Jag's Phantom 6 System | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phantom 6 System 1,540 replies

JAG Trade Journal 0 replies

JAG's Journal 0 replies

Phantom high, phantom low 0 replies

JAG fx Trade Journal 10 replies

Attachments: Jag's Phantom 6 System
Exit Attachments

Jag's Phantom 6 System

  • #41
  • Nov 15, 2010 12:42pm Nov 15, 2010 12:42pm
Quoting sidhujag
May be a bug in the code too... since it may not close all at once then it doesn't check again to close... but V2 I think is better in closing trades if they shouldn't be open. It sucks that testing takes so long to even get it going. We NEED to be able to test this quicker.

I emailed Dukascopy about the stupidity of having to go through email response to get a demo account and only lasting 15 days.
can u do backtesting on Dukas java platform? I've heard it can handle multiple Pairs during the strategy testing process.
  • #42
  • Nov 15, 2010 12:56pm Nov 15, 2010 12:56pm
I did V1 and V1x almost 12 hours ago,no trade at all.
regards
fred
Quoting sidhujag
Fred,

Which version?

V1 will open new basket only during LO or AO and V2 will open new basket during first few hours of LO/AO.

Jag
  • #43
  • Nov 15, 2010 1:27pm Nov 15, 2010 1:27pm
Quoting TradeRunner
I re-installed v1 and tried 2 different PrPct settings (deleting GV's) but the screen display reads $0.00 profit target with trying 2 different settings on 2 different accounts ($0.00 on both accounts)..

Try IBFX or some of the other brokers for demo. Takes about 5 minutes to get a demo and IBFX runs for as long as you use it.
Like I said you are OFF HOURS you will not get a profit target of anything else during off hours. This is due to the fact that if a basket was open it would try to break even now. I will fix this in the next release...

I know IBFX is unlimited. I mean Dukascopy because of the testing capabilities.
  • #44
  • Nov 15, 2010 1:28pm Nov 15, 2010 1:28pm
Quoting getting
can u do backtesting on Dukas java platform? I've heard it can handle multiple Pairs during the strategy testing process.
Yes the Java platform handles multi-currency backtesting with spread and its 99%
  • #45
  • Nov 15, 2010 1:30pm Nov 15, 2010 1:30pm
Quoting fredgold3434
I did V1 and V1x almost 12 hours ago,no trade at all.
regards
fred
Can you reset your global variables and reset the EA? Also depending on your broker it may try to trade asian session... in a few hours. Depends on your broker because most brokers don't support the pairs I have for asian. I use FXPRO.
  • #46
  • Nov 15, 2010 2:32pm Nov 15, 2010 2:32pm
Yay for Grumpy the clown!!

http://www.washingtonpost.com/wp-dyn...111102903.html
  • #47
  • Nov 15, 2010 2:49pm Nov 15, 2010 2:49pm
quote first post in phantom 6 system
"Every hour at minute 0, the EA adds a real trade for the pair that is the biggest winning phantom pair among the six pairs."

On "Jag V1" - I don't see EA opened trades with the same logic as above.
Does the winning pairs & basket(either buy or sell side) needs to be in the positive territory(winner basket) for trade to open on :00 or :30(attaching to M30 chart)
Happened so fast yesterday, forgot to pay attention.
thanks
  • #48
  • Nov 15, 2010 6:12pm Nov 15, 2010 6:12pm
Quoting sidhujag
Like I said you are OFF HOURS you will not get a profit target of anything else during off hours. This is due to the fact that if a basket was open it would try to break even now. I will fix this in the next release...

I know IBFX is unlimited. I mean Dukascopy because of the testing capabilities.
Got it ! Actually I'd forgotten all about the trading hours.
WAY too much testing.
  • #49
  • Nov 15, 2010 6:48pm Nov 15, 2010 6:48pm
Quoting getting
quote first post in phantom 6 system
"Every hour at minute 0, the EA adds a real trade for the pair that is the biggest winning phantom pair among the six pairs."

On "Jag V1" - I don't see EA opened trades with the same logic as above.
Does the winning pairs & basket(either buy or sell side) needs to be in the positive territory(winner basket) for trade to open on :00 or :30(attaching to M30 chart)
Happened so fast yesterday, forgot to pay attention.
thanks
Hi,

It only adds a trade to the winning phantom (starts a basket) at the start of the session (first two hours to find a entry).

Then it will add to the basket if it is best phantom and real trades of that pair are all in profit... this is v1x and v2.

I am in the process of removing the pair global vars and using mql iOpen to check in the history for our reference to use for phantom PL calculation. This is the more intuitive way to do it... and you will also not lose this data if you reset the EA by mistake or accident.

I am just testing it now will REUPLOAD V2 soon. Its basically the same but more housecleaning.
  • #50
  • Nov 15, 2010 7:08pm Nov 15, 2010 7:08pm
OK,

You can Redownload V2. I fixed V2 so now it doesn't use pricing info from globals... uses iOpen() to check the history. PrepLOHour and PrepAOHour are the times to use as reference for each respective session to calc phantom PL.

The hardest part of the EA will be handling multiple sessions with trades closed at BE from before.

When London or Asian session is on you can see the PL and manually confirm that this number is good (make sure you take into account spread).
  • #51
  • Nov 16, 2010 1:19pm Nov 16, 2010 1:19pm
I cleaned up the way V2 handles time(RE UPPED)... and wrapped the trading time functions so its cleaner... I got a trade on my home computer at 2 hours after session open (technically not allowed) and it ended up opening a loser basket. This was due to the fact that the broker and GMT time are not in sync (off by a few mins I think?) So I'm not using brokers time anymore at all... use GMT for everything.

