When running a back test this morning I noticed that a trade appears on some back-tests and not others. I have attached two graphics to highlight the differences. The strategy is live and the missing trade did occur in the live system. I ran the back-test multiple times and the results flipped between the two outcomes on an apparently random basis. This is not helpful. This is not a new strategy it has been live for several months. Also, in the graphics, I have graphed the flag which indicates when the market is open to trade (given that the Day and DayOfWeek parameters in PRT are of no use in Australia). You will note that when the trade does not occur that the flag indicates that the market closes at 8:10 AM Aus time rather than 6:00 AM Aus time which is on the other graphic and is correct. Is there something I can do to correct this? As is it it renders back-testing unreliable.