EA performs differently between brokers -- how is that possible if same historical data

 
I have demo accounts for 3 brokers on which I am testing an EA. The EA performs marvelously on FXDD but goes the other way with Interbank and StrategyBuilder. The all use the same imported data.

How is this possible if the data is the same? I get 90% modelling quality on all of them.

Rob
 
Check the symbol settings at the tester. When generating fxt files, the header takes data from the server, including spread. If spread was different for the different brokers, when you compiled your fxt file, the whole testing period will have different spread.

Otherwise write a simple EA to print account data (eg. spread, swap, margin calculation) and tick data (Bid, Ask, Time in seconds) to a file run it on all MT4 testers, and you will see what the difference is between the different fxt files.
Reason: