| This paper summarizes key lessons learned from using models from microstructure finance to explain and forecast exchange rates. The first section is an executive summary, which outlines seven lessons that pertain to how different transaction-flow measures (e.g., interbank flows versus end-user flows) perform in explaining concurrent returns and forecasting future returns. Section 2 addresses three overarching topics, including: (1) how various transaction-flow measures differ, (2) causality between transaction flows and returns and how to think about it, and (3) strategies for pinning down underlying flow drivers. Section 3 addresses empirical results underlying the seven lessons in section 1. |
Abstract
Full text |
