Art Phantom EA | Page 23 | 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

  • #441
  • Nov 25, 2010 2:53pm Nov 25, 2010 2:53pm
Quoting JulesvH
Just for fun and to understand the principle of hedging. Open, on a demo account, 0.1 lot of the following pair at the same time (or quick after each other)
Buy EURJPY
Sell EURUSD AND USDJPY

Look how much the starting loss is (because of the spread) Let it run for while and notice how the total loss changes only a bit, and how the profit/loss of the 3 individual pairs change.
But thank you.....
  • #442
  • Nov 25, 2010 3:11pm Nov 25, 2010 3:11pm
This strategy won't work you have to apply some kind of R:R
  • #443
  • Nov 25, 2010 3:24pm Nov 25, 2010 3:24pm
Quoting sidhujag
This strategy won't work you have to apply some kind of R:R
I have test it manually today and it works for me, please explain me what have I made wrong ?
  • #444
  • Nov 25, 2010 3:51pm Nov 25, 2010 3:51pm
Quoting Caillou
Because the most important principle of this EA is to use a perfect syntetic hedge to stop a loosing decision and to take the right one at this moment and again and again till we reach profit......and what is the most perfect REAL HEDGE ?
Perfect syntetic hedge happens when we buy all pair at the same time. So as i said in above post.. I trying this...
buy/sell all 3 pairs at a time and then balance this basket with this ea
  • #445
  • Nov 25, 2010 3:56pm Nov 25, 2010 3:56pm
Quoting Caillou
Yes, only ONE pair. For non U.S resident no problem in the same account but for U.S resident maybe 2 accounts is the way to do that.

Any coder could modify Art_Phantom_Full to trade only EUR/USD ? EUR/USD must be in Phantom Buy Group and in Phantom Sell Group. Then when we run the EA a Phantom Buy Trade and a Phantom Sell Pair for EUS/USD begin. Each 30 min (at 15 and 45) if the Phantom Trade is in profit and if all previous Real Trades are in profit then the EA opens a Real Trade (Buy or Sell).

As I' m not a U.S resident I have no problem...
Yaa me too desperately looking for this strategy. I was trying this for two pairs EU and UC.. but not able to code. so i created 2 diff eas.. but no trade till now!
  • #446
  • Nov 25, 2010 4:02pm Nov 25, 2010 4:02pm
Quoting 24hrs
Perfect syntetic hedge happens when we buy all pair at the same time. So as i said in above post.. I trying this...
buy/sell all 3 pairs at a time and then balance this basket with this ea
Correct, and when this perfect syntetic hedge arrives after a small loss, then the loss will not move till you break again this hedge....
  • #447
  • Nov 25, 2010 4:16pm Nov 25, 2010 4:16pm
Quoting sidhujag
This strategy won't work you have to apply some kind of R:R
Please, explain me what is wrong in this EUR/USD example movement:

1.- 8h00: 1,3600. Phantom Trades start ( Buy and Sell)

2.- 8h15: 1,3610. Real Trade Buy

3.- 8h45: 1,3620. Real Trade Buy

4.- 9h15: 1,3590. Real Trade Sell

5.- 9h45: 1,3640. Real Trade Buy

6.- 10h00: 1,3680. Close Basket

Results: trade 2 (+70) + trade 3 (+60) + trade 4 (-90) + trade 5 (+40) = +80 ( or wathever PP)

I know it is a bit late here in Spain and maybe I' m tired but " This strategy will not work" ..... why ?
  • #448
  • Nov 25, 2010 4:19pm Nov 25, 2010 4:19pm
Quoting Caillou
Yes, only ONE pair. For non U.S resident no problem in the same account but for U.S resident maybe 2 accounts is the way to do that.

Any coder could modify Art_Phantom_Full to trade only EUR/USD ? EUR/USD must be in Phantom Buy Group and in Phantom Sell Group. Then when we run the EA a Phantom Buy Trade and a Phantom Sell Pair for EUS/USD begin. Each 30 min (at 15 and 45) if the Phantom Trade is in profit and if all previous Real Trades are in profit then the EA opens a Real Trade (Buy or Sell).

As I' m not a U.S resident I have no problem...
I am with you on this one. It will work just the same as a three pair basket, but with less DD as the reversal is most efficient. Also less spread costs (especially if you do it on the E/U).

Further if we could code in the ATR filter I mentioned this would be more efficient that a straight - arbitrary - time filter that we have now.

Tim
  • #449
  • Nov 25, 2010 4:26pm Nov 25, 2010 4:26pm
Quoting timmy
I am with you on this one. It will work just the same as a three pair basket, but with less DD as the reversal is most efficient. Also less spread costs (especially if you do it on the E/U).

Further if we could code in the ATR filter I mentioned this would be more efficient that a straight - arbitrary - time filter that we have now.

