Nanningbob V5: Steve Hopwood's auto-trading robot | Page 77 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Nanningbob 4h trading system ver. 5.0 4,432 replies

Nanningbob 60x90 EA 3,920 replies

Help regarding NanningBob EA 27 replies

Nanningbob 4H trading system 2,017 replies

Help to improve profitable EA - 1H NanningBob 40MA 200 BB 10 replies

Attachments: Nanningbob V5: Steve Hopwood's auto-trading robot
Exit Attachments

Nanningbob V5: Steve Hopwood's auto-trading robot

  • #1,521
  • Oct 27, 2010 12:23am Oct 27, 2010 12:23am
Quoting Viv108
sorry to go on like a broken record... but the MT4 platform still crashes with the latest update.

Just thought i'd let you know!
Does the platform crash when running the bot, or does it crash on opening? I have found that I can run Max's EA without problems, but I cannot open MT4 with the bot preloaded. I have to open with blank charts, load the EA and start it running without closing the platform.

Cheers, Beastie.
  • #1,522
  • Edited 12:30pm Oct 27, 2010 6:55am | Edited 12:30pm
Quoting Viv108
sorry to go on like a broken record... but the MT4 platform still crashes with the latest update.

Just thought i'd let you know!
Hi
Take out all indicators and EAs from Metatrader that you don't need. Start Metatrader With “Expert Advisors” off. If it is not off when you start, click on the button the first thing you do. You have couple of seconds to do it in the start moment. Then don't touch anything until you see the trade tab in down left corner. If you make a mouse click before everything is “alive” than the Metatrader crash.
  • #1,523
  • Oct 27, 2010 8:31am Oct 27, 2010 8:31am
Take off the EA from most of the charts or better still from ALL the charts. Load them only after you have re-logged in.
  • #1,524
  • Oct 29, 2010 12:37am Oct 29, 2010 12:37am
I posted this in 'The Beast' thread, but it may be equally useful (potentially) here.


An idea struck me last night which I think could help any EA that incorporates a recovery system. I'm just thinking out loud really and have not had chance to carry out any tests, but anyway...

When a pair goes into recovery, we have the obvious problem of drawdown on the account equity. On top of this, that particular pair will not be earning any pips, all trading is effectively frozen until the BE line is reached. If we are currently long and waiting, then a lot of good short trades will be missed.

Then it struck me. Why not run long and short trades on seperate charts?
This way, when one chart is stuck in draw down, the other can still be earning. And of course both charts cannot go deep into recovery together, unless one of them is really too far gone.

The effectiveness is easy to test. Run a backtest with both long and short enabled and record the result. Then run seperate backtests with 'Only Long' and 'Only Short' and add the 2 results together.

I didn't get chance to try this last night, but I see no reason why the second test shouldn't produce better results.

There is one drawback in that a magic number would be required to keep the long and the short trades seperate. This may not be desirable.

Thoughts?

Cheers, Beastie.
  • #1,525
  • Oct 29, 2010 4:00am Oct 29, 2010 4:00am
that's already the case with MAXX bot
I have both CT & TT that takes trades in the both directions, not completely because for CT trades i stay stuck in one direction for recovery, but for example if a CT is long there is many chances that a TT trades occurs in the opposite direction.
But that's not the holy grail because profit trades are closed rather quikly and bad trades can lurk for long time.
I remember a bot I developed last year (basket EA) that works like you said (continue take trades even if a pair in large DD), moreover this bot used hedge, but that didn't solve the DD problem.
The problem with this concept is that your bot can overtrade rapidly, become out of control, your Balance continues to grow but not your Equity, and if you are not lucky, you add bad trades and DD increase.
Here is an example of my bot : currently one TT trade (short) and one CT trade (long) both in loss for moment
Attached Image
  • #1,526
  • Oct 29, 2010 4:27am Oct 29, 2010 4:27am
Quoting maxou888
that's already the case with MAXX bot
I know you bot takes both with trend and counter trend trades, but it is very difficult to predict and ride a trend. However, we know that the NB counter-trend strategy is sound. I believe there is some merit in allowing NB trades in the opposite direction when stuck in recovery.

