NanningBob V1 Trend auto trading robot by Steve Hopwood | Page 13 | 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 V1 Trend auto trading robot by Steve Hopwood
Exit Attachments

NanningBob V1 Trend auto trading robot by Steve Hopwood

  • #241
  • Jan 14, 2011 4:38am Jan 14, 2011 4:38am
Steve, I've noticed that quite a few trades run well past the TMA outer take profit lines, would you consider adding a facility to close a user specified part of the trade at the lines allowing the balance to run on for possible extra profits until stopped out.
  • #242
  • Jan 14, 2011 6:27am Jan 14, 2011 6:27am
Quoting SteveHopwood
Just to let everybody know, I have looked at adding Recovery into this robot.

It cannot be done. Everything I try causes problems. Solving them throws up different problems in an endless round of problem chasing.


Fair enough. Thanks for trying. Is it possible to turn the stop loss off then? I have put my stop at 99999 but it stops out in strategy test. We'll have the option to do a manual recovery then.
  • #243
  • Jan 14, 2011 5:22pm Jan 14, 2011 5:22pm
Quoting plateman
Steve, I've noticed that quite a few trades run well past the TMA outer take profit lines, would you consider adding a facility to close a user specified part of the trade at the lines allowing the balance to run on for possible extra profits until stopped out.
No. My experience of partial-closure-then-allow-to-run is merely that the partial closure happens, then the price retraces to be leaving lots of would-have-been-profits in the crims pockets. Then those trades that hit sl leave the profits from the partial closures looking silly.

Bob's strategy isn't really a trend trading strategy. It is scalping in the direction of the trend as ascertained by certain measurements. True trend trading needs massive stops and trades that run for days, weeks, sometimes even months and rides massive retraces. Partial closure may well have a role to play in a true trend trade, but not here.

  • #244
  • Jan 14, 2011 5:30pm Jan 14, 2011 5:30pm
Quoting Viv108
Fair enough. Thanks for trying. Is it possible to turn the stop loss off then? I have put my stop at 99999 but it stops out in strategy test. We'll have the option to do a manual recovery then.
Set VisualStopLoss to 0

Then find this line of code:
bool HideStopLossEnabled=true;//Removed the extern so that the user has no choice but I do not have to edit umpteen lines of code

and set it to false and recompile, or make it an extern so you can play around.

  • #245
  • Edited Jan 15, 2011 3:48am Jan 14, 2011 5:33pm | Edited Jan 15, 2011 3:48am
The bot had a much better trading day today.

Excellent.

  • #246
  • Jan 14, 2011 7:24pm Jan 14, 2011 7:24pm
My demo seems to be using the visual SL as the stop loss instead of the H/L plus 15 pips I have set. Anyone else experiencing this?
  • #247
  • Jan 14, 2011 7:55pm Jan 14, 2011 7:55pm
Quoting SteveHopwood
No. My experience of partial-closure-then-allow-to-run is merely that the partial closure happens, then the price retraces to be leaving lots of would-have-been-profits in the crims pockets. Then those trades that hit sl leave the profits from the partial closures looking silly.

Bob's strategy isn't really a trend trading strategy. It is scalping in the direction of the trend as ascertained by certain measurements. True trend trading needs massive stops and trades that run for days, weeks, sometimes even months and rides massive retraces. Partial...
Yeah I think your right we are making daily trades in the direction of the trend. If we want to ride out the retraces you have to use a higher time frame. I am looking at a 4H and daily charts but still thinking. I am about ready to do the same as I did with the 4H system. When you see profits take it. There seem to be plenty of trades every day to grab.
  • #248
  • Jan 15, 2011 1:16am Jan 15, 2011 1:16am
Steve, is there a simple way to add some logic that will only allow one trade per bar?
If you are trying to use a smaller stoploss, like half the swing, I am seeing it take a trade, get stopped out, then turn around and take another trade, get stopped out, etc. Limiting it to once per bar I think would help with some of this whipsaw.
Let me know?
Thanks
  • #249
  • Jan 15, 2011 4:14am Jan 15, 2011 4:14am
I just had to share this one with you guys. The 'commercial' ea rip-off merchants understand the principle, for sure.

Attached File(s)
File Type: zip Beauty-of-Mathematics.zip   641 KB | 11,855 downloads
  • #250
  • Jan 15, 2011 4:33am Jan 15, 2011 4:33am
Nice find Steve, a modern version of the old saying "Bullsh*t beats brains"
  • #251
  • Jan 15, 2011 6:09am Jan 15, 2011 6:09am
I have done some work on the bot this morning, to help us know where our hidden stops are:

 

  1. added a red, dashed hidden stop loss line to the chart
  2. displayed the hidden stop loss next to the open trade message
  3. coded the bot to adjust the hidden stop if we move the line
  4. included all of the above in the breakeven, jumping and trailing stop functions.

Be warned; 4) entailed a lot of copy/pasting/altering. Be afraid. Be very afraid.

You can watch all this at work in a visual backtest if you have the patience. It is worth turning off all the TMA reporting stuff if you do.

taz, I have not done anything about the next bar thingy. I do not see how your scenario can occur, so I want to wait until it happens on my demo before I do anything.

  • #252
  • Jan 15, 2011 7:33am Jan 15, 2011 7:33am
Steve,

This is the best compilation of different known truths.

You are brilliant both in thought and....u know what. However, I will advice that u only give 100% and not more pls.....lol



Quoting SteveHopwood
I just had to share this one with you guys. The 'commercial' ea rip-off merchants understand the principle, for sure.

