Phantom 6 System | Page 31 | 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

  • #601
  • Jan 23, 2011 9:08pm Jan 23, 2011 9:08pm
Quoting roundrock
thanks thorn.

oanda has these spreads

eurcad 5 pips
euraud 4
audcad 4
Sunday spread at oanda are a little high I would bet by monday morning those will be .5 to 1 pip lower than that. At least thats what I have seen on fxtrade. Maybe mt4 spreads higher? Not sure about that.

Edit: just got my mt4 up and running, looks like spreads are same on both mt4 and fxtrade. And the three you quoted are already lower.
  • #602
  • Jan 23, 2011 9:37pm Jan 23, 2011 9:37pm
Quoting mbkennel
Something I noticed with the Phantom series of EA's.

They are designed to open the same number of lots in each pair, e.g. 0.25, regardless of the value of each trade.

I thought the idea was that if you had exactly one trade open in each pair then you should have a full synthetic hedge, i.e. nothing going anywhere regardless of market movement.

But with one lot of each pair, you won't get that; you need to adjust.

Hypothetical example.

Suppose you had one lot of EURUSD and one lot of the (non-existent) USDEUR. Do those balance? No, because...
MBK,

Interesting, I like it. Especialy the phantoms not being reset on terminal shutdown. Mostly because Im a dummy and "X" out my terminal instead of minimizing it sometimes.

Question on adjusting the lot size. Does it make much difference when trading very small lots like .01. I could see where it would work well with a larger account trading bigger lots. So as an example if the EA trades .03 lots of eur/usd, a gbp/usd might only be .02 lots. Am I seeing this right?
I trade with Oanda and I know when you trade a % of your account you get different unit amounts for each pair thats not usd based. But since mt4 limits you to 1k lot sizes how much difference will it make? It does make sense thou to get you closer to a true hedge. Ill throw it on and see how it does.
  • #603
  • Edited 10:02pm Jan 23, 2011 9:46pm | Edited 10:02pm
[quote=mbkennel;4336883]Something I noticed .....
I thought the idea was that if you had exactly one trade open in each pair then you should have a full synthetic hedge, i.e. nothing going anywhere regardless of market movement.quote]


I think you are correct in that the original ea’s do not take into account the different $ values of the respective lots and therefore if you were trying to create a true hedged basket this would need to be the case
My understanding however is that the ea’s are not attempting to do this
A ‘ synthetic ‘ hedge is not a true hedged basket ( as I understand it )
MetaCoder has posted somewhere a description of a synthetic hedge

More importantly I believe is the fact we only trade the biggest winner from each hour and as we will have all seen that means we very rarely get equal numbers of open positions in each pair / supposed correlated pairs
So even if the respective $ values of each pair within the synthetic hedge are true hedged amounts it would suggest this is totally irrelevent within the basket we are actually trading as it will at no time be a true hedge in any event

As we add the biggest winner each hour we will again have all seen it is often the case for several hours in a row we are opening the same pair, which results in a basket that is in no way ‘ balanced ‘ and needs this pair ( or base currency ) to ‘pop’ and move enough to produce the required profit ... the basic idea behind the ea's
Often it is the case after being the biggest winner for many hours, but not getting the ‘ pop ‘ the particular pair turns around and tanks in the wrong direction for our basket
It is an extreme case of this, or a combination of pairs doing this, which will eventually result in the ‘ death ‘ basket
  • #604
  • Jan 23, 2011 10:20pm Jan 23, 2011 10:20pm
Quoting 4XWeezal
Thorn,
I understand the correlation aspect, but the ea itself picks the biggest winner of the three pairs on each side of the hedge. In a sense your picking three correlleated pairs (on the buy and sell side) to ensure a good basket ; but yet the pairs also have to be somewhat positive correlated within the group.? is that accurate.
Within the buy group or sell group, no they dont have to be correlated. Now they may be somewhat (eur/cad and eur/gbp in buy group), if euro is strong then both will probably being going up. What Im trying to do is counter balance any long trades with a short pair that moves the same direction as the long pair. And vice versa. I think the whole concept of the phantoms is to wait for a certain pair or two to make a big move and thats when you get your profit. when none of your pairs make any good moves thats when you get stuck with those 3-4 day baskets. Believe me I havnt found the perfect solution to trading these things, but the little adjustments here and there can make them worthwhile I think. When MC made these EA's he was creating a synthetic hedge of equal buys and sells of each currency. I just took it a little farther and looked at correlations. Also I didnt like trading gbp/chf and chf/jpy in the original 6, as to me they were just too choppy. But thats just my preference. There are alot of 6 pair combos you can use and still make a synthetic hedge.

