AshFX V2. SteveHopwood's latest trading robot - it works this time | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

AshFX Daily 2,650 replies

AshFX Daily (EA - MTaboneWeb) 250 replies

A system that works for a part time trader with minimum screen time 16 replies

SteveHopwood EAs 4 replies

If it works, it works... 3 replies

Attachments: AshFX V2. SteveHopwood's latest trading robot - it works this time
Exit Attachments

AshFX V2. SteveHopwood's latest trading robot - it works this time

  • #41
  • Oct 23, 2009 11:41am Oct 23, 2009 11:41am
Thanks
  • #42
  • Oct 23, 2009 2:21pm Oct 23, 2009 2:21pm
Quoting maya007
Yes, this is the reason, why I'm posting issues that I spotted. Please look at post #29, there is issue with bad SL value - it is taken from backtest so I hope you will be able reproduce it.
Thank you for your good work on this robot. Seems that is now working good.
Not even the remotest, dimmest, vaguest hint of the merest beginnings of the glimmering of a clue what caused this.

  • #43
  • Oct 23, 2009 5:15pm Oct 23, 2009 5:15pm
Quoting ivanyong
Thanks
Enjoy. Still a few bits of things to sort out, but looking good so far.

  • #44
  • Edited 12:48am Oct 24, 2009 12:37am | Edited 12:48am
Quoting SteveHopwood
Not even the remotest, dimmest, vaguest hint of the merest beginnings of the glimmering of a clue what caused this.



backtester Alpari UK, 5 digits, default settings, 1.1.2009 - 23.10.2009, EURUSD, Daily, Every tick

1 2009.04.09 00:00 sell 1 0.02 1.32657 21.32667 0.00000 0.00 10000.00 - initial stop loss bad
2 2009.04.10 00:00 close 1 0.01 1.31591 21.32667 0.00000 10.66 10010.66 - close half of trade
3 2009.04.10 00:00 sell 2 0.01 1.32657 21.32667 0.00000 0.00 10010.66 - remaining half
4 2009.04.10 00:00 modify 2 0.01 1.32657 1.32657 0.00000 0.00 10010.66 - modify second half SL to BE, SL is now OK
5 2009.04.13 14:30 s/l 2 0.01 1.32657 1.32657 0.00000 -0.01 10010.65 - stop at correct SL
6 2009.07.29 00:00 sell 3 0.02 1.41709 21.41719 0.00000 0.00 10010.65
7 2009.07.30 00:00 close 3 0.01 1.40365 21.41719 0.00000 13.43 10024.08
8 2009.07.30 00:00 sell 4 0.01 1.41709 21.41719 0.00000 0.00 10024.08
9 2009.07.30 00:00 modify 4 0.01 1.41709 1.41709 0.00000 0.00 10024.08
10 2009.07.31 10:27s/l 4 0.01 1.41709 1.41709 0.00000 -0.01 10024.07

Maybe problem is only in backtester... ? Have you same results?
Pavel
  • #45
  • Edited 3:48am Oct 24, 2009 3:15am | Edited 3:48am
Quoting maya007


backtester Alpari UK, 5 digits, default settings, 1.1.2009 - 23.10.2009, EURUSD, Daily, Every tick

1...
I was looking at how to lock in some profits at breakeven time and came across the solution to the above.

It is all to do with faulty code in the routine that calculates a stop loss when the swing hi-lo is too close to the market to be acceptable to the criminal.

Oops. I had not spotted this because I only ever backtested the hedge.

Nice one again Pavel. I am coding the fix right now.

  • #46
  • Oct 24, 2009 5:23am Oct 24, 2009 5:23am
Once I effectively turned of the wick filter by setting AllowableWickPercentage to 100, the robot started to take trades.

Looking good on the trading front, so far.

Attached Image (click to enlarge)
Click to Enlarge Name: open trades.gif Size: 62 KB
  • #47
  • Oct 24, 2009 6:07am Oct 24, 2009 6:07am
Hi Steve - as I see you are using the broker Alpari UK for testing your EA. I have the same all time. I'll turn of the wick filter by setting AllowableWickPercentage to 100 too ! So since monday we must have the same conditions - and the same trades - right ?
When could you post a new version of your EA after correction of the little bugs, above described here in the thread ? I would give a new try since monday then.

BTW: Whats your "CandleLengthOverAverage" setting now ? An important and sensitive parameter - you proposed earlier in a post to switch it off perhaps setting it to 1000 ?
  • #48
  • Oct 24, 2009 6:35am Oct 24, 2009 6:35am
Quoting Mandl2007
When could you post a new version of your EA after correction of the little bugs, above described here in the thread ? I would give a new try since monday then.
Before the markets open on Sunday night. I hope, later to day.

Quote
BTW: Whats your "CandleLengthOverAverage" setting now ? An important and sensitive parameter - you proposed earlier in a post to switch it off perhaps setting it to 1000 ?
I have left it at default for now. The default is pure guesswork.

