Art Phantom EA | Page 60 | 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

  • #1,181
  • Dec 10, 2010 9:22pm Dec 10, 2010 9:22pm
Quoting Caillou
Great report Vienna!

I agree with you, Single Pair Version is the best version of Phantom EA's in my opinion.

And I only believe in forwardtest, backtest has no credibility for me.....

Thank you Vienna
not yet. in my backtesting several unpleasant things have happened which can be easily validated with little commonsense.

to avoid big DD, i added Max Trade filter, lets say its 12. we have 9 buys and 3 sells. net 6 buys, if market goes in opposite direction (it did), its like a falling knife. soon it hit zero balance. now lets say we have 6 buys, 6 sells. net zero. but we are always stuck with the loss thats always LCOKED no matter what the price moves. lets say your start price is 1.2550. we have buys at 1.26, 1.2620, 1.2640, 1.2660, 1.2680, 1.27 and sells at 1.25, 1.2480, 1.2460, 1.2440, 1.2420, 1.24. now total net loss is 400 pips no matter where the current price is.

dropping basket whenever we reach Max Trade is more dangerous. loss is 10% per basket whereas profit is only 1%. and we get into this situation regularly (atleast once a month)

if market is trending this strategy works (like most of them). when its stuck in a range, we are dead

why should this strategy work ? whats the edge ? we are taking a random start price and adding a buy above it and short below it. if market doesnt move enough to reach PP, we will keep adding more trades and the situation gets worse. the basket reaches a stage where either you are stuck with a huge DD or stuck with a locked loss. if you have enough DD to wait for market to move in our direction then it works. for that we have to work with very very small lot, then profits also go down drastically

in case of multipair, we have several pairs and there is a chance atleast one or two will push it in our favor. and due to good mix, we dont get into huge DD situation as if one pair goes up some other will come down.

right now i have exhausted all ideas and filters to make single pair strategy profitable without a death trade
  • #1,182
  • Dec 10, 2010 10:30pm Dec 10, 2010 10:30pm
Quoting roundrock
in case of multipair, we have several pairs and there is a chance atleast one or two will push it in our favor. and due to good mix, we dont get into huge DD situation as if one pair goes up some other will come down.
I totally agree with you. I find that AddSix versions (2a,1a,3b) works much better due to the fact there is some pair or the other to bail us out.
This is what basket trading is meant to be according to me.
"Put all your efforts in becoming effortless"
  • #1,183
  • Dec 11, 2010 1:26am Dec 11, 2010 1:26am
Single pair phantom is working good for larger (1hr) timeframe (till yesterday). It is clossing avg 2-3 baskets with less trades (max 8 pairs) with controled dd..but because of EA problem I was not able test it for fully one week. But now testing with roundrock new EA. Now I want to see effect of weekend gap.

As far as multiple pair phantom is concerned I am demo trading APL with EU UC EC pairs but here also I faced some problem with ea. But yesterday i again work on EA and started testing... hope so it will work..

My idea of basket trading was..
1. Open three hedged pairs at same instance
2. Give them time settle down say 60m
3. And then add wining pair at interval of say 15m.
And don know how it will work.
  • #1,184
  • Dec 11, 2010 1:47am Dec 11, 2010 1:47am
Quoting 24hrs
Single pair phantom is working good for larger (1hr) timeframe (till yesterday). It is clossing avg 2-3 baskets with less trades (max 8 pairs) with controled dd..but because of EA problem I was not able test it for fully one week. But now testing with roundrock new EA. Now I want to see effect of weekend gap.

As far as multiple pair phantom is concerned I am demo trading APL with EU UC EC pairs but here also I faced some problem with ea. But yesterday i again work on EA and started testing... hope so it will work..

My idea of basket trading...
why do you need step 1 ? anyway you are going to add winning pairs.

also you mentioned till yesterday. is that forward testing or backtesting ? are you able to backtest for 1 full year ? whenever i do backtest i get different results for same EA and same settings..


you can use this EA for backtesting.
Attached File(s)
File Type: mq4 Single Pair Phantom.mq4   31 KB | 226 downloads
  • #1,185
  • Dec 11, 2010 1:59am Dec 11, 2010 1:59am
Quoting roundrock
not yet. in my backtesting several unpleasant things have happened which can be easily validated with little commonsense.

to avoid big DD, i added Max Trade filter, lets say its 12. we have 9 buys and 3 sells. net 6 buys, if market goes in opposite direction (it did), its like a falling knife. soon it hit zero balance....
Thank you Roundrock,

Maybe you are right, I don't know. Effectiveness in the hedge (and the hedge is the only real edge of these EA's) is bigger with one pair than with 3, 6, 14....

Only forwardtest will decide, but as I said this is only my opinion....
  • #1,186
  • Dec 11, 2010 3:57am Dec 11, 2010 3:57am
I still think that Floydian666 idea could be a very interesting mod. to test, basically it's the same as "Cut loss" strategy but instead of closing is hedging a losser trade. If a losser trade reaches a % PP (this could be an option, for example loss= $ PP) then the EA will hedge this trade (open a buy if we have a sell and the opposite).

