scooby-doo BOJ JPY pairs trading robot | Page 368 | 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

  • #7,341
  • Mar 9, 2010 7:14am Mar 9, 2010 7:14am
[quote=charvo;3533076]in fact, statistically/mathmatically, up 20% then drop to 6.8% is very bad for a month trading.

I totally am with you!

but with a RR 1 :10 if you get hit the percentage is going to drop like a rock so the incremental % increase you are looking for will be jagged not so smooth. I think the ability not have the original trade a buy early or a sell late in the session will really help to get that smooth line we are all looking for
  • #7,342
  • Mar 9, 2010 7:40am Mar 9, 2010 7:40am
Quoting Meister81
Same here, let it run on Alpari UK but no trades so far.
Best regards!
No trades with Char's EA yet but I did have a couple of THG 1.6 trades. Char's EA uses ADX for trend strength and Laguerre for entry (in addition to EMAs). So it probably doesn't trade as much as THG.

Invisible's 3TF EA continued to perform well yesterday.

Sandy
  • #7,343
  • Mar 9, 2010 7:56am Mar 9, 2010 7:56am
Quoting maxou888
525c performs well today again.
Max, thank you for the update, however I noticed that the break even is just that, break even. Could we have a couple of pips profit lock in please.

Tks.
  • #7,344
  • Mar 9, 2010 8:34am Mar 9, 2010 8:34am
Quoting sundog
Max, thank you for the update, however I noticed that the break even is just that, break even. Could we have a couple of pips profit lock in please.

Tks.
You can use TS included, but I didn't test it. However there is still many ways to close by profit.
Sorry, no time to watch at code actually.
  • #7,345
  • Mar 9, 2010 8:42am Mar 9, 2010 8:42am
Quoting maxou888
You can use TS included, but I didn't test it. However there is still many ways to close by profit.
Sorry, no time to watch at code actually.
Rgr, I have the TS set, but am unsure of the parameters. No problem I'll work it out.

Tks.
  • #7,346
  • Edited 9:01am Mar 9, 2010 8:50am | Edited 9:01am
Thanks to Elovv, Sundog, and Meister81.

On my end, HG v1.6 charvo is not trading, but original HG v1.6 is like toughest motherfucker ever lived, i'm shocked by its winnings today.

so i checked the difference. i found that v1.6 charvo enter on EMA20/EMA40/EMA120, while the original v1.6 enters on EMA20/EMA200.

As i mentioned i tried to follow Scoobs logic/comments on thread, so this difference may stem from that i lost myself somewhere. Sorry for this guys, although 3 EMAs may be safer, currently it seems killing too many entries.

Please download and overwrite last version, i compared line by line against original V1.6. should be working now.

Note that this is not my stupid guess of HG v2.x. This is a convenient/efficient version of HG v1.6.
Attached File(s)
File Type: mq4 The Holy Grail v1.6_charvo.mq4   15 KB | 496 downloads
  • #7,347
  • Mar 9, 2010 9:00am Mar 9, 2010 9:00am
Quoting sundog
You are free to reduce the SL %, as I mentioned to you in an earlier post, my SL% is less than 5 (4.5 actually), so far I haven't been stopped out prematurely.
Sundog, that's interesting.

I guess you have Hedging disabled then??

Also, when you say 'so far' you haven't been stopped out, what do you mean by 'so far'? How long have you been testing with this SL? (I noticed previously that you were using an SL of 3%. So presumably that was too tight?)

Cheers.
  • #7,348
  • Mar 9, 2010 9:10am Mar 9, 2010 9:10am
Maxou, I really get fascinated by your EA and want to help in anyway i can. I did spend sometime on it, but get lost in the middle.

maybe a few quick questions can help me a lot, many thanks in advance!

1. this is a reversal(to mean) EA, trading OS or OB conditions. (but i noticed that last Thur and Fri, your EA captured the big move around NFP, is it also capable of riding trend?)

2. hedge kicks in when an initial OB/OS reversal trade goes bad (market continue OB and OS, instead of retrace), like Steve's original BOJ.

3. if hedge also goes bad, what is the logic to exit the dire strait?

4. mode1, 2, 3, 4, and 5 are not really affecting each other, it is like 5 EAs working together on different pairs?

5. if some pairs are removed from a basket, for example, if i removed gbpchf, or nzdjpy, will that hurt the system's logic?

wow, so many dumb questions, you may just use yes or no,



Quoting maxou888
525c performs well today again.
  • #7,349
  • Mar 9, 2010 9:31am Mar 9, 2010 9:31am
Quoting Jezza
Sundog, that's interesting.

I guess you have Hedging disabled then??

Also, when you say 'so far' you haven't been stopped out, what do you mean by 'so far'? How long have you been testing with this SL? (I noticed previously that you were using an SL of 3%. So presumably that was too tight?)

Cheers.
Jezza, yep Hedging disabled, us crim.

