Nanningbob V5: Steve Hopwood's auto-trading robot | Page 12 | 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

  • #221
  • Aug 20, 2010 10:56pm Aug 20, 2010 10:56pm
I have been trading enough years to know that some trading systems work well for a while and then hit the skids, but the way NB's V5 system has started is pretty encouraging. In 5 weeks of trading this system (manually for the first couple of weeks then mainly relying on this EA) my account size has increased about 60% with no trades outstanding! And that is me being conservative with lot sizes and take profits! Now that I am getting more comfortable with letting the profits run a little after locking in breakeven, there is even more upside.

It's real early days, but if this current rate of return continued the account size would double every 9 weeks or so! Imagine if that happened... (dreaming at this point)...

I know I've said it before, but thanks again Steve and Simon for putting together an amazing EA, and huge thanks to NanningBob.
  • #222
  • Aug 21, 2010 12:49am Aug 21, 2010 12:49am
Quoting Invisible
@ Zurathustra

Thanks for coding this up.

One quick initial question: I cannot see the numbes 84 or 16 in the code.

Invisible
Ehh, for my typing ease i used 80 and 20, which I have since changed to a extern input so I can adjust them on the fly. I've done some more work on it since I posted the code earlier today. Like the code warns, it was very early theory testing.
  • #223
  • Aug 21, 2010 5:24am Aug 21, 2010 5:24am
Quoting f451
All,

This is the oft mentioned, somewhat promised modification to NB v5 which takes trades with the trend before stopping, reversing and taking the trade that NB's v5.3 system would normally take....
Boy oh boy, these are exciting times.

I have posted this robot into post 1 and copied the contents of this post into a pdf. This way, it does not get lost as the pages grow in this thread. When you release new versions, I shall replace the post 1 version.

The NB robot has proven a huge success on demo this week, so I shall restart on Sunday night and run a demo of NB Trend alongside it.

Cheers. This was great to get home to.

  • #224
  • Aug 21, 2010 5:26am Aug 21, 2010 5:26am
Quote
//| WARNING!!! THIS EA IS IN EXTREMELY EARLY DEVOLOPEMENT STAGES! |
//| THIS EA WILL BLOW ALL THE MONEY IN YOUR ACCOUNT LEAVING YOU |
//| BROKE, DIVORCED, AND HOMELESS. IT WILL THEN TRACK YOU DOWN AND |
//| KILL YOUR PETS AND ANALLY RAPE YOU! DO NOT USE IT ON A LIVE |
//| ACCOUNT UNDER ANY CIRCUMSTANCES!! YOU HAVE BEEN WARNED!! |
//
  • #225
  • Aug 21, 2010 5:56am Aug 21, 2010 5:56am
Quoting f451
All,

This is the oft mentioned, somewhat promised modification to NB v5 which takes trades with the trend before stopping, reversing and taking the trade that NB's v5.3 system would normally take. The philosophy here is that while NB's system is outstanding, its one weakness is the large draw downs which can occur during sustained trends. Bob uses the 2.4.2 recovery system to address this problem - this EA attempts to turn that weakness into a strength by trading with the trend (instead of against it) before taking the standard NB reversal trade...
I have tried backtesting this version last night and this morning however it loads the bollinger bands and then nothing else. I am using the default settings but no trades

Paul
  • #226
  • Aug 21, 2010 9:47am Aug 21, 2010 9:47am
Quoting paulbt38
I have tried backtesting this version last night and this morning however it loads the bollinger bands and then nothing else. I am using the default settings but no trades

Paul
Paul

Not sure what the issue is for you. Works fine in my environment. You will need a copy of the "Bollinger Bands %b" indicator posted in this thread... this is the only custom indicator.
BbSeparation = iCustom(NULL, 0, "Bollinger Bands %b", 25, 2, 0, 0, 0);

cheers
f451
  • #227
  • Aug 21, 2010 9:53am Aug 21, 2010 9:53am
Quoting SteveHopwood
Boy oh boy, these are exciting times.

I have posted this robot into post 1 and copied the contents of this post into a pdf. This way, it does not get lost as the pages grow in this thread. When you release new versions, I shall replace the post 1 version.

The NB robot has proven a huge success on demo this week, so I shall restart on Sunday night and run a demo of NB Trend alongside it.

