Jiva Dragon | Page 40 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Enter The Dragon - GBP/JPY 2 replies

Info on the Borihei-Dragon Scalping System? 15 replies

Jiva's 5,5 Trade Journal 5 replies

The Dragon 441 replies

***Dragon II*** 15 replies

Attachments: Jiva Dragon
Exit Attachments

Jiva Dragon

  • #781
  • Nov 5, 2010 2:04am Nov 5, 2010 2:04am
Hi everybody, today I didnt go to school, I passed all day backtesting and searching to a good .set file (Let me say this is a very very hard , time-eating work, thanks allot for every .set file you have done Jiva), and I came up with an idea, as settings may change on the future to feat better th eeach pair :

1) there is an indicator that is used in THV to place SL , called 3xATR (in attached picture you can see it at the bottom) , this is obviously dinamic, we could use this to choose a more accurate 50:50 SL and TP for the trade about to be placed, depending on the market's current feeling.


This could be done for eachone of the parameters (the most used variables in the EA's inputs) Jiva has been using more, to fit the curve on the backtest, this could help making the Ea more Dynamic to the markets condition.

The is the solution I can imagine at this moment, mybe it is nothing but if we all come with ideas, in others could be lighten the spark of the real solution.

Thanks Taiyakixz, Jiva , jm2110, kauaibobby and inso, Let's help to make this dragon get its' firts bones and joints.

_______-
Sequoia
Attached Image
  • #782
  • Nov 5, 2010 2:43am Nov 5, 2010 2:43am
Back test to 10 yrs, MetaQuotes data,
would love to get Duskacopy 99%. But it looks like a major task in and of itself to get that downloaded.
  • #783
  • Nov 5, 2010 2:49am Nov 5, 2010 2:49am
Quoting Taiyakixz
There are many advices but I want to see what they tried and how the result is.....
I hope they aren't just a critic....
Making suggestions, don't understand much about core operations to be a critic, what you guys do are selfless, kind of hard to criticize.
I do have version 11 on couple of charts, results are ok. SL to high for my liking based on trade frequency (this could be considered constructive criticism , now i can be accused of being a critic).
I will be messing around with MA when i get a chance thinking of SSMA 3, 7, 10, 14. Like the idea of having all these inputs. foward test lot slower than backtest.
  • #784
  • Nov 5, 2010 2:51am Nov 5, 2010 2:51am
Quoting Sequoia
Hi everybody, today I didnt go to school, I passed all day backtesting and searching to a good .set file (Let me say this is a very very hard , time-eating work, thanks allot for every .set file you have done Jiva), and I came up with an idea, as settings may change on the future to feat better th eeach pair :

1) there is an indicator that is used in THV to place SL , called 3xATR (in attached picture you can see it at the bottom) , this is obviously dinamic, we could use this to choose a more accurate 50:50 SL and TP for the...
Wow. Good stuff going on. We're gonna make this happen!!! Or maybe it is happening and we just have to see how to get on board, not get in the way and not get run over either....

jm2110 has mentioned ATR too. I'm not all that familiar with ATR. Maybe someone can explain how that might work.

For the 50-50 test we want as low a stoploss as possible and since the profits have to equal the stop it is not so much of an optimizing issue. Maybe a little bit for the 50-50.
20 pips would be great but that is a little thin. Looks like between 20 and 30 or so is the sweet spot for a 50-50 test.
  • #785
  • Nov 5, 2010 2:53am Nov 5, 2010 2:53am
Quoting junglist0000
Making suggestions, don't understand much about core operations to be a critic, what you guys do are selfless, kind of hard to criticize.
I do have version 11 on couple of charts, results are ok. SL to high for my liking based on trade frequency (this could be considered constructive criticism , now i can be accused of being a critic).
I will be messing around with MA when i get a chance thinking of SSMA 3, 7, 10, 14. Like the idea of having all these inputs. foward test lot slower than backtest.
The stop losses in v11 are way too high!!
It's a disaster waiting to happen.
v11 is being replaced by v14.2, stops are more like 30 pips...
We will be interested to see what you find with ma...
  • #786
  • Nov 5, 2010 4:07am Nov 5, 2010 4:07am
Results of long range back test is very encouraging!!
I backtested 10 years EURUSD for 50-50 stop loss 30 pips FXDD
11/4/00 to 11/4/10
optimized the channel and the moving averages.
No dss, not3, macd default.
then I took that set file and ran it for 11/4/09 to 11/4/10
very interesting..
here are 2 graphs
1st is the 10 year 50-50 optimized (win rate 53%)
2nd is 11/4/09 to 11/4/10 using the 10 year 50-50 set (win rate 68.90%)

