Historical issue with NG

[LightWeight]

Hello,

I'm having an issue with an automated NG strategy -- when I run it from Jan of 2009 to present it works fine in backtest, but if I set the start date back as far as Ninja goes, it then only shows trades for 2008, about 27 total. I have absolutely no idea why setting the start date as 2008 as opposed to 2009 would cause it to only test for 2008 and no further, any ideas?

Regards,

Andrew

[NinjaTrader_Bertrand]

Hi Andrew, are there any errors in the NT log tab when you over the 2009 start data range? How does the SampleMACrossover compare to that? Would it work normally? Do you work with any date time filter for the conditions in your script?

[LightWeight]

Hi Andrew, are there any errors in the NT log tab when you over the 2009 start data range? How does the SampleMACrossover compare to that? Would it work normally? Do you work with any date time filter for the conditions in your script?

There are no others in the log. I actually have other NG algos that run fine using the same dates, for some reason it's just this one (SampleMA returns other years fine as well). This algo doesn't have any day filters, but it does have a time filter for 4 PM to 7 PM PST (my other algos have similar filters).

I also noticed that if I put in 2009 till today it only shows trades for 2009-2012....so using a start date prior to 2009 will only give me 2008, and using 2009 for the start date only gives me 2009-2012. Can't figure it out at all.

[NinjaTrader_PatrickH]

Hello Andrew,

Thank you for your response.

Is this your own Strategy you developed or a third party?

Does the strategy utilize any logic to prevent trading after a certain amount of gain or loss (performance management)?

I look forward to your response.

[LightWeight]

Hello Andrew,

Thank you for your response.

Is this your own Strategy you developed or a third party?

Does the strategy utilize any logic to prevent trading after a certain amount of gain or loss (performance management)?

I look forward to your response.

This is a strategy I developed, and there is no logic regarding performance management.

Regards,

Andrew

[NinjaTrader_Bertrand]

Thanks Andrew, would then need to be isolated out via debugging. Try to go as simple as possible and comment advanced sections out and focus on getting for example the main conditions to trigger as expected, just printing each affected variable state should be helping understanding why it would not process as you would expect in those times.