Ninjatarder Backtest of a Strategy

[⭐ laser1000it]

It seems the Backtest in Ninjatrader 7 are all wrong. I've tested some strategy and I've seen the final resut is big different from reality.

 

I'll explain it better: Monday I've launched in RealMode a strategy and stopped at Friday. I got a final result. Today (Saturday) I start a backfill and the result is big different than the Real Mode.

 

Where am I wrong? it seems impossible this mistakes for a professional platform as NT. Have you also seen these differences?

 

PLS share your experiences regarding this issue.

 

 

NB: in Real Mode I 've used timeframe 5 min and/or Renko

[⭐ nadjib]

you can use https://strategyquant.com much batter

[⭐ laser1000it]

you can use https://strategyquant.com much batter

 

Thanks but I need testing some NT strategy. It does not seem to me that SQ can help me.

[⭐ JDizzle22]

You should load tick data in the strategy and then place the orders on the tick data instead of the main time frame

[orfila]

Was the result you got after the backfill better than real mode?

 

a few ideas :

are you basing the strategy on a higher time frame and/or larger renko bars?

are you using the open for entries?

Is your system redrawing?

[⭐ laser1000it]

Was the result you got after the backfill better than real mode?

 

a few ideas :

are you basing the strategy on a higher time frame and/or larger renko bars?

are you using the open for entries?

Is your system redrawing?

 

Many thanks Orfiila for your replay,

 

I have take this strategy: https://www.sendspace.com/file/hkonpo

 

Whit this CS I've got in this week a quite positive return. But how you can see the backfill show a result big different

 

tested on FGBL 09-18 Renko 9 ………..defaut strategy parameter except: V1=25 ; Min Bar Required = 2

 

 

 

 

 

I am aware this strategy will most likely be a loser after a severe test ... but I do not understand the difference between backfill and realmode.

[lbf4223]

...................... Edited June 26, 2018 by lbf4223

[⭐ JDizzle22]

You didn't mention how much slippage you're giving your trade entry signals in backtesting. If it's not to the level of what you observed in real-time, that's most likely the reason for the difference.

 

That won't give you better results with your current issue--you need to load a secondary data series, 1 tick data, and then when you use your enter (ex. EnterLongLimit), the BarsInProgress needs to be set to the 1 tick data series. This will give you pretty close to realistic.

[lbf4223]

............... Edited June 26, 2018 by lbf4223

[⭐ JDizzle22]

Not in a backtest. You are completely wrong.

[⭐ laser1000it]

Thanks for your replay...in anycase I have understood nothing. PLS clarify with screenshot example as I can proceed.

So PLS give me an example of s*tupid proof that I will copy. At the end next Friday night I can launch the backtest which must be as similar as possible to realmode

[⭐ laser1000it]

This is Monday 25 June 2018 real mode result

 

This is what I see when I today open the Platform.

 

 

The different is BIG.

 

At this point after all the above issue I am still.

[mike83]

which indi u backtesting?

[⭐ laser1000it]

which indi u backtesting?

 

See post #6 (strategy not ind)

 

Update: related to post #12 I referred to what I sse in the chart the day after ...I have not talked about anybody backtest

 

I don't know how you cantrade with Ninjatrader but from what I find out it's a platform to throw in the trash

Edited June 26, 2018 by laser1000it

[lbf4223]

........... Edited June 26, 2018 by lbf4223
duplicate post

[lbf4223]

laser1000it, this is the example you want to look at from the NT website.

 

https://ninjatrader.com/support/forum/showthread.php?t=6652