This isn't exactly a walk forward test. I will do some of those later.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: eu11-4-00to11-4-10.gif Size: 9 KB
Click to Enlarge Name: eu11-4-09to11-4-10.gif Size: 8 KB
Attached File(s)
File Type: zip eu11-4-00to11-4-105050after.zip   2 KB | 232 downloads
  • #787
  • Nov 5, 2010 4:08am Nov 5, 2010 4:08am
ah.....many comments about my previous post....

I just mean that some of advices are which they can try and see the results....
I don't critisize constructive criticism. They are welcome...
  • #788
  • Nov 5, 2010 4:49am Nov 5, 2010 4:49am
Quoting jiva34
The stop losses in v11 are way too high!!
It's a disaster waiting to happen.
v11 is being replaced by v14.2, stops are more like 30 pips...
We will be interested to see what you find with ma...
i am backtesting V14.2 and try to force SL that look reasonable and then optimize, I will let you know if there is good results.

i have an idicatore that look very promissing in a live chart, it maby can replace MACD in the EA and give more accurate entrys.
attached pic and indicator
many thanks for your work Jiva & Taiykixz
Attached Image (click to enlarge)
Click to Enlarge Name: nl.gif Size: 20 KB
Attached File(s)
File Type: mq4 BB_MACD_CCT.mq4   4 KB | 240 downloads
  • #789
  • Nov 5, 2010 5:04am Nov 5, 2010 5:04am
nahhhhhrrr Jiva, hadn't enough time last few weeks, will be back on Monday!
Then we'll rock that shit : D

see ya!
Hero calls followed by margin calls...
  • #790
  • Nov 5, 2010 5:19am Nov 5, 2010 5:19am
Quoting FxHope
i am backtesting V14.2 and try to force SL that look reasonable and then optimize, I will let you know if there is good results.

i have an idicatore that look very promissing in a live chart, it maby can replace MACD in the EA and give more accurate entrys.
attached pic and indicator
many thanks for your work Jiva & Taiykixz
Thank you for testing, but 4H chart is different by broker's timezone as jiva mentioned at the 1st post, be careful of it.
I checked BB_MACD_CCT code, but the calulation of the dots are the same as MACD. You just mean the looks are good...?
  • #791
  • Nov 5, 2010 5:55am Nov 5, 2010 5:55am
Taiyakixz
The pic is 4h but i work in 1h my mistake.

it look very good, yep. but I know that what you see its not what you get. so your decide. http://cdn.forexfactory.com/images/icons/icon7.gif
  • #792
  • Edited 3:08pm Nov 5, 2010 2:53pm | Edited 3:08pm
Very nice looking short entry EURAUD.
Smart, very smart...expert adviser...
This is the old series v11. Huge stop loss 250 pips..

1st edit........... first take profit reached, 2nd take profit on this old v11 series is 890 pips.
trailing stop is 77 pips, so either EURAUD makes a big move or the end position stops out at the trailing stop, falls back to break even for 0 pips, closes above the channel hi which will act like a trailing stop or reaches 890 pips...

2nd edit...........
could go all the way. v11 had some things going for it.... big stop losses and all.... It also had big profit targets...
2nd profit target is 890 pips!!! with trailing stop 77....
Looks like it is going to get there....
With a positive rollover along the way... currently +12.80

Just a note to illustrate a bit how the ea is designed to be able to catch a big move... The 3 types of exits....

Attached Thumbnails http://cdn.forexfactory.com/attachme...1&d=1288855751 http://cdn.forexfactory.com/attachme...1&d=1288901606
Attached Image (click to enlarge)
Click to Enlarge Name: euraud4.gif Size: 62 KB
  • #793
  • Nov 5, 2010 4:09pm Nov 5, 2010 4:09pm
Quoting Taiyakixz
Thank you for testing, but 4H chart is different by broker's timezone as jiva mentioned at the 1st post, be careful of it.
I checked BB_MACD_CCT code, but the calulation of the dots are the same as MACD. You just mean the looks are good...?
Hi Taiyakixz what do you think about using the indi I posted before to place the 50:50 (SL & TP) ? I see the common ATR indicator is like an oscilator.

