Phantom 6 System | Page 52 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Jag's Phantom 6 System 295 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: Phantom 6 System
Exit Attachments

Phantom 6 System

  • #1,021
  • Feb 26, 2011 4:03am Feb 26, 2011 4:03am
The problem I see with cutting a loser based on size or floating PnL is that you have no idea whether the pair you're cutting is the pair that is going to bring your basket to profit.

Why not use the phantom logic to chose which pair to cut? Eg:
(a) do a reverse of the phantom buy - cut 1 pair based on the worst performing pair in the basket(*)
(b) keep a record of the phantom buys, and cut the pair that was a phantom buy last out of all the pairs

(*) has this been suggested before?
  • #1,022
  • Feb 26, 2011 5:10am Feb 26, 2011 5:10am
Cutting the worst performing is very similar to the cut loser, which sometimes is OK but often isn't.

Only difference is "when?".
  • #1,023
  • Feb 26, 2011 10:05pm Feb 26, 2011 10:05pm
Quoting IamTHAT
A good post outlining the major issues which need to be addressed.
Can you please elaborate on your manual system? Do you enter based on any particular indicator or logic? How do you close your orders? All at once?
Very Interesting concept. Do you use any stoploss or just hedge with any other pair?
Yikes..It would be a long post. LOL.
Only testing for 2 weeks. Let me go a month or so and if its still decent Ill put something together. This is the phantom thread for EA trading so I dont want to get off topic from that. Maybe Ill start a new thread at some point. Plus, I have the entry criteria set, but Im still tweaking some things like trade size, exit target, etc.
  • #1,024
  • Feb 27, 2011 1:26am Feb 27, 2011 1:26am
Quoting thorn
Yikes..It would be a long post. LOL.
Okay, Thanks, I will be waiting..
"Put all your efforts in becoming effortless"
  • #1,025
  • Feb 27, 2011 3:27am Feb 27, 2011 3:27am
Hi FraserD,

I am missing your meticulous testing results. Do you ahve any updates to supply us with.

Would you mid telling us what Cut Loser percentage you are using on your tests?

Look to hear from you
  • #1,026
  • Feb 27, 2011 3:31am Feb 27, 2011 3:31am
Hi JCS,

I believe you may have gone live on the EA.

I was wondering what kind of effect the difference in price feed between demo and live accounts was having on the frequency of the basket pops.

Are you trading with Alpari UK?
Are you trading 0.01 lots?

Would you mind enlightening us on this?

Many thanks in anticipation
  • #1,027
  • Feb 27, 2011 6:49am Feb 27, 2011 6:49am
Quoting Nicastro
Hi JCS,

I believe you may have gone live on the EA.

I was wondering what kind of effect the difference in price feed between demo and live accounts was having on the frequency of the basket pops.

Are you trading with Alpari UK?
Are you trading 0.01 lots?

Would you mind enlightening us on this?

Many thanks in anticipation
No I am NOT trading live
I still do not feel comfortable with the levels of DD

Not demo testing at the moment, as I guess we know just about most things the current ea's can and can't do and repeating same demos does not add to our knowledge

Waiting and hoping for Roundrocks revised ea with margin protection in mind
  • #1,028
  • Feb 27, 2011 9:50am Feb 27, 2011 9:50am
this kind of function in fact should be "complicated decision", in other words, either manually performed or very complex subroutine.

however, if want a simple solution, then the answer to "when" might be "early London afternoon/New York morning".

Quoting mbkennel
Cutting the worst performing is very similar to the cut loser, which sometimes is OK but often isn't.

Only difference is "when?".
  • #1,029
  • Feb 27, 2011 2:18pm Feb 27, 2011 2:18pm
Quoting charvo
this kind of function in fact should be "complicated decision", in other words, either manually performed or very complex subroutine.

however, if want a simple solution, then the answer to "when" might be "early London afternoon/New York morning".
Isn't that a typical time for turnarounds? Can you motivate?

And what would be the criterion? If you are in margin danger then the trading session could really screw you well before the once-per-day time, unless this is for a different reason.

Cutting the "least distrurbance to ongoing basket when in margin danger" is easily implementable now that I think about it.

You need only one (not two as I previously believed) floating point global variable.

This variable accumulates the total (signed) P&L of all trades it has closed without a basket close. When in margin danger, choose the open trade (different from most recently added symbol) whose P&L, when added to the sum of previous P&L's (which can be zero) gives the smallest absolute value. Close it, and update the ongoing P&L upon a successful close.

Clear the ongoing P&L variable upon basket close.
  • #1,030
  • Feb 27, 2011 2:26pm Feb 27, 2011 2:26pm
Quoting mbkennel
Cutting the worst performing is very similar to the cut loser, which sometimes is OK but often isn't.

Only difference is "when?".
I realize I was wrong in the above. It's similar, but not very similar.

