scooby-doo BOJ JPY pairs trading robot | Page 119 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading correlation pairs by using the other pairs 843 replies

Export All pairs or selected pairs to .csv with script 3 replies

Pairs of Currency Pairs 4 replies

Trending Pairs / Ranging Pairs 0 replies

Robot trading 120 replies

Attachments: scooby-doo BOJ JPY pairs trading robot
Exit Attachments

scooby-doo BOJ JPY pairs trading robot

  • #2,361
  • Nov 16, 2009 6:34pm Nov 16, 2009 6:34pm
Ok, sorry to sound like a mini-Hitler here, but I have warned people.

Recent askers of daft questions/arguers with clearly experienced and expert contributors are barred.

I shall continue to do this until people get the message.

I am trying to keep this thread open to all but it is becoming increasingly difficult to do so. I now understand why some threads get moved to boards inaccessible to new readers.

'Tis up to you guys. Keep on asking the stupid questions and I will keep on barring you. I hate myself for doing this, but the contributions of those with expertise here are far too valuable to risk losing in favour of those who cannot even be bothered to read post 1 properly.

  • #2,362
  • Nov 16, 2009 6:37pm Nov 16, 2009 6:37pm
In regards to news events, I don't think it can ever be fully automated.

The FF News Event Calendar isn't exactly 100% accurate, and as such to build it into the EA might be turning a blind eye. Just my opinion.

I think it will be better if there's some form of user input as to what events at what time should be taken into account. Maybe at the start of the week, the user loads up a text file which reads the timings of important events of the upcoming week, and thus the EA automatically switches off at those times.

With an already automated system, surely it can't be such a bother for everyone to do a little bit of week at the start of the trading week that will only take up half an hour at most?
  • #2,363
  • Nov 16, 2009 6:40pm Nov 16, 2009 6:40pm
Quoting Skylar Locke
Sorry if this is a little off-topic.

Scooby, I've read that you're possibly trying to incorporate support and resistance into the EA in some form.

The THV EA (http://www.forexfactory.com/showthread.php?t=168347) has incorporated this into their EA taking into account support and resistance points in terms of entry and exit points, so maybe this is something that can be looked into.

I'm not too sure about this, as I know nothing of MT4 coding, but they have done this to a high degree of complexity by the looks of it. So maybe by using...
Thanks for this. Trouble with the kind of thread you link us to is this: scoobs doesn't need to read it; I don't have time to read it.

If the substantive points can be coded then we can think again. Can you post the substantive points?

There is much that can be done to program s/r into an ea. Sing out if you can help. Don't worry if you cannot.

  • #2,364
  • Nov 16, 2009 6:40pm Nov 16, 2009 6:40pm
I feel like such a newbie and you will probably agree. I have downloaded your EA and the files. I have FFCAL Scoobs, Forex Market Hours and RS Group Jpy strength index in Indicators folder. I put the EA on a 15m Gbp/Jpy chart along witht the RS Group indi. I opened a 1h chart gbp/jpy and put RS Group on it. My broker is 5 digit so I set criminal ECN true.

I am getting error "FFCal Scoobs: invalid filename for FileOpen function"

Please help the newb. LOL
  • #2,365
  • Nov 16, 2009 6:41pm Nov 16, 2009 6:41pm
Quoting Skylar Locke
Sorry if this is a little off-topic.

Scooby, I've read that you're possibly trying to incorporate support and resistance into the EA in some form.

The THV EA (http://www.forexfactory.com/showthread.php?t=168347) has incorporated this into their EA taking into account support and resistance points in terms of entry and exit points, so maybe this is something that can be looked into.

I'm not too sure about this, as I know nothing of MT4 coding, but they have done this to a high degree of complexity by the looks of it. So maybe by...
You can build in support/resistence (from your broker level) for a single currency but i am not sure how you do this for a combined/group of currencies.

It would mean us having a single average support line and a single average resistence line but i am not sure that would be of any use to us.
  • #2,366
  • Nov 16, 2009 6:46pm Nov 16, 2009 6:46pm
Quoting Skylar Locke
In regards to news events, I don't think it can ever be fully automated.

The FF News Event Calendar isn't exactly 100% accurate, and as such to build it into the EA might be turning a blind eye. Just my opinion.

I think it will be better if there's some form of user input as to what events at what time should be taken into account. Maybe at the start of the week, the user loads up a text file which reads the timings of important events of the upcoming week, and thus the EA automatically switches off at those times.

With an already automated system,...
This is a very good point.

Those of us who have been around for a while recognise that 'automated' systems need monitoring most of the time. As a programmer, I am guilty of trying to program everything possible - a matter of pride, I suppose.

It could be that including the news filter gives users a false sense of security.

Food for thought, for sure.

