scooby-doo BOJ JPY pairs trading robot | Page 141 | 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

  • #2,801
  • Nov 19, 2009 6:16am Nov 19, 2009 6:16am
Quoting scooby-doo
Steve

If you build in a 200 SMA and 5% hedge into the 24H robot then you can just ignore all news.

Scoobs.
For all the trades in the basket?

  • #2,802
  • Nov 19, 2009 6:41am Nov 19, 2009 6:41am
Took a hit this am on the v 2.8 it was the buy side no sells last night
my sell side is 88.89% my buy side is 71.43% lost 10% of acct last night

FXDD no hedge because I live in a country where I'm considered too stupid to think for myself

DX how did the Hedge far for you
  • #2,803
  • Nov 19, 2009 6:51am Nov 19, 2009 6:51am
Just checked my 24hr 1.5 total tank lost 70% since yesterday afternoon
only 16% wins on the buy side

Might try turning sell only on to see what happens
  • #2,804
  • Nov 19, 2009 7:05am Nov 19, 2009 7:05am
Quoting dryclean
Took a hit this am on the v 2.8 it was the buy side no sells last night
my sell side is 88.89% my buy side is 71.43% lost 10% of acct last night

FXDD no hedge because I live in a country where I'm considered too stupid to think for myself

DX how did the Hedge far for you
Hit the SL... 30% gone
The hedge did only small good as the main losers where AUD, CHF and GBP...

Will see over my risk settings again

Just as an addition:
I wrote an EA some years ago where an entry filter was even more strict then scooby's: It only entered trades when the distance of pricecurrent was 100 pips above 200SMA for a buy and 100 pips below for a sell.
Still using it privately with success...
"The only successful substitute for brains is silence."
  • #2,805
  • Nov 19, 2009 7:34am Nov 19, 2009 7:34am
Quoting SteveHopwood
For all the trades in the basket?

No, just GBPJPY
  • #2,806
  • Nov 19, 2009 7:53am Nov 19, 2009 7:53am
I've had two buy baskets within the last 1 1/2 hours hit TP, but got slaughtered last night.
  • #2,807
  • Nov 19, 2009 8:07am Nov 19, 2009 8:07am
Quoting Delage
I've had two buy baskets within the last 1 1/2 hours hit TP, but got slaughtered last night.
Same here. What was interesting last night is after the firstBuy basket stopped out, the Scooby 24H EA almost immediately opened a second Buy basket --- despite the down trend.
  • #2,808
  • Nov 19, 2009 8:08am Nov 19, 2009 8:08am
Everyone

You need to step back and think what you are trying todo here.

The original method works well when there is BOJ intervention. However, trying to adapt it to run 24/5 means a different approach entirely.

You should never go against the trend unless you know what you are doing. i.e. Many years of experience.

The GBPJPY has the highest value and range, therefore this is your base currency for calulating starting lot size and for checking the 200 SMA for the entire basket.

If GBPJPY bid price is below the 200 SMA and XXXJPY oversold then you should only be looking for sell trades, if it is above the 200 SMA and XXXJPY overbought then you should only be looking for BUY trades.

This applies to the 24/5 Robot only.

Cheers,

Scoobs.

PS. We get away with sometimes going against the trend with the original method because we know the BOJ will intervene after or shortly before their official session time.
  • #2,809
  • Nov 19, 2009 9:30am Nov 19, 2009 9:30am
Quoting zznbrm
There is a strategy out there (presumably traded by scooby-doo for quite some time) that has an edge (BOJ intervention during Tokyo Open/Close). It appears to me that the main goal of this thread is to create an EA to automate the trading of this strategy. A second EA has also been added to trade the strategy 24/5.

In the real programming world, we define requirements, then we design/construct/test.

I simply don't understand why the continuous tinkering (adding/removing news, changing RSI levels, adding hedging, etc). Why all the changes?

Would...
The BIGGEST problem here is trying to convince people that you need equity to be able to run these methods. And I am not talking micro accounts trading 0.01 or such.

