Ichimoku cloud burst ea | Page 42 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Renko and Ichimoku Cloud Cross 132 replies

ichimoku cloud breakout strategy 29 replies

Need an Ichimoku Stop Loss behind the cloud - EA 0 replies

The Ichimoku Cloud 4 replies

Attachments: Ichimoku cloud burst ea
Exit Attachments

Ichimoku cloud burst ea

  • #821
  • Apr 6, 2012 8:40am Apr 6, 2012 8:40am
The only reason i say I RB should be 4 -9 range is bcoz i think with low RB sl will be tight. if we have a big RB sl will be loose and if it does hit even once it hurts a lot.
  • #822
  • Apr 6, 2012 8:42am Apr 6, 2012 8:42am
Any Idea how to prevent this huge spike buy order that opened below the cloud? It was looking correctly for the next clean candle but due to the news release and slippage it didn't open until below the cloud and didn't produce a stop loss or a take profit.

http://dl.dropbox.com/u/49530125/audjpy.gif
  • #823
  • Apr 6, 2012 8:43am Apr 6, 2012 8:43am
Quoting kumetto
Bad bad bad.EA missed those.AUS/JPY missed also.
This is due to the NFP effect , i doubt even if the ea fires the broker will fill these orders, it is better to ignore this as it is not a normal condition, EA not firing in these circumstances is infact a blessing in disguise.
  • #824
  • Apr 6, 2012 10:36am Apr 6, 2012 10:36am
Quoting kumetto
Bad bad bad.EA missed those.AUS/JPY missed also.
Same here. EA did not fire on multiple charts. Caught all the losses and almost none of the wins
  • #825
  • Apr 6, 2012 10:44am Apr 6, 2012 10:44am
Quoting d4amenace
Same here. EA did not fire on multiple charts. Caught all the losses and almost none of the wins
Maybe the spread is too high at the time.
Follow me to the Promised Land
  • #826
  • Apr 6, 2012 11:08am Apr 6, 2012 11:08am
kumetto: There was NFP and some other major news out at 13:30 GMT. I had 6 positions open because I forgot to close them and shut down the EA before the release. All of them was in profit, and all of them were stopped out just before the NFP release making a huge loss overall.
I think it's not fair to blame the EA, since almost all the traders say that you should stay out of trading at Friday, and during news release, especially NFP (except if you have specialized strategy to trade the news).
  • #827
  • Apr 6, 2012 11:17am Apr 6, 2012 11:17am
Quoting xlitang
Maybe the spread is too high at the time.
I figured it out. My ping is too high from my home computer. I need a VPS with low latency to my broker server.
  • #828
  • Apr 6, 2012 11:24am Apr 6, 2012 11:24am
So this week I've used 6 pip range bars on all the 9 pairs I trade. There were 2 big losses this week: at the beginning of the week, when lots of pairs opened with a relatively big gap, and the NFP release. Before the NFP the overall gain was about 3%, same as the previous week.

I've made a spreadsheet from the exported statement, you can find it in the attachment.

The biggest loser pairs was:
EURJPY, GBPJPY
The biggest winners:
EURUSD, GBPUSD, USDCHF

Next week I'll run the EA on 2 accounts. I'll set the range bar sizes on the 2 different method seller9 and unimak suggested (based on 15 min 480 ATR vs. averaging and multiplying the ranges on 1 min chart).
I will wait for tuesday London open with the start (this week I started the EA when the market opened), since it will be easter and bank holidays almost everywhere.
Attached File(s)
File Type: xls ichi_ea_apr2-6.xls   70 KB | 473 downloads
  • #829
  • Apr 6, 2012 11:32am Apr 6, 2012 11:32am
Quoting Zool
So this week I've used 6 pip range bars on all the 9 pairs I trade. There were 2 big losses this week: at the beginning of the week, when lots of pairs opened with a relatively big gap, and the NFP release. Before the NFP the overall gain was about 3%, same as the previous week.

I've made a spreadsheet from the exported statement, you can find it in the attachment.

