Basket 14 EA System | Page 100 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

BASKET TRADE Management: Using multiple trade management EA's on basket trades 11 replies

Basket Knights - A Basket Trading Round Table 1,313 replies

Relative Strength Basket Trading System 3,479 replies

Simple Basket Trading system 61 replies

TMS basket: Dashboard for a basket of MA-Filtered RSI Signals 135 replies

Attachments: Basket 14 EA System
Exit Attachments

Basket 14 EA System

  • #1,981
  • Dec 1, 2010 6:47am Dec 1, 2010 6:47am
Interesting how many people all of a sudden jumped on the new 3 pair thread (already at 58 pages) and left this one standing in its tracks... Human nature is quite silly at times.
  • #1,982
  • Dec 1, 2010 4:36pm Dec 1, 2010 4:36pm
Quoting Saidar
Interesting how many people all of a sudden jumped on the new 3 pair thread (already at 58 pages) and left this one standing in its tracks... Human nature is quite silly at times.
Silly? I call it adapting, what's the point of staying on the merry go round? I'd rather watch the merry-go-round... from afar
  • #1,983
  • Dec 1, 2010 6:27pm Dec 1, 2010 6:27pm
Saidar,

In my opinion, there is nothing "silly" about going to another thread which is exploring potentially very profitable concept.
At the same time, we all are very grateful to MetaCoder and Frazerd for all their work. Without them the other thread would not exist.

  • #1,984
  • Dec 1, 2010 6:29pm Dec 1, 2010 6:29pm
Quoting Saidar
Interesting how many people all of a sudden jumped on the new 3 pair thread (already at 58 pages) and left this one standing in its tracks... Human nature is quite silly at times.
I'm not convinced that the 3 pair versions are better than the AddSix_02a. I'm getting regular closures with the 02a. So far, that's the one I'd most likely go live with.
  • #1,985
  • Dec 1, 2010 8:42pm Dec 1, 2010 8:42pm
First time in a long time that I've seen any of you guys on this thread. Welcome back. I am monitoring the Art thread pretty closely and using some of the-game's mods on the 6 pair versions. I would start testing some of the 3 pairs, but I don't want to discard my AddSix and Phantom6 demos and I don't think my VPS can handle any more MT4 platforms.

So for the time being I am reading all the posts and trying to keep up with all the new developements. Sounds like folks are getting some good results with the full and light versions, but a few don't seem to be so lucky.
Frazer
  • #1,986
  • Dec 1, 2010 11:45pm Dec 1, 2010 11:45pm
Quoting the-game
Silly? I call it adapting, what's the point of staying on the merry go round? I'd rather watch the merry-go-round... from afar
What I mean is the basic human instinct to find that new "thrill". I see it happen all the time people get bored and if something new pops up they abandon the old one and flock to the new one.

Just an observation
  • #1,987
  • Dec 2, 2010 1:41am Dec 2, 2010 1:41am
Hehe, I'm still keeping an eye on this tread, but nothing new happens. Don't get me wrong, anything new doesn't have to happen! I think it is testing time. MC, Frazerd and others have put a great effort getting this thread and the Phantom thread at such a high level. With EA's that work! (Still looking forward on your weekly resultsupdate, Frazerd!)
I had Phantom Addsix03bx running on a 500:1 account with MMTU 0.2 and PP 0.25. I did nicely, but after a while I got in a drawdown. I waited for 3 days, but it didn't seem to recover. Maybe I was too impatient..... So I'm testing Art Phantom Light now, with the only mod to trade during the Europe markets. Just to see how it performs. I'm curious if the 3 pairs and trading window result in less DD. I only aim for 1% profit per day. I may even be a little less, if the DD is acceptable.
Maybe I should mod the Addsix03bx version with the Cut Looser (CL) option and give that one another try.
When using that mod, the MMTU, PP en CL_multiplier should be in a well tested/thought balance. (Like MC mentioned before with the MMTU and PP)

