Art Phantom EA | Page 29 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phantom 6 System 1,540 replies

Phantom 6 - Manual System 8 replies

Phantom Entry on ODL Japan. 2 replies

Tradeview CAD/JPY "Phantom" spike 0 replies

Phantom high, phantom low 0 replies

Attachments: Art Phantom EA
Exit Attachments

Art Phantom EA

  • #561
  • Nov 27, 2010 11:04am Nov 27, 2010 11:04am
Quoting roundrock
i rewrote the code so that we can backtest. this is a single pair phantom strategy, you can drop to any chart any timeframe. it will trade that pair and that timeframe. if you drop it on EU 1 Hr chart, then it will trade EU pair and adds are done on start of the hour. if you drop it on 30min chart then it trades every 30 min at start of bar

when you start EA , it remembers the current price. then at start of the next bar, if the pair has moved up (with respect to start price), and if all prev Long trades are in profit then it will add a Long....
And for Forwardtest on Monday is it better this version or your last one?
  • #562
  • Nov 27, 2010 11:10am Nov 27, 2010 11:10am
Quoting roundrock
i rewrote the code so that we can backtest. this is a single pair phantom strategy, you can drop to any chart any timeframe. it will trade that pair and that timeframe. if you drop it on EU 1 Hr chart, then it will trade EU pair and adds are done on start of the hour. if you drop it on 30min chart then it trades every 30 min at start of bar

when you start EA , it remembers the current price. then at start of the next bar, if the pair has moved up (with respect to start price), and if all prev Long trades are in profit then it will add a Long....
I just tried it now backtest, it is not working, it is not closing all the trades when it has to, just randomly
  • #563
  • Nov 27, 2010 11:16am Nov 27, 2010 11:16am
Quoting vanupips
I just tried it now backtest, it is not working, it is not closing all the trades when it has to, just randomly
before you start backtest, you need to hit F3 and delete the global varaibles. it closes when the PL reaches the PP.

you can see the print statements in journal.

only the last cycle will be closed by the backtester as it reaches the end of testing period
  • #564
  • Nov 27, 2010 11:18am Nov 27, 2010 11:18am
Quoting Caillou
And for Forwardtest on Monday is it better this version or your last one?
i would suggest this one, as its more convinient and we can try different TF and different pairs.

my initial testing reveals 30min is better than 1 hour

i tested this for full year 2010 on H1, it survived !!! it ended up 70% profit
  • #565
  • Nov 27, 2010 11:25am Nov 27, 2010 11:25am
Quoting roundrock
i would suggest this one, as its more convinient and we can try different TF and different pairs.

my initial testing reveals 30min is better than 1 hour

i tested this for full year 2010 on H1, it survived !!! it ended up 70% profit
I imagine you run for backtest 24h with above settings? So if we load on M30 it checks at min 0 and min 30?
On H1, not the better TF and without any mod it recovers any DD during 2010? Veeeeeery promising results
  • #566
  • Nov 27, 2010 11:27am Nov 27, 2010 11:27am
Quoting Caillou
Hi Roundrok, excellent work as always. I' m not at home now. Do you have more results: 6 months, this year, ....

Would it be very hard to code with Rate and Charvo' s mod in order to backtest it?

Thank you
%

Caillou, my focus right now is to have a basic EA with original idea that we can backtest to see if it survives !!! once we prove that it can survive, then we can add tweaks to improve profitability. so lets backtest the basic idea and coemup with results

if it survives, i would be happy with a consistent 5% monthly return (without any fear of losing the account) with simple system rather than trying for a 10% return with tweaks that may be complex and/or inconsistent

my immeadiate goals are

1. prove it survives

2. arrive at optimum settings (MMTU, PP for given leverage, account balance)

3. build confidence in the system to go live (after cleaning up the code)
  • #567
  • Nov 27, 2010 11:33am Nov 27, 2010 11:33am
Quoting roundrock
before you start backtest, you need to hit F3 and delete the global varaibles. it closes when the PL reaches the PP.

you can see the print statements in journal.

only the last cycle will be closed by the backtester as it reaches the end of testing period
did it , still not working
  • #568
  • Nov 27, 2010 11:34am Nov 27, 2010 11:34am
Quoting roundrock
%