To be honest, Im actually leaning more toward a 4 pair phantom now instead of 6. Im still in early stages of testing but looks good so far. I just started a new one on Oandas new mt4 platform. gotta love those lower spreads.
  • #605
  • Jan 23, 2011 10:30pm Jan 23, 2011 10:30pm
thorn:
what's the link to oanda's new mt4? I can't find it on their website
  • #606
  • Jan 23, 2011 10:40pm Jan 23, 2011 10:40pm
[quote=jcs;4336962]
Quoting mbkennel
Something I noticed .....
cription of a synthetic hedge

More importantly I believe is the fact we only trade the biggest winner from each hour and as we will have all seen that means we very rarely get equal numbers of open positions in each pair / supposed correlated pairs
I understand that. But if you trade only once per hour, why should you be more affected by a GBPxxx trade at 9:00 than a USDxxx trade at 10:00?

To me it seems as though they should "count" equally. My mod makes that happen, at least on average.

Quote
So even if the respective $ values of each pair within the synthetic hedge are true hedged amounts it would suggest this is totally irrelevent within the basket we are actually trading as it will at no time be a true hedge in any event

It's not completely irrelevant.

Also consider if you choose once-per-hour trades based on $ p&l it favors GBPxxx over USDxxx. Is that desirable?
  • #607
  • Jan 23, 2011 11:02pm Jan 23, 2011 11:02pm
[quote=mbkennel;4337050][quote=jcs;4336962]
I understand that. But ..... quote]

I think my use of the word ' irrelevent ' was a bad choice

Having the pairs balanced in monetary value is very sensible ... all pairs being equal etc
What I was suggesting is that I don't think it will have much effect on the outcome of an ' unbalanced ' basket, which is what we always seem to have
  • #608
  • Jan 23, 2011 11:15pm Jan 23, 2011 11:15pm
As I have said before, I believe the need to preserve margin availability is essential if we are to ensure we can continue to trade during the extended basket periods
This to me suggest we either stop adding biggest winners at some point .... not in the best interests of the basket
Or we remove ‘ redundant ‘ trades from the basket at some point
In a prolonged basket we will always reach a point where we have multiple trades of a same pair

What do I mean by ‘ redundant ‘ ?
I question the need to keep multiple open positions of the same pairs within the basket when they have not produced the ‘ pop ‘ at the time they were expected to ... ie when they were obviously the strongest pair

It has been said here previously that closing positions at some point may be the way to go, but the difficult part of this is which pair ( or pairs ), or combination of, would need to be closed so as not to disturb the ' balance ' of the basket
Well I do not think in the first place we ever have a ‘ balanced ‘ basket, so maybe this is not something to worry over
I also doubt it is in any way possible to answer which pair / pairs to cut as we simply do not know which of the pairs will eventually make the required move to close the basket, so again maybe this is something not to worry about

So my thoughts are ....... When we have at least 3 or more open trades of the same pair we close 2 of the positions, providing always of these 2 positions one is in loss and one is in profit and with a slightly higher profit ( to ensure we are always adding to the account balance )

This way the account balance always remains proportional but margin availability is replenished for adding further trades ( which in turn may again become ‘redundant’ in the future )
By only doing this when there are minimum 3 trades of the same pair open we ensure we keep at least one position of the pair to contribute to any potential ‘pop’
By doing this with every pair as the opportunity arises we should be able to keep the overall number of trades at any point in control and reduce the drawdown on the margin

As we can not always guarantee to get one losing and one winning trade in each of the pairs it is difficult to predict how this would pan out over the period of any basket
It obviously needs to be coded and tested

Any thoughts ?
  • #609
  • Jan 24, 2011 4:15am Jan 24, 2011 4:15am
Quoting tazmaan
thorn:
what's the link to oanda's new mt4? I can't find it on their website
pl visit oanda forums, you will get the link and instructions. there is a great discussion on the platform issues.

in a nutshell, create a subaccount and while doing that go to advanced, check MT4 option. now down load the mt4 and login with username as account number and password.
  • #610
  • Jan 24, 2011 10:44am Jan 24, 2011 10:44am
Quoting zznbrm
There are plenty of coders up to the challenge. However, programmers like detailed requirements. Supply details and you will probably get some interest.

Hi,

As requested here are some details: -

1) Keep Phantom Trades concept as it stands though if necessary substituting pairs on either the buy side or the sell side.

2) Construct one synthetically hedged sub-basket (Basket A) consisting of Buy E/J, Sell U/J and Sell E/U. At each 00 or 15, or 30 (at user’s option) past the hour trade the pair or pairs that are in profit provided the aggregate of the previous pairs are also in profit. That is to say, with the X Factor as currently existing.

3) Construct a further synthetically hedged sub-basket (Basket B) consisting of Buy G/U, Sell A/U and Sell G/A. At each 00 or 15, or 30 (at user’s option) past the hour trade the pair or pairs that are in profit provided the aggregate of previous pairs are also in profit. Different hedged currency pairs could be the subject of Basket B at a user’s individual choice.