__________-
Sequoia
Attached Image
  • #794
  • Edited 4:46pm Nov 5, 2010 4:29pm | Edited 4:46pm
Quoting Sequoia
Hi Taiyakixz what do you think about using the indi I posted before to place the 50:50 (SL & TP) ? I see the common ATR indicator is like an oscilator.

__________-
Sequoia
I'm not sure we want these values to be changing in 50-50. For the series we are developing now with v14.2 we are not looking to capture big moves, like v11 could do. We are looking for high win rate, with small stop loss and overall profit.

Maybe Taiyakixz has some thoughts about ATR. I am seeing finding s/l and t/p for a 50-50 test is more or less a given. Mostly I see 30/30/30 is the sweet spot.

Right now, at the present moment, I don't see how to incorporate ATR. I will put some time into experimenting with it. The idea of being dynamic with the market as it moves is very attractive. Usually it takes me a little while to assimilate.

Thank you posting the indi, now it is on file..and we won't have to go searching all over everywhere to try to find it when we are ready. jm2110 also likes the ATR. So let's see where this goes.

Have a good weekend everyone, and we will try to be as ready as possible for testing live market when it opens...
  • #795
  • Nov 5, 2010 4:46pm Nov 5, 2010 4:46pm
Quoting jiva34
I'm not sure we want these values to be changing in 50-50. For the series we are developing now with v14.2 we are not looking to capture big moves, like v11 could do. We are looking for high win rate, with small stop loss and overall profit.

Maybe Taiyakixz has some thoughts about ATR. I am seeing finding s/l and...

Thanks Jiva, you too have a great weekend .
enjoy the weekend everybody
  • #796
  • Edited 8:36pm Nov 5, 2010 5:14pm | Edited 8:36pm
Okay forget the weekend. I've got work to do...
Here is the first in a series of walk test I will be doing.
Very, very nice. Ol' jm2110 does it again!!!

Evidently, backtesting can be effective if it is done over a long enough period of time. We will be experimenting with different lengths of time. Should it be optimize every month for 5 years optimized? every week 10 years optimized???

Something that is catching my attention is the possibility of a weighted optimization. Like there are exponential moving averages, weighted moving averages, etc. That give more significance in the end data to the later values in the series.... hmmm.

How this test was set up.
v14.2
EURUSD FXDD 1hr
5 years from 11/4/00 to 11/4/05
Optimized 50-50 take profit/stop loss setting for channel and moving averages.
Use macd default for crossing
no dss, no t3ma.

These optimized settings for 11/4/00 to 11/4/05 were then used to run a backtest for the next year, 11/4/05 to 11/4/06.

Here are the results...

appx 1trade/week
$5374 net on $10000 opening
win rate 69.57%
13.48% max drawdown
30 take profit 1/30 take profit 2/30 stops

edit: as usual I may have put on the party hat a little too early. Results of some of the subsequent walks forward in this series of tests are not so enlivening... We know what we are looking for with this, so we will keep looking...

edit: this parameter is not right. This was the only forward test that was profitable out of 5 such walks. 5 years back to test 1 year forward is not statistically sound.
I am starting a new series of tests: 5 years back 1 month forward.
Attached Image (click to enlarge)
Click to Enlarge Name: walk11-4-5to11-4-065050.gif Size: 9 KB
Attached File(s)
File Type: zip walk11-4-5to11-4-06.zip   1 KB | 264 downloads
  • #797
  • Nov 6, 2010 1:08am Nov 6, 2010 1:08am
Quoting Sequoia
Hi Taiyakixz what do you think about using the indi I posted before to place the 50:50 (SL & TP) ? I see the common ATR indicator is like an oscilator.

__________-
Sequoia
Thank you for your hard work.
You didn't go to school...was it okay...? Please do not fail...

In regard to 3xATR, I know it is told good generally. I checked some charts but at the moment I don't find how it increases the performance with the EA.
I think that you can check the chart manually and see how 3xATR works.
Would you please check the chart and when you find that it is usuful, tell us how it works...?
Of course you don't have to check all chart in many years, enough as you have confidence.

This is what I want to say at my old post.
Not only with this thread, I receive many PMs that xxxx is good, oooo is good, good idea to make an ea.....
But they were not profitable. With most of them, it was very easy to find that they were not good.
I am little tired of checking them.....