Caillou, my focus right now is to have a basic EA with original idea that we can backtest to see if it survives !!!...
Absolutely agree. Let' s test the pure version and later, with optimum MMTU and PP we could add mod if needed.....I know it is early yet but as I said before the recovery potential of this EA is amazing....
  • #569
  • Nov 27, 2010 11:36am Nov 27, 2010 11:36am
Quoting Caillou
I imagine you run for backtest 24h with above settings? So if we load on M30 it checks at min 0 and min 30?
On H1, not the better TF and without any mod it recovers any DD during 2010? Veeeeeery promising results
as i mentioned i tested on hourly chart and M30 Chart. this EA places orders at start of the bar. that means for M30, at 0, 30 min. on hourly chart at 0 min

regarding recovering DD, i didnt analyse the results in deep. but it survived and closed the year in 70% profit. we can see the report and figure out how many max trades between each pop. lot of times its around 10 trades. but since some are long and some are short the total DD is not much. you get max DD when the current price is near the start price.
  • #570
  • Nov 27, 2010 11:42am Nov 27, 2010 11:42am
Quoting vanupips
did it , still not working

can you try for just 2 days ? and then see the jornal, report and see whats happening ?

when you say not closing, what does it mean ? is it not closing when it hits PP ? or when it hits PP, its not able to close all the trades ? lets say we have 10 trades when it reaches closure, instead of closing all 10 trades is it closing only few of them ?

pl give more details. pl note after a pop, it places 1 buy, 1 sell at same price ... dont know why but thats what they mentioned. pl give more details..
  • #571
  • Nov 27, 2010 12:02pm Nov 27, 2010 12:02pm
Quoting zznbrm
charvo,

Please see the attached spreadsheet. First sheet is a summary of the last 3 months. Subsequent sheets are details of a few test cases.

Test criteria
------------------------
- MBTrading; 1.0 lot = 10,000 units; commission = 0.0000295, Leverage = 50
- Starting equity of $10,000 for each basket
- Started new basket at 0200 and 0800 (New York time) regardless of result of previous baskets
- Used 2.0 lots per trade, Emergency SL = $5,000, TP = $50
- New trades will not be opened when Margin Level % is below 200%
- Assumed spread...

Hi Zznbrm,

A few observations:

1.- You mention Charvo twice, this is not Charvo' s thread, this is not Caillou' s thread, it is the thread of every person interested in Art_Phantom EAs, so if you want to speak only to Charvo, please send him a pm.

2.- Obviously, your test doesn't reflect Art_Phantom EAs's operating mechanism, so your results might be better, worst but sure not the same than Art_Phantom EAs so please, don't confuse anyone.

3.- If you have rewrote Pure Version, please share it or otherwise don't sell us something that you are not going to share.

Thank you very much
  • #572
  • Nov 27, 2010 12:30pm Nov 27, 2010 12:30pm
Quoting Caillou
1.- You mention Charvo twice, this is not Charvo' s thread, this is not Caillou' s thread, it is the thread of every person interested in Art_Phantom EAs, so if you want to speak only to Charvo, please send him a pm.
charvo is the one who specifically asked for the test. So, I addressed him but posted here so that others could see the results.

My results show the "problem" areas - the points in time where the strategy has potential for some serious drawdown. I think that can be very useful information. In fact, you can use this information to analyze the market during these potentially bad times in order to adjust the strategy for maximum longterm profits.

Quote
2.- Obviously, your test doesn't reflect Art_Phantom EAs's operating mechanism, so your results might be better, worst but sure not the same than Art_Phantom EAs so please, don't confuse anyone.

There are a quite a few variations of the Art Phantom EAs...too many in fact. I think the thread is quite confusing anyhow. But point taken, I will not post test results again.

Quote
3.- If you have rewrote Pure Version, please share it or otherwise don't sell us something that you are not going to share.

Like I stated, I have every intention of sharing the version I created once the thread goes in 1 direction, not multiple directions like it is now.

Well, now that I think about it, it appears I have overstayed my welcome. I wish you much success with your strategy.

FYI, the opposite of "win" is "lose", not "loose".
  • #573
  • Nov 27, 2010 12:37pm Nov 27, 2010 12:37pm
Caillou, I'm sorry but I do not understand how to load the EA and in those charts...Can you help me ?

Roberto
  • #574
  • Nov 27, 2010 12:53pm Nov 27, 2010 12:53pm
Quoting zznbrm
charvo is the one who specifically asked for the test. So, I addressed him but posted here so that others could see the results.

