only'? In the manual it suggests using 'every tick mode' if you are using a trailing stop - but as this can take well over a week to complete using 6 months of data I have used 'open price only'. However this is not representative of what we are trading.

2. If I run optimization on 'open price only' and do this 5 times - I get five different results. Why - any idea?

3. Of these five different results which one do I chose? Is the first one the most accurate; or do I just chose the best result and use those?

4. However if I then run a back-test over the last six months using the best results from the optimization - but change 'model' to 'every tick mode' from 'open price only' I get totally different results (obviously). But the best results from the optimization can lead to losing results on backtest. Is this right? Have you had this?

5. So, therefore, do I take the 5 results of optimization and backtest all of them and then chose the best result from the backtest? This is what I did for this week, but the results are not good.

I just ran the optimization again (first run since the weekend) from 1 Sept 2008 to 21 Feb 2009 and produced a result with it. I then ran a backtest from 23rd Feb 2009 to to today and the backtest to today was good.

I then ran the optimization again - same parameters - and got a second set of results. I then ran these on the same backtest from 23rd Feb 2009 to to today and the backtest was terrible - losing trade all round.

This just doesn't make sense.

Does Harre or anyone with experience in optimising have any ideas on this?

[Harre]

Hey Jilly. I suggest you post this question on the forex-ai page as Gary and co normally answers really quickly [EDIT: I see you already have, good girl cause I'd like to hear as well /EDIT]. That the results are slightly different I guess is because there's probably some function built in to prevent entries from being the exact same time. FAPT has a little randomizer like that to avoid causing x thousand entry signals the exact same tick.
I haven't done the optimisation for every tick myself but will start one tonight on my VPS since I'm off skiing tomorrow. I'll let it crunch some numbers and see what happens.
So basically I am using the previous optimization I did but with trailing stop at the moment which isn't ideal even though it has been profitable.

Of the optimization results you get you'll find that a lot of them give the same low draw down. Select one with low draw down and good profit and give it a go. Chances are nobody is using the exact same settings but the logic behind it is still sound. 4 indicators on the pair you're trading and 4 indicators on the alternative pair which has to confirm the trade.

[JillyB]

Hi Harre,

If you're around could you help me with some questions please.

I'm still a bit perplexed with the weekly training.

1. Still unsure if I set the optimization going using 'every tick mode' or 'open price only'? In the manual it suggests using 'every tick mode' if you are using a trailing stop - but as this can take well over a week to complete using 6 months of data I have used 'open price only'. However this is not representative of what we are trading.

2. If I run optimization on 'open price only' and do this 5 times - I get five different results. Why - any idea?

3. Of these five different results which one do I chose? Is the first one the most accurate; or do I just chose the best result and use those?

4. However if I then run a back-test over the last six months using the best results from the optimization - but change 'model' to 'every tick mode' from 'open price only' I get totally different results (obviously). But the best results from the optimization can lead to losing results on backtest. Is this right? Have you had this?

5. So, therefore, do I take the 5 results of optimization and backtest all of them and then chose the best result from the backtest? This is what I did for this week, but the results are not good.

I just ran the optimization again (first run since the weekend) from 1 Sept 2008 to 21 Feb 2009 and produced a result with it. I then ran a backtest from 23rd Feb 2009 to to today and the backtest to today was good.

I then ran the optimization again - same parameters - and got a second set of results. I then ran these on the same backtest from 23rd Feb 2009 to to today and the backtest was terrible - losing trade all round.

This just doesn't make sense.

Does Harre or anyone with experience in optimising have any ideas on this?


Gary's reply to my questions

"Hi,

1. We use "open price only", because "every tick' would take forever" Open price only is not accurate when you use "T" trail S/L, so we change it to "P" during optimization. The point for the optimization is to find good parameters for the NN to enter trades. The exits are not optimized if you use P for optimization then change it to T during trading. We only just started using the Trailing S/L for a few weeks as this is a new feature. Our past results were achieved using Periodic S/L.

2. There are over 1 billion combination for the optimizer. When you start the optimizer using genetic algorithm, it will pick an initial set and keeps tuning that for about 10,000 times. When you run it again, it will pick another set and keep tuning that out of the possible 1bill combination. The results are always going to be different because we only test 10K out of 1Bill. It is OK, because we want the NN to get general idea how prices have moved in the past 6 month. It is not an exact science.