So you didn't get rid of me! Still hanging around.
Once again, thank you all, especially MC and Frazerd!
  • #1,988
  • Dec 2, 2010 6:05am Dec 2, 2010 6:05am
Quoting JulesvH
Hehe, I'm still keeping an eye on this tread, but nothing new happens. Don't get me wrong, anything new doesn't have to happen! I think it is testing time. MC, Frazerd and others have put a great effort getting this thread and the Phantom thread at such a high level. With EA's that work! (Still looking forward on your weekly resultsupdate, Frazerd!)
I had Phantom Addsix03bx running on a 500:1 account with MMTU 0.2 and PP 0.25. I did nicely, but after a while I got in a drawdown. I waited for 3 days, but it didn't seem to recover. Maybe I was too...
The AddSix_02a started out shaky for me with no closure for 8 days. I did see it in good positive territory during that time several times but it couldn't quite reach my TP. It had opened more than 70 trades in that basket. Since that time with a slightly higher lot size I haven't seen it go more than 20-25 trades in a basket with most baskets having 15-20 trades. Sure wish this could be back tested to see a worse case. I think someone had a script to do that on one of these phantom threads but I cant find it again. If I could see the 02a survive a 2-3 year back test I'd probably go live with it now.
  • #1,989
  • Dec 2, 2010 6:59am Dec 2, 2010 6:59am
Quoting JulesvH
Hehe, I'm still keeping an eye on this tread, but nothing new happens. Don't get me wrong, anything new doesn't have to happen! I think it is testing time. MC, Frazerd and others have put a great effort getting this thread and the Phantom thread at such a high level. With EA's that work! (Still looking forward on your weekly resultsupdate, Frazerd!)
I had Phantom Addsix03bx running on a 500:1 account with MMTU 0.2 and PP 0.25. I did nicely, but after a while I got in a drawdown. I waited for 3 days, but it didn't seem to recover. Maybe I was too...
I coded the basic version of Addsix in MT5 and I'm running some optimization tests, will show you how dangerous this EA can be if you don't manage it very carefully.

The problem is sooner or later you will hit a draw down that never recovers, I'm seeing this over and over and over and over again on backtests.

Some take a month, others two five years, but the end results is always the same.

I have seen a linear relationship between risk and "time until fail". The higher the risk, the more money it makes in the short run but the sooner it fails. The lower the risk the longer it stays active but the less money it makes over time.

Will post the code and charts soon to explain what I mean.

Don't get me wrong this EA is an excellent strategy and concept it just needs some filters and risk management that will cut the huge drawdowns somehow.

Hope MC will help regarding this once I post the MT5 code
  • #1,990
  • Dec 2, 2010 7:00am Dec 2, 2010 7:00am
Quoting Toothman
The AddSix_02a started out shaky for me with no closure for 8 days. I did see it in good positive territory during that time several times but it couldn't quite reach my TP. It had opened more than 70 trades in that basket. Since that time with a slightly higher lot size I haven't seen it go more than 20-25 trades in a basket with most baskets having 15-20 trades. Sure wish this could be back tested to see a worse case. I think someone had a script to do that on one of these phantom threads but I cant find it again. If I could see the 02a survive...
I'm on it with MT5 will post soon.
  • #1,991
  • Edited 7:36am Dec 2, 2010 7:07am | Edited 7:36am
Ok to get an idea here is a graph of an excellent test, but look what happens after a while... BOOM!

If we can get some people to study WHY this happens, and what is the CAUSE of this we might actually find a way to counter it. My firs guess is big news events.

Like I said it does not matter what settings you use, this crash happens sooner or later, the only difference is good settings will last longer like the attached graph which lasted two years.
Attached Image (click to enlarge)
Click to Enlarge Name: TesterGraphReport2010.12.02.jpg Size: 67 KB
  • #1,992
  • Dec 2, 2010 7:37am Dec 2, 2010 7:37am
Quoting Saidar
Ok to get an idea here is a graph of an excellent test, but look what happens after a while... BOOM!

If we can get some people to study WHY this happens, and what is the CAUSE of this we might actually find a way to counter it.

Like I said it does not matter what settings you use, this crash happens sooner or later, the only difference is good settings will last longer like the attached graph which lasted two years.
Saidar,
Thank you for the back test results. Could I ask a favor of you? In my opinion, the 02a may be the most consistent of the AddSix EA's. Could you run a test on the same period with the AddSix_02a EA with these settings? .4 MMTU, PP .5 using 50:1 leverage. I'm working with a $40,000 demo account. Your question on why the crash happens. The only answer is that market must have gone through a long period of stagnation to get you into a DD that couldn't recover. In your test how many trades were closed by margin call before the whole account crashed? What was your leverage? We may be able to control the death basket by eliminating the biggest loser trade in the basket as the other thread is experimenting with. thank you
  • #1,993
  • Dec 2, 2010 7:49am Dec 2, 2010 7:49am
Quoting Toothman
Saidar,
Thank you for the back test results. Could I ask a favor of you? In my opinion, the 02a may be the most consistent of the AddSix EA's. Could you run a test on the same period with the AddSix_02a EA with these settings? .4 MMTU, PP .5 using 50:1 leverage. I'm working with a $40,000 demo account. Your question on why the crash happens. The only answer is that market must have gone through a long period of stagnation to get you into a DD that couldn't recover. In your test how many trades were closed by margin call before the whole account...
The test results you saw are only the BASIC phantom addsix. It buys or sells the biggest winner from the two baskets every hour. The phantoms are reset every x hours.

I will code versions for the AddSix02 and 03 soon just need to figure out what makes them different.

Metatrader 5 is not the same as Metatrader 4. Metaquotes revamped the entire order handling system.

In MT4 you can have more than one order for a currency pair. In MT5 you can't.