I know that Roundrock doesn't think this could be an advantage but if a coder could add this mod to Single_Pair_Phantom version from post 998, it would be great......

Thank you
  • #1,187
  • Edited 4:06am Dec 11, 2010 3:59am | Edited 4:06am
Quoting hopfi2k
Nicastro,

........................

Coming back to your suggestion (sorry for the small "excursion" above), I'm not sure to how that would be beneficial. E.g you open a sequence of trades, unluckily, just before a trend change. Now you are slowly facing DD getting bigger and bigger. If QQE would start delayed, the only thing it will do is preventing opening of more trades, because they are against the trend. Whith the current version, it would not have opened a sequence of trade from the beginning, as long as they would have violated the current...

Hi andy,

Thank you for taking the time to respond to my post.

I have been testing Phantom Baskets for some 6 weeks and have whittled down the most beneficial as being Art Phantom Light. My observations show that this EA is profitable during the active period of 7AM to 2PM GMT. A Trade sequence opened during that time closes at PP within 150 minutes at a very respectable DD of less than -6%.

In my experience, the opening of a trade sequence during this time window is not an issue. Art Phantom Light does so efficiently. Using QQE to filter opening of a trade sequence would be like gilding the lilly, and to our detriment, missing an opportunity that will not come round for another 60 minutes.

With Art Phantom Light, seeking to trade beyond this time window is courting a DD problem. It is a bigger DD problem if you are using Art Phantom Full and exponentially more so if you should seek to use the Extreme version.

With Art Phantom Light, a sequence opened after 2PM lingers for longer and the longer it takes to close, the greater the DD and the greater the risk that the basket will drift for hours on end. Often until the next day where the opportunity cost of not being in the maket at its most active period (7AM -2PM) becomes considerable

There is of course no way of knowing when a drifting basket will arise, but as sure as eggs are eggs, a drifting basket will arise and hence it is important to have measures in place to control the risks posed by them.

A multitime frame QQE would be a very effective risk control measure. A 4H, a 1H and a 15M QQE, would definitely ensure that trades are with the trend. I see that you have already coded a 4H and 15M QQE indicator. Can you please add to it, as an option, the 1H QQE suggested so that if acivated all 3 indicators would need to be in agreement?

I am pleased that your modification at #1142 gives the EA the facility to reduce PP after a set number of trades. This will hopefully get us to a quicker closure of the basket and move onwards and upwards from there.

As you can probably tell, I am not a coder. I would not know where to start- even on a simple cut and paste operation. Would it be possible for you to code that:-
a) QQE would set in only after a trade sequence contains a user defined number of "X" trades
b) That MMTU can be altered to "Y" after a set number of trades
c) To set a user defined total maximum number of orders that may be comprised within a trade sequence

With all these in place, I think we can keep the DD beast on a leash

I hope that these can help fellow traders

Sincerely

Giuseppe
  • #1,188
  • Edited Dec 12, 2010 5:45am Dec 11, 2010 12:36pm | Edited Dec 12, 2010 5:45am
Quoting Caillou
I still think that Floydian666 idea could be a very interesting mod. to test, basically it's the same as "Cut loss" strategy but instead of closing is hedging a losser trade. If a losser trade reaches a % PP (this could be an option, for example loss= $ PP) then the EA will hedge this trade (open a buy if we have a sell and the opposite).

I know that Roundrock doesn't think this could be an advantage but if a coder could add this mod to Single_Pair_Phantom version from post 998, it would be great......

Thank you

here you go......
Attached File(s)
File Type: mq4 Single Pair Phantom.mq4   36 KB | 301 downloads
  • #1,189
  • Dec 11, 2010 12:43pm Dec 11, 2010 12:43pm
Quoting roundrock
here you go......
Oh, Roundrock, your generosity has no limits.

I can't say how gratefull I am. Thank you very much.

I think the only way to test a mod is doing forwardtest so next Monday I will start with this one...

Thank you.
  • #1,190
  • Edited 2:29pm Dec 11, 2010 2:01pm | Edited 2:29pm
Quoting roundrock
here you go......
Sorry Roundrock, just one question: if the trade A has a loss=%PP, then the EA will open an opposite trade B but if later the same trade A becomes a winner and then reaches again a new loss=%PP, the EA will open another opposite trade or not ? Because once a trade A is hedged, we don't have to hedge it again anymore, neither A nor B.

I don't know if I have explained well, or if it is possible to code like that, sorry.

Thanks again.
  • #1,191
  • Dec 11, 2010 9:10pm Dec 11, 2010 9:10pm
Quoting Caillou
Sorry Roundrock, just one question: if the trade A has a loss=%PP, then the EA will open an opposite trade B but if later the same trade A becomes a winner and then reaches again a new loss=%PP, the EA will open another opposite trade or not ? Because once a trade A is hedged, we don't have to hedge it again anymore, neither A nor B.

I don't know if I have explained well, or if it is possible to code like that, sorry.