3. You run it first and have a look at the results. If you have a freak good result say 5%DD and 50K profit and the other results are 8-9%dd, you don't pick the best result. Run it again until you find several good results. Then pick one from the lower range of the DD and the best profit. Look also the Profit ratio. If it is more than 4, I usually don't pick it. It means it had a couple of good trades and maybe a lot of losing one. We want the EA to pick good entries most of the time, not shooting for a couple of large profitable trades and many losing ones. We typically write down the best result (if any) from each run and pick the best one out of the 3-5 runs.

4. We run the backtest for January 2009 on the account we achieved 65% profit and used the same parameters we have used for live trading. The "backtest" showed a break-even instead of the 65% profit we have achieved. The MT4 always looks at the worst case scenario and often report incorrect results. Even when you use "every tick" it actually uses minute chart not tick. It only uses the minute chart however if you have 6 month worth of minute data available. Most brokers however only allow about 1-2 month 1 minute data. You can download 1 minute data from Metaquotes through the history center, but that data will be very different to your brokers data. To cut the long story short backtesting that way is very unreliable. Pick the results based on point 3, not based on how it worked based on backtesting.

5. No. Pick the best based on DD and profit.

Gary"

[carloss]

useful reading for those who want to see how this EA works in the background:

http://translate.google.com/translate?u=http%3A%2F%2Farticles.mql4.com%2Fru%2F 289&langpair=ru%7Cen&hl=ru&newwindow=1&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools

[JillyB]

I'm going to take this as my first real week trading the Forex AI, because last week I still had questions about the optimising and I didn't start it running at the beginning of the week.

So this week I set it going last night and it opened a trade at midnight (11am GMT) - it uses broker time, not GMT. First trade was a sell trade which closed this morning at 8.00am for +13 pips. I was interested to see that it actually closed the trade itself and reversed the position to a buy. It didn't let the sell trade get stopped out and then reverse to a buy. So now I am in a buy trade and very slightly under water with it, but not by much. I will update the thread during the week as the trades progress.

[agathos23]

I'm going to take this as my first real week trading the Forex AI, because last week I still had questions about the optimising and I didn't start it running at the beginning of the week.

So this week I set it going last night and it opened a trade at midnight (11am GMT) - it uses broker time, not GMT. First trade was a sell trade which closed this morning at 8.00am for +13 pips. I was interested to see that it actually closed the trade itself and reversed the position to a buy. It didn't let the sell trade get stopped out and then reverse to a buy. So now I am in a buy trade and very slightly under water with it, but not by much. I will update the thread during the week as the trades progress.

I didn't get the EA up and running until this morning. I was late in sending the e-mail to obtain the permanent license. It's up now. Let's see how it does. I'm using the trailing stop method this time around.

[JillyB]

Well the buy trade it entered yesterday morning ended up being stopped out yesterday evening. The darned thing is that it was stopped out by 6.6 pips before the price changed direction and headed north! So far this week I've had a +£8.91 and a - £46.56. Total for week is -£37.65. That's letting it trade at 3% risk on a £4,400 account. It choses the lot size and it's currently at 0.1 lots.

There was a trade last week which did exactly the same thing and I ended up having a losing trade. The AI was on a sell trade and I got stopped out (this time by only 2 pips) before price turned south and would have made it a profitable trade.

I have to question the fact that twice now the AI has got out at the top (or bottom) of the market. This is totally due to the trailing stop loss which is set at 90 pips - and I wonder how arbitary is this? From what I've seen so far 100 pip trailing sl would have turned two losing trades into winning ones. Question is - where do you stop? 100 pips is just as arbitary as 90 pips and there are bound to be trades that only just get stopped out when you use this amount.

The system is now on a sell trade which it opened at 8.00am. However I'm not altogether comfortable with this as when price headed north (just after I got stopped out last night) it also took out a rising trend line. And the logic in this EA is.........??

[Harre]

I was away skiing and missed out on sending their support email about license. Should have it tonight so will start it up then. I had left optimization with every tick running while I was away but it was only 30% through... I will use one of those value sets to give it a try but if the result isn't great I'll just go back to normal optimization.

The trailing stop is an optional setting and profit is higher without it in the back tests I've done. It was added because it was a feature request. I prefer it at the moment because I am risking about 9% instead of 3.6 so anything that can keep the draw down to a minimum is fine by me, even if it decreases the profit slightly.