Cheers. This was great to get home to.

Thought you'd enjoy a little welcome back present.

cheers
Simon
  • #228
  • Aug 21, 2010 10:15am Aug 21, 2010 10:15am
I would like to bring your attention to a very important point, about the reentry system.
Like every martinguale system, there is a day where the account suffers big losses.
Macman in post #29 described this very well (look the 4 charts).
I have coded in the past many martinguale/grid systems, with very sophisticated entry points, that eat many pips during many years but one day....Sometime backtesting for 10 years with 1 pair never showed catastrophic event so the result was wonderfull. But with another pair I got 1 day (sometimes only 1 time in 10 years!) which kill my account on a strong trend.
It's not because a pair didn't have any problem on backtest during 10 years that we are safe from get troubles (margin call) one day or another.
So, recalling Macman post #29, fine tunes the params for each pair is a wrong safety sentiment. It's only luck. I have done this enough with anothers systems (very smart reentry systems) that I know what I am speaking about.
1rst picture is a backtest GbpUsd from 01/01/2007 until now.
Everything is defaut except my code modif to calculate some params from daily ATR, I also set emergencySL=500.

There is 2 important points on this picture :
A : Not lucky, the 3th level rentry point failed to save the bad trades
B : Lucky, the 3th level rentry point succeded to save the bad trades.

Note the C point is also very lucky, strong reversal with 3th level rentry trade.
So I think in order to have a clear view about the NanningBob system profit, we need to disable the recovery system, because every martinguale systems show outstanding results (and often fakely save bad strategies) during long time but failed one day or another.
I will conduct some tests without recovery system.
Attached Image (click to enlarge)
Click to Enlarge Name: Image 8.jpg Size: 171 KB
  • #229
  • Aug 21, 2010 10:22am Aug 21, 2010 10:22am
Quoting f451
Paul

Not sure what the issue is for you. Works fine in my environment. You will need a copy of the "Bollinger Bands %b" indicator posted in this thread... this is the only custom indicator.
BbSeparation = iCustom(NULL, 0, "Bollinger Bands %b", 25, 2, 0, 0, 0);

cheers
f451
Thanks. I have all the indicators loaded as I have been backtesting the original ea. I have even tried this on 2 different machines with no joy and different currency pairs. I will check my backtest data as there may be a problem in it.

Its bizarre as with the main ea when I backtest it on CHFJPY for instance since 01-2009 it works great and uses full 242 recovery and then continues to work. Yet if I go further back to say beginning of 2007 and backtest it stops working after hitting level 3 recovery.

Its all got me baffled (although that doesn't always take much)

Paul
  • #230
  • Aug 21, 2010 10:28am Aug 21, 2010 10:28am
Quoting maxou888
I would like to bring your attention to a very important point, about the reentry system.
Like every martinguale system, there is a day where the account suffers big losses.
Macman in post #29 described this very well (look the 4 charts).
I have coded in the past many martinguale/grid systems, with very sophisticated entry points, that eat many pips during many years but one day....Sometime backtesting for 10 years with 1 pair never showed catastrophic event so the result was wonderfull. But with another pair I got 1 day (sometimes only 1 time in...
Thanks Max, but 2.4.2 is not Martinguale; nor does it allow leaving umpteen trades open with gradually increasing lot sizes. There are very good explanations of the 2.4.2 in Bob's post 1 zips.

This robot has a strictly limited 2.4.2 recovery function, as described in the user guide.

  • #231
  • Aug 21, 2010 10:41am Aug 21, 2010 10:41am
Yes I know (I read the explanation), the limit is level 3, but the problem is the same, sometimes we have to take the loss and this loss can be huge.

Here is a test for GbpUsd without recovery system and hard emergency SL =500 pips.
The good news is that the system is profitable (+11% in 3 years)
Working on fine tune the SL will be smart.
I will try with SL is a multiplier of DailyATR, but may be someone could try others SL system ( BB, or SAR points or other idea)
Attached Image (click to enlarge)
Click to Enlarge Name: Image 9.jpg Size: 164 KB
  • #232
  • Aug 21, 2010 10:55am Aug 21, 2010 10:55am
Quoting paulbt38
Thanks....
Hi Paul,
I'm having the same problem, no trades during backtesting. Am just working my way through the EA to see if I can find out why it's not placing any trades.