Thanks again.
hedge is placed only ONCE per trade. also hedged trade can be hedged. lol. i added a cap for number hedges, sometimes market goes up and down in a range, then we keep getting real trades and hedged trades, soon DD multiples. CLM must be above 3, else you get too many hedges.

pl redownload the EA
  • #1,192
  • Edited Dec 12, 2010 3:20am Dec 11, 2010 9:48pm | Edited Dec 12, 2010 3:20am
10k , 200:1 MMTU 0.2, PP 2, lot size 0.1, H1, CLM 3, CLCap 15, since jan1 2010 --> closing balance 17800

10k , 200:1 MMTU 0.2, PP 1, lot size 0.1, H1, CLM 3, CLCap 15, since jan1 2010 --> closing balance 18100 Max DD 48%

10k , 200:1 MMTU 0.2, PP 0.5, lot size 0.1, H1, CLM 3, CLCap 15, since jan1 2010 --> closing balance 15100

so higher PP is looking better. also CL Hedge concept is looking better (thanks caillou)

without CLM

10k , 200:1 MMTU 0.2, PP 1, lot size 0.1, H1, since jan1 2010 --> closing balance 15700 Max DD 61%
10k , 200:1 MMTU 0.2, PP 2, lot size 0.1, H1, since jan1 2010 --> closing balance 15600
  • #1,193
  • Dec 11, 2010 11:00pm Dec 11, 2010 11:00pm
dd on those roundrock?
  • #1,194
  • Dec 11, 2010 11:21pm Dec 11, 2010 11:21pm
Its confusing...
which ea i should test from monday.. lots of versions
  • #1,195
  • Dec 12, 2010 1:19am Dec 12, 2010 1:19am
Quoting roundrock
hedge is placed only ONCE per trade. also hedged trade can be hedged. lol. i added a cap for number hedges, sometimes market goes up and down in a range, then we keep getting real trades and hedged trades, soon DD multiples. CLM must be above 3, else you get too many hedges.

pl redownload the EA
Hi Roundrock,

So now each time a trade A reaches loss=%PP, we hedged with trade B but trade A and trade B could be hedged more than once? Is this correct? And CAP limits the numbers of times the EA hedges a trade?

So it's not possible once a trade A is hedged by trade B to bookmark trade A and B in order to not get hedged anymore?

If all that is correct, I think it could be easier to just add "Cut loss" mod. (classic version) to Single_Pair_Phantom.

I prefer the hedge mod but I think that to make it works properly we need that once a trade A is hedged by B, A and B could not be hedged anymore.

Thank you Roundrock for all your time and effort.
  • #1,196
  • Dec 12, 2010 3:21am Dec 12, 2010 3:21am
Quoting tazmaan
dd on those roundrock?
i updated my results
  • #1,197
  • Dec 12, 2010 3:27am Dec 12, 2010 3:27am
Quoting Caillou
Hi Roundrock,

So now each time a trade A reaches loss=%PP, we hedged with trade B but trade A and trade B could be hedged more than once? Is this correct? And CAP limits the numbers of times the EA hedges a trade?

So it's not possible once a trade A is hedged by trade B to bookmark trade A and B in order to not get hedged anymore?

If all that is correct, I think it could be easier to just add "Cut loss" mod. (classic version) to Single_Pair_Phantom.

I prefer the hedge mod but I think that to make it works properly we need that once a...

Lets A is long trade and is in loss, reached CLM. Now a hedge short trade B is placed. from now on there will not be any hedge trades on A. But hedged trade B can run into loss, another hedge trade C is placed on B.

for hedging we take original trades as well as hedged trades.

CutLoserCap is for entire basket. not for one trade
  • #1,198
  • Dec 12, 2010 3:36am Dec 12, 2010 3:36am
Quoting roundrock
Lets A is long trade and is in loss, reached CLM. Now a hedge short trade B is placed. from now on there will not be any hedge trades on A. But hedged trade B can run into loss, another hedge trade C is placed on B.

for hedging we take original trades as well as hedged trades.

CutLoserCap is for entire basket. not for one trade
Thank you, all clear now.......

So is it possible to not hedge trade B? Because trade A and trade B are hedged and the loss is locked forever so if we do that maybe we don't need CAP mod.....
  • #1,199
  • Dec 12, 2010 4:16am Dec 12, 2010 4:16am
Quoting Caillou
Thank you, all clear now.......

So is it possible to not hedge trade B? Because trade A and trade B are hedged and the loss is locked forever so if we do that maybe we don't need CAP mod.....
done. both A, B will not hedged again, pl take EA again
  • #1,200
  • Dec 12, 2010 4:25am Dec 12, 2010 4:25am
Quoting roundrock
done. both A, B will not hedged again, pl take EA again

Thank you Roundrock.

Like this, trades hedged will be less, we are going to lock a loss when it reaches CLM and we are going to maximize the authentic potential of this EA: the HEDGE.

Now I can't run any backtest, I'm at home with Macs, and on my VPS a backtest is extremely slow. Tomorrow at the office with PCs I'll do.......

1 trader viewing now