Elsevier

Energy Economics

Volume 76, October 2018, Pages 115-126
Energy Economics

Uncertainties and extreme risk spillover in the energy markets: A time-varying copula-based CoVaR approach

https://doi.org/10.1016/j.eneco.2018.10.010Get rights and content

Highlights

  • There exists negative dependence between energy returns and uncertainty changes.

  • The risks of clean energy and oil returns are more sensitive to financial and energy uncertainties.

  • The impact of economic policy uncertainty on energy returns is relatively weak.

  • The upside and downside CoVaRs have asymmetric effects in response to uncertainty risks.

Abstract

In this paper, we explore the impact of uncertainties on energy prices by measuring four types of Delta Conditional Value-at-Risk (∆CoVaR) using six time-varying copulas. Three different measures of uncertainty (economic policy, financial markets and energy markets) are considered, and the magnitude and asymmetric effects of their influence are investigated. Our results suggest that there generally exists negative dependence between energy returns and changes in uncertainty. The risks of clean energy and crude oil returns are more sensitive to uncertainties in the financial and energy markets, while the impact of economic policy uncertainty is relatively weak. The upside and downside CoVaRs and ∆CoVaRs demonstrate significant asymmetric effects in response to extreme uncertainty movement. Our findings therefore have important implications for energy portfolio investment.

Introduction

The energy market is one of the most important components in the economic system whose prices have been increasing in volatility in recent years with high risk and large uncertainty (Ji and Zhang, 2018). In the aftermath of the global financial crisis, uncertainty has played a fundamental role in influencing energy prices, which has been documented by several previous studies. For instance, Joëts, 2014, Joëts, 2015 and Yin and Han (2014) indicate that increased volatility in the level of uncertainty will lead to increased volatility in energy prices as well as an alteration in their driving mechanism. Therefore, such volatility can strongly influence the real economy and hamper the stability of the financial system, potentially even triggering a systemic risk in the global financial markets (Bilgin et al., 2015).

The driving factors behind the energy market are complex and diversified. Hence, energy prices cannot be fully explained by the framework of supply and demand. Further, increasing uncertainty in the energy market has rendered the price dynamics even more complicated (Mellios et al., 2016; Zhang, 2017). Uncertainty has therefore become an important factor affecting both price volatility and market risk. Thus, identifying the most appropriate means of measuring the impacts of uncertainty has become the focus of considerable research attention (Baker et al., 2016; Bloom, 2014; Jurado et al., 2015; Pástor & Veronesi, 2013; Zheng, 2014).

In order to differentiate between the different measures of uncertainty, Bloom (2014) argues that the respective volatility of stock markets, bond markets, exchange rates and GDP growth all rise steeply during recessions. The author concludes that almost every macroeconomic indicator of uncertainty appears to be countercyclical, that is, the volatility of the indicator rises when the overall economy is slowing down. Pástor and Veronesi (2013) question how political uncertainty affects market prices. They conclude that there is a small but growing body of work concerning the theoretical effects of government-induced uncertainty on asset prices, while there is currently only a modest amount of empirical work relating political uncertainty to the equity risk premium. In any economy that is growing fundamentally normally, politicians prefer to largely maintain their current policies, whereas politicians are tempted to experiment in an attempt to alleviate economic policy uncertainty during turbulent periods. Jurado et al. (2015) hence investigate the importance of time-varying economic uncertainty and the role it plays in macroeconomic business cycle fluctuations. The above-mentioned studies all shed new light on the dynamic relationship between uncertainties, volatility and energy price movement.

In any economy, uncertainty fundamentally appears to have both a short-term and a long-term component (Barrero et al., 2016). Barrero et al. (2016) show that investment is significantly more sensitive to long-term uncertainty, while employment responds equally to both short-term and long-term uncertainty. At the same time, they find that oil uncertainty is particularly closely related to short-term uncertainty, while policy uncertainty is particularly related to long-term uncertainty. Therefore, we consider measures of both short-term and long-term uncertainties in order to investigate the impact of such uncertainties on energy markets. It has become a stylised fact that the business cycle between asset returns is not constant over time, since it can generate an economy that displays the asymmetric features of information (Nalebuff & Scharfstein, 1987).

