The Unsexy Side of Machine Learning

A lot of this stuff we were doing anyway, we didn't even know it was machine learning technically. We've got a toy example of a managed future strategy that has a very low correlation to the managed futures index by design. So can I interrupt and ask a question? Most of these big firms are doing some form of trend following, the very obvious opposite would be some form of a mean reversion. Right. Have you found with that that it did fall into that simple framework? Are you still able to do a decorrelated trend process?