4) The Profit Pct setting (PP) in the EA applies to the entire basket. The entire basket consists of Basket A and Basket B. Close the entire basket once PP is reached.

5) MMTU applied, as currently existing, to each trade in either Basket A or Basket B, with option for there to be different MMTU for each sub-basket.

6) The idea is that by having 2 synthetically hedged sub-baskets within a larger entire basket we get to closure (ie PP) quicker without incurring extensive DD or the basket lingering on for extended periods.

7) There will invariably come a time when 1 pair in either sub basket will go seriously against us taking us further and further away from PP.

8) In this event, if the aggregate of any particular pair in Sub-Basket A should be at a loss equal to a user defined percentage of Last Flat Equity, then the EA is to enter a recovery trade at the ratio of 1:1.1 to freeze the loss and help us to PP when the retrace occurs as surely it will. By way of example, if Sub- Basket A has 3x Long E/J trades totalling 0.9 standard lots showing a combined loss of say 10% of Last Flat Equity, then the EA would enter an immediate Short E/J trade of 0.99 standard lots.

9) That will then arrest the decline to leave the other pairs in the 2 sub-baskets to do their work and take us to PP.

10) Ditto as in 8 for any pair comprised within sub-basket B.

11) Only ONE PAIR in each sub-basket can be the subject of a recovery trade.

12) We need to have user defined settings of when a recovery trade is to be implemented:- eg

a) ONLY after X no. of trades for any pair AND
b) Once the currency pair exceeds Y% of Last Flat Equity

13) If a recovery trade is in place, The EA is to allow the user to alter the PP to anything up to 0.01 (or even to a minus number) so that a user can close the current basket and hopefully move onwards and upwards. Altering the PP, if manually, must not lead to a resetting of the phantoms. The phantom re-setting problem does not occur if an alternate PP is set at the outset of a basket trade sequence.


14) Likewise, the user should be able to, in identical circumstances to13 above, alter the MMTU in the hope of quickly getting out of a dull basket. Again, altering the MMTU, if manually, must not lead to a resetting of the phantoms. The phantom re-setting problem does not occur if an alternate MMTU is set at the outset of a basket trade sequence.


15) The EA will also need to retain the current mods available, ie CL, TH, PS and of course the X Factor that so many are using in testing and live trading.


16) While the basic concept works fine, and trades profitably in testing, there is still the lingering threat of a death basket. Implementation of these ideas should prevent the death basket from ever occurring.

17) These ideas need to be forward tested - all that is needed is a coder to code them……………..

I hope a coder can rise to the challenge
  • #611
  • Jan 24, 2011 3:25pm Jan 24, 2011 3:25pm
Here are my results from a Phantom6_01.mq4 (from the 1st post) forward test from 12 Dec 10.

IBFX demo, standard account, 100:1 leverage (ea says 50:1???), $3,000 account opening
MMTU = 0.5
PP = 0.75
Max floating DD: -10.02%

Results: +29.56%
Statement attached.

Pretty amazing results, right? Interestingly I restarted my VPS every Sunday but it didn't seem to mess anything up.
Attached File(s)
File Type: zip DetailedStatement.zip   29 KB | 386 downloads
  • #612
  • Edited 10:25pm Jan 24, 2011 6:10pm | Edited 10:25pm
Quoting thorn
Within the buy group or sell group, no they dont have to be correlated. Now they may be somewhat (eur/cad and eur/gbp in buy group), if euro is strong then both will probably being going up. What Im trying to do is counter balance any long trades with a short pair that moves the same direction as the long pair. And vice versa. I think the whole concept of the phantoms is to wait for a certain pair or two to make a big move and thats when you get your profit. when none of your pairs make any good moves thats when you get stuck with those 3-4 day...
Thanks, for your insight.
Also, very nice ----The maxx. Looks good. Gives us all lot of ideas.....
...can anyone explain in laymens term what synthethic hedge means? (even after reading the thread from beginning..it is still confusing). It shows crossing off some pairs but doesn't really help me understand the logic why its set up this way. I think it means equal dollar value on each side of the trade of the pairs with similar lot sizes....
Also, I like the correlation response; this will help a lot in setting up different baskets to test. It makes sense but I think there is a little more to add to it somewhere. If you just correlate; than you may not be on equal value of trade so it could be out of balance.
  • #613
  • Jan 25, 2011 3:55am Jan 25, 2011 3:55am
Quoting Nicastro
Hi,

As requested here are some details: -

1) Keep Phantom Trades concept as it stands though if necessary substituting pairs on either the buy side or the sell side.

[font=Times New Roman]2) [size=3]Construct one synthetically hedged sub-basket (Basket A) consisting of Buy E/J, Sell U/J and Sell E/U. At each 00 or 15, or 30 (at user’s option) past the hour trade the pair or pairs...