The term orders is not used anymore, in MT5 you talk about positions. You can only have ONE position open per currency pair. But you can add to the existing position. Think of a basket with eggs. The basket is the position, the eggs are the orders you placed. You cannot manage the orders individually only the basket as a whole.

So it is not possible to close the biggest losing order in MT5, what is possible is to close the biggest losing position, but that position includes all orders opened for the currency pair.

Hope you understand what I'm trying to say.

This is actually a good thing for this strategy because MT does not need to close 50 trades and take 5 minutes. MT5 only has max 6 positions to close which is obviously a lot faster.

But even if it were possible it won't work. The problem is where do you draw the line? When do you tell if a order is ready to be closed? This will change the dynamics of the system to the core and it might not work anymore.

What we are needing is to PREVENT it from happening, not sure if it is even possible.

One idea that I had a while ago is to balance all 6 pairs out in case of an emergency, that is to make a perfect hedge. This will enable your account to stay alive and wait the storm out. As soon as the market settles you start to unbalance the hedge again.

This is of course very difficult, linear algebra iterative operations needs to be programmed to find the lots to buy or sell to make a perfect hedge.
  • #1,994
  • Dec 2, 2010 8:32am Dec 2, 2010 8:32am
See attached for version 0.1 of MT5 addsix.

This is the basic version, it adds a trade every hour on the biggest winning pair, and resets the phantoms every x hours.

Metacoder, I'm not sure if this is an exact mirror of your technique because I did not really use your code I programmed it according to my style, but so far the backtests show that it is working ok.

Will post new versions if I detect any problems.

If you want to backtest this, use 100 000 dollars as you cannot use a mini account while backtesting. Also use "open prices only", do NOT use every tick.

The tester is still broken and I have code that more or less fixes the problem, it is not 100% perfect but the results are ok.

If you want to run it on demo you will NEED TO REMOVE THE CODE that the tester uses, otherwise it won't work! Go to the source code and remove the indicated code, remember to also remove the remaining bracket at the end of the OnTick() class.

Also remember to show all currencies in Market Watch before doing anything.
Enjoy!
Attached File(s)
File Type: zip Phantom_AddSix_0.1.zip   23 KB | 476 downloads
  • #1,995
  • Dec 2, 2010 9:24am Dec 2, 2010 9:24am
Quoting JulesvH
Maybe I should mod the Addsix03bx version with the Cut Looser (CL) option and give that one another try.
When using that mod, the MMTU, PP en CL_multiplier should be in a well tested/thought balance. (Like MC mentioned before with the MMTU and PP)
I already have that going on my Tadawal 500:1 demo with MMTU 0.4 PP 1.0 CutLoserMultiplier(CLM) 2.. see attached
Attached File(s)
File Type: xls TD_AddSix_03bx_500-1_MMTU-0.2_PP-0.5.xls   126 KB | 411 downloads
Frazer
  • #1,996
  • Dec 2, 2010 10:43am Dec 2, 2010 10:43am
Quoting Saidar
One idea that I had a while ago is to balance all 6 pairs out in case of an emergency, that is to make a perfect hedge. This will enable your account to stay alive and wait the storm out. As soon as the market settles you start to unbalance the hedge again.
Very very tricky tactically, but if you've thought it all through, I'd be interested.

I'm very grateful for your translation to MQ5! Thank you!
Open to new approaches.
  • #1,997
  • Dec 2, 2010 11:55am Dec 2, 2010 11:55am
MC,
At the beginning of the phantom 6 system you seemed to be favoring the 02a. Are you still testing that one? I'm getting very good results and considering going live with it. thanks
  • #1,998
  • Dec 2, 2010 1:43pm Dec 2, 2010 1:43pm
Quoting MetaCoder
Very very tricky tactically, but if you've thought it all through, I'd be interested.

I'm very grateful for your translation to MQ5! Thank you!
I will update the expert if I find any bugs, as you can see it is not as perfect as your MT4 version is but for now it is working fine.

Also you can modify the 6 pairs if you want to they are not static.

About balancing the trades to form a perfect hedge, as you said very tricky indeed. Every detail of each order will need to be saved in an array of some sort to make this happen as each order was placed at different currency values. For some reason my calculations always gives me an extra variable with no equation so that means I have two variables unaccounted for so will need to program some kind of iterative algorithm to find a solution.

As far this is the only solution that I could come up with as nobody can see into the future so preventive action is basically impossible.
Will see what I can come up with.
  • #1,999
  • Dec 2, 2010 1:46pm Dec 2, 2010 1:46pm
About backtesting, like I said the backtester is still broken.

You can check this, test on different currency pairs you will get different results.

My advice would be to use it as an indication if your settings will work or not, do not use optimized parameters and think they will make you rich.
  • #2,000
  • Dec 2, 2010 4:03pm Dec 2, 2010 4:03pm
Can someone code the ea for MT5
7 traders viewing now