Trades being eaten by NT7.

[Frederic]

Hello,
I use windows 8.1 X64 and use NT7 X64 bits, in demo, using CQG account. I get the problem that often my trades disappear- I explain: I open a trade in 6E, 6A, CL... and close for example 6A...on control center, generate... no 6A?? executions, I can see 6A... then wait a little... again generate... nothing.... close NT7, restart it, and nothing, still nothing, but 1 day later, it is to be seen in the report... generate from date to date.

this problem occur often, I close every times NT7, clean cache and so on, but still same problem.

today I had another problem. I open my first trade in 6E, and close it, and I got 3 executions.... 2 times executions BUY and 1 execution SELL.... 8 ticks winning,, and a profit from 2940USD ????

I check the log, and can not see any error. Again, i clean cache, reset simulator....

what can be problem for this crazy behaviour?

[NinjaTrader_Jason]

Hello Frederic,

Can you please send me all your log and trace files. Go to Control Center-->Help-->Mail to Support in NinjaTrader. Put 'Attn Jason, Trades being eaten by NT7.' in the subject field and enable 'Log and trace files'.

Could it be the orders are mismatched? NinjaTrader performance reporting is simply a set of calculations of fills that have a time stamp within the date range specified. The first step in the process is to filter the fills and then pair them to form trades. If positions are carried over night then surely you can have unexpected results.

Example:
Enter 1 June 4th
Exit 1 June 5
Enter 1 June 5
Exit 1 June 5

In the fill flow above, if you generated a report from June 4 through 6, you would see two trades reported (Entry June 4th and Exit June 5th) and (Entry June 5th and Exit June 5).

If you generated a report from June 5th through June 6th, you would only see one trade reported where the exit on June 5th is the trade entry and the entry on June 5th is the exit.

As you can see, the result would yield different results. NinjaTrader has no way of knowing if a fill is truly an entry or exit thus it has to build trade pairs starting with the first fill as entry in the date filtered dataset.