Time-Frequency Spillovers and the Determinants among Fossil Energy, Clean Energy and Metal Markets.

• Author: Ding, Qian

1. INTRODUCTION

   Metals are an energy-consuming resource that require fossil energy for their extraction, production, smelting and recycling processes; therefore, fossil energy is an important input for the operation of metal-extraction systems. To alleviate global climate change and ensure energy supply security, countries worldwide regard clean energy as the main pathway to low-carbon development (Bahn et al., 2019; Bernai, 2013). However, compared with traditional fossil energy, clean energy is more metal intensive and consumes more metals in terms of type and quantity, especially strategic metals (de Koning et al., 2018; Grandell et al., 2016; Sovacool et al., 2020; Zhou et al., 2019). Metals have energy-enabling properties and are important materials for clean energy technology-related equipment. Emerging technological revolutions are triggering a strong demand for the required metals (Tokimatsu et al., 2017). The global energy system is rapidly shifting from carbon-based to metal-based resources and the enhanced connections between clean energy and metals have a huge impact on the global energy transition (Wang et al., 2019). Clarifying the relationships among fossil energy, clean energy and metal markets is important to promote green investments and secure a low-carbon economy.

   Clean and fossil energy sources are intuitively substitutable. When fossil energy prices increase, the incentive to further develop clean energy sources also increases, which potentially leads to higher stock prices for clean energy firms (Xia et al., 2019). However, the clean energy sector is highly dependent on technologies and innovations powered by traditional fossil energy sources, which may lead to a complementary relationship between clean and fossil energy sources (Jiang et al., 2021). Metal is an important raw material in the clean energy industry; therefore, supply and demand changes in clean energy prices may affect metal prices because the availability of metals for clean energy technologies and the possibility of their supply disruptions often lead to dramatic and unexpected price increases (Leader et al., 2019). The development of clean energy stimulates the consumption for metals leading to increasing metal prices, which may in turn increase the pressure on the clean energy sector. However, this is not conducive to facilitating low-carbon transitions (Shao et al., 2021). In addition, fossil energy is consumed in the mining and smelting of metals; therefore, changes in fossil energy prices also directly affect the cost of mining and refining metals (Zhang and Tu, 2016), which in turn affects metal prices.

   The interrelationship between fossil energy and clean energy assets is well documented (Kyritsis and Serletis, 2019; Naeem et al., 2020; Saeed et al., 2021; Song et al., 2019; Umar et al., 2021; Wen et al., 2014). Fluctuations in fossil energy prices, particularly oil prices, can significantly affect stock market returns (Khalfaoui et al., 2019; Sarwar et al., 2019a, 2020; Waheed et al., 2018), especially for clean energy stocks. Uncertainty due to the volatility of fossil energy markets may be an important factor in accelerating or impeding the transition to sustainable energy systems (Balcilar et al., 2019). The consensus in the literature is that market fluctuations (i.e., price changes) in the fossil energy industry may have a strong effect on the clean energy industry's development, especially related to the investments and returns for clean energy in the capital market. There is also a strong correlation between fossil energy and metals. At the supply and demand level, Gutowski et al. (2013) show that the availability of fossil energy is an important constraint on the sustainable supply of metal resources. Vidal et al. (2017) find that fossil energy consumption in the metal minerals sector increases as the level of metal minerals decreases and demand increases. Oil is closely related to the metal markets in terms of financial linkages. Shao et al. (2021) observe a causal relationship between oil price uncertainty and fluctuations in Chinese metal stocks, and Liu et al. (2021a) show that both shocks and jumps in oil prices have an impact on the Chinese non-ferrous metals market. Mensi et al. (2021a) indicate that there is lower tail dependence and higher tail independence between the oil and non-ferrous markets. After examining the hedging effects of oil and gold on commodities, Naeem et al. (2022) suggest that oil is a safe haven for metal markets.