To get the usuable lot sizes corrrect then you need to divide your balance by 25000 units for the base lot size, also we need 2 decimal places. So, the minimum to start trading is 0.10, which means a minimum of 2500 of base currency. This is even before you start talking about hedging if things go wrong etc.

These methods are sound and used by many trading floors. I guess the misconception here is that they could be adapted for retail traders with no experience that just want a 24/5 fully automated system with no user intervention.

Personally, i would just stick with the original method which relies on the fact that we can buck the trend with the help of BOJ and make some small profits. I said it many times, if you want to make more money then you have to use a bigger account balance, no ifs or buts.

I said I would research if the BOJ could be adapted to run outside of Asia session and would report back at the end of January 2010. Then we find another Robot pops up on the thread and everyone goes off in a completely different direction. I am not knocking this but its causing confusion and in anycase, maynot actually be viable for retail traders/brokers.

Scoobs.
  • #2,810
  • Nov 19, 2009 9:58am Nov 19, 2009 9:58am
Quote
These methods are sound and used by many trading floors. I guess the misconception here is that they could be adapted for retail traders with no experience that just want a 24/5 fully automated system with no user intervention.

I would NEVER run a 24/5 robot with no user intervention. Maybe I'm a little more conservative than other retail traders around here.

Quote
Personally, i would just stick with the original method which relies on the fact that we can buck the trend with the help of BOJ and make some small profits. I said it many times, if you want to make more money then you have to use a bigger account balance, no ifs or buts.

I said I would research if the BOJ could be adapted to run outside of Asia session and would report back at the end of January 2010. Then we find another Robot pops up on the thread and everyone goes off in a completely different direction. I am not knocking this but its causing...

Ok, good stuff scooby-doo. I hope in the future this thread will concentrate on the original EA only.
  • #2,811
  • Nov 19, 2009 10:27am Nov 19, 2009 10:27am
24H 1.7 in post 1. SMA code included.

  • #2,812
  • Nov 19, 2009 10:34am Nov 19, 2009 10:34am
Quoting zznbrm
There is a strategy out there (presumably traded by scooby-doo for quite some time) that has an edge (BOJ intervention during Tokyo Open/Close). It appears to me that the main goal of this thread is to create an EA to automate the trading of this strategy. A second EA has also been added to trade the strategy 24/5.

In the real programming world, we define requirements, then we design/construct/test.

I simply don't understand why the continuous tinkering (adding/removing news, changing RSI levels, adding hedging, etc). Why all the changes?

Would...
The 'proper' robot has the design structure you look for; it is a proven trading strategy. Some of the 'enhancement' arose from ideas we had; hedging was always in my mind but was not necessary for the original bot.

The 24H versions is basically a bit of fun. If we end up making it work, great. If not, nothing lost apart from a bit of my coding time. Develoing on an ad-hoc basis, constant change is inevitable.

Good things can come out of developing trading systems on an ad-hoc basis; the KG Range robot is a great example. It was a long time in development but has changed little in recent months.

  • #2,813
  • Nov 19, 2009 10:38am Nov 19, 2009 10:38am
Quoting SteveHopwood
24H 1.7 in post 1. SMA code included.

Steve,

The SMA rule does not apply prior to Asia/Tokyo start. Only after.

So, if we are in the situation whereby the SMA rule applies then the hedge should kick in at 5% if there is a potential trend reversal.

The reason we dont need to worry about the SMA prior to Tokyo open is because we take advantage of the fact that the BOJ will buck the trend in the short term for their own needs, all we do is ride their tail coats. We need the SMA rule after as then we are probably approaching London, Franfurt and New York sessions whereby the BOJ are less preverlent.

Also, the Bid should be greater than (SMA 200 + 0.200) for BUY or Bid should be less than (SMA 200 - 0.200) for SELL on a 5 digit broker's.

I havent looked at the latest code but the above needs to apply.

Cheers,

Scoobs
  • #2,814
  • Nov 19, 2009 11:04am Nov 19, 2009 11:04am
Quoting scooby-doo
Steve,


Also, the Bid should be greater than (SMA 200 + 0.200) for BUY or Bid should be less than (SMA 200 - 0.200) for SELL on a 5 digit broker's.