If we run a NB backtest allowing LOng only, it will be profitable. Same for Short only. If we run both together, they will each produce profit when the other is in recovery, thus helping offset the drawdown. I realise that the profit will not offset the DD completly, but it will at least capture the opportunities that would otherwise be missed.

I shall try some tests when I get chance and see how it fares.

Cheers, Beastie.
  • #1,527
  • Oct 29, 2010 8:55am Oct 29, 2010 8:55am
how about a basket for breakeven?

If recovery = true
let the trade continue

If DD = BE basket
close trade
  • #1,528
  • Oct 29, 2010 9:28am Oct 29, 2010 9:28am
Quoting Beastie Boy
I know you bot takes both with trend and counter trend trades, but it is very difficult to predict and ride a trend. However, we know that the NB counter-trend strategy is sound. I believe there is some merit in allowing NB trades in the opposite direction when stuck in recovery.

If we run a NB backtest allowing LOng only, it will be profitable. Same for Short only. If we run both together, they will each produce profit when the other is in recovery, thus helping offset the drawdown. I realise that the profit will not offset the DD completly,...
Yes, far from stupid idea, but not easy to manage in code
We shoud have a different magic number for opposite trades. Working with 2 charts is not not easy because that means pass info thru global variables, that means 2x20+ global variables.
I will think about your idea but I am afraid I need a full day to code that and for moment no free time unfortunatly.
  • #1,529
  • Oct 29, 2010 12:12pm Oct 29, 2010 12:12pm
Quoting maxou888
I will think about your idea but I am afraid I need a full day to code that and for moment no free time unfortunatly.
Thanks for considering it Max. I must admit, the problem regarding Global Variables never occured to me. I simply changed the magic numbers, disabled a part of code and ran the EA on 2 charts, one set Long Only, the other Short Only.
I ran it all through today seemingly without problems. Will the GV only cause problems if the EA has to be restarted?

Incidently, as I write this, I have just finished my first test using NB trades only:-

Long and Short Trades - 2573 Profit
Short Trades only - 2064, Long Trades only 1358, totaling 3422.

That's a 33% profit increase. Not enough of a test to be conclusive, but promising.

Cheers, Beastie.
  • #1,530
  • Oct 29, 2010 12:25pm Oct 29, 2010 12:25pm
Beastie, keep continuing down that path, its worth pursuing. I've been doing this with ea's for years now, usually playing with martingale methods, but I think it has its place here also. Any time trades are not closed for a loss for whatever reason (martingale, recovery, hedge), always makes sense to me to let it trade the other side while current trades are stuck in wait mode (dd mode).
Curious as to what you come up with
  • #1,531
  • Oct 30, 2010 4:31am Oct 30, 2010 4:31am
Quoting tazmaan
...always makes sense to me to let it trade the other side while current trades are stuck in wait mode (dd mode).
Curious as to what you come up with
My thoughts exactly. Unfortunately, weekends are not good for testing with Metatrader due to the spread problems. I am interested to see what benefits this can bring though.

Cheers, Beastie.
  • #1,532
  • Oct 30, 2010 9:35am Oct 30, 2010 9:35am
Quoting Beastie Boy
My thoughts exactly. Unfortunately, weekends are not good for testing with Metatrader due to the spread problems. I am interested to see what benefits this can bring though.

Cheers, Beastie.
I used few hours to modify my bot for working with your idea, unfortunatly there are bugs and it doesn't work
  • #1,533
  • Oct 30, 2010 12:55pm Oct 30, 2010 12:55pm
Quoting maxou888
I used few hours to modify my bot for working with your idea, unfortunatly there are bugs and it doesn't work
How are you trying to implement this Max? I can imaging that adding the long and short NB functionality to your existing bot would be very difficult.
Would it help to try and get it working with just the NB trades, whilst retaining all the other features (time management, variable recovery lots etc) that you have built into the current EA.
Trying to manage NB, 4H, RSI and KG all together must be mind blowing!

