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

  • #1
  • First Post: Edited Dec 14, 2009 5:34am Oct 17, 2009 11:24am | Edited Dec 14, 2009 5:34am
Attached is the trading robot for Ash's system.




Updates:
Attached File(s)
File Type: pdf AshFX V2 robot user guide.pdf   88 KB | 5,069 downloads | Uploaded Oct 24, 2009 11:02am
File Type: mq4 AshFX V2 auto-trading robot by SteveHopwood.mq4   66 KB | 2,806 downloads | Uploaded Dec 14, 2009 5:29am
  • #2
  • Oct 17, 2009 11:49am Oct 17, 2009 11:49am
Quoting SteveHopwood
Attached is the trading robot for Ash's system. There are a variety of new features, but the biggest change is that this one works. I know, cos it has been working happily on demo for weeks.

Good luck guys.

I will load it on my VPS first thing monday.

Thanks for your ongoing commitment, Steve!

Regards

FOrex37
  • #3
  • Oct 17, 2009 11:53am Oct 17, 2009 11:53am
Quoting Forex37
I will load it on my VPS first thing monday.

Thanks for your ongoing commitment, Steve!

Regards

FOrex37
My pleasure. My frustration with the previous version was the result of problems caused by my not fully understanding Ash's system. I thought I did, but had some fundamental mis-conceptions.

Fat lot of use me insisting that people using trading robots should first understand the system they trade if I fail to follow my own advice.

Hey ho.

  • #4
  • Oct 17, 2009 12:02pm Oct 17, 2009 12:02pm
Hi Steve ! Nice to see you here again, and much better: your EA is working now ! Could you us the links to the basic stuff of the system your EA is based on ? Pdf, forum threads .... so on.
  • #5
  • Oct 17, 2009 12:18pm Oct 17, 2009 12:18pm
Looked into this system a while back - is it not just slow trend following? What's so special about it?
  • #6
  • Oct 17, 2009 12:48pm Oct 17, 2009 12:48pm
Quoting Mandl2007
Hi Steve ! Nice to see you here again, and much better: your EA is working now ! Could you us the links to the basic stuff of the system your EA is based on ? Pdf, forum threads .... so on.
There is a link right at the start of the pdf user guide in post 1 that will take you to where you will find everything you need to know. Ash is there offering invaluable insight into his trading methods.

  • #7
  • Oct 17, 2009 12:50pm Oct 17, 2009 12:50pm
Quoting Sim
Looked into this system a while back - is it not just slow trend following? What's so special about it?
Yes, it is. What makes it so helpful to less experienced traders is the willingness of the system developer, Ash, to provide advice, experience, support and expertise in his posts. If you are already a successful trader, you probably don't need it. On t'other hand, the robot can sit quietly churning away on an account, slowly generating a few extra pips; this might appeal to you.

  • #8
  • Oct 17, 2009 1:09pm Oct 17, 2009 1:09pm
Quoting Sim
Looked into this system a while back - is it not just slow trend following? What's so special about it?
Guess, there`s nothing special about systems that make money (manually traded) - specially when nice robots (from Steve) do the slave work. Nothing special really, plenty of money making systems around......And all of them well tested and documented..... and all of them for higher TFs....

Forex37
  • #9
  • Oct 17, 2009 3:31pm Oct 17, 2009 3:31pm
Quoting Forex37
Guess, there`s nothing special about systems that make money (manually traded) - specially when nice robots (from Steve) do the slave work. Nothing special really, plenty of money making systems around......And all of them well tested and documented..... and all of them for higher TFs....

Forex37
Sensing quite a sarcastic undertone there mate, which is pretty lame in how unnecessary it is, but carry on with your mindset that I was stating that systems such as these are "nothing special" when I didn't even mention the phrase, let alone the rest of your comment digressing into nonsense about timeframes and the concept of EAs doing slave work. Save it for someone who actually asks you about it.