Tim
Thank you Tim,

I was thinking I was going crazy. I will never understand why someone with his own thread instead of trying to improve his EAs must go to other threads to say "this will not work" and leave...... He must not have too much confidence in his EAs.
  • #450
  • Nov 25, 2010 4:36pm Nov 25, 2010 4:36pm
Quoting sidhujag
This strategy won't work you have to apply some kind of R:R
What are you on about? If you make a statement like this please provide your reasoning.

It will work no different than running 3 instruments (pairs/crosses if you like) versus just one instrument to do the same job - and more efficiently.

This EA lets price action be the ultimate decision maker on which way the it goes to profit. Doing this with one instrument or three it does not matter - just more efficent with one pair.

The key for us is to get a good filter that allows trade sequences every chance of completing within a market with some momentum.

As mentioned put an ATR(6 period) on your 1hr Euro/USD chart and attach over this a SMA(20 period). Everytime the ATR(6) moves = or > than the SMA then you have potentially some momentum. So this could be our volatility/momentum filter.

The time filter we have now will ensure that we don't start trade sequences within the Asian time zone as well.

And with a reasonable level of risk we can make ~5% per week (trading, say once per day for a 1% pp) and the EA will dig itself out of most flat periods with only a small pip target if the risk is set correctly - should the expected market momentum not eventuate for a particular day.

Finally an overall emergency Close all at % DD can allow us to sleep well.

Time will tell what level this would best suit. But I'd likely place it at ~20% to start with...to catch to big daddy death trade.

Any coders keen to convert the full version with these mods (could convert the-games latest full version with time filter & close out worst trade).....

I am sure this will give us a powerful EA that efficently trades the net movement of the Phantom pair within the basket, that has an edge by only trading during volatile periods and never duringthe usual quiet Asian session. The other 'edge' for this EA is that it uses pure price action to define its direction - no bullshit indicators (the Momentum indicator I mention above only indicates market volatility and possible momentum).

Tim
  • #451
  • Nov 25, 2010 4:47pm Nov 25, 2010 4:47pm
Quoting timmy
What are you on about?...
Well said. Hope a coder could help us.......
  • #452
  • Nov 25, 2010 4:55pm Nov 25, 2010 4:55pm
Good evening, I' m going to sleep. See you tomorow.....
  • #453
  • Nov 25, 2010 5:02pm Nov 25, 2010 5:02pm
Quoting Caillou
Good evening, I' m going to sleep. See you tomorow.....
Sleep well Caillou

I assume everyone understands how to put a SMA(20) over the ATR(6). If not, just drag the SMA over the ATR then click on the "Apply to" and click on 'Fist indicators data'... and you have the moving average of the Average True Range on the same indicator.

For the SMA I use either period 20 or maybe 12.

Have a look at it - it breaches up over the moving average nicely ahead of the daily move - on most days. It won't tell you which direction the market will go (so it is a useful indicator for us - no bull$hit prediction of the daily move), the EA will then use price action to get us into the right direction and take us to profit.

Cheers
Tim
  • #454
  • Nov 25, 2010 5:17pm Nov 25, 2010 5:17pm
Quoting timmy
What are you on about?...
Well the R:R is not good enough given the edge it provides.

Based on what people have been showing backtests/forward tests... the 20% mark is easy eclipsed before you could actually make 20%.

If the EA provides a stable equity curve long term through different types of market environments... ranging to trending back to ranging then it will be on to something.

That being said.. the amount of time to test it is not worthwhile as it is...

If someone uses this idea to provide a vehicle to create long term equity gains I will invest 100k in them that's how confident I am that you are just wasting time.

That's not to say the idea isn't good. You may apply R:R to actually increase your odds... but then its not the same EA... the idea of using a basket is what I got out of all of this. The more instruments you look at to base the same decision the better your "edge" is. I am just telling you so you don't get false hope and waste time.

If you really want to get somewhere with this strategy or any strategy you have to find a way to test it better.. I've been saying this forever but alot of effort is wasted on just tweaking on times and other features that may or may not get a better EA... if you can backtest you can prove what you have is actually good through different types of markets aswell as the limitations to know what you should expect.

Mt4 is obviously easy to implement but its also easier to lose your account aswell than to do something proper and figure out a way to backtest either through Dukascopy or some other platform that allows multicurrency backtesting.

FWD testing is just half the battle and you can't purely rely on it with this type of EA because we as humans love to temper and test different settings as soon as we think it will help for a certain instance of time without understanding the full implications of the change you just made.

So I would love for my own EA to prove me wrong but even as its up 10% DD got up to 10%... that to me is not good odds esp with the state of the markets and what they can do in one day even with the best money management.