As for the logic of the EA your guess is as good as mine. If both the currency pair and the alternative indicators point in one direction, that's the trade it will make. I guess with the settings you've got it sees a downward trend continuing. You can see the two numbers at the top of your chart. If both numbers are positive it will buy at the next opportunity. If both numbers are negative, it will sell. If there's one of each it will do nothing except obviously use SL and TP. If you're in a trade and both indicators change to the opposite, i.e. if you're in a long position and both indicators are negative, that's when you will see the EA exit the current trade and enter a new one in the other direction. Only happened to me once so far.

The only time this EA seems to hurt is when we're in a seesaw day that's big enough to take out the SL's. There's something being developed to cope better with these days I read on the forum.

[JillyB]

Hi Harre,

Hope you had a great time skiing - never been myself; far too cold! Brrrr!!! Given a choice I'd much rather be lounging on a caribbean beach.

I stand corrected about the EA with this mornings trade. It did know what it was doing and I did manage to leave it alone, so it's still in the trade and has now locked in 36.2 pips. So at least I know that this trade can't be a losing one.

Take your point about using the trailing stop Harre. I would imagine that not using it is more profitable, but I don't think the swings from +ve profit to -ve profit would do my head in slightly. I'd much rather get so far with a trade and know that I've locked in at breakeven - then I'm on a free trade and it can do what it wants as far as I'm concerned.

As you say though, just having a stop-loss that doesn't move (or does it at the start of the next 4 hour candle?) would have meant those previous 2 losing trades I had, would have ended up in profit.

[Harre]

Another SL hit today for -87 so approx 440 pip's up since going live. A slightly optimistic entry this morning I thought. Even with a downward trend it looked like it was likely to get a bit of a retrace upwards.

[JillyB]

It looks like we have different trades triggering on the AI, Harre. My version entered a sell trade yesterday morning at 8.00am and was closed out via the trailing stop at 1.07pm today for +£72.52. Which put's it up on the week and slightly down overall (last week was a bit of a disaster as I kept messing with it).

Since it closed the trade out earlier today it hasn't entered another one and the settings are showing -ve for eur/usd and +ve for gold, so it won't enter until they are both the same.

[Harre]

...and was closed out via the trailing stop at 1.07pm today for +£72.52. ....

Sweet, glad to hear it's starting to make some dosh for you. I'm going to re-do the optimization this weekend since I am currently testing a set I got from the optimization with every tick that I did last week but only completed about 30% of. Will go back to the bar entry one cause it's been giving good results so far.

[JillyB]

Sweet, glad to hear it's starting to make some dosh for you. I'm going to re-do the optimization this weekend since I am currently testing a set I got from the optimization with every tick that I did last week but only completed about 30% of. Will go back to the bar entry one cause it's been giving good results so far.

Yes that's what I'm doing. One run at the optimisation and if the results it gives are ok - I use them. I'm going to be running it on a weekly basis I think - run it on a Saturday ready for the open on Sunday evening.

[JillyB]

Forex AI opened a sell trade at 8.00am this morning. Entry at 1.25890.

[Harre]

I got my entry last night for 1.2634. Like you I'll run the optimization over the weekend and have been doing so weekly, but I won't start up the AI until Monday morning around 8 GMT. Simply too risky to dabble with the potential gaps the first couple of hours for me to like it. I'm sure there's plenty to be gained from it but needless risk I reckon.

Edit: 124 profit on that one so the total up to approx 560 /Edit

[Harre]

Got a long now at 1.2546

[JillyB]

No trades open for me.

[Harre]

Closed for another 124 pips so total up to 680 in profit :-) Another good week for the AI after the lackluster start.

[Harre]

Got a long this morning when I started it up and it's been closed out for -63. Total profit down to 620. *starting to sing Boomtown Rats: I don't like Monday's*

Edit: Another long entered, this time with a better price at 1.2565 /Edit

[JillyB]

Starting balance for AI for this week +£0.95

00.00 9th March long trade opened 1.26761- closed at 6.16am for -£29.24

08.00 9th March long trade opened 1.26417 - closed at 11.04am for -£48.12

Total for Forex AI so far = -£76.41

Hope it starts to get better soon!

[Harre]

Closed for +40 and total stands at +660.

[Harre]

A short early this morning was stopped out for -62 pips. Total down to +600.

