Phantom 6 System | Page 35 | 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

  • #681
  • Jan 28, 2011 10:50am Jan 28, 2011 10:50am
Quoting frazerd
jcs, could you post all of the settings you are using on the 500:1 account? I'm going to put this one on a Tadawul 500:1 demo and see how it works out there. Alpari UK is not accepting US clients, but Tadawul is.

EDIT: BTW, I have been monitoring these latest mods and they look very promising. Thanks to Nicastro and jcs for the great ideas and to roundrock for the coding.
Thanks for the praise FraserD. We will need a multiparty effort to bring these improvements to a successful conclusion. All contributions are welcome!
  • #682
  • Jan 28, 2011 10:57am Jan 28, 2011 10:57am
Don't for get that even the base version of this system is very profitable. I'm up 35% on demo since Dec 12 on Phantom6_01.mq4 from the first post.
  • #683
  • Jan 28, 2011 11:00am Jan 28, 2011 11:00am
Quoting jcs
No closures whilst I have been asleep
Rather than try to explain what did or did not happen I attach a screenshot of the open trades at this time
Just as I thought - nothing but drift and DD over the slow periods. It is much better to be out of the market altogether

Ideally, we should aim to be in a position where we have closed the last basket during the US morning session and then wait for the London open to start again.
  • #684
  • Jan 28, 2011 11:40am Jan 28, 2011 11:40am
Quoting jcs
If you look at the screenshot of my current 500:1 basket, it is sorted by symbol to make it easy to see how many trades there are in each pair
( because recovery is set to true for testing here there are a couple of hedge trades in there )
I know this is an extreme test as far as MMTU etc, but just ignore that here and concentrate on the pairs and Margin

the Margin Level is at 175%
[font=Arial][size=3]I will...
What you are suggesting is what some traders are doing manually. i dont think anyone has yeat tried to codify such an approach.

Why dont we solve the problem by trading only sub-basket A or sub-basket B, with trade count for recovery set to 3 and recovery lot ratio set to 0.95.

In that way, if you are only trading say sub-basket A, only 2 pairs will ever ever need to use recovery (due to the correlation inherent in the sub-baskets see e.g. EU and EY).

So that makes 8 trades (3 trades plus 1 recovery trade for each of the 2 non -ve correlated pairs). Add to that 1 or at max 2 trades taken by the 3rd pair in the sub-basket. You thus have at max 10 trades in a perfect synthetically hedged basket waiting for 1 or more of the pairs to break and lead you to a pop of the basket.

Under these circumstance, I envisage that the basket will pop by the time it gets to 15 trades.

Would that solve the margin problem?
  • #685
  • Jan 28, 2011 11:49am Jan 28, 2011 11:49am
Hi,Achab:
Thanks for your insights and trading reports! Great Job!

2ax = addsix_2a + add more pos for a pair only when previous pos are positive.

and you also know addsix_2a reset all phantoms every 8 hours (by default in external setting)

that's the difference, imho, from 3ax that does not reset every X hours.

the coder who edited all versions of 3x, 3ax, or 3bx used the entry logic pretty distant from MC's original, so use them very cautious, I'm not quite sure what those 3... versions aim at.

Quoting achab
On this page (http://www.forexfactory.com/showpost.php?p=4237873) it's possible to find, among others, the "2ax" version and the "3ax" too (talking about the "X" factor).
Now, my question is: what is(are) the difference(s) between the "2ax" and "3ax" version?
  • #686
  • Jan 28, 2011 3:04pm Jan 28, 2011 3:04pm
Quoting IamTHAT
AddSix 2ax is also coded but not tested by anyone extensively.
I uploaded the AddSix_02ax some time ago on this thread and I have been running it on a demo since 29 Nov. I also am using the CL option on that demo. It is not the best performing AddSix EA among my demos. AddSix_01x_CL has produced almost 3 times as much profit over the same time period. I am running 5 AddSix demos, each has its own tab in the attached spreadsheet.

There is a lot of difference in how the EAs perform for various individuals. The one that works best for me is not always the one that works best for someone else. As I have stated before, it depends a lot on when you start the EA. No one has figured out why or what the best time is to start them. For some unknown reason the start time will either put you into a favorable cycle with the EA closing baskets often or someone who starts his just a few minutes earlier or later gets into a unfavorable cycle. I wish we could come up with the solution to entry timing. I think it is a big factor affecting performance.
Attached File(s)
File Type: xls IBFX AddSix.xls   939 KB | 412 downloads
Frazer
  • #687
  • Jan 28, 2011 7:04pm Jan 28, 2011 7:04pm
Quoting charvo
following Thorn's accurate explanation, "PipStep" was introduced after this "X" factor, so that all the current positions of a pair have to be more than "PipStep" (an integer of pips) of floating profit before the EA can place another new position.

my experiences (2 months) gives me a feeling that "X" is definitely needed, "PipStep" enables you to quantify the risk appetite. However, "PipStep" should not be too large. 6~10 pips (4 digit broker) seems a good range.
Thanks to you and Thorn for the great info!

