nonetheless – Your image has just highlighted something to me about the way PRT presents the results of its walk forward tests. Your image shows a WF line that is at an almost identical angle to that of the in sample tests and yet the WF efficiency is only 54.85%. I had always assumed that if a OOS test was as good as a IS test then it would be 100% efficiency. If it was twice as good it would be 200% efficiency and if it was half as good it would be 50% efficiency but it appears that I have been thinking incorrectly (not for the first time and I very much expect not for the last time!)

What we really need now is a 200K back test with some decent OOS.