[JillyB]

Long trade just entered at 11.00am gmt at 1.26983.

Not too sure about this one as there seems to be resistance immediately above where price is.

[Harre]

Stopped out again, this time for full SL -92. Profit since start down to 510 and the week not too cool so far.

[JillyB]

Long trade just entered at 11.00am gmt at 1.26983.

Not too sure about this one as there seems to be resistance immediately above where price is.

This one did move in my favour - so perhaps I should take it all back.

The AI has now locked in profit - so I'm on a free trade, whatever it does now this one can't be a losing trade for me.

[JillyB]

Trailing stop took me out at +33.4 pips for +£20.04

Total for Forex AI so far = -£76.41 + £20.04 = - £56.37

Still in negative territory - so I'm not convinced as yet. I thought after two weeks I would at least be showing a small profit. Hmmmm........

[Harre]

Still in negative territory - so I'm not convinced as yet. I thought after two weeks I would at least be showing a small profit. Hmmmm........

Short term doesn't interest me at all. When I started off with this on demo I think I had 4 losing trades on the trot before it started to get some back. Look at their live results, they've been pretty poor lately but over longer term it's paid dividends. In 2 months time, when it's been running for a quarter, I'll make my first "review" of this EA, not before. As far as I'm concerned it may still be a waste of time, or it may be a sound system generating nice steady profit, no way of telling yet. One thing is for sure, I have more belief in this than I ever had with FAPT!

[JillyB]

Sell trade entered this morning at 12 noon at 1.27100 - Stopped out for -£59.46

Total now since start = -£115.83

Another sell trade entered at 4.00pm at 1.27809, lets see how this one goes.

[Harre]

I'm short 1.2759

[JillyB]

Stopped out of that last trade for -£39.42

Total since start = -£155.25 EEK!!!!
Not good. But one consolation is that the Forex AI chaps aren't doing any better. I see from their live trading results that they've had a string of losses as well.

[Harre]

Stopped out for -78 and profit since start approx 430. Turning out to be a fairly horrible week but nothing much to do when there's no direction and whip saw all around. Personally I am looking at shorting when I've got confirmation that eurusd has hit the resistance.

EDIT: Seems like the AI decided to do it before me... New short this time at 1.2843 which seems reasonable. Would be nice to recoup some of this weeks losses before the weekend :-) /EDIT

[JillyB]

Sell trade entered at 4.00am this morning at 1.28240.

Stop is now at +1.2 pips - so again I know that this can't be a losing trade.

[Harre]

Mine was stopped out for a couple pips profit but am now in another short which has locked in some profit and is currently standing at 95 pips up. Let's see where that ends

[JillyB]

Mine was stopped out for a couple pips profit but am now in another short which has locked in some profit and is currently standing at 95 pips up. Let's see where that ends

Yes I got stop out for 0.8 pips and the grand total of +£0.58.

Total since start is now -£154.67.

No further trades opened for me. Eur/usd is -ve and Gold is +ve at the moment.

[Harre]

Got a short at 1.3015 which looks like a decent entry

[agathos23]

I've suspended trading this EA for now. When I get back with it it'll be exclusively in demo. I haven't had the time to run the optimizations and I've been playing around and testing other systems.

[Harre]

Since last update I've had 2 small profits and 1 loss and still +430 pips since live.

Currently in a short at 1.3029

Edit: Stopped out for -30 and down to +400 since start.
New short at 1.3117 /Edit

Edit2: That got utterly screwed by the fed's decision to pump in 2 trillion dollars in the economy. The dollar just disappeared in a black hole.... Only up 310 since live now. Edit2

[pavlus]

Hello,guys.I'm using ForexAI for a month already and getting pretty much the same trades as you using GOLD like alt.symbol.Since monday I've started another account with forexAI using USDCHF instead of gold.So far there were only 2 trades : one short for -55.00 and....one long gor +403.:D Account with gold didn't catch this buy trade.Maybe It was just luck?Who knows.I will continue testing account based on USDCHF.

[Harre]

Been away a few days and yesterday I didn't have the AI running. Switched it on this afternoon and just had a short exit for 40 pips. In total up 350 since going live. Hope to get back to winning ways after last week which was expensive.

[pavlus]

I have e-mailed Gary(author of EA) about these bad results we have with ForexAI for few weeks and he said he is working on new version to be profitable in such conditions.New version will be available in 2 weeks or so.
Let's wait and see.

