Tracking global economic uncertainty: implications for the euro area

This paper should not be reported as representing the views of the European Central Bank (ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB.

Abstract

This paper sheds light on the impact of global macroeconomic uncertainty on the euro area economy. We build on the methodology proposed by Jurado et al. (2015) and estimate global as well as country-specific measures of economic uncertainty for fifteen key euro area trade partners and the euro area. Our measures display a clear counter-cyclical pattern and line up well to a wide range of historical events generally associated with heightened uncertainty. In addition, following Pi er and Podstawski (2018), we estimate a Proxy SVAR where we instrument uncertainty shocks with changes in the price of gold around specific past events. We find that, historically, global uncertainty shocks have been important drivers of fluctuations in euro area economic activity, with one standard deviation increase in the identified uncertainty shock subtracting around 0.15 percentage points from euro area industrial produc-tion on impact.

Keywords: Uncertainty, Forecast Errors, Stochastic Volatility, Proxy SVAR, Economic Activity.

JEL-Classiication: D81, C11, C55, E32, F41, F62.

ECB Working Paper Series No 2541 / April 2021

Non-technical summary

The outlook for euro area activity has become increasingly uncertain in recent years, reflecting global and domestic headwinds such as escalating trade conflicts, moderating Chinese demand, (geo-) political tensions and stress in several emerging economies. More recently, the Covid-19 pandemic has triggered an unprecedented rise in uncertainty, globally and in the euro area, with elevated uncertainty frequently singled out by the Eu- ropean Central Bank (ECB) and international institutions as an important driver of euro area economic developments. This paper sheds light on the impact of fluctuations in global uncertainty on the euro area economy. We follow a two-pronged approach where we estimate global as well as country-specific measures of economic uncertainty for fifteen key euro area trade partners and the euro area. The measures, which can be tracked over time and relate to fluctuations in real economic activity, are based on a specific notion of uncertainty. Specifically, in our set-up high uncertainty is associated with an inherent difficulty to predict economic outcomes. Our measures line up well to a wide range of historical events generally associated with heightened uncertainty and display a clear counter-cyclical pattern. In a second step, we estimate a Proxy Structural VAR to gauge the impact of global uncertainty shocks on the euro area economy, employing changes in the price of gold around specific past events as an external instrument. In addition to our uncertainty indicator, the model, estimated over the period July 2000 to December 2019, also includes a set of monetary, real and nominal variables. We find that, historically, global economic uncertainty shocks have been important drivers of fluctuations in euro area activity. A one standard deviation increase in the identified global uncertainty shock adversely affects euro area industrial production for up to 10 months. The biggest decline is observed on impact, when the dent to industrial production amounts to around 0.15 percentage points. Results suggest that economic uncertainty has been an important determinant of euro area industrial production also in recent years.

ECB Working Paper Series No 2541 / April 2021

Introduction

By demanding “rigidly defined areas of uncertainty and doubt” philosopher Vroomfondel in Douglas Adam’s “The Hitchhiker’s Guide to the Galaxy” inadvertently captures one of the key features of uncertainty: it seems familiar to everybody but its quantification can be elusive. This reflects the intrinsically diffuse nature of the concept, which originates from a multitude of sources along different geographical dimensions. While the majority of existing studies analyse the effects of uncertainty in individual countries, financial globalization and the diffusion of new technologies have prompted the emergence of a recent literature which classifies instead uncertainty as a global phenomenon and assesses its effects on the global economy and on individual countries. Our paper feeds into this recent literature, and contributes to it in two ways. First, we develop econometric estimates of global uncertainty that are independent from any structure imposed by specific theoretical models as well as from single (or small number of) observable economic indi- cators. Second, we are the first to assess the impact of global macroeconomic uncertainty on the euro area economy by relying on a Proxy VAR, thereby offering an analytically sound identification of uncertainty shocks.

Several proxies of global uncertainty have been proposed in the literature. Berger et al. (2017) measure global uncertainty through the lens of a dynamic factor model with stochastic volatility covering 20 advanced economies. Their results indicate that uncertainty rose sharply in the past, such as in the early 1970s, in the 1980s and during the great recession. At the broadest degree of aggregation, Carriero et al. (2020) develop a measure of international macroeconomic uncertainty by means of a large vector autoregression to assess its bearing on a comprehensive set of economic and financial variables in advanced economies. Shen et al. (2017) apply the methodology proposed by Jurado et al. (2015) (henceforth JLN) to explore the effects of global uncertainty shocks on an array of global macroeconomic variables as well as on commodity markets. Compared to us, however, they rely on a more limited set of series. In assessing the impact of an increase in domestic uncertainty on the UK economy, Redl (2017) constructs a global measure which serves as a control variable for the effect of global uncertainty shocks. Carri`ere-Swallowand C´espedes (2013) survey cross-country heterogeneities in the reaction of investment and private consumption to global uncertainty shocks with a focus on differences between developed and emerging economies. At a more granular level,

ECB Working Paper Series No 2541 / April 2021

Mumtaz and Theodoridis (2017) and Ozturk and Sheng (2018) consider both common (global) and country-specificmeasures of uncertainty while Mumtaz and Musso (2019) add a regional angle. Bonciani and Ricci (2020) construct a measure of global financial uncertainty and study the impact on 44 economies. They find that sudden increases in global financial uncertainty depress both real and nominal variables, while the magnitude of the response varies with country specific characteristics and the state of the economy. Pfarrhofer (2019) analyses the impact of international macroeconomic uncertainty shocks using a Bayesian model with drifting coefficients and stochastic volatility in the errors, and finds that the impact of such shocks are sizable across countries. Cuaresma et al. (2019) investigate the macroeconomic consequences of international uncertainty shocks on G7 countries with a large-scaleBayesian vector autoregression featuring factor stochastic volatility and discover large effects in all economies considered. Finally, by scrutinising spillovers of uncertainty shocks in the United States and the euro area to third countries, Belke and Osowski (2019) introduce a cross-borderperspective.

In this paper, we extend the methodology developed by JLN for the US to a large group of advanced and emerging economies covering 74 per cent of global output. Specif- ically, we estimate global as well as country-specific measures of economic uncertainty for fifteen key euro area trade partners and the euro area. Our indicators of uncertainty can be tracked over time and relate to fluctuations in real economic activity. Differently from alternatives based, for instance, on dispersion of forecasts from professional forecasters or on news, the approach used by JLN yields an objective measure of uncertainty. In line with JLN, we start from the premise that what matters for economic decision making is not whether particular economic indicators have become more or less variable or dispersed, but whether the economy has become more or less predictable. Compared to existing studies, our analysis is grounded on a much more sizeable data-set, affording robustness, with a notable rise or fall in uncertainty only occurring in the event of shifts in the conditional volatility of a large number of series. This is in line with uncertainty- based theories of the business cycle which typically predicate the existence of common (and often counter-cyclical) variation in uncertainty on a large number of series. To quote JLN – one “would expect to find evidence of an aggregate uncertainty factor, or a common component in uncertainty fluctuations that affects many series, sectors, markets, and geographical regions at the same time”.
Source: ECB – European Central Bank