My results show the "problem" areas - the points in time where the strategy has potential for some serious drawdown. I think that can be very useful information. In fact, you can use this information to analyze the market during these potentially bad times in order to adjust the strategy for maximum longterm profits.



There are a quite a few variations of the Art Phantom EAs...too many in fact. I think the...
This last correction tells me a lot about you.......
  • #575
  • Nov 27, 2010 1:12pm Nov 27, 2010 1:12pm
Quoting Caillou
This last correction tells me a lot about you.......
Glad I could finally post something useful to you.
  • #576
  • Nov 27, 2010 1:19pm Nov 27, 2010 1:19pm
Quoting zznbrm
Glad I could finally post something useful to you.
You are right, FINALLY.......
  • #577
  • Nov 27, 2010 1:41pm Nov 27, 2010 1:41pm
Quoting roundrock
can you try for just 2 days ? and then see the jornal, report and see whats happening ?

when you say not closing, what does it mean ? is it not closing when it hits PP ? or when it hits PP, its not able to close all the trades ? lets say we have 10 trades when it reaches closure, instead of closing all 10 trades is it closing only few of them ?

pl give more details. pl note after a pop, it places 1 buy, 1 sell at same price ... dont know why but thats what they mentioned. pl give more details..
opened 4 pos, then closed 2 of them (PP 0.5)
then opened 6 new pos ,closed 1 of the first and 4 of new for 0.4 p
then opened 2 new pos -not on right time- same time as last was closed,
then closed 3 positions-with loss and opened same time 2 new-...
opened 1 new then closec 3 with 1.8 P
it is just not working properly...
  • #578
  • Nov 27, 2010 2:13pm Nov 27, 2010 2:13pm
Quoting roundrock
i rewrote the code so that we can backtest. this is a single pair phantom strategy, you can drop to any chart any timeframe. it will trade that pair and that timeframe. if you drop it on EU 1 Hr chart, then it will trade EU pair and adds are done on start of the hour. if you drop it on 30min chart then it trades every 30 min at start of bar

when you start EA , it remembers the current price. then at start of the next bar, if the pair has moved up (with respect to start price), and if all prev Long trades are in profit then it will add a Long....
Hi roundrock,cailou,

Thanks for your EA. I shall test this latest version.

I have a question. Can i use this EA in different pairs in same account?

I mean in one 5000 $ account, i will use this EA in euro/usd ( 30min) and euro/jpy ( 30 min ) ? or only in one pair?

Thanks.
Intraday only.
  • #579
  • Nov 27, 2010 2:38pm Nov 27, 2010 2:38pm
Quoting navk
Hi roundrock,cailou,

Thanks for your EA. I shall test this latest version.

I have a question. Can i use this EA in different pairs in same account?

I mean in one 5000 $ account, i will use this EA in euro/usd ( 30min) and euro/jpy ( 30 min ) ? or only in one pair?

Thanks.
Let's wait for Roundrock, but if we have one balance for both pairs, maybe the 2 EAs will close all trades when the General Equity Target is reached.....
I' m not sure......
  • #580
  • Nov 27, 2010 3:12pm Nov 27, 2010 3:12pm
Quoting sidhujag
What you guys should be concentrating on is statistics on win/loss ratio over a long period of time.

IE: If we decide 20% loss is the cutoff then we need atleast 20 wins in a row at 1% PP... multiple times over a long time to be profitable.

So if we have all the changes we need now we should be concentrating on slapping them on demo tests to see if the strategy is sustainable..

If we can backtest thats even better. But I prefer not to have just one symbol since that defeats the purpose of the original idea... to distribute your money while...
Hi Jag,

I don't understand why you're always talking about a 20% SL, this EA has not a SL, no MC's or mine version has never had a SL, we are just testing an idea and I prefer concentrate my time in how we can help this EA in recover a DD.

Maybe I'm wrong, maybe I'm right but what I' m sure is that this EA will always recover if we are able to maintain the DD below a fixed level. This level and what to do in order to not to exceed it is what I'm going to concentrate. If we achieve this, who needs a SL?

And why the original idea is the best one? If we follow the original idea why are we talking in this thread? An hedge is possible with 6, 3 or just 1 pair but yes, is differeent, with just one pair the correlation is 100% always, not 90% or 80%.
0 traders viewing now