Update:pre-release of new version is available

[Harre]

New version is out so I'm suspending trading until I've downloaded it and done the optimization. Just skimmed the email but it sounded like a very good improvement :D

[Harre]

Had some probs with the pre release v4 so going back to v3. Got a long at 1.3246 which looks promising at the moment 55 pips in profit.
The theory behind v4 sound ok, just needs sorting out and the optimization needs clarified. When I tried it I just got a lot of small SL's and then it missed the big move last week.

EDIT: Pity that version 4 isn't quite ready for live yet cause it would have closed this trade for 70-80 pips profit instead of 4 measly pips... /EDIT

[Harre]

One trade last night which closed 40 pips profit so 390 up since live.

Still using v 3 since I haven't had time to optimize and backtest v4 properly yet.

[JillyB]

One trade last night which closed 40 pips profit so 390 up since live.

Still using v 3 since I haven't had time to optimize and backtest v4 properly yet.

Hi Harre

Are you still using Gold as the ALT symbol? I had a try with usd/chf, as the correlation seemed higher, but I didn't really get any trades triggering.

[Harre]

Yup still using gold. Got a long with 120 pip profit locked in. It opened Friday and normally I would have closed it before the weekend but forgot :-)

[Harre]

The long went in for 120 and after that a loss for 34 and gain for 27 so all in all about 500 profit since going live. Still using v3 since I haven't had time to optimize v4 properly and give it a test run. V3 is doing ok tho so no rush.

[Harre]

I've had one loss for 39 and one profit trade for 45 so still about 500 up in total. Need to get v4 optimised and in demo at least soon.

Edit/ Just had a loss for 50 pips so 450 total profit. /Edit

[Harre]

Haven't been updating over Easter but 5 trades have been executed since last time.
3 losses of 50, 39 and 42 pips about 130 all in all, and
2 wins of 189 and 79 pips about 270 all in all.
That puts the total profit since live up to approx 590 so things are looking rosey again :-)
Still running v 3 and considering changing to the set SL rather than trailing. On a number of occasions lately the trailing stop has been hit killing what would have been a profitable trade. I will re-run the back tests but if I remember correctly they also showed a better profit for the fixed SL compared to trailing. I did choose to use the trailing stop to lower the risk in the beginning but wonder if it's time to revise that decision.

[Harre]

2 losses 94 and 56 pips and 1 win of 79 pips. In total about 520 up since live. Still running v3 with trailing stop

[jhansen]

2 losses 94 and 56 pips and 1 win of 79 pips. In total about 520 up since live. Still running v3 with trailing stop

I've been following this thread a while...

No recent posts from anyone that actually purchased the system (?)

Is Forex AI just a fly-by with losses equal to profits or have you guys been busy with other things?

I'd love to see a live MT4Stats from a live account running this system profitable :D

Kind regards

Jhansen

[Harre]

I have been away quite a bit but recently started using v4 on my live account. Haven't had a trade since though so need to check why that is. If you want live results have a look at the AI results page. It displays all their trades and unlike FAPT they don't all of a sudden remove an account when things go pear shaped...
I think they've made less profit recently than before but still doing ok.
Jilly, you still using this?

[jhansen]

I have been away quite a bit but recently started using v4 on my live account. Haven't had a trade since though so need to check why that is. If you want live results have a look at the AI results page. It displays all their trades and unlike FAPT they don't all of a sudden remove an account when things go pear shaped...
I think they've made less profit recently than before but still doing ok.
Jilly, you still using this?

Hi Harre:-)

The EA seems interesting enough, I just wish that there was a little more "solid" proof that Forex AI does actually make more than it loses (profits) over a longer term...

I guess maybe Jilly has moved on since this thread seems to be pretty inactive - I had somewhat high hopes for this ea to have something that trades during day (GMT time)

How long is it since you had trades with this beast, Harre ?

Kind regards

Hansen

[Harre]

My problem is that Startforex decided to put up the minimum deposit for EA trading to $1000 and without telling me. This despite the fact that they agreed to allow me a deposit of $600 when I started. This is the reason why my Forex-ai hasn't been trading.

I've withdrawn what I put in so what's in my account is pure profit but I can only trade manually until I fund my Broco account.

Look at the results on the forex-ai page, then try it yourself. There are no fake results there.

[jhansen]

ok thank you harre,
i'll consider it:-)