In the attached document are the details for 3 trades that would have only been successful if the sub time frame bars were processed in the reverse time order.
This bug has plagued me by indicating successful back test results which should not have been and vice-versa. Also, leading me down avenues of code which would not work in the real world.
If I am wrong about my findings, I'd like to get an explanation as to why the back-test process produces such anomalous results.
Check out the document here...

Comment