Currencies that are more exposed to US monetary policy yield positive average excess returns. This result holds both for pure monetary policy shocks and for central bank information shocks, identified via sign restrictions on interest rate surprises using high-frequency data. Currency characteristics help explain the heterogeneity of these exposures across currencies and time. We then build exposure indices to gauge this effect around policy announcements. Long-short trading strategies that condition on such exposure indices display significant excess returns after controlling for dollar, carry and momentum factors.

We document that U.S. monetary policy surprises have large, persistent and asymmetric effects on government bond yields worldwide. Moreover, the impact of Fed policy on sovereign debt markets has experienced a structural break around the Global Financial Crisis (GFC). Treasury term premiums increase persistently following Federal Reserve easing shocks until 2007, but show a protracted decline in the post-GFC sample. Advanced and emerging market economy yields essentially mimic the Treasury market's asymmetric response to Fed surprises. While the break of the term premium response to easing shocks around the GFC is consistent with a change in the net duration of primary dealers' Treasury positions, intermediary balance sheet constraints cannot explain the asymmetric effects of monetary policy on yields. The persistent and asymmetric global yield responses are broadly in line with the dynamics of mutual fund

ows following U.S. monetary policy shocks.

This paper studies the feedback between stock market fluctuations and wealth inequality dynamics. We do so by means of a dynamic consumption-based general equilibrium model with endogenous asset returns and a non-degenerate wealth distribution for a continuum of households. Households are heterogeneous in risk aversion and thus choose different expected portfolio returns and portfolio return volatilities, generating time-varying wealth inequality. We show how to solve the model analytically in terms of a cumulant generating function, which encodes information about all the moments of the distribution of risk aversion. With this result, we recover the nobservable distribution of risk aversion using time variation in the slope of the observable equity term structure. We also confront the model with US data on the wealth distribution to recover a second estimate of the distribution of risk aversion. By comparing the two estimates, we show quantitatively that there is significant feedback between stock price dynamics and wealth distribution dynamics