  • #49
  • Oct 24, 2009 9:12am Oct 24, 2009 9:12am
I have uploaded the fixed version to post 1. Also added a couple of inputs - details in the updates section.

  • #50
  • Oct 24, 2009 11:36am Oct 24, 2009 11:36am
I have forgotten to do anything about slippage. Nag me if I don't do anything before Monday.

  • #51
  • Edited 12:58pm Oct 24, 2009 12:08pm | Edited 12:58pm
Quoting SteveHopwood
I have uploaded the fixed version to post 1. Also added a couple of inputs - details in the updates section.

Thank you for quick correction. It seems that new version takes more trades than previous.
For example this two are new in interval 1.1.-30.4.2009:
1 2009.02.10 00:00 buy 1 0.02 1.30028 1.27462 0.00000 0.00 10000.00
3 2009.03.09 00:00 buy 2 0.02 1.26763 1.24560 0.00000 0.00 9948.69

First had quite big drawdown, so I will look if they were placed correctly.

Edit: Both trades are ok.
Pavel
  • #52
  • Oct 25, 2009 1:48am Oct 25, 2009 1:48am
And Steve, please look at this. I have done manual backtesting EURUSD Daily. Robot missed these trades from 1.1.2009 (all profitable):
30.3. sell, 21.5. buy, 8.6. sell, 29.7. sell, 2.9. sell, 9.9. buy, 25.9. sell
Why?
Ash's conditions were met. Maybe candle/wick's filter can affect this?
Pavel
  • #53
  • Oct 25, 2009 3:25am Oct 25, 2009 3:25am
Quoting maya007
And Steve, please look at this. I have done manual backtesting EURUSD Daily. Robot missed these trades from 1.1.2009 (all profitable):
30.3. sell, 21.5. buy, 8.6. sell, 29.7. sell, 2.9. sell, 9.9. buy, 25.9. sell
Why?
Ash's conditions were met. Maybe candle/wick's filter can affect this?
Yep. Decisions about these lengths are up to the individual to decide until we arrive at a consensus. My settings for these are the defaults.

  • #54
  • Oct 25, 2009 3:25am Oct 25, 2009 3:25am
I have added a slippage allowance to the trade sending function. Post 1 as usual.
  • #55
  • Oct 25, 2009 5:45am Oct 25, 2009 5:45am
Hi Steve ! Is the "slippage allowance" hidden inside the EA without possible change of this parameter ? Earlier I saw some EA's where I could change the value of slippage pips allowed.
  • #56
  • Edited 7:15am Oct 25, 2009 6:58am | Edited 7:15am
Hi Steve - glad to have now the new corrected version of your EA. So beginning from monday I'll start a new demo test. Attachted is a folder with the EA used now (plz check it - is this really your last version I'm using now ?). Included is the template and the set file.
Main differences (from the default) of these set file are: lot size is 0.2 (not 0.02) and an important change is this: in accordance with Steve's earlier post #31 I set the "CandleLengthOverAverage" to 1000 to switch this parameter off.
"AutoCalculatePartClose" is "false" in my set file and "UseHorizontalLineSR" is "false". "BnAllowablePipsAway" is "500" - not "1000" and "CurrencySymbol" is $.

Broker is Alpari (UK).

So lets start again !


Btw: Everybody can watch this Demo acc on Alpari (UK) - Acc Nr is 1192260 and password is 1kakgne !
Attached File(s)
File Type: zip SteveHopwoods EA test 1 25_10_2009.zip   45 KB | 512 downloads
  • #57
  • Oct 25, 2009 7:07am Oct 25, 2009 7:07am
I decided to open a second Demo acc with Alpari (UK) using the SteveHopwoods default set file for comparison of EA trading results:

Acc: 1209746

Password: 6eyjwmu

Template is the same as in the post above. Unchanged EA and set file are from post 1 of this thread.
  • #58
  • Oct 25, 2009 11:05am Oct 25, 2009 11:05am
Quoting Mandl2007
Hi Steve ! Is the "slippage allowance" hidden inside the EA without possible change of this parameter ? Earlier I saw some EA's where I could change the value of slippage pips allowed.
Yep. It is hidden.

You can have it as a visible input if you want.

  • #59
  • Oct 25, 2009 1:03pm Oct 25, 2009 1:03pm
I would like to start by thanking Steve for this amazing work!

I have been experimenting with this EA for about 2 weeks now and I see that you have now made some changes!
I will be trying the newly posted EA this week in hopes of it putting in some trades. (The older EA did not put in a single trade over last week...only 4 pairs though and not much was tweeked)

I am very anxious to see the new results! I will keep ya'll posted!

Thanks again to AshFX and Steve for making this happen!
  • #60
  • Oct 25, 2009 6:10pm Oct 25, 2009 6:10pm
amazing EA from steve, looking for a test with the default of steve.
Hope can get a good result and post up once after run it sometimes.

Steve is the allowablewickpercentage put it 100 ?
or got other thing need to set for it?
cos i put around 1 hours haven't seen any trade yet.
thanks
0 traders viewing now