Cutting the worst performing according to phantoms bases P&L on the open point of the phantoms and not the individual trade's opening point.

I don't think this has been suggested before. It might be better than cut losers based on actual open trade P&L

I'm not sure though that there is any means to choose the best pair to eliminate---random choice may be as good as any.

The strategy works by exploiting short-term momentum. If previous trades went bad then it was because the momentum didn't persist long enough. The information that "there was a short term trend back then which caused a trade" is no longer relevant; only relevant information now might be just "you have many trades open".

Maybe the best strategy is to give up, close all and reset if there is no profit by a certain amount of time and move on.

Alternatively if you believe that information decays, cutting the oldest trade might be fine.
  • #1,031
  • Feb 28, 2011 11:15am Feb 28, 2011 11:15am
Hi, MB, check below please:

Isn't that a typical time for turnarounds? Can you motivate?

And what would be the criterion? If you are in margin danger then the trading session could really screw you well before the once-per-day time, unless this is for a different reason.

my assumption is London noon is the most active hour, at that moment, if we cut a loser of the worst performing pair, we will have 2 outcomes:
1. we cut it wrong, the market reverse, but the EA will quickly add to winning pairs and we will be better if the market moves fast. the cut itself won't make our margin condition worse (right?).
2. we cut it right, the market continues, the margin gets better.

the only danger is the market reverse, then ranging, without reaching a closure ---- that's why i feel the good time to do "cut a loser" is during active market hours.

You need only one (not two as I previously believed) floating point global variable.

This variable accumulates the total (signed) P&L of all trades it has closed without a basket close. When in margin danger, choose the open trade (different from most recently added symbol) whose P&L, when added to the sum of previous P&L's (which can be zero) gives the smallest absolute value. Close it, and update the ongoing P&L upon a successful close.

Clear the ongoing P&L variable upon basket close.

this is absolutely smart logic for coding, when we apply the margin-release function. I manually do this everyday around NY 7~10 pm, tired of the manual crap, but it is a must, since my LIVE account's margin is always low. but as you see, i feel this function should work in Asian session.

Thanks for your coding logic again, are you going to implement it?
  • #1,032
  • Feb 28, 2011 11:26am Feb 28, 2011 11:26am
Again, please see RED font

I realize I was wrong in the above. It's similar, but not very similar.

Cutting the worst performing according to phantoms bases P&L on the open point of the phantoms and not the individual trade's opening point.

I don't think this has been suggested before. It might be better than cut losers based on actual open trade P&L

no, seemingly not mentioned before. however there is one implicit assumption of your idea----the EA does not reset every X hours. at most, the EA may reset once every 1 or 2 days so that your cutting logic may work well. I feel it may be quite helpful to phantom_01.mq4 (theMaxx used it). then it'll be more "cut the loser fast" fundamentally, since the phantom pl is more convincing as trend indicator.

I'm not sure though that there is any means to choose the best pair to eliminate---random choice may be as good as any.

absolutely
  • #1,033
  • Mar 6, 2011 3:10am Mar 6, 2011 3:10am
Hi All, and in particular Roundrock,

I am reporting on the DD EA Basket Closures for the week ending 4th March 2011:-



OOTB settings on EA Phantom6_01DD
100:1 Leveraged Account.
Running only during London and NY sessions.


1st Basket:
Basket held over the preceding week-end
First trade entered at 14.55 GMT on 24 Feb.
Basket closed at 8.15 GMT on 1st Mar.
Trades: 89
Largest Floating DD -16.37%



2nd Basket: - MMTU now increased to 1%
First trade entered at 13.47 GMT on 1st Mar.
Basket closed at 17.26 GMT on 1st Mar .
Trades: 32
Largest Floating DD -16.37%



3rd Basket: - MMTU now increased to 1%
First trades entered at 10.21 GMT on 3rd Mar.
Basket closed at 12.44 GMT on 4th Mar - Closed EA for the weekend and NFP .
Trades: 76
Largest Floating DD -22.44%



The EA Phantom6-01xR had a basket open at 13.15 GMT on 18th Feb. This basket finally closed on Tuesday 1st March at 9.01 GMT.
Since then it has managed to close 5 baskets on Tuesday each with a PP of 0.8
Final basket opened 16.18GMT on 1st March
Basket still open.
NFP would have lead to margin call - closing 2 trades.
500:1 leveraged account with MMTU of 2.1% and PP of 0.8%


It is clear that Phantom6-01xR needs JCS's closure rules incorporated into the EA.

The introduction of a 3rd Basket could avoid the lingering basket.

Roundrock, may we please please ask that you devote sometime to this.

