I was running a strategy on GBP/USD. On 2009.05.11 02:07 the strategy submitted a trade with a SL=1.5211.0
The trade and SL were correctly registered.
The SL was reached 1st time at 2009.05.11 03:53:30.

However the SL was not triggered and executed by the JForex-engine until more than 3 hours later at 2009.05.11 07:06:58.045.

When debugging this behaviour I found out that the tick data was missing and thus no ticks were processed by the tester-engine between Mon 2009.05.11 03:13:54.670 and Mon 2009.05.11 07:06:58.045

The chart shows bars on all periods except the ticks for the missing 3 hours.

I accidentally stumbled across this gap as the SL of the trade was not executed. I now question, if this happens more often unrecognized. This could of course lead to false trading signals and results in tester.

The question now is why
1) JForex didn't download the data and
2) didn't detect the data gap

JForex could detect inconsistencies in the data by checking bar-data against the last processed tick and take measure (reload data), should there be an inconsistency (i.e. time of last tick processed is before start time of bar, but bar close price not identical with tick price).

The following logging shows the tick-gap:


Best RR.

I would expect, that the close prices (ask, bid) of a bar that gets formed are equal to the prices of the last processed tick.

I've done a run on GBP/USD for the period 5.1.2009 to 24.11.2009 and did check in every onBar call, if the ask and bid price of the last tick processed in onTick is equal to the close prices of the bar. The resulting log has nearly 22.000 entries with differences between the close data of the bars and the tick-data, showing that the tick and bar-data are quite often not in sync!

The log is attached to this posting.

Could you please shed some light on this?

Thanks in advance, RR.