Looking at the code of the one I hacked on (it has the "BAL"/balanced lot size mod), it appeared that using PipStep overrode the "All Trades Must Be In Profit" criterion, i.e. if you have pipstep on, then only that mattered, and the profit-trade criterion was ignored.

Was that intentional?
  • #688
  • Jan 28, 2011 9:11pm Jan 28, 2011 9:11pm
MB, I looked into your code, as usual, the pipstep is considered in "allsymboltradesinprofit()", so that means "all trades in profit of 'pipstep' pips".

yes, that's right, if pipstep is switched on, 'pipstep' will decide how much profit is the criteria.

in fact, i coded it in a natural way, i feel.

Quoting mbkennel
Looking at the code of the one I hacked on (it has the "BAL"/balanced lot size mod), it appeared that using PipStep overrode the "All Trades Must Be In Profit" criterion, i.e. if you have pipstep on, then only that mattered, and the profit-trade criterion was ignored.

Was that intentional?
  • #689
  • Jan 28, 2011 10:04pm Jan 28, 2011 10:04pm
Quoting Nicastro
What you are suggesting is what some traders are doing manually. i dont think anyone has yeat tried to codify such an approach.
Would that solve the margin problem?
For sure it would reduce margin use, because we are trading less pairs and of course using the recovery which is another new feature
Sounds like a totally different strategy, which may or may not work, but not really what I was looking to achieve here

The balance of the 6 phantom pairs obviously works well overall, all are reporting profitable closed baskets. It is only on the extended baskets with many open trades do we get in trouble, as much with our own emotions of fear and panic as with the large DD and reducing margin level and looming possible margin call ... it is this I am trying to find a way to avoid/reduce

Less open trades equals less margin use
The CutLoser strategy goes some way to reduce the number of trades, but it also reduces the balance/equity and hence requires a larger pop to recover these closed losses
I was looking to keep the balance/equity stable throughout the basket so we only ever need the same sized pop to close a basket and to do this we need to close both a profit trade and a loss trade equally

I too have recently been doing this manually, once margin level has reached 250% I have been looking to cut the 2 trades. This 250% level still leaves a good number of trades in the basket to look for the pop.
A larger or smaller level may in time prove to be better
My choice of which trades to close is fairly random and has no real logic applied to it, I try to close pairs that have multiple open positions if possible so as not to remove any pair from the basket totally

My baskets are still closing successfully
Are they closing any better, or quicker than those left without any manual intervention ? ..... I have no idea I think I need to set up a test on this
BUT ...... I am not constantly in fear of the margin call and feel more in control of the situation ( this may of course just be an illusion )

The question I am now looking to find an answer to is which pairs is it most beneficial to close ? If I close a EURUSD in profit which pair would be best to close in loss, if I close a EURJPY in profit which pair to close in loss etc etc ( or does it even matter )

Do you or anyone else have any thoughts on this
  • #690
  • Jan 28, 2011 11:48pm Jan 28, 2011 11:48pm
Quoting frazerd
I uploaded the AddSix_02ax some time ago on this thread and I have been running it on a demo since 29 Nov. For some unknown reason the start time will either put you into a favorable cycle with the EA closing baskets often or someone who starts his just a few minutes earlier or later gets into a unfavorable cycle. I wish we could come up with the solution to entry timing. I think it is a big factor affecting performance.
I was missing your regular xls sheets for quite sometime.It is so helpful.

About timing factor, I have a feeling that, using Pipstep alone (say 25 ) and doing away with additions based on timing could be one solution.
Whenever there is sudden change in direction of a pair, there is no way we can catch the move early, and it hurts in the present scenario with any single pair has many positions open in one direction. So we enter whenever there is movement in terms of pips rather that wait for fixed period to lapse.
I don't know whether this goes against the principle of hedging.
"Put all your efforts in becoming effortless"
  • #691
  • Edited 1:50am Jan 29, 2011 1:42am | Edited 1:50am
Quoting jcs
The last basket has just closed, with a 18 $ loss ? ..... EDIT This may well be due to movement in price during time period needed to close all trades, at this lot level it only needs a few pips iether way to make a big difference ... is there no way to close out trades quicker, or is this just another MT4 limitation ?

I attach screenshot of closed trades so you may analyse

Nicastro .... Has this done what you intended ?
Please let us have your analysis

Attachment
Hi JCS
Your edit answers your query. I know of this problem. We will need in time to speeden up the Close All function if we can.

On looking at the screenshot, I noticed what looked like 2 recovery trades for both GU and EY. There should only have been 1 recovery trade entered for each pair.

