ncandle and reversal filtered strategy

Viewing 15 posts - 1 through 15 (of 27 total)

1 2 →

Author

Posts
05/18/2017 at 11:27 AM #35968 quote
Francesco78
Participant

Master
Dear all

I thought it would be worth opening a discussion regarding the strategy I recently posted.

USDJPY 3candles and reversal strategy with ADX and VOL filter.

There has been some issues for the code for some of you which I think could be somehow related to different platform used by different people. I followd the suggestion by Nicolas and mad slighly change in the code that hopefully will make it working for everyone who is interested.

I have been looking for other currency pairs/other timeframes that can be suitable for the strategy and wanted to share some of the results.

In particular you find attached :
- CADJPY time frame 1hr
- USD CAD time frame 1hr
- USD JPY time frame 1hr
- USD JPY timeframe 15 mins.
Nicolas, shephinc and Juan Salas thanked this post

usdjpy_1hr.png usdjpy_15min.png usdcad.png cadjpy.png
05/18/2017 at 11:28 AM #35973 quote

Francesco78
Participant

Master

I also attach the codes

shephinc thanked this post

usd_jpy_ncandlereversion_15min.itf cadjpy_ncandlesreversion.itf USDCAD_ncandlesreversion.itf

06/11/2017 at 4:22 PM #38065 quote

noisette
Participant

Veteran

Hi Fancesco78,

I tried to backtest 2 USD:JPY strategies but no position is open.

Did you send the right codes?

Regards

06/11/2017 at 6:20 PM #38071 quote

Francesco78
Participant

Master

Hi Noisette ,this is a problem many have had, I think is because of the condition ATR min

I think you need to devide by 100 the atrmin because you might have a different quote or just multiply the atrmin by pointvalue.

Let me know how it goes

thanks

Francesco

06/11/2017 at 6:22 PM #38072 quote

Francesco78
Participant

Master

line 11

atrmin= 10*pointvalue // take poistion only if there is enough vol

same for the others

06/11/2017 at 9:46 PM #38084 quote

noisette
Participant

Veteran

Hi,

Still doesn’t work. I’ll spend some time to find the solution.

Regards

Francesco78 thanked this post

06/12/2017 at 11:14 PM #38163 quote

noisette
Participant

Veteran

Hi,

It’s very strange: It sometime works and sometime not!

But when it works, I don’t good result as your!

06/13/2017 at 11:46 AM #38214 quote

Francesco78
Participant

Master

Realy weird, do you mind to send me a screenshot of what you see?

Thank you

06/15/2017 at 10:47 PM #38394 quote

noisette
Participant

Veteran

Hi,

As you said, I think that there is a problem with ATR: ATR value is most of the time under 1 so it ATRMIN should be a value around 0,05.

And also the size of position is very important.

Regard.

06/26/2017 at 10:27 PM #39150 quote

DarioMazza
Participant

New

Doesn’t work… im so srry

06/29/2017 at 1:50 PM #39403 quote

Francesco78
Participant

Master

HI Dario, I’m sure we can make it work. Check carefully the code, check that the order of magnitude of the number are comparable among eachother

07/04/2017 at 1:02 PM #39781 quote

kg6450
Participant

Average

Hi @Francesco78, I’ve been looking at a few of your codes and they all work really well, I’m just a bit curious about your position sizing code.

I understand the logic behind it – open larger positions when volatility is low and vice versa but in practice don’t you find that the system is overall more profitable with a fixed position size?

For example I have been testing the DAXMNandBK strategies and I found that its far more profitable to just open at £5/point which gives a much, much greater profit and a lower drawdown. What do you think?

07/04/2017 at 1:25 PM #39786 quote

Francesco78
Participant

Master

Hi kg6450.

Thank you for spending your time on my codes.

Regarding your question, I think you are generally right, I was using that position sizing at the beginning of my experience on coding, I thought it would have made sense but in practice yes, when I put my system live I tend to use fixed sizes now. I changed personally my attitude, at the beginning I was looking for big profits and few strategies, so I was tempted to over optimize them and try to add something “special” on them, like for instance your position sizing. Now I am comforted with the idea that there are really a lot of strategies that you can use live so I tend to keep small and constant size but many strategies at once.

AlgoAlex thanked this post

07/04/2017 at 2:22 PM #39792 quote

kg6450
Participant

Average

I know what you mean, when you think about the idea of position sizing based on volatility it seems like a solid idea, but when it ran and opened at £10/point I found I would feel uncomfortable and end up closing out early.

More systems at smaller position sizes I think is definitely the safer method!

07/04/2017 at 2:31 PM #39793 quote

Francesco78
Participant

Master

It is not really about the size, you can scale it down even if size is linked to volatility.

I think is about trying to do logical things vs “artistic” thing. I have the impression a genral rule that might be useful when building an algorithm is:

try to come up with a logic idea and then twist it a little bit in an illogical way to make it work
Author

Posts