You are correct, previously I had it down to 3%, but I had a couple of losses, while (I dont think) were not due to the low SL, it did have the effect of reducing the total amount, so I thought it prudent to increase the SL up to give it some more room to breathe.

I have had this running on a live micro for 9 weeks now, however it shares it with other EA's so is not responsible for all wins/losses.

regards
  • #7,350
  • Mar 9, 2010 10:06am Mar 9, 2010 10:06am
Quoting maxou888
You can use TS included, but I didn't test it. However there is still many ways to close by profit.
Sorry, no time to watch at code actually.
Yes, it's doing well for me since dropping it on yesterday as well. I do agree though, from what i've seen so far, I see a nice move in our favor, then just to have it close at be later. All I see in the inputs is whether to use TS or not, but nowhere to set the amount of TS.
In any case, it's performing nicely as is, up 7.5% in less than a day. We'll see how it gets through the week
  • #7,351
  • Edited 11:48am Mar 9, 2010 10:34am | Edited 11:48am
Hello All

This is a modified version of the 2TF Scooby bot mot.

 

  1. It only enters a trade when H1 and H4 are all in the same trend according to the 20EMA, 34EMA and 100SMA.
  2. M15 and H1 have to be OB/OS RSI.
  3. M5 RSI has to be coming out of the OB/OS zone (>80 or <20).
  4. M5 5EMA and 8EMA have to be in the direction of the trade.

This seems to improve entries further and keeps floating DD low.

So, to give an example, for a buy to initiate:

 

  1. 20EMA > 34EMA > 100SMA on H1 and H4 = trend is up
  2. M15 and H1 RSI are both <20 = oversold
  3. M5 RSI >20 = M5 is coming out of oversold region
  4. 5EMA >8EMA on M5 = M5 trend is turning

The latter 2 conditions are new for this version. As before I am also using MPTM (same set as before) to manage trades.

The basic premise is lifted from the Holy Grail, with the addition of the MA on the M5. The point is to not only spot when a reversal in the direction of the trend is likely to take place (OBOS), but also to use both the RSI and the MA on the M5 to spot when it is actually beginning to start. I think Scooby's Laguerre RSI implementation on the Holy Grail he is testing now is likely to do this better, but this is my crude attempt to achieve a similar effect.

I only had sells since the start of the week, as I forgot to reverse the "<" in one part of the code, which meant that buys would never be initiated. However, I am very happy with its performance. As a reference I have been running my original mod in parallel. This new version avoided the bad trades and has a 100% record so far with exceptionally small floating DD.

So put it on DEMO and see how it goes.

Invisible

P.S. I have a feeling that all this will end up being is some entertainment while we wait for the full Scooby Holy Grail, but if this does turn into something valuable in its own right, that's great too.

P.P.S. If there are any ideas about how to lock in profit better, that would be appreciated also, as I think the exits can be improved a lot.

Attached File(s)
File Type: mq4 Invisible's Scooby 3TF MA 20 34 100 - RSI 80 - 20 - M5 RSI and MA entry filter.mq4   7 KB | 435 downloads
  • #7,352
  • Mar 9, 2010 11:22am Mar 9, 2010 11:22am
Quoting Invisible
It only enters a trade when M15 H1 and H4 are all in the same trend according to the 20EMA, 34EMA and 100SMA.
Nicely done, Invisible!

Check the trend calculation as it appears the M15 MAs are missing.
  • #7,353
  • Mar 9, 2010 11:27am Mar 9, 2010 11:27am
Quoting Invisible
Hello All

This is a modified version of the 3TF Scooby bot mot.

 

  1. It only enters a trade when M15 H1 and H4 are all in the same trend according to the 20EMA, 34EMA and 100SMA.
  2. M15 and H1 have to be OB/OS RSI.
  3. M5 RSI has to be coming out of the OB/OS zone (>80 or <20).
  4. M5 5EMA and 8EMA have to be in the direction of the trade.

This seems to improve entries further and keeps floating DD low.

So, to give an example, for a buy to initiate:
[list][*]20EMA > 34EMA > 100SMA on M15 H1 and H4 = trend is up[*]M15 and H1 RSI...


Thanks Invisible. The original 3 TF did wel llast night. 13 trades and 11 winners. Will be interested to see how the new version compares.

Sandy
  • #7,354
  • Mar 9, 2010 11:40am Mar 9, 2010 11:40am
Quoting charvo
Thanks to Elovv, Sundog, and Meister81.

On my end, HG v1.6 charvo is not trading, but original HG v1.6 is like toughest motherfucker ever lived, i'm shocked by its winnings today.

so i checked the difference....
Thanks Charvo. I have dowloaded the knew version and put it on my live mini Ibfx account. I think I have read somewhere that Scoobs EA:s dont take trades on Ibfx. I prefer mini live before demo. Lets see what happens.

