Renko Intraday Backtesting Problem

[starcd]

There is a major difference between intraday realtime and backtesting with renko bars.
In backtesting the last bar of the day does only reflect the close of the day.
So it could be that a trade that starts on close of the bar before will end as zero tick trade with no win or loss even if it has... See Screenshots.

Is there anything I can do to resolve that issue?

[NinjaTrader_Bertrand]

starcd, the formation of the bars historical vs realtime could be unfortunately different, especially for those tick based types - the most accurate Renko backtesting would be seen when running a Market Replay simulation for the day, as then the intrabar formation would be known. For your backtest, I think a time filter in the strategy could help as well, to eliminate the entries close to the end of the session.

[starcd]

Thank's Bertrand,
is there a way to set the time filter for entries in Renko at a fixed time e.g.190000? Because in Renko you don't know at what time the last correct bar will occure ...

[NinjaTrader_Bertrand]

Correct, you would not know as those types would not be time based - one thing you can do is adding a finer time based series, for example 1 minute one and then using those timestamps for your filtering via Times -

http://www.ninjatrader.com/support/helpGuides/nt7/times.htm?zoom_highlightsub=Times

[starcd]

Thank's Bertrand,

I looked at the example. But how can I compare a custom time e.g. 190000 with my second time based series?

int a=190000;

if (a < Times[1][0].Ticks) {
...
}

[NinjaTrader_Bertrand]

You don't have to go to the ticks level, if you want to compare to an int convert the Times return with ToTime - http://www.ninjatrader.com/support/h...ightsub=ToTime