Right now looks like the way V2 checks for phantom winners causes more drawdown at first... V1 seemed to close quicker... but that's because at Frankfurt open the trend was set and retraced at MID London... normally this isn't the case its the other way around... Frankfurt sets up London and then it goes until NY.... we'll see what happens.
  • #52
  • Nov 16, 2010 3:52pm Nov 16, 2010 3:52pm
Is someone running V2 that we can see using myfxbook or something?

I'm trying a new idea in V3 to have two versions running together trying to achieve 1 pct profit.

The two versions would be as follows:
-latest V2
-V2 using V1x times for phantom reset.

Will see how this goes.
  • #53
  • Edited 7:01pm Nov 16, 2010 6:45pm | Edited 7:01pm
Fixed yet another bug! Damn.

A new bar wasn't being properly detected based on the timeframe of the chart. Now whatever chart the EA is placed it will trade based on that timeframe... it was hardcoded to every 0 minute.

There was also a bug with spread with some pairs... causing them to have higher PL than they should... AUDJPY was factoring in only 0.5 pip spread! Damn mql API doesn't give proper info...

For some reason it MODE_SPREAD gives 50 for GBPJPY but 5 for AUDJPY... they both have the same spread... and both are 2 digit pairs. Weird.

BTW when updating to the newest version on V2 you can simply replace your current EA with this one on the chart and you don't have to worry... it should detect the state of the EA properly... unless your EA has a basket open with some closed trades already from BE.
  • #54
  • Nov 16, 2010 7:27pm Nov 16, 2010 7:27pm
Did you fix V1 as well, or just V2 ?
  • #55
  • Nov 16, 2010 7:43pm Nov 16, 2010 7:43pm
just V2... V1 is fine because I rewrote the core in V2. The bugs in V1 were the ones present from the original EA I think.
  • #56
  • Nov 18, 2010 10:52am Nov 18, 2010 10:52am
Quoting getting
any other pairs to substitute GBPNZD?
Hey Getting,

Wonder why V1 took EU sells at 9:30/10. It shouldn't have??

PrepLOHour is 8 and LOHOur is 9 and at GMT 8 there was a spike up on EU... why did it sell after this? Only way it should have sold is if it was the biggest winner and it was lower than what it started off with at GMT 8.. and that didn't happen.

EDIT: I think I know what happened! There were two 1% equity pops during first hour of London session. I let the EA open new baskets only at LOHour and one EJ popped at 9 and then it reset the phantoms when you close all in profit... from 9 to 9:30 EU was actually going down and it sold this since the hour is still LOHour.. and then did another 1% equity pop!

wow... not sure if it should reset the phantoms like that but lets just go with it.
  • #57
  • Edited 12:54pm Nov 18, 2010 12:41pm | Edited 12:54pm
[quote=sidhujag;4182831]Hey Getting,

Wonder why V1 took EU sells at 9:30/10. It shouldn't have??

I don't know - u tell me....
great job!!!

looking good so far after five days of testing.
some of the early advantage over other phantom In my opinion:
-reduced in "Avg trade length"
-reduced in "Trades per day"
as more data comes in I'd have a better judgment, but for now I'm loving it!!!

another interesting result from Caillou:
http://www.myfxbook.com/members/Sergioar
  • #58
  • Nov 18, 2010 2:10pm Nov 18, 2010 2:10pm
This only worked last 5 days for you because since start of frankfurt one direction has been picked. Asian session seems to be good because you're using a very high lot size for the size of account... its probably going to blow up on you but lets see how far it goes... it only need like 15 pips to close 1%

From what I deduce about the way the phantoms work between different times... we have V2 which has a longer time between phantom reset and real trade... (longer MA) and we have V1/V1x which trade in shorter term (shorter MA).... longer MA is taking more time to close because as NY nears price has been changing direction. While the shorter MA wins because right at start of a session it will take the current trend.

I have an idea in V3 to allow two bots to run on same account for 1% total profit.

Idea is to use one V3 with shorter MA and one with longer MA... when MA's are going in the same direction you win... when they are going opposite you add DD.. .but incase the trend actually changes... you have a headstart.

I think the fact that we choose good times and the currencies go on a good run (less headfakes) means this strategy may be better because the chances of both MA's to go in the same direction is higher at this time.

I'm running it on my work comp...up 2% in 2 days so far starting with AS yesterday.

EU was in DD but EJ helped out.
  • #59
  • Nov 18, 2010 5:10pm Nov 18, 2010 5:10pm
Hi Sid.......... I'd hoped to post some results but I couldn't get it (V1) to trade again unless I used primarily default settings. I was going to try some other time/pair settings but the EA didn't like the changes. May try again later.

Thanks,
TR
  • #60
  • Nov 18, 2010 5:37pm Nov 18, 2010 5:37pm
[quote=sidhujag;4183247]This only worked last 5 days for you because since start of frankfurt one direction has been picked. Asian session seems to be good because you're using a very high lot size for the size of account... its probably going to blow up on you but lets see how far it goes... it only need like 15 pips to close 1%

Kaboommm!!!
I'll reduced MM to used to 1 when the dd sink below previous low of float DD
Attached Image (click to enlarge)
Click to Enlarge Name: pic.PNG Size: 31 KB
0 traders viewing now