If you're happy with the fact that any day you may lose your account hoping to double up so you can have a free ride then go for it... but I warned you in any case. Some of you may understand where I'm coming from but to the ones that don't then disregard me and continue on... its better proving someone wrong now that you got motivation (and no that's not reverse physcology if that's what you're thinking)

I really do hope for everyones prosperity... The big dogs have all the money and we would be happy with their crumbs... I would rather one of you have it instead of them since the greedy bastards are what got the countries in this mess in the first place.

So what's your stance? You think I'm wrong? Please elaborate...
  • #455
  • Nov 25, 2010 5:29pm Nov 25, 2010 5:29pm
Quoting timmy
What are you on about?...
Sure you let PA be the dictator but at what odds? How do you define odds with this system for expectancy of profits? The only way is to FWD test and make sure your 20% DD won't get hit before your 20% profit.
  • #456
  • Nov 25, 2010 5:36pm Nov 25, 2010 5:36pm
Quoting timmy
I am with you on this one. It will work just the same as a three pair basket, but with less DD as the reversal is most efficient. Also less spread costs (especially if you do it on the E/U).

Further if we could code in the ATR filter I mentioned this would be more efficient that a straight - arbitrary - time filter that we have now.

Tim
ATR filter may work and volume may be even better but there usually is a "set up" time when the big boys will lure in the pigs and buy at a discount or sell when it is expensive. This is done a few hours into the session.

With your EU example yes it will be less DD.. but at what cost? Won't it be harder to reach your PP? This comes back to the original problem and you're running in circles.
  • #457
  • Nov 25, 2010 5:39pm Nov 25, 2010 5:39pm
Quoting Caillou
Please, explain me what is wrong in this EUR/USD example movement:

1.- 8h00: 1,3600. Phantom Trades start ( Buy and Sell)

2.- 8h15: 1,3610. Real Trade Buy

3.- 8h45: 1,3620. Real Trade Buy

4.- 9h15: 1,3590. Real Trade Sell

5.- 9h45: 1,3640. Real Trade Buy

6.- 10h00: 1,3680. Close Basket

Results: trade 2 (+70) + trade 3 (+60) + trade 4 (-90) + trade 5 (+40) = +80 ( or wathever PP)

I know it is a bit late here in Spain and maybe I' m tired but " This strategy will not work" ..... why ?
Sure Caillou that's great... my EA with the same concept also closed in profit today... but give it a few weeks and you'll see. Set your DD cutoff to say 20% as Timmy says and in 2 months lets see the balance. Then try on a real account with minimum money say $500 and see the same result. That's 4 months wasted and $500 cleaned out. I've done this before on ideas that seem good at first... and I've looked at basket trading and other strategies for a long time... but it ALL COMES BACK TO MM AND R:R.

You will never win with martingale.. see LTCM. And with the kind of Risk to Reward this system offers the long term equity curve will be decreasing.

If you don't believe me then ask a 10+ year trader on here... Hanover had a good spiel about why Martingale sucked even if your system hasn't lost in 10 years.

A little story... I made a scalping EA for EURCHF... called it the super scalper it hadn't lost since 1999 in backtest and FWD test for 6 months. It made 1000% every year because of big lot sizes... however the SL was set to something like 20% of my account. I went live and it hit SL 2 weeks after I started. I couldn't handle the fact that an SL was hit so I doubled down (that was the Greece incident)... and eventually had to manually intervene to get myself out after like 80% of my margin used.
  • #458
  • Nov 25, 2010 6:04pm Nov 25, 2010 6:04pm
Quoting sidhujag
Well the R:R is not good enough given the edge it provides.

Based on what people have been showing backtests/forward tests... the 20% mark is easy eclipsed before you could actually make 20%.

If the EA provides a stable equity curve long term through different types of market environments... ranging to trending back to ranging then it will be on to something.

That being said.. the amount of time to test it is not worthwhile as it is...

If someone uses this idea to provide a vehicle to create long term equity gains I will invest...
Firstly let me say that this is all an experiment - and that the reality here is that the initial EA will likely produce a negative expectancy. I am not blindly doing this assuming that it can be profitable in the basic form. I have been trading for a living for many years and would not have survived if I couldn't find a system with a true edge. (I mainly trade news after the event)

The idea of using a basket simply allows price action to dictate which currencies are strongest and to let the market dictate what the mix of pairs will take us into a net profit position.

To create a positive expectancy we need an edge for this system. The EA is simply a tool to take advantage of price movement in its simplest form - and the more pairs we trade with the greater the potential DD as the more unrelated currencies take longer to react to each other, than say two primary pairs and there associated cross. (The three pairs approach as we have it currently)

The 20% max DD is relative to risk and purely arbitrary. If I trade small contacts it may not be hit for a long time. Now that is not to say that I have made more than 20% every time it gets hit -BUT - if I am able to apply an edge to this process then I have a chance of doing so (positive expectancy). Maybe applying a volatility/ momentum filter will help with this.

