Flow Rider: Tradable Ecosystems' Relative Entropy of Flows As a Determinant of Relative Value

While the predictability of fund flows based on prior-period returns is empirically well- established in the literature, we extend its insights to tradeable macro ecosystems, where there is anecdotal evidence of distribution divergences between flows and prior-period returns. We study the relative entropy of flows to capture an exogenous price signal, which we hypothesise to represent the aggregate directional positioning of informed investors. We construct and assess factor portfolios and a spectra of unconstrained systematic long- only portfolios to test our hypothesis, in pursuit of demonstrating informativeness and the additionality of the factor in augmenting traditional weighting schemes to produce superior systematic portfolios and in adding to the discretionary investor's toolbox a method for identifying the positioning of supposed "smart money" in the aggregate. Our findings indicate that the relative entropy signal demonstrates relevance in predicting future returns and identifying relative value within our designated tradeable macro ecosystem. Within certain portfolio configurations, the signal can contribute meaningfully to portfolio construction outcomes, based on historical evidence.