Recently, the impacts of various uncertainties on both the real economy and the associated macroeconomic variables have been the focus of increased research attention due to the consequences of the global financial crisis. Since there is currently no consensus regarding the best measure of uncertainty, many distinct proxies have been employed in the prior literature. According to Fernandes et al. (2014) and Mele et al. (2015), the Chicago Board Options Exchanges (CBOE) implied volatility index (VIX) is the most commonly used proxy for global financial market uncertainty.1 However, the VIX may not be linked to true economic uncertainty, since it might be driven by factors associated with time-varying risk-aversion rather than economic uncertainty (Bekaert et al., 2013). Baker et al. (2016) thus develop an alternative measure of economic policy uncertainty (EPU), which has been widely used to explore its impact on both monetary policy and financial markets (Wisniewski and Lambe, 2015; Antonakakis and Floros, 2016; Aastveit et al., 2017; Demir and Ersan, 2017; Li, 2017). There also exist other alternative measures of uncertainty. For example, Bachmann et al. (2013) take disagreement into account and propose a proxy of uncertainty as measured by the cross-sectional disagreement of economic agents, while Jurado et al. (2015) use the common variation in uncertainty across hundreds of economic series as an alternative measure of economic uncertainty. As we know, only a limited number of prior studies have focused on the impact of uncertainty on the energy market. Aloui et al. (2016) employ a static copula to determine that higher uncertainty as measured by equity and economic policy uncertainty can only increase oil returns during certain periods of time. Chen and Kettunen (2017) investigate the impact of uncertainty within the strictness of carbon policy on power generation firms' capacity investment decisions and find that it is optimal for firms with higher risk aversion to invest more in renewable technologies than their less risk-averse rivals.

Although various types of economic uncertainty have been measured thus far, little evidence has actually been provided to verify the effective influence of uncertainty on energy prices which is more closely linked to finalisation. In order to extend the current research on the relationship between energy prices and various uncertainties, we further consider the CBOE's crude oil volatility index (OVX) as a proxy for energy market uncertainty (Aboura & Chevallier, 2013; Ji and Fan, 2016). Therefore, three uncertainty measures, namely economic policy uncertainty (EPU), the VIX and the OVX, are selected in this study as being representative of economic policy, financial market and energy market uncertainty, respectively. These measures can facilitate a comprehensive analysis of the different diffusion channels through which uncertainty influences energy prices.

Our study contributes to the current literature in three main respects. First, three different measures of uncertainty (EPU, the VIX and the OVX) are chosen so as to compare their distinct influences on both fossil and clean energy prices. Using this approach, we can identify the risk transmission channel from uncertain market conditions to different energy returns. Second, the non-linear and dynamic dependence relationships between energy price returns and uncertainties are investigated using six time-varying copula models. Finally, the value-at-risk (VaR) of energy price returns conditional on the VaR of uncertainties is measured by proposing four types of Delta Conditional Value-at-Risk (∆CoVaR) to explore the influence of extreme uncertainty movement on energy price returns. This study therefore presents the first model of risk comovement between energy returns and uncertainty changes from the perspectives of extreme market risks. It hence provides a new analysis tool for financial investors and risk managers seeking to control their trading risks during extreme periods.

The remainder of this paper is structured as follows. The next section describes the empirical methodology of our study, while Section 3 explains the utilised data. Section 4 presents our results, while Section 5 details our conclusions and the associated policy implications.

Section snippets

Methodology

In this section, in order to investigate the impacts of different uncertainties on energy market risks, a copula-based CoVaR approach is employed. First, four types of CoVaR and ∆CoVaR are measured to quantify the risk spillover effects of extreme upward and downward uncertainty changes on extreme energy market returns. Furthermore, some necessary statistical tests are employed to verify the existence of risk spillover effects from uncertainty to the energy markets as well as asymmetric effects

Data and preliminary analysis

We consider daily data for the S&P 500 Global Clean Energy Index (CEX), the crude oil prices (OIL) and the natural gas prices (GAS) covering the sample period from May 10, 2007 to April 13, 2017 (a total of 2591 observations). Our sample period includes the global financial crisis of 2008–2009 and the Eurozone sovereign debt crisis of 2010–2012. In order to capture the uncertainty, we consider three different measures of uncertainty, namely economic policy, financial markets and energy markets.

Estimation of the time-varying copula-GARCH model

Since all series exhibit the autocorrelation and conditional heteroscedasticity, ARMA-GARCH-t model is employed to construct the conditional marginal distribution. Table 2 presents the estimated coefficients for energy returns and uncertainty changes. The values of the freedom degree of the t distribution measured by υ range from three to eight, which indicates that the error terms were not normal. In addition, the conditional volatilities of all the variables followed a GARCH (1, 1) process

Conclusion

This study empirically investigates the dynamic dependence between energy price returns and uncertainties, including economic, financial and energy uncertainties. Our results suggest that the risks seen in the energy sources in response to the uncertainties are heterogeneous rather than homogenous in structure. The clean energy and oil sources are more strongly influenced by financial and energy uncertainties than by economic policy uncertainty. Financial and energy market uncertainties have a

Acknowledgments

We are grateful to faculty members of the University of Vaasa, Stavanger University, Warsaw School of Economics and Linköping University as well as seminar participants at other institutions for helpful discussions. We particularly thank Panu Kalmi, Juha Junttila, Anupam Datta and Michal Rubaszek for valuable comments. The first author acknowledges support from the National Natural Science Foundation of China (Grant No. 71774152, No. 91546109) and the Youth Innovation Promotion Association of

References (47)

Cited by (0)

View full text