Multi-Instruments Synchronization Problem

[freewind]

Hi All,

I'm running a 5 min strategy that load many instruments and trade only one of them.
I use

CalculateOnBarClose = false;

and I only run on the first tick for BIP==0

protected override void OnBarUpdate()
        {
            //Calculate only once.
            if (!FirstTickOfBar || BarsInProgress != 0) return;
...

My problem is that I have different results when I'm running live and backtesting.
From my investigation I think that the problem is when I'm trying to refer to the other instruments previous 5 minute close Closes[index][1] I'm getting different results in backtest and live.
How can I refer to the previous close of for example BIP == 1 and getting the same result in live and backtesting?

TX in advance for any assistance.

[NinjaTrader_AdamP]

freewind,

There are known discrepancies in live trading vs. backtesting.

http://www.ninjatrader.com/support/helpGuides/nt7/index.html?discrepancies_real_time_vs_bac.htm

Here is a link to a way to mitigate some of the backtesting discrepancies : http://www.ninjatrader.com/support/f...ead.php?t=6652

Please let me know if I may assist further.

[freewind]

Hi Adam,

I'm aware of everything you mentioned but I found out that when I'm referring to Closes[index][1] in real time I'm getting identical tick and not the previous five minute close price!!!!

I can send you privately real time logs that demonstrate the problem.

Waiting for your reply.

[freewind]

NT Support,

When I'm using the setup I mentioned in my first entry I get the following results:
Close[0] ==> Current Tick
Close[1] ==> 1 Tick backward
Close[2] ==> 5 Min backward
Close[3] ==> 10 Min backward

If you need I can send you printouts of live connection.
BTW I'm using NT 7.007 connected to eSignal.

[NinjaTrader_AdamP]

freewind,

Close[] is a data series that won't work as you expect without BarsInProgress filtering. Its better to use Closes[][] in multi-instrument or multi-timeframe strategies.

Closes[0][X] is the primary data series, Closes[1][X] is the secondary, etc. I notice you are mentioning something about this in a prior post. How are you using this series?

[freewind]

I know that.
I just wrote it for the example I gave.
I think there is a bug in NT when addressing Closes[0][1] when the system is set to CalculateOnBarClose=false.

[NinjaTrader_AdamP]

freewind,

If you would like me to check into this further,

Please :

1) Outline specifically the issue you are experiencing,
2) List steps to reproduce the issue

and/or 3) provide me with code that reproduces the issue.

Unfortunately I am unable to replicate this with my own examples I have tried using information available in this thread.

I look forward to helping you resolve your issue.