By using a one pair 'basket' (eg E/U ) we can backtest filter ideas to see how they impact on the functionality of the EA. You may well be right in that three pairs may work out to be more contructive as a basis for the EA to work with.

Of course this EA will fail in certain markets as it stands - you are not telling us anything we don't know already. Currently it will always fail under a fantail pattern (ie ever increasing range at both ends - with price bouncing back and forth). So being able to backtest ideas with a one pair basket will help to manage this. it maybe that the failure pattern can be 'contained through filtering and that the EA is then able to make profits greater than those losses at the fail safe point (what ever we deem this as).

Explain what you mean by applying R:R to increase the odds. You do not present any constructivity to your argument. Certainly if you push out the target pp to a larger number and filter it on the basis of a longer term trend, then it can be done with the EA pretty much as it is. For example with another filter such as a simple SMA of longer period. When the price action crosses up over a long term SMA filter and the EA moves into a net profit based position to the buy side (which it would do) then the EA simply maintains the pp as open and will look to close all when the price action reverses back over the long period SMA. This will provide excellent R:R using the base EA we have developed so far but with these type of additional filters.

This is all future development ideas. I agree with the sentiment of the other comments you make, so feel free to bring something constructive to the table.

Tim
  • #459
  • Edited 6:26pm Nov 25, 2010 6:14pm | Edited 6:26pm
Quoting timmy
Firstly let me say that this is all an experiment - and that the reality here is that the initial EA will likely produce a negative expectancy. I am not blindly doing this assuming that it can be profitable in the basic form. I have been trading for a living for many years and would not have survived if I couldn't find a system with a true edge. (I mainly trade news after the event)

The idea of using a basket simply allows price action to dictate which currencies are strongest and to let the market dictate what the mix of pairs will take us into...

By the time SMA crosses the ATR the best setup has already passed in alot of cases.

Feel free to go to my thread to see what I'm talking about regarding the R:R.

You wait for a breakout then get in with min 10:1.

We again come back to testing... if we can test multiple pairs we can identify if its better to even have multiple pairs vs single pairs through backtesting. Yes single pairs will allow you to dictate when reversals happen quicker but then again you will fall for more headfakes aswell. Applying multiple pairs is like averaging your position.. by doing this you are hoping to catch the trend but at what cost we don't know yet. If we can conclude that the cost isn't as high as the benefit 51/100 times then sure we can have an edge... but at this point through fwd testing all I'm saying is we are wasting time that we can spend on actually improving our success in forex by focusing on the testability of our idea.

Whenever you program something you have to have a way to test its parts and test it as a whole... we are missign half the picture of testing the idea. I think by doing this we will see glaring problems with the logic we deduced to put in the EA even though it may seem likes to be according to everyone's best interest.

I did put something constructive ont he table... if you see it then it may be of benefit... but I'm certainly not taking anything away other than those who think it may be a nuisance to tell others what's right and what's wrong.. and I'm only doing so because I've been here and done that and gone full circle before.... you learn from your mistakes so why not teach others before they make the same.

You don't want to limit your exposure to testing just because Mt4 doesn't allow it... this is not acceptable and shouldn't be for anyone who is not serious. If you see a problem with this then you will have a way harder time being successful in this environment. Being able to test single pairs may be all you need to test this strategy but then its really the same strategy we started with and not one I am commenting on.
  • #460
  • Nov 25, 2010 6:23pm Nov 25, 2010 6:23pm
Quoting sidhujag
ATR filter may work and volume may be even better but there usually is a "set up" time when the big boys will lure in the pigs and buy at a discount or sell when it is expensive. This is done a few hours into the session.

With your EU example yes it will be less DD.. but at what cost? Won't it be harder to reach your PP? This comes back to the original problem and you're running in circles.
I looked at your thread and never posted because you spent a lot of time speculating about "when the big boys come into the market", and what the market does, and how you have set your EA to work with this etc, etc. The market is just all the price movement - sometime it is Hedge funds, sometimes it is sovereign funds, sometimes it is news events ...whatever makes the price action is the market. Speculation is a waste of time. Patterns you think you see are just the human brain trying to make order out of chaos. Your story on your Euro/CHF scalper is proof of this.

As to your second point. Why would it be harder to reach the pp - if you have three buys and 1 sell then price moves up - you will hit your pp with a one pair EA. Same if you have a buy 3*E/U, Buy 1*U/J and Sell 1* E/J, its just the synthetic net resulting pair that moves you to your pp (the example is not exact only to prove a point). It might be quicker or slower since the timed 15 and 45 min entry period effects which pairs also open and when.

Both setups get effected by the fantail failure pattern (although more complex for the three pair model).

It is all about the EA simply managing the price action, with smart filters built around it - this is why this EA has potential.

Tim
4 traders viewing now