I havent looked at the latest code but the above needs to apply.

Cheers,

Scoobs
It didn't, but does now. Also included a UseSma input so users can turn it off in the run up to Tokyo open.

  • #2,815
  • Nov 19, 2009 11:18am Nov 19, 2009 11:18am
Quoting SteveHopwood
It didn't, but does now. Also included a UseSma input so users can turn it off in the run up to Tokyo open.

While you are on a roll............

There is a way to make the Overbought and Oversold settings dynamic.

Example:

When we first start we dont know really where to set them, so we set them to 80/20 or 90/10. And now for the clever bit. In the code, Oversold = Old.Bar if Old.Bar < Oversold, then on a new bar if Old.Bar <= Oversold and New.Bar > Oversold we have a trade. Likewise for Overbought.

This means the system will change as the market changes.

Understand ???????
  • #2,816
  • Nov 19, 2009 11:32am Nov 19, 2009 11:32am
Steve

Not sure I explained that very well so I will try and do it with pesudo code...

Over.Bought = 80
Over.Sold = 20

if(Old.Bar < Over.Sold)
Over.Sold = Old.Bar
else
Over.Sold = 20 //reset

if(Old.Bar > Over.Bought)
Over.Bought = Old.Bar
else
Over.Bought = 80 //reset

BUY XXXJPY if(Old.Bar >= Over.Bought && New.Bar < Over.Bought)

SELL XXXJPY if(Old.Bar <= Over.Sold && New.Bar > Over.Sold)
  • #2,817
  • Nov 19, 2009 12:17pm Nov 19, 2009 12:17pm
Just closed sell basket after 1 1/2 hours at TP. Some worrying dd, moved very slowly until the last few minutes.
  • #2,818
  • Nov 19, 2009 12:23pm Nov 19, 2009 12:23pm
Quoting Delage
Just closed sell basket after 1 1/2 hours at TP. Some worrying dd, moved very slowly until the last few minutes.
As I have stated before, its not a good idea to open trades between 4.00pm GMT and 9.00pm GMT unless you are very experienced. The markets are usually dead.

Scoob.
  • #2,819
  • Edited 1:23pm Nov 19, 2009 12:31pm | Edited 1:23pm
Hi Steve

I think there is something wrong with the settings in FFCal (getting FF calendar and/or processing the calendar data).
My VPS is located in the US in the EST time zone, actual time is 13:20 or 01:20 PM. The 24H Ea V.1.7 is mentioning that trading is suspended because of news. NewsImpact is set to 3 (High).
According to FF calendar there isn't any high impact news concerning JPY pairs between 12:00 and 13:30 EST time. Or i'm i wrong ?

Thanks for checking

Edit : I think i found something. FFCal downloads the XML with the wrong time zone. Instead of downloading it with the local time zone it is downloaded with the GMT 0 time zone. Attached you find my actual XML. There you see that it has the info that today 19.11.2009 at 13:30 there is USD Unemployment Claims. But 13:30 is for GMT 0 and not my local time zone. FFCal should download the XML with the local time zone or the EA should consider the local time zone when checking the XML file.....

I hope i was clear with this information's, else let me know
Attached File(s)
File Type: xml 11-19-2009-GBPJPY15-FFCal.xml   15 KB | 353 downloads
  • #2,820
  • Nov 19, 2009 1:02pm Nov 19, 2009 1:02pm
Quoting scooby-doo
While you are on a roll............

There is a way to make the Overbought and Oversold settings dynamic.

Example:

When we first start we dont know really where to set them, so we set them to 80/20 or 90/10. And now for the clever bit. In the code, Oversold = Old.Bar if Old.Bar < Oversold, then on a new bar if Old.Bar <= Oversold and New.Bar > Oversold we have a trade. Likewise for Overbought.

This means the system will change as the market changes.

Understand ???????
This would take away the need for globals and possibly get us into trades earlier.

I assume that by 'bars' you mean separate 15 min bars, not tick by tick.

4 traders viewing now