Sincerely

G Nicastro
  • #1,034
  • Mar 6, 2011 3:40am Mar 6, 2011 3:40am
thanks for the report nicastro

1st Basket:
Basket held over the preceding week-end
First trade entered at 14.55 GMT on 24 Feb.
Basket closed at 8.15 GMT on 1st Mar.
Trades: 89
Largest Floating DD -16.37%



2nd Basket: - MMTU now increased to 1%
First trade entered at 13.47 GMT on 1st Mar.
Basket closed at 17.26 GMT on 1st Mar .
Trades: 32
Largest Floating DD -16.37%

1st basket closed on 8:15 gmt and 2nd one closed at 17:26 GMT. But there are 32 trades. is it opening every 1 hour or 15 min ?

i didnt understand what u mean by jcs closure ? can you explain? are you talking about closing when margin dips to dangerous levels ? i didnt understand what jcs mentioned. can you explain with an example ?

btw, why is that you have 100:1 and 500:1 ? can u test both with same settings ?
  • #1,035
  • Edited 5:06am Mar 6, 2011 5:04am | Edited 5:06am
Quoting roundrock
thanks for the report nicastro

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


1st basket closed on 8:15 gmt and 2nd one closed at 17:26 GMT. But there are 32 trades. is it opening every 1 hour or 15 min ?

[font=Times New Roman][size=3]i didnt understand what u mean by jcs closure ? can you explain? are you talking about closing when margin dips to dangerous levels ? i didnt...
Hi Roundrock,

I respond to your points as follows:-

1) The DD Ea opens trades every 15 minutes.
2) Yes. JCS closure is set out in his post 965. JCS will you please provide further examples of how you would like to see the margin protector operate so that Roundrock can code the necessary
3) I have 2 demo accounts. 1 at 100:1 leverage and the other at 500:1 leverage. Do you wish me to test both EAs with the same leverage settings? That will not be a probelm. Which EA? At which leverage?
4) The third sub-basket idea is set out in my post 929

Sincerely

G Nicastro
  • #1,036
  • Mar 6, 2011 7:49am Mar 6, 2011 7:49am
last week i posted a EA for DD that trades every hour not every 15 min. i think u didnt replace it. i had every 15 min to test the bugs (i can speed up with 15 min)

regarding leverage, make sure u test both with same settings. doesnt matter if its 100:1 or 500:1
  • #1,037
  • Mar 7, 2011 1:42am Mar 7, 2011 1:42am
Quoting roundrock
last week i posted a EA for DD that trades every hour not every 15 min. i think u didnt replace it. i had every 15 min to test the bugs (i can speed up with 15 min)

regarding leverage, make sure u test both with same settings. doesnt matter if its 100:1 or 500:1
Sorry Roundrock,
I did not see the 1hr DD Ea. Can you please post it again so I can start testing at 500:1 too so we can get a comparative

Thanks in advance
  • #1,038
  • Mar 7, 2011 10:42am Mar 7, 2011 10:42am
Quoting Nicastro
Sorry Roundrock,
I did not see the 1hr DD Ea. Can you please post it again so I can start testing at 500:1 too so we can get a comparative

Thanks in advance
This one opens every hour:
http://www.forexfactory.com/showthre...73#post4396373

This one opens every 15 minutes:
http://www.forexfactory.com/showthre...39#post4412439
  • #1,039
  • Mar 7, 2011 7:25pm Mar 7, 2011 7:25pm
here it is. this one opens every 1 hr
Attached File(s)
File Type: mq4 Phantom6_01DD.mq4   56 KB | 309 downloads
  • #1,040
  • Mar 7, 2011 10:36pm Mar 7, 2011 10:36pm
Need you guys do me a favor, if possible.

i was thinking how to determine if our pairs selection are good, or
is there another set of pairs that are better?

i must say i'm lazy and slow, and just realised that i could do this to gain a preliminary idea.

i use TheMaxx's statements uploaded on Feb 10 (1135 trades). import into excel, and do these:

i count how many trades each pair has produced. it is the 2nd row of the table below. but the # of pairs cannot really tell if a pair is very constructive to profit winning. a big # of positions by xxxyyy may very likely be due to the xxxyyy pair tend to reverse, and trap the whole basket.

so i count how much $ each pair made or lost, then i got the 3rd row of the table below. it looks a little clear that chfjpy and eurchf are two sleepers in this workforce. apparently, they might need to be replaced. In other words, eurchf placed 175 trades to contribute only $20 profit, which implies eurchf is usually the one who slows down the whole basket. when a basket closes in profit, it is usually not because of eurchf. and chfjpy is further worse.

this in fact strengthen my guess that usdjpy, eurusd might be a better substitute.

however, i don't have statistics to prove it. in case any of you guys tried different pairs config, could you please share your demo/live trades statements so that we can further this analysis? i remember thorn and Mbkennel seemingly tried different pairs.
Attached Image (click to enlarge)
Click to Enlarge Name: excel.jpg Size: 29 KB
8 traders viewing now