Working Paper: CEPR ID: DP13835
Authors: Mark P. Taylor; Ilias Filippou
Abstract: We evaluate the cross-sectional predictive ability of a forward-looking monetary policy reaction function, or Taylor rule, in both statistical and economic terms. We find that investors require a premium for holding currency portfolios with high implied interest rates while currency portfolios with low implied rates offer negative currency excess returns. Our forward-looking Taylor rule signals are orthogonal to current nominal interest rates and disconnected from carry trade portfolios and other currency investment strategies. The profitability of the Taylor rule portfolio spread is mainly driven by inflation forecasts rather than the output gap and is robust to data snooping and a wide range of robustness checks.
Keywords: foreign exchange; currency risk premium; Taylor rules; data snooping bias
JEL Codes: F31; G11; G15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Taylor rule signals (C22) | currency excess returns (F31) |
| inflation forecasts (E31) | currency valuations (F31) |
| Taylor rule signals (C22) | future interest rate movements (E43) |
| future interest rate movements (E43) | currency valuations (F31) |
| output gaps (E23) | currency valuations (F31) |