Research

Structural VARs are often identified by using instruments derived from the residuals of auxiliary regressions (e.g., Romer and Romer (2004)). We derive analytical characterizations of the instrument’s correlation with the true shocks and evaluate the procedure’s finite-sample performance using Monte Carlo experiments based on the model of Smets and Wouters (2007). We find that such instruments are meaningfully correlated not only with the monetary policy innovation, but also with other structural shocks, leading to substantial biases and variation in estimated impulse responses. We then examine several proposals from the literature designed to mitigate VAR issues and propose a correction based on instruments for other structural shocks.

I exploit the empirically documented comovement in households’ expectations to identify what I call sentiment shocks. These disturbances shift the entire system of their economic perceptions causing deviations from fully rational expectations. The estimated shock series is correlated with consumer sentiment measures yet remains distinct from standard macroeconomic shocks. A Structural VAR analysis shows that sentiment disturbances have large and long-lived effects on household consumption, with the response of durable goods spending being especially strong. Sentiment shocks account for over 30% of the volatility in durable consumption and around 20% in non-durables at horizons of one to two years. I extend an otherwise standard New Keynesian model by introducing sentiment shocks that trigger a deviation of expectations from the rational benchmark. I present analytical results demonstrating that, depending on parameter values, a positive sentiment shock can generate fluctuations of either sign in output, inflation or the interest rate. The parameter estimates suggest that the high persistence of sentiment disturbances gives rise to prolonged effects on the model economy, consistent with the empirical impulse responses. Based on the estimated parameters, the quantitative results imply that the equilibrium effects of sentiment shocks on output and inflation are primarily driven by expectations of future interest rate changes. The latter reflects the anticipated monetary policy reaction to expected output fluctuations arising from sentiment disturbances.

In this paper, I present empirical evidence that the effects of asset purchases conducted by the Federal Reserve depend on the state of macroeconomic uncertainty. I find that when aggregate uncertainty is elevated, a tapering of asset purchases has counterintuitive effects: industrial production steadily increases, and unemployment declines. However, in a state of normal uncertainty, a reduction of asset purchases yields expected contractionary outcomes, as industrial production falls, and unemployment exhibits an upward trend. I interpret these empirical findings as reflecting the informational content of the Federal Reserve’s announcements on quantitative easing, which becomes particularly relevant in highly volatile economic conditions. By scaling back asset purchases under elevated uncertainty, the Federal Reserve reveals that its own economic outlook has strengthened relative to its prior projections. This information improves market expectations and acts as a stimulus to economic activity.