Global and Country‐Specific Output Growth Uncertainty and Macroeconomic Performance

Abstract

We identify global and country-specific measures of output growth uncertainty for a large OECD country sample by means of a dynamic factor model with stochastic volatility. We find evidence for major bouts of global uncertainty in the early 1970s and late 2000s, and a number of periods with elevated levels of either global or national uncertainty, particularly in the early 1980s, 1990s and 2000s. VAR impulse responses of national macroeconomic variables to our estimated measures of uncertainty reveal that global uncertainty is the major driver of macroeconomic performance in most countries, whereas the impact of national uncertainty is small and frequently insignificant. We also find that uncertainty is transmitted primarily through investment and trade flows rather than through consumption demand.

Measuring the international dimension of output volatility

“…Del Negro and Otrok, 2008;Berger et al, 2016b). Overall, results and interpretation remain qualitatively unchanged but we find that the high frequency noise present in quarter-on-quarter data tends to blur the correlation structure in the growth rates, i.e.…”

Assessing International Commonality in Macroeconomic Uncertainty and Its Effects

“…While our paper shares with Berger, Grabert, and Kempa (2016) the focus on e↵ects of shocks to macroeconomic uncertainty as most commonly conceived, we believe our approach o↵ers a couple of advantages. One, as noted above, is that we consider a larger dataset (in part of our results; we also consider a GDP-only dataset, as do they).…”

“…They find the global component to be an important driver of time-varying volatility. Using GDP growth for 20 countries, Berger, Grabert, and Kempa (2016) estimate a factor model with stochastic volatility components common to the world and specific to each country; in a second step, for each country, they estimate VARs with other variables and uncertainty to assess the e↵ects of uncertainty. Carriere-Swallow and Cespedes (2013) and Gourio, Siemer, and Verdelhan (2013) also use simple, small VAR approaches, measuring uncertainty with the volatility of stock returns.…”

“…In the existing studies that have assessed the e↵ects of uncertainty on macroeconomic fluctuations across countries, uncertainty has commonly been measured and assessed using a small set of variables for each country. For example, Berger, Grabert, and Kempa (2016) assess the e↵ects of uncertainty using seven variables per country, and Cross, Hou, and Poon (2018) use three variables per economy, pairing one large economy with one small economy. In our larger data set, we have roughly 25 variables for each of the U.S. and E.A.…”

“…One, as noted above, is that we consider a larger dataset (in part of our results; we also consider a GDP-only dataset, as do they). The other is that our approach assesses uncertainty and its e↵ects with an internally-consistent one-step approach, rather than the two-step approach used in Berger, Grabert, and Kempa (2016) and common in other studies in the uncertainty literature. In the two-step approach, a measure of uncertainty is estimated in a preliminary step and then used as if it were observable data in the subsequent econometric analysis of its impact on macroeconomic variables.…”