Could I please ask Roundrock to look at the code again
  • #692
  • Jan 29, 2011 4:01am Jan 29, 2011 4:01am
Quoting Nicastro
Just as I thought - nothing but drift and DD over the slow periods. It is much better to be out of the market altogether

Ideally, we should aim to be in a position where we have closed the last basket during the US morning session and then wait for the London open to start again.
Interestingly only 2 of the trades in the basket would have been opened if StopAfterNoTrades had been set to true at the times you suggested
So looks a good idea and I hope Roundrock can introduce this into the ea as we have requested

As far as waiting for London to open before turning it back to false I think I would do this a few hours earlier, so I have some trades in the basket and ready to move at the London open as there is often a move at that time
  • #693
  • Jan 29, 2011 4:35am Jan 29, 2011 4:35am
Quoting Nicastro
Looks fine so far achab. At this pace it will beat any type of return available elsewhere.
Lets see how it continues to perform
It's performing very well:
P/L: +674.20
Floating loss: -418.20
Actual balance: +256.00
Monthly return: +40.9%

Next Friday I'll post the new complete results (available here: http://www.mt4i.com/users/achab-phantom6_01xr).
  • #694
  • Jan 29, 2011 4:39am Jan 29, 2011 4:39am
Quoting charvo
Hi,Achab:
Thanks for your insights and trading reports! Great Job!

2ax = addsix_2a + add more pos for a pair only when previous pos are positive.

and you also know addsix_2a reset all phantoms every 8 hours (by default in external setting)

that's the difference, imho, from 3ax that does not reset every X hours.

the coder who edited all versions of 3x, 3ax, or 3bx used the entry logic pretty distant from MC's original, so use them very cautious, I'm not quite sure what those 3... versions aim at.
So, after reading a lot here and your last answer, do you suggest me to use the version "2ax" because of the importance of "X" factor?
/achab
  • #695
  • Jan 29, 2011 4:41am Jan 29, 2011 4:41am
Quoting frazerd
I uploaded the AddSix_02ax some time ago on this thread and I have been running it on a demo since 29 Nov. I also am using the CL option on that demo. It is not the best performing AddSix EA among my demos. AddSix_01x_CL has produced almost 3 times as much profit over the same time period. I am running 5 AddSix demos, each has its own tab in the attached spreadsheet.

There is a lot of difference in how the EAs perform for various individuals. The one that works best for me is not always the one that works best for someone else. As I have stated...
Another difference is what Provider is using. In fact, I am testing many different Brokers and, even if the demos are identical and the EA started in the same time (well, same minute), the results are always different.

I am studying which provider is more profitable and the final result will be one of the reasons to make a choice where to open a live account.

Suddenly, can you confirm me that Phantom_AddSix_02a works only with floating platforms (and doesn't work is the trading platform uses fix spread)?

Nice weekend.
/achab
  • #696
  • Jan 29, 2011 6:22am Jan 29, 2011 6:22am
Quoting jcs
Roundrock

Can Nicastro’s suggestion on restricting the hours when new trades are allowed to be opened be automated please
With ‘ StopAfterNoTrades ‘ set to true can we have additional user parameters to set hours to stop and start this feature and also to reduce the ProfitPct during these hours

Thanks

here you go .....

the start, finish hours are as per your broker times
Attached File(s)
File Type: mq4 Phantom6_01xR.mq4   60 KB | 293 downloads
  • #697
  • Edited 7:43am Jan 29, 2011 6:59am | Edited 7:43am
Quoting roundrock
here you go .....

the start, finish hours are as per your broker times
Thank you ..... EDIT ... there is no parameter to change the Profit % when StopAfterNoTrades is set to true .. can we have one please
  • #698
  • Jan 29, 2011 12:18pm Jan 29, 2011 12:18pm
Thanks roundrock! awesome programming...stupid question.... but I am assuming that a recovery trade is an attachable ea that we should use in addition to the phantom that is already running that has gotten in trouble from market conditions?
  • #699
  • Jan 29, 2011 9:18pm Jan 29, 2011 9:18pm
yes, i have traded LIVE using versions that DO NOT reset for every X hours. unfortunately, the performance is much worse than Addsix_02a which resets every 8 hours.

so i think, reset is important,

then, "X" factor is also important, it is apparently wrong(imho) to open positions when previous pos are negative. but "pipstep" allow you to quantify exactly how much price distance for each added position, instead of "X" factor that only requires " > 0 ". my own feeling is pipstep is great to be 6~10 pips. that is, all previous pos have more than 6~10 pips floating profit.

so, reset, plus X factor (or pipstep), it is 2ax, then.

Quoting achab
So, after reading a lot here and your last answer, do you suggest me to use the version "2ax" because of the importance of "X" factor?
/achab
  • #700
  • Jan 29, 2011 9:28pm Jan 29, 2011 9:28pm
results will be definitely different, even if two identical trades on one same broker.

MT4, now i believe is designed to be flawed in real-time trading.
server pause, your home internet, EA hold off, broker manual backstage pause (one person pause your entries for 5 secs, don't be surprised)...... all these factors make your LIVE trading a stochastic performance that will be unique.

but one thing i'm sure is that the more reputable/honest the broker is, the easier this EA will trade.

This EA is waiting for "big moments", that's the basic truth. if your capital is sufficient to sustain till that moment, you'll win. if your broker plays dirty tricks so that you missed that moment, you have to make sure your capital is sufficient to sustain till another "that moment". that's why, an honest broker is more important.

it seems not quite related with ECN or market makers those crap...

just cheap 1 cent...

Quoting achab
the results are always different.

I am studying which provider is more profitable and the final result will be one of the reasons to make a choice where to open a live account.

/achab
4 traders viewing now