  • #253
  • Jan 15, 2011 9:35am Jan 15, 2011 9:35am
Hi Steve

Just ran a strategy test in visual mode with default settings.. as you can see, there are 3 occasions when shorts were entered (look at the 3 crosses)... but closed early (this happens regardless of the playing around with the stop loss distance entered).

If we were not stopped out so early, there were some serious pips to be made.
Attached Image (click to enlarge)
Click to Enlarge Name: Steve's work.jpg Size: 144 KB
  • #254
  • Jan 15, 2011 11:57am Jan 15, 2011 11:57am
I have been running some visual backtests this afternoon.

These showed up some faults in the programming; these are fixed in post 1.

I am still not sure the hidden trailing/be/jumping stop functions are correct. I shall see next week when the bot works with open trades.

I have added a straight take profit input.

I have added cancel the setup code when the 5MA wanders the wrong side of the middle TMA line - visual backtesting showed this to be necessary.

No more for now. We have to let the bot do its job next week to see what else needs fixing.

  • #255
  • Jan 15, 2011 12:19pm Jan 15, 2011 12:19pm
Quoting SteveHopwood
taz, I have not done anything about the next bar thingy. I do not see how your scenario can occur, so I want to wait until it happens on my demo before I do anything.

Hi, Steve. I had the same thing occur once on Friday in the EURGBP during the 1200 GMT hour. A Sell trigger occurred at 12:02 @ .8426, price moved to the 15 pip BE, SL moved to 3, price reversed and stopped out at 12:55 @ .8423 and then the bot immediately opened a new trade @ .8420 because the trigger conditions were still valid. I get lost in the code; at what point does the GV reset? And what happens if price is above the high TMA and below the 5x5MA? Will a trade always be triggered? And if price moves 15 pips in profit, sets BE to 3 and price reverses, will it always re-enter if still below the 5x5MA? And is that a good thing or bad?

Side note: a similar thing happened on AUDUSD on Thursday but did not re-enter a trade immediately because price was still below the high TMA. Once price moved above and then started down again, a new trade was entered as it should. And then I got stopped at the 3 pip BE just before a massive 100+ pip down drop! Yeesh.

P.S. Thanks for all the work!
  • #256
  • Jan 16, 2011 8:54am Jan 16, 2011 8:54am
Hey steve.. i just ran another back test with your latest update and default settings.

Something is not right....even when i play around with the inputs. I get the similar errors as shown my message a posts above (with the 3 'x's).

What if we reversed buys and sells LOL
Attached Image (click to enlarge)
Click to Enlarge Name: backtest.jpg Size: 152 KB
  • #257
  • Jan 16, 2011 2:01pm Jan 16, 2011 2:01pm
Ran the robot this whole week, while it wasn't great and I did loose cash on the demo account I did find some interesting tweaks to ti and I am working with the stop loss and take profits to make sure a positive trade doesn't turn negative.

I am hoping also since this is still early in the year and while I have only been trading for around 4 years I never trade this early in January. So the markets are a bit crazy and not trading the way they should so we will see how this week goes. I attached last weeks report to show the trades and everyone knows what the EA did on ym IBFX demo account.

Thanks cant wait for this week to start I wont start until Tuesday due to MLK day in the United States.
Attached Image (click to enlarge)
Click to Enlarge Name: January 10-14 Report.png Size: 38 KB
  • #258
  • Jan 17, 2011 7:20am Jan 17, 2011 7:20am
i can tell you with 100% that if the bot traded as it should have been last week. on the 15 minute chart like peter porter did,we had made around 100 pips a day.

problem is people run backtests etc. but how many do actually check if those trades that are done, are right following nanningsbob rule set?

you see last week i had 68 trades. some trades were not opened while they should, some trades that went bad should not have opened at all.

now steve has updated his Ea, so i am now busy testing to see if those bugs still appear in the EA. if they still appear i will add my excel sheet in here so steve or any other coder can simply check which trade was bad or should not have been made. that way it can be fixed in code.

i believe nanningsbob idea is the awesome and if the bot does the trades as it should, it will will be profitable.

and if you backtest fine, but at least check if the bot does the trade following the ruleset of nanningbob.

thank you
  • #259
  • Edited 8:14am Jan 17, 2011 8:03am | Edited 8:14am
Quoting trade4fun
i can tell you with 100% that if the bot traded as it should have been last week. on the 15 minute chart like peter porter did,we had made around 100 pips a day.

problem is people run backtests etc. but how many do actually check if those trades that are done, are right following nanningsbob rule set?

you see last week i had 68 trades. some trades were not opened while they should, some trades that went bad should not have opened at all.

now steve has updated his Ea, so i am now busy testing to see if those bugs still appear in the EA....
I am looking forward to your testing of Steves EA
I am testing another method of the TMA in conjunction with pending trades , overall trend and if I come up with anything good I will report back .
As I said before, I see no harm in trying different strains of the strategy and in a way it stops all of us testing the same thing
  • #260
  • Jan 17, 2011 8:47am Jan 17, 2011 8:47am
Quoting trade4fun
problem is people run backtests etc. But how many do actually check if those trades that are done, are right following nanningsbob rule set?
Hi buddy - 'people' may run backtests blindly - but i ALWAYS run them in visual mode. If you look at some of my posts in the past, you will see how I have commented on how some trades were taken whilst others were not. Some trades were closed out too early and had they stayed open, we would have made some money.

Its all part and parcel of the evolution of the EA. The system clearly works, the EA needs some work.

Regards
0 traders viewing now