  • #2,367
  • Nov 16, 2009 6:48pm Nov 16, 2009 6:48pm
http://www.forexfactory.com/showthread.php?t=204276

That's the link to the Support and Resistance indicator coded by Hanover. I believe that that is the most sophisticated piece of support and resistance coding that you can ever find available at the moment. I'm not sure as to how to integrate it into your EA, it might be a little bit hard because of the various calls etc.

Pretty much, use the indicator to check if price is at a key support or resistance level, and filter out trades that way. We can maybe say that if there's a buy signal, and price is within 3 pips of a key support or resistance level, then not to take the trade unless the key support or resistance level is below the buy price, which makes it a very high chance of a nice bounce into take profit level.

The problem is that different currency pairs has different support and resistance levels, so it will have to be filtered according to currency pairs and add/remove accordingly before the trade is placed.

This is the basic idea I have in my head right now, so I will need to flesh it out a little bit more. Will let you know if I can get around the problem, and will post it then.
  • #2,368
  • Nov 16, 2009 6:48pm Nov 16, 2009 6:48pm
Quoting Skylar Locke
In regards to news events..........
I think it will be better if there's some form of user input as to what events at what time should be taken into account. Maybe at the start of the week, the user loads up a text file which reads the timings of important events of the upcoming week, and thus the EA automatically switches off at those times.

With an already automated system, surely it can't be such a bother for everyone to do a little bit of week at the start of the trading week that will only take up half an hour at most?
Good idea.
  • #2,369
  • Nov 16, 2009 6:49pm Nov 16, 2009 6:49pm
Steve, do you mind if I talk code with you right now?

In your EA your calculating RSIbasket only, which in KG's indicator is the Red line Buffer0. Why are you grabbing the RSIbasket values to use for the White line Buffer1?
LowerTfWhiteLine = GetRSIbasket(Period(), 0);

Am I wrong? Just curious, thanks
24 hours in a day – 24 bottles of beer in a case – Coincidence? I think not
  • #2,370
  • Nov 16, 2009 6:51pm Nov 16, 2009 6:51pm
Quoting Skylar Locke
http://www.forexfactory.com/showthread.php?t=204276

That's the link to the Support and Resistance indicator coded by Hanover. I believe that that is the most sophisticated piece of support and resistance coding that you can ever find available at the moment. I'm not sure as to how to integrate it into your EA, it might be a little bit hard because of the various calls etc.

Pretty much, use the indicator to check if price is at a key support or resistance level, and filter out trades that way. We can maybe say that if there's a buy...
Understood but the idea behind basket or group trading is that you send the whole basket and the + and - work themselves out. If we start cherry picking which ones to send and which ones not too we end up going back to trading single pair's.

But i stand to be corrected.
  • #2,371
  • Nov 16, 2009 6:55pm Nov 16, 2009 6:55pm
I agree, Scooby. That's the flaw in the idea, in that we will not be able to get a universal support and resistance level for the currency pairs in the basket.

Though at the moment, I think we will need at least one more filter if this EA is to be changed to a 24/5 EA.

I'm cracking my head trying to find out what it is.
  • #2,372
  • Nov 16, 2009 6:57pm Nov 16, 2009 6:57pm
Quoting scooby-doo
Where do these people come from ?
Perhaps they were sent by BOJ to create havoc in your thread.....
  • #2,373
  • Nov 16, 2009 6:57pm Nov 16, 2009 6:57pm
Quoting Skylar Locke
I agree, Scooby. That's the flaw in the idea, in that we will not be able to get a universal support and resistance level for the currency pairs in the basket.

Though at the moment, I think we will need at least one more filter if this EA is to be changed to a 24/5 EA.

I'm cracking my head trying to find out what it is.
One area i am looking at is picking the dominating/highest value pair being GBPJPY and using that as an indicator for S/R etc. It will work if correlation with the other pairs holds. But that is a big assumption.
  • #2,374
  • Nov 16, 2009 6:59pm Nov 16, 2009 6:59pm
Quoting matrixebiz
Steve, do you mind if I talk code with you right now?

In your EA your calculating RSIbasket only, which in KG's indicator is the Red line Buffer0. Why are you grabbing the RSIbasket values to use for the White line Buffer1?
LowerTfWhiteLine = GetRSIbasket(Period(), 0);

Am I wrong? Just curious, thanks
It is the white line that shows the relative strength of the trigger currency, in this case JPY.

To be honest, I do not know how the red line is calculated. As originally posited, the idea was this:

 

  1. white line > red, JPY is strong against the other currenceys.
  2. white line < red, JPY is weak against the other currenceys.
  3. trades were taken once the white line was on the correct side of the 50 line.
  4. diff is the difference between white and red.

Therefore, for this system the red line is irrelevant.

Does this answer your query?

  • #2,375
  • Nov 16, 2009 7:00pm Nov 16, 2009 7:00pm
Quoting scooby-doo
One area i am looking at is picking the dominating/highest value pair being GBPJPY and using that as an indicator for S/R etc. It will work if correlation with the other pairs holds. But that is a big assumption.
And we all know what happens to correlations between currencies. Eventually, if not sooner.