Best wishes
Elovv
  • #7,355
  • Mar 9, 2010 11:47am Mar 9, 2010 11:47am
Quote
Check the trend calculation as it appears the M15 MAs are missing.

You are right! This ain't no 3TF bot. At least not in terms of overall trend detection. What was I thinking????!!??!?

OK so it checks the MAs on H1 and H4 for the overall trend. Not M15 as well.

You know what, it is still doing pretty good for a 2TF trend checking bot. Much better than the original Invisible Mod. So perhaps it is serendipity.

I will leave it as is and see how it plays out.
  • #7,356
  • Edited 2:07pm Mar 9, 2010 11:51am | Edited 2:07pm
Quoting charvo
Maxou, I really get fascinated by your EA and want to help in anyway i can. I did spend sometime on it, but get lost in the middle.

maybe a few quick questions can help me a lot, many thanks in advance!

1. this is a reversal(to mean) EA, trading OS or OB conditions. (but i noticed that last Thur and Fri, your EA captured the big move around NFP, is it also capable of riding trend?)
It does both. I mean, first, it trades on the way of the trend. The question is: how to determine the trend. Actually I use MACD BB that gives, in my opinion, the best results. Trend determination is very difficult. If we use fast changing trends, we have more reversal, but react faster, if we use long trend detection (higher Tf), we are less often on the bad way, but we need much more time to detect a trend reversal and we can continue opening trades on the wrong side during many days ! After many test I feel these MACD BB are the best.
To enter the trade, it uses OB &Os levels. Mode 1 : StochRsi (it's an internal function same like indic StochRsiOscilator) or LaguerreRsi, and KG for other modes (it's an internal function same like indic "RS group Kg")

Quoting charvo
2. hedge kicks in when an initial OB/OS reversal trade goes bad (market continue OB and OS, instead of retrace), like Steve's original BOJ.
It's a little more complicated
The hedge is opened if the loss is > PairTarget*4 AND the trend had reversed (so the hedge is a trade in the new trend way) and a KGStatus signal occurs (only KG mode 2,3,4,5)
We continue to open hedge if we are still on this new trend AND a new signal occurs (mode1) or a new KGStatus signal occurs (KG mode 2,3,4,5)

Quoting charvo
3. if hedge also goes bad, what is the logic to exit the dire strait?
We cut the hedge on Opposite trend (NO mode 1) or opposite KGStatus Signal (KG mode 2,3,4,5)

Quoting charvo
4. mode1, 2, 3, 4, and 5 are not really affecting each other, it is like 5 EAs working together on different pairs?
Yes and no. They open trades independently, except it (mode1) controls globally the pair exposure, and, very important, the functions currency target and global target close all the trades for all the modes. The idea is to try to close profit trades with losing trades if there is too many trades opened in one currency, or globally.
Quoting charvo
5. if some pairs are removed from a basket, for example, if i removed gbpchf, or nzdjpy, will that hurt the system's logic?
Absolutly not, except functions currency target and global target that will have some tought times to close trades if where is not enough currencies. but if you delete only few pairs no problem
  • #7,357
  • Mar 9, 2010 12:10pm Mar 9, 2010 12:10pm
Maxou: many thanks! this is one of gems in this thread.

[quote=maxou888;3535423]It does both. I mean, first, it trades on the way of the trend. The question is: how to determine the trend. Actually I use MACD BB that gives, in my opinion, good results.
  • #7,358
  • Mar 9, 2010 12:33pm Mar 9, 2010 12:33pm
@invisible

I want to put your newest modification to demo. Could you say me how to setup? Which Pair(s) and TF??
  • #7,359
  • Mar 9, 2010 12:49pm Mar 9, 2010 12:49pm
Our friend Boris Schlossberg`s latest article "How To Scalp Fundamentally" can be found here:
http://www.investopedia.com/articles...tner=fxweekly3

and a 10 minute video can be found at youtube posted 3 days ago.
kind regards
  • #7,360
  • Mar 9, 2010 1:19pm Mar 9, 2010 1:19pm
Quoting SteveHopwood
All this gave me an idea.

I have added a couple of inputs to BOJ Original 2.4 to allow users to control whether they want the robot to take buy and sell trades:

 

  1. LongTradesAllowed
  2. ShortTradesAllowed

If you do not want to trade against a trend in case BOJ does not intervene, simply turn off the direction you do not want. This applies to both wave 2 & 3 trades but not separately; for example turning off sells will stop the bot trading wave 3 short at Tokyo open and wave 2 pre close.

Steve,

As I understand this modification, the user would need to change the setttings prior to each trading window after manually determining the trend direction.

Would coding in a trend filter, such as the 200SMA as Scooby seemed to suggest, that could be enabled or disabled make the bot more automatic for those users who may not be able to access their computer at the appropriate times to change a manual setting for LongTradesAllowed and ShortTradesAllowed?

JR
5 traders viewing now