I hope it will not shock experimental physicists too much if I say that we do not accept their observation unless they are confirmed by theory.
- Arthur Eddington
We should make this very simple. If there is no explanation or theory for the numbers in a systematic model, there should be no faith in their ability to be repeatable. Statistics without theory have no meaning. Investors should be able to get comfortable with back-tested data provided the work is done carefully and if the results can be described through a theory.
If you are a trend-follower, you should have an explanation for why markets trend. If there is an alpha strategy for hedge funds, you should be able to explain why the alpha is exists and whether it can be arbitraged away. You should be able to differentiate between skill, luck, structural advantage. We can go so far as to say that skill without a story should not be believed. Repeatable results are only those that can be explained with a theory.
No comments:
Post a Comment