This is built on top of Phantom6_01X with TH, ST, CLM features.
I have Basket A and Basket B with symbols externalized (you can chose whatever you want, but make sureeach basket forms synth hedge)
Pl note we ALWAYS buy first symbol and sell next 2 symbols. This is true in both baskets. You cant have 2 buys and 1 sell. This is
fixed.
Whenever a pair has total loss more than RecoveryPct , this is percentage of last flat account balance, AND the number of trades is
more than TradeCountForRecovery, then recovery process kicks in. Now EA places opposite order for lotsize equal to RecoveryLotRatio times the TradeCountForRecovery. Each basket can have its own recovery process, but is limited to only one pair in each basket.
If you dont want all this recovery process, simply set RecoveryTradeAllowed to false, then it acts like normal phantom EA
I havent tested this, i request Nicastro to forward test this first and post bugs here before everybody starts testing. I will try to fix bugs. There
is lot of code change, so bugs are expected
I also modified the code to handle changes to MMTU, PP at runtime, but not sure it works. pl test
Attached File(s)
File Type: mq4 Phantom6_01xR.mq4   59 KB | 259 downloads
  • #614
  • Edited 9:31am Jan 25, 2011 9:25am | Edited 9:31am
Quoting roundrock
This is built on top of Phantom6_01X with TH, ST, CLM features.
I have Basket A and Basket B with symbols externalized....................



Thanks Roundrock.
This is great.
I have downloaded and will start testing. Will start tomorrow with first phantom set for 8.15GMT.

Acc. Size = 1140,
Margin 100:1
MMTU 2.1 for both,
PP of 0.8
Recovery PCT= 10%
Trade Count for Recovery= 5
Recovery Lot Ratio = 1.1
Phantom reset = 8Hrs
I will manually set Stop after no Trades to True after 3PM GMT
CLT = False

I will keep the tread posted on forward demo results.
  • #615
  • Jan 25, 2011 9:30am Jan 25, 2011 9:30am
nicastro, one question


12) We need to have user defined settings of when a recovery trade is to be implemented:- eg

a) ONLY after X no. of trades for any pair AND
b) Once the currency pair exceeds Y% of Last Flat Equity


is it AND or OR ? What if the total loss exceeds the pct but the number of trades are less. then there will not be any recovery. why do you need the AND condition here ?
  • #616
  • Jan 25, 2011 9:37am Jan 25, 2011 9:37am
I decided to go live again with this EA and I love it when things work out like this. Two basket closures in a day and a half. $463 in the bank. Only 4 trades in the last basket , 8 in the first basket. My strategy is, since these EA's like volatility, I'm turning it off (and closing basket) if I'm in profit around noon on fridays and back on monday mornings.
  • #617
  • Jan 25, 2011 9:38am Jan 25, 2011 9:38am
Quoting Nicastro
Thanks Roundrock.
...

MMTU 2.1 for both,

...
Sorry for my question, but MaxMarginToUse set to 2.1 isn't too much?

/achab
  • #618
  • Jan 25, 2011 10:10am Jan 25, 2011 10:10am
Quoting achab
Sorry for my question, but MaxMarginToUse set to 2.1 isn't too much?

/achab
way too much, 0.3 to 0.5 is reasonable, but again it depends on the appetite for risk.
"Put all your efforts in becoming effortless"
  • #619
  • Jan 25, 2011 10:12am Jan 25, 2011 10:12am
Quoting Nicastro
Thanks Roundrock.
This is great.
I have downloaded and will start testing. Will start tomorrow with first phantom set for 8.15GMT.

Acc. Size = 1140,
Margin 100:1
MMTU 2.1 for both,
PP of 0.8
Recovery PCT= 10%
Trade Count for Recovery= 5
Recovery Lot Ratio = 1.1
Phantom reset = 8Hrs
I will manually set Stop after no Trades to True after 3PM GMT
CLT = False

I will keep the tread posted on forward demo results.
By the way, I started a test with a demo account I already had (see attached picture). The online statements and stats can be reached here: http://www.mt4i.com/users/achab-phantom6_01xr/stats.

To see/load the current settings, just download the file named "phantom6_01xr-setup_01.set.tx".

/achab
Attached Image (click to enlarge)
Click to Enlarge Name: Phantom6_01xR.jpg Size: 234 KB
Attached File(s)
File Type: txt phantom6_01xr-setup_01.set.txt   1 KB | 221 downloads
  • #620
  • Jan 25, 2011 11:37am Jan 25, 2011 11:37am
Quoting Nicastro
Hi,

As requested here are some details: -

[font=Times New Roman]1)...
I also suggest a simple improvement to this EA: to open a log file in append for writing the date, time and the lowest margin.
It would be perfect to study the old tradings to optimize the personal TS.

I hope you agree with me.
Good evening.
/achab
2 traders viewing now