Changing expected returns can induce spurious autocorrelation in returns. We show why this happens with simple examples and investigate its prevalence in actual equity data. In a key contribution, we use shifts in ex ante expected return estimates from options prices, factor models, and...