Empirical essays on uncertainty and fiscal shocks

Abstract

The thesis empirically investigates the macroeconomic effects of uncertainty and fiscal shocks, particularly focusing on the role of nonlinearities in the transmission mechanisms of shocks. Chapter 1 investigates the effects of global uncertainty shocks in open economies and explores the role that different levels of country risk exposure might have in the transmission of the shock. I employ an Interacted VAR model and examine the dynamic responses of output and exchange rate to a global uncertainty shock for a group of developed economies. The I-VAR model allows to account for time variation in country relative riskiness, through the computation of state conditional impulse response function. As a measure of global uncertainty, I take the realized volatility of the MSCI World Index log returns. Taking the U.S. as the reference country, I consider the spread between each country’s interbank rate and the Federal Funds rate as my measure of relative riskiness. I define a country as being in high risk regime when its interbank rate is lower than the FFR (negative spread), whereas I define the country in low risk regime when its interbank rate is higher than the FFR (positive spread). The spread reflects different levels of precautionary savings in the two countries. My results show that countries’ relative risk exposure to aggregate risk plays a relevant role in the transmission of a global uncertainty shock and provide evidence for the presence of nonlinear effects in the transmission of the shock to economic activity. Indeed, the dynamic responses of output in high risk are estimated to be negative and statistically larger than those in low risk. Chapter 2 estimates time-varying spending multipliers, to assess the evolution of U.S. fiscal policy effectiveness over time. I ask whether and how the monetary policy stance affects the transmission of fiscal policy shocks and devote special attention to the size of spending multipliers when the economy is at the zero lower bound (ZLB). To this aim, I model a set of standard macro-fiscal variables over the period 1954Q3-2015Q3 by using a time-varying parameter VAR model with stochastic volatility, which allows both shocks and the fiscal transmission mechanism to change over time. Unconventional monetary policy in the last part of the sample is explicitly accounted for by including the shadow rate estimated by Wu and Xia (2016), which captures relevant information about the monetary policy stance from movements in the term structure of interest rates. Estimated spending multipliers display substantial variability over the investigated period. Particularly, the effects of changes in government spending on economic activity are estimated to be larger in periods characterized by passive monetary policy, namely when monetary policy is not strongly reacting to inflation, and at the ZLB. Chapter 3 is based on joint work with Giovanni Caggiano, Efrem Castelnuovo and Tim Robinson. It estimates time-dependent finance-uncertainty multipliers in the post-WWII U.S. sample. We model a number of macroeconomic indicators with a time-varying parameter VAR along with measures of financial uncertainty and credit spreads. Uncertainty shocks are identified by relying on the measure of financial uncertainty recently constructed by Ludvigson, Ma, and Ng (2016). Gilchrist and Zakrajsek (2012)'s credit spread measure is decomposed in two orthogonal components - the exogenous excess bond premium and the endogenous part of the spread - to control for credit supply shocks while considering the amplification effect due to financial frictions following an uncertainty shock. Our results point to substantial variations in the finance-uncertainty multipliers over the investigated period. Uncertainty shocks always have recessionary effects, although with time-varying persistence, and those effects are amplified in periods of financial distress.