   More importantly, many studies have considered the supply-demand connections between clean energy and metals (Choi et al., 2016; Grandell and Thorenz, 2014; Liu et al., 2019; Shammugam et al., 2019). However, with the evolution of supply-demand relationships and strengthening financial integration, the financial connectedness between clean energy and metals is becoming increasingly important. Some studies investigate the financial connections between a single metal category and clean energy markets. For instance, Baldi et al. (2014) develop a multi-factor market model to analyse the effects of rare earth prices on the performance of clean energy markets and find that rare earth prices are negatively correlated with clean energy markets during periods of price increase. Zheng et al. (2021) study the risk transfer between China's clean energy and rare earth markets at the firm level and find strong links. Yahya et al. (2020) analyse the dependence between the non-ferrous metals and clean energy markets based on the cross-quantilogram correlation approach, and suggest that the asymmetric dependence is time-varying.

   Although these studies provide good insights for analysing the relationships between fossil energy, clean energy and metal markets, some limitations must be resolved. First, scholars generally focus on the pair-based relationships between fossil energy, clean energy and metal markets, and fail to integrate these three markets into the same systematic framework for comprehensive analysis. Second, scholars do not use multiple time horizons to consider the connections between these markets. The economic entities interacting in different markets have heterogeneous beliefs, preferences, goals, information reception and risk tolerance (Chen et al., 2020; Ding et al., 2021; Wang, 2020). Therefore, the spread of economic and financial shocks across markets causes different frequency responses in these markets. Third, scholars exploring the relationship between clean energy and metal markets usually only consider a single type of metal and the total index for the entire clean energy industry, which ignores the heterogeneity of different types of metals and clean energy markets in various sub-industry investment areas. Finally, scholars only explore the intensity and scale of spillovers between energy and metal markets without performing an in-depth analysis of their impact mechanism, let alone their determinants.

   To supplement the literature, we evaluate the time-frequency spillovers among fossil energy, different types of metals and clean energy markets. We use the Barunik and Krehlik (2018) (BK) spillover framework to solve the problem of measuring the connectedness between various markets in the frequency domain. We comparatively analyse the spillover effects among fossil energy, different types of metals and clean energy sub-sectors to contribute to formulation of effective diversification strategies. Moreover, we construct spillover networks to determine the transmitters and receivers in the spillover system. More importantly, we expose the determinants of system-wide spillovers at multiple levels.

   This study contributes to the empirical literature in four important ways. First, we break through the limitations of previous pair-based relationship studies and comprehensively explore the energy-metal nexus by integrating fossil energy, clean energy and metal markets into the same systematic framework based on metals' energy-consuming and energy-enabling properties. Second, the spillovers may vary with different time frequencies because of market participants' different expectations and reactions. Our innovative consideration of spillover analysis frameworks within a multi-period time horizon framework (short- and long- term) provides new insights for analysing the connectedness among fossil energy, clean energy and metal markets. Third, this study considers the heterogeneous interactions between different types of clean energy and metals markets, and also considers the role of fossil energy market. This comprehensive industry-wide analysis not only enriches the traditional research literature on energy-metal relationships, but can also help policymakers to formulate rational energy policies. Finally, this study investigates the contribution of multiple uncertainties and economic and financial factors to system-wide spillovers in different frequencies. The analysis of spillover impact mechanisms may provide policy implications for policymakers to promote the development of the clean energy sector.

   This study is organised into the following parts. Section 2 analyses the influence mechanism of spillovers. Section 3 provides the methodology and data. The empirical findings are analysed in Section 4. Finally, Section 5 concludes the paper and proposes relevant policy implications.

2. INFLUENCE MECHANISM ANALYSIS

   Multiple influencing factors affect the close connections between the fossil energy, clean energy and metals markets following the acceleration of the clean energy transition. We explore the most important factors affecting spillovers among these markets based on the literature, i.e., economic fundamentals, financial and uncertainty factors (Ding et al., 2021;Saeed et al., 2021; Tan et al., 2020; Wang, 2020; Xiao et al., 2019; Yang et al., 2021), as shown in Figure 1.

   2.1 Economic Fundamentals

   During economic booms, the demand for energy saving...