This has been a problem since I started using Zorro, and now that some more months have past I thought I would ask about it. I just downloaded historical data so that I could run a Z12 and Z3 test and see how it should have performed in September and October 2013.

The end date for the Z3 and Z12 test is 26/07/2013. The Z3 chart ends at this date but the Z12 chart ends at about December 2012. The test trade log file for Z12 is 28/12/2012, confirming the chart is correct. Z3's last trade is the indicate test end date in July 2013.

Am I missing something? Do you want me to post anything?

Your observation is correct. The Z systems are WFO optimized. That means the simulation period is divided into equal cycles. The test must end with the last cycle, which was in July. If the test period were variable, all cycles would become longer and thus would not fit to the parameter sets anymore.

I am aware of this. What I want to do is get the test results with the current settings for September and October so that I can compare that to my live account. I fear that when 1.20 is released, a retrain may change the settings or the end date will be the same (26/07.) Is there a way to do a test with the current settings up to the current date?

No, the Z systems store the end date of the last optimization and use it for the simulation end date. When 1.20 is released, it will contain new parameters sets and thus a new end date.

FYI - the bit above where I say the last trades in the file are from December 2012, well for some reason that occurs in the Zorro instance that I ran the data Download script on. It does not occur in the Zorro instance which I reserve for Zs testing.

jcl, are you saying that since the dates are hard-coded the Z systems cannot be reTrained with the latest price data? ReTraining would, however, use updated AssetsFix.dta and .ini data?

Thanks.