PeeJay
  • #233
  • Aug 21, 2010 11:20am Aug 21, 2010 11:20am
I haven't managed a trade in backtesting either - Alpari UK.

I cannot see anything in the code to prevent trading, so maybe it is just a tester thingy.

  • #234
  • Aug 21, 2010 3:41pm Aug 21, 2010 3:41pm
Quoting SteveHopwood
I haven't managed a trade in backtesting either - Alpari UK.

I cannot see anything in the code to prevent trading, so maybe it is just a tester thingy.

The same here no trades taken on GU between 01.02.2010 and 30.07.2010. Do any of the indicators need to be placed on the chart in order for EA to work properly?
  • #235
  • Aug 21, 2010 4:27pm Aug 21, 2010 4:27pm
Quoting maxou888
Yes I know (I read the explanation), the limit is level 3, but the problem is the same, sometimes we have to take the loss and this loss can be huge.

Here is a test for GbpUsd without recovery system and hard emergency SL =500 pips.
The good news is that the system is profitable (+11% in 3 years)
Working on fine tune the SL will be smart.
I will try with SL is a multiplier of DailyATR, but may be someone could try others SL system ( BB, or SAR points or other idea)

I worked on another Nanningbob Sixth sytle EA in another thread where we used Martingale Style money management, its its what eventually made me stop working on it. This is really one of the reasons why I recommend using W%R as the trigger to exit trades. Since the core principle behind the Sixth Screen indicator is to enter a trade and collect some pips on the retrace, it makes more sense to me at least to quickly get out of trades right when they start to not perform as desired, rather then continue to pile more lots onto them. (I should say "actively" working on it, because of course I continued to pursue the concept and how to make it better, which is what led me to the W%R discovery)

By watching the correlation between the W%R and the Sixth Screen indicator, you can see that as long as W%R is above/below the -16 -84 levels, then the trend hasn't finished setting its new high or low, and therefore is not a good time to enter. Yeah, you will exit a LOT of trades at 1-5 pips of loss, but but keeps you from taking those big hits while still maintaining the profits potential of the Sixth Screen system.
  • #236
  • Aug 21, 2010 4:40pm Aug 21, 2010 4:40pm
Quoting krunofx
The same here no trades taken on GU between 01.02.2010 and 30.07.2010. Do any of the indicators need to be placed on the chart in order for EA to work properly?
No, they do not.

To explain, ea's cannot read things on the screen. The extract values by sending calls to the indi the displays the data/graphic on the screen and receiving replies that equate to the screen display.

  • #237
  • Aug 21, 2010 4:59pm Aug 21, 2010 4:59pm
Quoting Zurathustra
I worked on another Nanningbob Sixth sytle EA in another thread where we used Martingale Style money management, its its what eventually made me stop working on it....
I popped a W%R onto an eu chart. Ok, so this is hardly an exhaustive investigation, but this is what I saw: what you say is true, but by the time there is a trade trigger for V5, W%R is also showing the trade is 'correct' i.e. the trend has exhausted itself, for the time being at least.

Therefore this is merely adding an unnecessary indi that risks complicating what is essentially a simple confirmation system of these facts: prices go up; then they go down; then they go up; then they go down.................

  • #238
  • Aug 21, 2010 5:14pm Aug 21, 2010 5:14pm
Same problem for me. No indication of any problems in the strategy tester, but no trades either. Tried changing settings, but never got one that worked....
  • #239
  • Aug 21, 2010 10:14pm Aug 21, 2010 10:14pm
Quoting Invisible
Same problem for me. No indication of any problems in the strategy tester, but no trades either. Tried changing settings, but never got one that worked....
Mmmm. I just ran the backtester on the pairs that the EA has taken trades on over the last 5 days (EURCHF, GBPCAD, EURCAD), and nothing. No trades. It works on forward demo, just not in the backtester.

Bother. Not sure why. Certainly nothing obvious. Closed market?

Simon
  • #240
  • Edited 11:32pm Aug 21, 2010 10:57pm | Edited 11:32pm
Not sure but

"bool DoesTradeExist(int magicnumber)" there are no function or inputs associated with it, should it have "extern int magicnumber =0;"
0 traders viewing now