Here’s the deal. For may forex robots doesn’t even matter because they’re not sensitive to the time of day the robot trades.

However, FAP Turbo is so GMT offset is important. Your question about which GMT offset to use when back testing, the one before the change over or after, is something most people don’t even think to ask.

If you run your back test over a period of time that extends over the daylight savings switch over dates then you’re going to get invalid results. For people who don’t consider this they simply look at the results and say “hey that sucks” without realizing the test isn’t valid to begin with.

There are two ways to handle this.

The more difficult way is to go into the historical data and shift the time stamps on the data over by an hour after the daylight savings switchover so that it matched up with the same data time period as before the switch over.

The easiest way for most people is to simply run multiple tests and combine the results.

It’s actually not difficult. here’s what you do. Start at the beginning of the time period you want to test and run until the daylight savings change date.

Then do the next test from that date until the next daylight savings change date. ect … etc …

The thing to remember is that when you finish the first test be sure to take your closing balance and use that as your starting balance for the next test.

In the end you don;t actually get the same nice report with only one graph, but the results are more valid which is the important thing. Also, to get the total wins/loses you simply need to add them up from each test report.

I know it’s a bit of a pain but that’s the world of trading. You’re always at the mercy of your data and if the data doesn’t correctly reflect reality then you have to mess with it to get real valid results.

hope that helps.

I’ll be going into it a bit more deeply in the guide.

cheers,
rob

When I want to test a strategy, should I use new GMT offset or the old one?
The test period includes before and after GMT offset change.

Thank you.