The biggest loser pairs was:
EURJPY, GBPJPY
The biggest winners:
EURUSD, GBPUSD, USDCHF

Next week I'll run the EA on 2 accounts. I'll set the range bar sizes on the 2 different...
zool did you run the ea on all days all the time or run it selectively as EJ is shown as loss with or without nfp ?? something is wrong pls chek.
  • #830
  • Apr 6, 2012 12:12pm Apr 6, 2012 12:12pm
Quoting unimak
zool did you run the ea on all days all the time or run it selectively as EJ is shown as loss with or without nfp ?? something is wrong pls chek.
I've run it all the time. The EJ catched a good trade after the NFP release, that's why the loss is smaller including the NFP trades. Looking back at the chart the problems was:
The EA didn't catch the big down move at the beginning of monday, because there was a spike before that stopped out the previously opened sell trade, and the EA was looked a buy trade after that. That was about 150-160 pips missed. You can see this stopped out short trade in the spreadsheet - this was the first trade the EA made.
Then there was a bouncing price movement at Apr 3 afternoon, where there was 4 losing trades in series.
These 2 things caused the loss made on EJ. A better range bar size would have filtered out the bouncing up and down through the kumo, and looking for both side trades would let the EA catch the big short trade at monday.
  • #831
  • Apr 6, 2012 12:21pm Apr 6, 2012 12:21pm
Check the GJ at 2012.04.02 12:07 that was the trade the EA missed on the EJ (these two pairs move in a very similar way).
Theres something strange on the GJ: there was no trades between monday morning and afternoon,but there was about 5 entry opportunities looking at the chart. After that there was a trade opened without stop loss, this happened 2 times this week, I don't know the reason for this. maybe it's my broker. There was a missed big down move at thursday, also caused bay a previously closed short trade.
  • #832
  • Apr 6, 2012 12:53pm Apr 6, 2012 12:53pm
seller9: Will you release the u9 version before the starting of the next week? I don't want to change versions in the middle of the week, because that would make the testing results inaccurate.
  • #833
  • Apr 6, 2012 1:06pm Apr 6, 2012 1:06pm
Quoting Zool
kumetto: There was NFP and some other major news out at 13:30 GMT. I had 6 positions open because I forgot to close them and shut down the EA before the release. All of them was in profit, and all of them were stopped out just before the NFP release making a huge loss overall.
I think it's not fair to blame the EA, since almost all the traders say that you should stay out of trading at Friday, and during news release, especially NFP (except if you have specialized strategy to trade the news).
I'm not blaming the Ea Just shouted what happened
Trade with 60
  • #834
  • Apr 6, 2012 5:37pm Apr 6, 2012 5:37pm
Everyone,

I notice there is still a lot of discussion and consideration for the optimal settings for this EA. It no doubt has loads of potential, but of course the key is correct parameterization as the market takes on different personalities..

I found this article that describes how to optimize any EA settings using realtime data at predefined intervals using the strategy tester of another MT instance communicated with through a text file. You tell it what variables of your EA to match and it will optimize for you. ->

http://articles.mql4.com/506

How Automated Optimizer Works
The automated optimizer works on the basis of using Terminal-Tester to optimize parameters of the Expert Advisor attached to the chart in Terminal. For this, the program sends to the Terminal-Tester a file containing optimization parameters (optimise.ini) and launches the Terminal Tester in the optimization mode. Then it copies the obtained results of "FileReport........htm" back to the Terminal and filters the best values from the obtained results.


@seller9 do you think this could work with either the rangebar EA or the ichimoku one?

Quoting kumetto
I'm not blaming the Ea Just shouted what happened
Skype: heliosphan187
  • #835
  • Apr 6, 2012 6:18pm Apr 6, 2012 6:18pm
Unimak please give feedback if I'm doing it the way you thinked about the range bar size. I saved the 1 minute data from my broker as a csv file, then load it into open office as a spreadsheet (excel can't handle more than 65535 rows). Then I subtracted the low value from the high value in each row, to get that minutes range. Then I put an average function on this column to get the average range of the 1 minute chart. Then I multiplied this with 1.75, 1.681 and 2.681 (these latter two are Fibonacci levels - I think you referred to the Fibonacci levels with the 61,8%). The 75% and 61,8% gave almost the same value, after rounding it's actually the same.

Here is a link to the Openoffice document I made for this:
http://db.tt/HSKiHfZE

The other method for the range bar size, is that I put an ATR indi on a 15 minute chart with 480 set as the period (1 business week), and the range size is the indi's value at the last candle today at market close.

So here it is (name - atr value - 168.1% avg - 268.1% avg):
AUDUSD - 9 - 4 - 6
EURAUD - 10 - 4 - 6
EURJPY - 12 - 4 - 6
EURUSD - 9 - 4 - 6
GBPJPY - 14 - 4 - 7
GBPUSD - 9 - 4 - 6
USDCAD - 6 - 3 - 4
USDCHF - 6 - 3 - 5
USDJPY - 8 - 2(?) - 3

This week I set all the pairs at 6 pip value. I gave the summarized results earlier today, so I think I won't test the 268,1% settings.