  • #2,376
  • Nov 16, 2009 7:03pm Nov 16, 2009 7:03pm
Another area I'm looking at in terms of the 24/5 EA is the reward and risk ratio.

With the original method, I think it's safe to say that it works perfectly the way it is.

I'm not sure if the same concept can be applied to a 24/5 EA in terms of the take profit and stop loss factors. I'm just thinking out loud here.

For a 24/5 EA to work well, the key idea is to cut the losses short, and let the winners run.

So maybe, we close out half the position at our take profit multiplier level, and shift the rest to break even level and let the ball run. Add in a trailing stop with either take profit multiplier divide by 2 or 4 for a tighter trailing stop, and maybe we will be able to ride those breakout points and yet have our position safe still without running into drawdown.

The stop loss is good as it is, with positions being closed as soon as we're 2-3% into account drawdown.

Just a thought.
  • #2,377
  • Nov 16, 2009 7:10pm Nov 16, 2009 7:10pm
Quoting SteveHopwood
It is the white line that shows the relative strength of the trigger currency, in this case JPY.

To be honest, I do not know how the red line is calculated. As originally posited, the idea was this:

 

  1. white line > red, JPY is strong against the other currenceys.
  2. white line < red, JPY is weak against the other currenceys.
  3. trades were taken once the white line was on the correct side of the 50 line.
  4. diff is the difference between white and red.

Therefore, for this system the red line is irrelevant.

Does this answer your...

OK, you just prompted me to look at the indicator code again and I did have the White Line and Red Line buffers backwards, Damn;
SetIndexBuffer(0,t3Array);
SetIndexBuffer(1,rsiArray);

So yes, I knew that the Red line was irrelevant with this system but thought that's what RSIbasket was calculating, woops.

Thanks
24 hours in a day – 24 bottles of beer in a case – Coincidence? I think not
  • #2,378
  • Nov 16, 2009 7:10pm Nov 16, 2009 7:10pm
Still having that TP close error on the 24hr EA. (Screenshot attached).
Or has this been fixed in V1.2 ?
Attached Image (click to enlarge)
Click to Enlarge Name: TP Close Error.jpg Size: 230 KB
  • #2,379
  • Nov 16, 2009 7:12pm Nov 16, 2009 7:12pm
Quoting Skylar Locke
Another area I'm looking at in terms of the 24/5 EA is the reward and risk ratio.

With the original method, I think it's safe to say that it works perfectly the way it is.

I'm not sure if the same concept can be applied to a 24/5 EA in terms of the take profit and stop loss factors. I'm just thinking out loud here.

For a 24/5 EA to work well, the key idea is to cut the losses short, and let the winners run.

So maybe, we close out half the position at our take profit multiplier level, and shift the rest to break even level and let the ball...
With all due respect to Steve and other expert coders here, my trading floor employes 14 programmers who also have extensive experience of financial markets. Part of their remit is to research and try to fully automate trading scenarios. Their dept budget would make you cry with envy. But, still they cannot provide anywhere near a 24/5 solution to a single or group of currencies. The system generates a buy, sell or close signal, its then MY job to agree or override. I wont even go into the paperwork and explanations I have to make if I decide to override its decisions/advice.

But what is not possible this hour maybe possible come Tokyo open hour tomorrow. Now, when exactly is that ?

Scoobs
  • #2,380
  • Nov 16, 2009 7:14pm Nov 16, 2009 7:14pm
Letting winners run defeats the purpose of basketing. You might as well pick tops and bottoms and hope for the best.

The idea of this is to jump onto a liquidity move and make take it off the table, and wait for the next move.

As the indicator is showing an index and is used on multiple time frames, Scalping this method is probably alot smarter than trying to trade a swing with it.

Equity stop loss is by far superior to a regular stop with a basket, as I have encountered a stop out on g/j but the other pairs hit my basket level of profit so it was basically a 5$ return trade instead of a 200$ one. It worked out pretty good.

Settings are very important and you can't scalp too little or make too much otherwise you will find yourself on the receiving end of an ugly drawdown.

I feel this method will do well in trending periods, because as has already been shown, if the trend is going, the indicators on 1h and 15m do not match thus you don't get a trade until a plateu period and usually its in the direction of the trend.
Thats just what I have observed looking at the 24/5 EA.

I believe overfiltering something that already focuses on a central currency rather than currency pairs is silly and will be counter productive. News time filters etc are all necessary as well as the option of trading on friday or not. But I don't think adding indicators would help. Support and resistance may help, but as we know, support and res is made to be broken

This EA will be succsessful with sound money management as well as a little manual intervention. The wins far outweigh the losers and will outweigh them over the longer term in my opinion.

The other thing I observed was that the basket is alot safer if you remove aud/jpy or gbp/jpy from the mix gbp/jpy being the better of the two to remove

Just my 2cents..

P.S what time settings do I use for sydney? Excuse my humour.. sorry.
2 traders viewing now