stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Trendy Gary
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=911
Page 14 of 16
Author:  dudest [ Fri Feb 08, 2013 11:44 am ]
Post subject:  Re: Trendy Gary

sponn wrote:
dudest wrote:@Cuz: for the fun of it, decided to try out TG one last time but without the SuperTrend indi (removed it from experts\indicators)

And guess what?....(as kinda expected), it now takes only buys (all JPYs currently) and is 100% profitable... (Half-jokingly) Wonder if there's a mod for TG to also only take sells, then we could run the two side-by-side across many pairs... :)

Cheers

Could you please share how are your results now after nice retracements on jpy pairs...
It's in drawdown (as expected )
Hence my comment about a 'sell-only' version being great

Cheers

https://www.myfxbook.com/portfolio/tren ... ndi/481912
Author:  garyfritz [ Sat Feb 09, 2013 2:37 am ]
Post subject:  Re: Trendy Gary

garyfritz wrote:
xfr8dog wrote:So what do you guys think about stacking the winners a la some kind of caterpillar input with a basket trailing stop?
That's a VERY interesting idea. Like you said, SuperTrend catches big long moves, but you get killed during sideways periods, and that tends to eat all the profit from the big moves. But if you increase the profits in the big moves by stacking... hmmm...
I tried a simple-minded test and it doesn't look that promising.

Here's a simple SuperTrend reversal system running on D1 EURUSD: when ST switches under the price, go long, when it switches over, go short:
ST1.gif
(EURUSD was the only pair that did that well, BTW.)

Then I added a simple stacking approach: every time the market moved X ATRs in my favor, I added another contract. Most settings of X made less money than the non-stacking approach. The ones that made more, did it with far more risk and volatility. For example, here's the result of stacking every 0.5*ATR:
ST2.gif
It made twice as much profit over the 5-year test, but it was almost entirely due to one monster trend back in 2010. Most of the time it was LOSING money. Except for that one huge trend, the simple non-stacking approach worked a lot better.
Author:  sponn [ Sat Feb 09, 2013 6:31 pm ]
Post subject:  Re: Trendy Gary

dudest wrote:
It's in drawdown (as expected )
Hence my comment about a 'sell-only' version being great

Cheers

https://www.myfxbook.com/portfolio/tren ... ndi/481912
Great.. but who will choose when to use buy only or sell only versio ? EA must trade both directions. In 2011 ther were some guys who was so smart - they thought that they can only buying gold...

very nice video to watch
http://www.youtube.com/watch?v=yNw2d-ilFXY
Author:  Wiggy1965 [ Sat Feb 09, 2013 11:25 pm ]
Post subject:  Re: Trendy Gary

sponn wrote:
dudest wrote:
It's in drawdown (as expected )
Hence my comment about a 'sell-only' version being great

Cheers

https://www.myfxbook.com/portfolio/tren ... ndi/481912
Great.. but who will choose when to use buy only or sell only versio ? EA must trade both directions. In 2011 ther were some guys who was so smart - they thought that they can only buying gold...

very nice video to watch
http://www.youtube.com/watch?v=yNw2d-ilFXY
Sponn,

You need to read the entire thread, as this has been discussed at length...however there is no real solution...!

Regards
Wiggy
Author:  mobthehop [ Sun Feb 10, 2013 2:19 am ]
Post subject:  Re: Trendy Gary

Out of curiosity - could those of you who only have long trades try the attached EA / indicator combo please.

Set most of the standard options in EA to false, removed "(1)" from indicator itself and amended iCustoms call in EA accordingly - did a quick logic test with Strategy Tester on 2013 data for 6 pairs and XAG and there were buy /sell orders for all....

For full list of changes, please use http://www.text-compare.com/ and copy in original version as well as my "modded" version

Cheers

Mop

PS: all quick tests ended with a loss for first 6 weeks of 2013.....
Author:  xfr8dog [ Mon Feb 11, 2013 2:44 am ]
Post subject:  Re: Trendy Gary

Then I added a simple stacking approach: every time the market moved X ATRs in my favor, I added another contract. Most settings of X made less money than the non-stacking approach. The ones that made more, did it with far more risk and volatility.
Thanks for testing that Gary.

Looking at the pure stacking approach that makes sense, because at the end of the runs, there is always going to be some trades that goes into negative, negating some or a lot of the wins from the early entries.

I'm still a little curious how it would work with a basket trailing stop for the long runs, and/or a limit on how many stacking trades to enter, but that adds a lot more complexity I would assume.

It probably doesn't make a bit of difference anyways...

Anders
Author:  garyfritz [ Mon Feb 11, 2013 2:49 am ]
Post subject:  Re: Trendy Gary

I tried trailing a stop, to minimize the damage at that "end of run going negative" phase, but it didn't help that much.
Author:  Mike [ Tue Feb 12, 2013 9:18 am ]
Post subject:  Re: Trendy Gary

Update
Since Jan 13th, when I started my demo with cuz-settings, I have only winning buy trades, my Equity is right now 20315 CHF starting with 16290 CHF, up 24%. Biggest DD 16%, Lotsize 0.1 except for gold 0.01, all trade comments "TG 0.1 1Hr true daiPiv"
Author:  AGT [ Tue Feb 12, 2013 8:56 pm ]
Post subject:  Re: Trendy Gary

Hi Mike ! To make it clear: could you attach your MFXB link ? ;)

Here is mine:
https://www.myfxbook.com/portfolio/tren ... s-4/442844


You may find it in my "EA Portfolio":

http://www.myfxbook.com/members/TGT_FX
or
http://www.stevehopwoodforex.com/phpBB3 ... =36&t=1374 :mrgreen:
Author:  Mike [ Tue Feb 12, 2013 11:07 pm ]
Post subject:  Re: Trendy Gary

I do not have any
All times are UTC Page 14 of 16