The USDJPY with the 2 pip range setting actually will be like a renko chart, because on 2 pips it's impossible to have wicks.
  • #836
  • Apr 6, 2012 7:26pm Apr 6, 2012 7:26pm
The big losses on the EJ and GJ bothered me enough, to check the charts how can it be made profitable. I only went through quickly on the GJ, it's really late here (2:00 am) so I cannot post a screenshot and such things, but check it on your charts.

So I set the range bar size to 14 pips, and the Ichimoku settings to 7,10,34. I thinking of not using the BEJS but a fixed take profit of 25-30 pips, or closing with the cloud-following stop loss when the price enters the kumo. There would be numerous re-entries on bigger moves (if seller9 will implement the "auto" feature). There are some negative stop-loss hits, because the GJ loves to swing around the kumo sometimes, but at a quick glance I think the gains would be bigger than the losses.

Please check it, and share your thoughts.
  • #837
  • Edited Apr 7, 2012 12:03am Apr 6, 2012 7:53pm | Edited Apr 7, 2012 12:03am
Things added

1) AutoIchiBreakEvenPips: Set to true to use this along with the "AutoIchiBreakEvenPipsMultiplier". This mutiplies the pips/bar by this multiplier to arive at your BE setting if used. Set to false to use the "IchiBreakEvenPips" setting. So if your RB is 4 pips and your multiplier is 2.5 your BE will be 10.

2) AutoTakeProfitPips: This uses the Daily Range (20)(last 20 day average) multipled by the "AutoTakeProfitPipsMultiplier" to set a TP that is sent to the Broker. Set to false to use the "TakeProfitPips" instead. So if the DR(20) is 94 and the multiplier is 0.50 the TP will be set at 47 pips.

3) AutoBuySell: Set Buyer and Seller to false when using.

Attached File(s)
File Type: ex4 ichimoku_u8.ex4   42 KB | 8,581 downloads
  • #838
  • Apr 6, 2012 11:10pm Apr 6, 2012 11:10pm
Quoting Zool
Unimak please give feedback if I'm doing it the way you thinked about the range bar size. I saved the 1 minute data from my broker as a csv file, then load it into open office as a spreadsheet (excel can't handle more than 65535 rows). Then I subtracted the low value from the high value in each row, to get that minutes range.


Then I put an average function on this column to get the average range of the 1 minute chart.

>>>> once you get the difference between buy and sell you just add all the cells and divide by the number of colums you...
The other method for the range bar size, is that I put an ATR indi on a 15 minute chart with 480 set as the period (1 business week), and the range size is the indi's value at the last candle today at market close.

>>> this method is not that accurate as the one done using raw data .it will do the job but it is not a killer .

the values that you have posted has some mistake i am getting 3.5 pip for eu so you need to multiply 1.75 % to it to arrive at the rb figure . now to this figure we may have to add 1 pip or subract 1 pip to arrive at the magic number .

so suppose you get 9 then either it is 9 or 8 or 10 in these 3 numbers like the correct settings.

now using these three if we can get some strategy tester to work accuractely we will know the winning percentage if we run all the time this ea mon to fri or
if we can get a specific time span in gmt where the hit is high then it will too good.

best of luck and wish you all the success in your testing . keep it up

Then I put an average function on this column to get the average range of the 1 minute chart.

>>>> once you get the difference between buy and sell you just add all the cells and divide by the number of colums you get the avg and then to that avg your multiply by 1.75 % or 1.68.1 % (yes this fib level)

Then I multiplied this with 1.75, 1.681 and 2.681 (these latter two are Fibonacci levels - I think you referred to the Fibonacci levels with the 61,8%). The 75% and 61,8% gave almost the same value, after rounding it's actually the same.


>>> yes you need round off to the nearest value on the higher side
  • #839
  • Edited 1:15am Apr 7, 2012 12:57am | Edited 1:15am
The RangeBar EA's doesn't generate the offline charts when the market is closed. Am I right?
  • #840
  • Apr 7, 2012 12:58am Apr 7, 2012 12:58am
Quoting ^Sam
The RangeBar EA's doesn't generate the offline charts when the market is closed. Am I write?
Exactly...you have to wait for server feed during market hours
0 traders viewing now