I really thank for your hard work and your idea.
I just want to say if you show me how it works, it would be great help.
Please understand......
  • #798
  • Nov 6, 2010 3:02pm Nov 6, 2010 3:02pm
Quoting Taiyakixz
Thank you for your hard work.
You didn't go to school...was it okay...? Please do not fail...

In regard to 3xATR, I know it is told good generally. I checked some charts but at the moment I don't find how it increases the performance with the EA.
I think that you can check the chart manually and see how 3xATR works.
Would you please check the chart and when you find that it is usuful, tell us how it works...?
Of course you don't have to check all chart in many years, enough as you have confidence.

This is what I want to say at my old...
thanks for your consern about school, no the reason I didn't go was because no classes that day (festivities in my country).

Puff yes I know should be very anoying to be a very good MT4 programmer, evert newbie (like me) must have "The profitable idea, Holy grial" lol, and asks you for help, I dont want to be anoying, please forgiveme if I am being .

Well about ATR : is an indicator that measures volatility , for example , in the THV system they sugest to place the SL at the distance that 3XATR(20) (ATR(20) multiplied by 3) suggests, this is because the market's current volatility. So for scalping (and most trading styles) you might want to place a good SL :
1) if your trade was good but the SL was too close, SL will be trigered before actually it makes profit, because of too much volatility.

2) if you place a very distant SL, and your trade goes against you, you would loose too much. (beacuse of too small volatility)

So this ATR indicator, works to know the actual volatility and place this SL , if the system has a good winning/loosing percentage, (Like this one ) We could be seeing more profitable trades, because the volatility changes, and this way we could not be stopped by the SL before the trade goes to TP.

I dont know how they did to change this ATR oscilator value into a logic number-distance to place SL , but if you want I can ask them, or search more about this, because many of the THV indicators code is closed.

I hope I made a clear post, as you know my english is not the best ,if there is anything not clear enough please tell me, I will be pleased to clarify it.

Thanks for your interest

_________-
Sequoia
  • #799
  • Nov 7, 2010 1:50am Nov 7, 2010 1:50am
Quoting Sequoia
thanks for your consern about school, no the reason I didn't go was because no classes that day (festivities in my country).

Puff yes I know should be very anoying to be a very good MT4 programmer, evert newbie (like me) must have "The profitable idea, Holy grial" lol, and asks you for help, I dont want to be anoying, please forgiveme if I am being .

Well about ATR : is an indicator that measures volatility , for example , in the THV system they sugest to place the SL at the distance that 3XATR(20) (ATR(20) multiplied by 3) suggests,...
oh I see, I thought you cut the class.....In Japan many students often skip the class which they don't think important....

Below is one example of 3xATR stoploss.
The yellow line is stoploss. ATR at the opened position is 0.0033, so ATRx3 = 0.0099.
This trade would be loss with 3xATR. I saw some trades like this.
This is the reason why I don't code it at the moment. Of course I don't mean all trades are so, but I want to see some good trades with 3xATR before I code it.

I think you can do the same way and check some trades by looking the chart. I want to see how 3xATR works and makes the EA more profitable.
More trades better, but it is enough as possible.
You could show me 3xATR's good working, then I will code it.

Attached Image (click to enlarge)
Click to Enlarge Name: dragon_eurusd_h1_1011071504_m.gif Size: 21 KB
  • #800
  • Edited 12:11pm Nov 7, 2010 7:48am | Edited 12:11pm
Debut!!
v14.2 EURUSD 1hr
30 pip stop
50/50 edge test win rate 84.62%
12 month back test
Around 7 trades/month
max drawdown 7.85%
+$25,957 on $10,000 initial deposit
trading 1 standard lot.
This is all rearview mirror.
12 months of testing forward walk tests show consistent profit.
So let us see how the performance is for going forward.
There will a few losing trades. But stop loss is 30 pips so life will go on...

I have used another macd indi in the template, you will need to download it in order to be able to use it. The calculations are the same, I believe, just nicer to look at.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 11-4-09to11-4-10stops.gif Size: 8 KB
Click to Enlarge Name: eurusd.gif Size: 61 KB
Attached File(s)
File Type: zip 11-4-09to11-4-10stops.zip   3 KB | 422 downloads
File Type: mq4 MACD with crossingblack.mq4   3 KB | 442 downloads
0 traders viewing now