Anyway Steve yeah thanks I see that it is Ash's additional aid on top of the system which makes it stand out above similar systems, and this is why you instructed all to read the relevant material before going live...I brought it up mainly to make sure that it can withstand the ranging times on h1 which seem to kill a lot of EAs I have tested and even got developed. The sheer variability from one day to another is quite radical, and I noted this when testing Ash's system previously, which is why I didn't take it further. So I'm just wondering if you came accross this problem because even a time filter can help.
  • #10
  • Oct 17, 2009 4:01pm Oct 17, 2009 4:01pm
Quoting Sim
Anyway Steve yeah thanks I see that it is Ash's additional aid on top of the system which makes it stand out above similar systems, and this is why you instructed all to read the relevant material before going live...I brought it up mainly to make sure that it can withstand the ranging times on h1 which seem to kill a lot of EAs I have tested and even got developed. The sheer variability from one day to another is quite radical, and I noted this when testing Ash's system previously, which is why I didn't take it further. So I'm just wondering if...
One of the points Ash makes a lot is that he trades his system on the D1 chart. Others experiment with lower time frames, but he appears not to, so H1 ranging is not an issue.

Following his posts makes it clear that his trade-describing pdf file (attached here simply to help illustrate what I am saying) is merely the start of his trading considerations.

On top of the info in the pdf, it rapidly becomes clear when reading his posts, that Ash also considers:

 

  1. Stoch movement: too choppy kills the trade.
  2. Trade trigger candle length: too long kills the trade
  3. Trigger candle wick length: too long kills the trade
  4. Support/resistance: too close to a major sr point kills the trade.

Frankly, almost everything kills a potential trade.

My robot offers as many opportunities to follow Ash's thoughts as I am able to offer. I have found them to be accurate with one exception, and I was not the one to spot this.

moondog did some research that showed that a long wick is actually a good thing - not hard to work out why, as I wrote in the user guide. My robot went for about 3 weeks without placing a trade. Consistently, it spotted the setups that manual traders were spotting, then cancelled them for all the reasons the manual traders were cancelling them.

Last week, I told it to ignore wick lengths. It subsequently took 4 trades; 3 have already closed half the position and moved the sl to be. The fourth was going bad so I closed it; a mistake that, as I forgot I had hedging not stop loss set. Hey ho. Can't remember everything. Even if the remaining trades retrace to close out at BE, the account will still be nicely into profit.

So, there is much more to Ash's system than meets the eye on first reading about it. Even worse, he relies a lot on 'feel' (i.e. trading experience) when deciding whether to take a trade or not. People are kind about my programming, but not even I can programme that into a robot.

That was the long answer. The short answer is: no idea what happens on the 1H tf - the system is not about this and I haven't tested it. My guess is that a D1 system taken down to a 1H chart will get murdered, but I don't actually know in this case.

Attached File(s)
File Type: pdf AshFx.pdf   99 KB | 2,219 downloads
  • #11
  • Oct 17, 2009 4:11pm Oct 17, 2009 4:11pm
"http://cdn.forexfactory.com/images/attach/pdf.gif AshFx.pdf (99.5 KB, 0 views)"

"Could you please attach that template which is inside that
.pdf file example?"

Thank you.
Viewer Discretion Advised: Shall we shag now or should we shag later? :-)
  • #12
  • Oct 17, 2009 4:14pm Oct 17, 2009 4:14pm
Great answer, thanks for the time you spent to reply.

I don't know why I said h1 I meant d1 because I've mainly tested similar EAs on d1 and that sorta adds to cause losing trades to be big due to the stop reverse aspect of cutting false signals when in negative, and SLs are also of the same nature as have to be wide on D1

Nonetheless I commend you for your work, I do wonder how you managed to code the stoch filter from being too choppy? What logic did you use for that?

Lastly can I just ask what you do personally use for the EA to judge what trades to take? I brought the initial question up because I had an incline you had a smart way of filtering the d1 chop, and you have My final want would be for you to simply give me a rough gauge of weekly profitability, because my history centre is well and truly buggered from so much testing On all three brokers I use to test... I dunno what's happened. Redownloading the data doesn't fix it.

Thanks again
  • #13
  • Oct 17, 2009 4:35pm Oct 17, 2009 4:35pm
Quoting SkyzerFX
"http://cdn.forexfactory.com/images/attach/pdf.gif AshFx.pdf (99.5 KB, 0 views)"

"Could you please attach that template which is inside that
.pdf file example?"

Thank you.
Nope - the template has moved on to include a really useful dashboard, so mine does this. Follow the link in the robot user guide and download it from there. Read the thread and learn about the dashboard, then add it to your armoury.