I would love to help, but your bot has gone well beyond my meagre capabilities.

Cheers, Beastie.
  • #1,534
  • Nov 1, 2010 8:52pm Nov 1, 2010 8:52pm
--------------------
  • #1,535
  • Nov 2, 2010 8:16am Nov 2, 2010 8:16am
Look at this one
I followed BaestieBoy idea to don't stay sticked
I completly rewrited the core code. Much more simple.
Added close by profit, very important
Play with this one in BT, very funny. More martinguale style, but not non-sense
Attached File(s)
File Type: mq4 MAXX_v150_LS10.mq4   177 KB | 625 downloads
  • #1,536
  • Nov 2, 2010 8:54am Nov 2, 2010 8:54am
Thanks Max. Will give this a go tomorrow morning.

  • #1,537
  • Nov 3, 2010 1:06am Nov 3, 2010 1:06am
Wow Max, what can I say. I have run a test, and it is insane! I haven't a clue what was happening, but the result was good. Attached is the backtest graph for 1 year on EU. It lost some towards the end, but it was fun to watch.
I have also tried GU which started well, but unfortunately blew the account. Perhaps I need more conservative lot sizing.

I shall have some fun with this.

Cheers, Beastie.
Attached Image (click to enlarge)
Click to Enlarge Name: Max v150.jpg Size: 59 KB
  • #1,538
  • Nov 3, 2010 3:58am Nov 3, 2010 3:58am
Quoting Beastie Boy
Wow Max, what can I say. I have run a test, and it is insane! I haven't a clue what was happening, but the result was good. Attached is the backtest graph for 1 year on EU. It lost some towards the end, but it was fun to watch.
I have also tried GU which started well, but unfortunately blew the account. Perhaps I need more conservative lot sizing.

I shall have some fun with this.

Cheers, Beastie.
Like I said , this one is a little more "martinguale" style, so more dangerous, and sometimes blew the account. We could add some feature to limit that.
I also strong beleive that master EA will often close all trades End of Day so will also limit the overtrading problem.
Anyway if you have some ideas let me know
  • #1,539
  • Nov 3, 2010 7:22am Nov 3, 2010 7:22am
Quoting maxou888
Like I said , this one is a little more "martinguale" style, so more dangerous, and sometimes blew the account. We could add some feature to limit that.
I also strong beleive that master EA will often close all trades End of Day so will also limit the overtrading problem.
Anyway if you have some ideas let me know
First of all a big thanks for your generosity in to want to participate in all the affiliate of the forum, the results of your job, and particularly of your EA!

As well you know, the mole of the EA is the big exposure that can jeopardize the good results precedents if not, at times, the real initial capital: I am confident however that a solution will be found!

Meanwhile, I would like to know however if the release 150 support the orders already opened by the preceding version 148; then I have a lot of curiosity for the "martingales style" and to this intention I have had the opportunity of verifying, I believe, the concrete application of this new method: this morning they have been closed 5 orders from 0.04 lots every of USDCAD; they were all orders open sell as it results from the enclosure statement.

I would like therefore to ask you some information in more on as this "martingale style" works and if they were had to use the parameters of the General Protections (particularly: UseVolatilyProtection; CheckCurrencyExposure; UseScoobsATRfilter that, of default, they are false), the drawdown could be limited in some measure.

Thank you and best regards.

Frankalde
Attached File(s)
File Type: doc Martingales_Style_3nov10.doc   36 KB | 372 downloads
  • #1,540
  • Nov 3, 2010 10:57am Nov 3, 2010 10:57am
BaestieBoy

Please do some BT with this one
you have 2 params for play :
TightenTrendDetection from 0 to 4, better use from 2 to 4
TightenTrendConditionForOpen 0 or 1
Also can choose Breakeven 0 or 1, TrailingStop 0 or 1
Attached File(s)
File Type: mq4 MAXX_v205.mq4   175 KB | 610 downloads
3 traders viewing now