**2. Data**

As for crude oil prices, we used the nominal daily data derived from the US Energy Information Administration (EIA, https://www.eia.gov/dnav/pet/hist/RWTCD.htm (accessed on 1 May 2021)) for West Texas Intermediate (WTI). After computing the daily log-returns, we obtained quarterly overall, upside ("good"), and downside ("bad") realized variances by taking the sum of the daily squared returns, positive returns only, and negative returns only over a specific quarter. As a robustness check, we also analyzed the quarterly *RV* of the Brent crude oil returns, which, in turn, was also sourced from the US EIA (see https://www.eia.gov/dnav/pet/hist/RBRTED.htm (accessed on 1 May 2021)).

Figure 1 plots the *RV* (and its "good" and "bad" counterparts, as defined in Section 3) of both the WTI and Brent crude oil returns. During the GFC, sharp fluctuations in *RV*s were observed over 2020 associated with the COVID-19 outbreak, thus, highlighting the importance of our question.

Uncertainty is a latent variable and, hence, requires methods to measure it. As documented by Gupta et al. [27], there are three broad approaches to quantify uncertainty, apart from the various ones associated with financial markets (such as implied-volatility indices, like the popular VIX, realized volatility, idiosyncratic volatility of equity returns, and corporate spreads): (1) a text-based approach, with the main idea to construct indices from searches of key words or terms related to (economic and policy) uncertainty in major newspapers or country-reports; (2) using stochastic-volatility estimates from various small and large-scale structural models (related to macroeconomics and finance) to derive measures of uncertainty; and (3) using the dispersion of professional forecaster disagreements to obtain uncertainty estimates.

For our metric of uncertainty, we used the first approach outlined by Ahir et al. [28], mainly because it is not model-specific, as it does not require any complicated estimation of a large-scale model to generate it in the first place. In addition to the uncertainty data, the associated spillover of the G7 economies and China to other economies in the world,

are available publicly for download (https://worlduncertaintyindex.com/data/ (accessed on 1 May 2021)).

**Figure 1.** Realized volatility. For better readability, the vertical axes of the panels are not the same for all time series.

From frequency counts of "uncertainty" (and its variants) in the quarterly Economist Intelligence Unit (EIU) country reports from 1996 for 8 of the 143 countries, Ahir et al. [28] constructed quarterly indices of economic uncertainty for 37 countries in Africa, 22 in Asia and the Pacific, 35 in Europe, 27 in the Middle East and Central Asia, and 22 in the Western Hemisphere. The EIU reports provide an analysis and forecasts of political, policy, and economic conditions, as well as a discussion of significant political and economic developments in each country. These reports are compiled by a central EIU editorial team from work done by country-specific teams of analysts.

In order to make the uncertainty indexes comparable across countries, the raw counts were scaled by the total number of words in each report. In addition to the uncertainty indexes of each of the 143 countries, the dataset of Ahir et al. [28] also provides the uncertainty spillover metrics for the G7 and China, which, in turn, determine the choice of the countries in our paper, and the quarterly sample period of 1996Q1 to 2020Q4, which was the latest available data at the time of writing this paper. Specifically, the eight (G7 plus China) uncertainty spillover indexes of one of these particular countries to the remaining 142 countries was computed by counting the percent of word "uncertain" (or its variant) mentioned within a proximity to a word related to a particular G7 country or China in the EIU country reports.

The spillover index was then rescaled by multiplying 1,000,000 with a higher number suggesting higher uncertainty related to the specific country involving the G7 or China and vice versa. For further details regarding the words related to the G7 and China that were used, the reader is referred to Ahir et al. [28]. We used the cross-sectional sum over time to obtain the total uncertainty spillover (on to the remaining 142 economies) indexes of each of these eight countries.

Understandably, since we aimed to contribute to the oil *RV* forecasting literature by analyzing whether accounting for spillovers of uncertainty of the G7 countries and China to the rest of the world mattered over and above the uncertainty of these economies, we relied on the method of Ahir et al. [28] for a matter of consistency and similarity in how both these indexes are derived, even though alternative ways of constructing country-level uncertainty indexes, though not spillovers, are available in the public domain (see, for example, the indexes avilable at: http://policyuncertainty.com/ (accessed on 9 July 2021) based on the work of Baker et al. [29]).

Figures 2 and 3 plot the uncertainty time series and the international spillover effects. While the uncertainty series tended to fluctuate consistently over time, the spillovers had sudden massive spikes from the country(ies) of origin of the GFC, the European sovereign debt crisis, "Brexit", and the outbreak of the Coronavirus pandemic.

**Figure 2.** Time series of uncertainty. For better readability, the vertical axes of the panels are not the same for all time series.

**Figure 3.** Time series of international spillovers. For better readability, the vertical axes of the panels are not the same for all time series.