  • #14
  • Oct 17, 2009 4:47pm Oct 17, 2009 4:47pm
Quoting Sim
Great answer, thanks for the time you spent to reply.

I don't know why I said h1 I meant d1 because I've mainly tested similar EAs on d1 and that sorta adds to cause losing trades to be big due to the stop reverse aspect of cutting false signals when in negative, and SLs are also of the same nature as have to be wide on D1
Easy to hit a wrong key and therefore ask a different question to that which we intend. We all do this.

Quote
Nonetheless I commend you for your work, I do wonder how you managed to code the stoch filter from being too choppy? What logic did you use for that?
This is one of those areas where the robot cannot match a humans experience. I have made no attempt to code chop-spotting into the robot. So long as the Main-Line is heading in the right direction and is not reading overbought/sold then the robot considers the Stoch filter to be ok.

Interestingly, moondog's testing also shows that the Stoch filter works better when using the Signal line rather than the Main line, but I haven't done anything about this yet.

Quote
Lastly can I just ask what you do personally use for the EA to judge what trades to take? I brought the initial question up because I had an incline you had a smart way of filtering the d1 chop, and you have My final want would be for you to simply give me a rough gauge of weekly profitability, because my history centre is well and truly buggered from so much testing On all three brokers I...
I am a rubbish manual trader. For me, robot trading is the only way to go, so I use it to trade, not to trigger my manual trades. It is not yet trading live because: the KG Range ea is doing so fantastically that I do not actually need another trading robot; the Ash robot has proven its worth in rejecting prospective trades but has yet to prove itself with actual taken trades.

As to filtering D1 chop, I think the system already does that; nothing clever added on my part.

  • #15
  • Oct 17, 2009 5:18pm Oct 17, 2009 5:18pm
Amazing I see, thanks you cleared a lot up in my mind - I didn't mean to sound doubtful with that "what's so special about this one" question but indeed I was blunt, and you have told me what is special about it BTW Where in UK ya from?

One more thing what is this kg range EA?

Thanks again
  • #16
  • Oct 17, 2009 5:31pm Oct 17, 2009 5:31pm
Quoting Sim
Amazing I see, thanks you cleared a lot up in my mind - I didn't mean to sound doubtful with that "what's so special about this one" question but indeed I was blunt, and you have told me what is special about it BTW Where in UK ya from?
Misterton, a village in the middle of nowhere on the A161 that you can find if you trace it down from the Humber estuary.

Quote
One more thing what is this kg range EA?
Something no trader should be without. Read about it at http://www.forexfactory.com/showthread.php?t=152341

Mind, you will learn some lessons about being blunt. I take no prisoners.

  • #17
  • Oct 17, 2009 7:56pm Oct 17, 2009 7:56pm
Steve,

Still no trades for me when backtesting on H4 or Daily

See attach for EA settings.
Attached File(s)
File Type: zip test.zip   2 KB | 575 downloads
  • #18
  • Oct 17, 2009 9:15pm Oct 17, 2009 9:15pm
hello you need the .set file to show settings not the same .mql

Haha steve interesting I remember trying to make sense of that thread ages ago also, without any luck - but thanks. There's my job for tomorrow!

Haha wow you're quite near Leeds I went there last weekend for my friend's 21st, at Birmingham Uni but born in London
  • #19
  • Oct 17, 2009 10:09pm Oct 17, 2009 10:09pm
Great work from Steve again, Many many thanks!

I am using a 4-digit broker-mig, the stradegy test doesn't show any trades, why's that? I have adjusted the parameters related with digits already. Anyone has same problem?
  • #20
  • Oct 18, 2009 4:18am Oct 18, 2009 4:18am
There was an unexpected problem for backtesters using mt4's strategy tester.

I have come to the conclusion that the more info we can hide from the criminals, the better. The original robot had a pseudo-random MagicNumber generation function based on the Birthday input. This seems not to work in the tester.

I have changed this so that we can use a default magic number of 12345 on demo accounts. If users put the robot onto a live account, the robot will demand a change before it will work.

Should the robot turn out to have legs and end up going on live accounts, users will have different magic numbers, making it just a tiny bit harder for the criminals to detect the robot.

Download the new version from post 1. I know it works in ST because it is doing so right now - just taken its first trade.


3 traders viewing now