The Impact of Energy Market Uncertainty Shocks on Energy Transition in Europe.

• Author: Balcilar, Mehmet

1. INTRODUCTION

   Since 1973, when the first oil price shocks occurred, energy market shocks have led to recession for many world economies and hampered sustainable growth. An important feature of energy markets is the high volatility in prices, arising from demand and supply shocks, and other events, such as wars and internal political conficts. As Kilian (2008a) pointed out, oil price shocks have been the distinguishing characteristic of the world economies since 1970s, and it is significantly linked to the economic performance.

   Consequently, policy makers have been incited to look for alternate energy sources, not only to address the increasing environmental damages associated with traditional energy sources, but also to protect their economies from these recurrent shocks. However, the increasing use of renewable energy sources still remains a slow process. The European Union (EU) is developing new strategies and setting targets for member countries. For example, all EU countries have agreed on a new EU renewable energy target, according to which renewable sources in the final gross energy consumption share should reach at least 20% by 2020 and 27% by 2030. While the penetration rates of low carbon production in Europe have increased, the transition has been much slower than the policy goals, and there are significant differences between individual countries.

   The most important effect on the low carbon transition comes from the increased use of non-exhaustible renewable energy in the energy mix because renewable energy does not generate any carbon dioxide (C[O.sub.2]). Nuclear power is an alternative source of clean energy. Indeed, more than one-quarter of electricity depends on nuclear power in the EU. However, the growth in nuclear power in the EU has uncertainties. First, low load factors (5,000-6,000 hours per year) and reduced wholesale prices of renewables input will make the nuclear plants disadvantaged. Second, some member states are strongly anti-nuclear. For this reason, this study concentrates on the share of renewable energy in the total energy consumption as a definition and measure of "energy transition". Renewable energy is derived from naturally replenished resources, including solar, wind, hydro, and geothermal.

   Uncertainty in energy markets has both direct and indirect effects on energy transition. First, its detrimental effects on economic activity slow energy transition due to investment concerns and the shift in energy changes. Uncertainty will hold decision makers from adopting other forms of energy more confidently. Second, uncertainty will slow the adoption of policies because of the influence of producers and consumers, who may be unwilling to change their technology or current energy use behavior. Energy transition is a complex social and economic transformation process involving the coordination of many actors (Blazquez et al., 2018).

   Energy market uncertainties may discourage all actors. Economic analysis has extensively studied the uncertainty in economic environment since the theoretical works of Friedman (1968) and Bernanke (1983). Later studies (see e.g. the recent studies: Durnev, 2010; Julio and Yook, 2012; Brogaard and Detzel, 2015; Handley and Limao, 2015) support the adverse economic effects of uncertainty in macroeconomic aggregate variables on economic growth and investment.

   But there are only a few studies focused on the economic effects of energy market uncertainty. Based on the studies of Bernanke (1983) and Pindyck (1991, 1999), an effect of uncertainty relates to changes in energy prices that can create uncertainty about the future of energy prices. This uncertainty effect causes consumers and producers to postpone irreversible investments. Additionally, there will be the effect of precautionary saving in response to an increase in the price of energy.

   In this study, we consider uncertainty relating to oil price shocks, supply and demand shocks in energy markets, and residual shocks to energy prices (1) in a panel data stochastic volatility model. We use data on 28 member states of the European Union (EU-28) from 1995-2015 to estimate the effect of uncertainty on energy transition. Our study contributes to the literature on energy transition in several ways.

   First, to the best of the authors' knowledge, there are no studies investigating the direct effect of energy market uncertainty on the energy transition. Previous studies were all focused on the effects of oil price volatility on economic activity. The effect of energy market uncertainty has significant consequences on economic growth as previously discussed (Mork, 1989; Mork et al., 1994; Mory, 1993; Lee et al., 1995; Hamilton, 1996, 2003), but whether it has any significant effect on energy transition is not investigated. Our study directly estimates this effect using a regression where the dependent variable is the percentage of renewable energy in the total primary energy production.

   Second, our study examines not only the effect of energy price shocks (oil and residual price shocks) but also estimates the effects of energy supply and demands shocks, while the previous research on the effect of energy price shock and energy price volatility has only considered the effect of price shocks (Hamilton, 1983; Burbridge and Harrison,1984; Gisser and Goodwin, 1986; Mork, 1989; Mork et al., 1994; Lee et al., 1995; Ferderer, 1996; Papapetrou, 2001; Jimenez-Rodriguez and Sanchez, 2005; Lardic and Mignon, 2006; Cunado and de Gracia, 2005; Balcilar et al., 2015).

   Third, our study also contributes to the literature by estimating energy market uncertainty using a panel stochastic volatility model with time varying parameters (PTVP-SV). Empirical studies in economics and finance generally use generalized autoregressive conditional heteroskedasticity (GARCH) models to estimate volatility as a measure of uncertainty. The volatility specification in GARCH models is deterministic and does not allow volatility shocks. Thus, volatility estimates from GARCH models is an improper measure of uncertainty. Compared to GARCH models, volatility in the SV model used in our study is specified as a latent stochastic process that allows volatility shocks, and, therefore, is a superior measure of uncertainty.

   Fourth, we allow a time-varying impact from the energy market demand, supply, and energy price uncertainties to energy transition. Energy markets are subject to frequent large changes leading to structural breaks, so assuming a constant parameter model leads to misspecification errors. We incorporate structural breaks or regime shift by specifying a time-varying parameter model. The models are estimated using the Bayesian particle marginal Metropolis Hastings (PMMH) algorithm.

   Fifth, in addition to energy market related uncertainties we examine effects of several other factors on renewable energy transition. We consider levels of C[O.sub.2] emission, energy demand, energy intensity, energy price, energy supply, energy tax ratio, heating degree days, oil price, and share of fossil fuel based energy in final energy consumption as factors that may affect energy transition. In addition to the PTVP-SV model, effects of these additional factors are also examined by estimating reduced form static and dynamic panel data models.

   The results show the vulnerability of energy transition to changing economic conditions. For instance, the subprime mortgage crises induced breaks in all parameters of the model relating to energy transition. Moreover, the empirical results indicate that countries with higher levels of C[O.sub.2] emission, energy demand, energy intensity, energy price, and fossil fuel based energy share adopt low carbon economy policies slower while countries with high levels of inland energy production, energy tax rate, and heating degree days have faster renewable energy transition. The results for the EU-28 contribute valuable information on policy-making and show the need for policies that increase the robustness of renewable energy markets, particularly related to energy prices and supply conditions.

   The rest of the paper is organized as follows. In Section 2, we examine the energy market uncertainty and the channels through which it affects energy transition. Section 3 discusses the energy transition in the EU. Section 4 presents the data and explains the methodology, while Section 5 presents the empirical results. Finally, Section 6 concludes with policy implications.

2. ENERGY MARKET UNCERTAINTY AND ITS EFFECTS

   Previous studies on energy market uncertainty examined either effects of oil price shocks or oil price uncertainty on economic activity. Energy market uncertainty, arising more broadly from energy demand, energy supply and energy price shocks, has a direct impact on households, firms, and policymakers who take the initiative for clean energy transition. It also has a significant effect on renewable energy investments and technology adaptation both by firms and consumers.

   However, ambiguities exist on whether energy market uncertainty helps energy transition or hampers it. Lieberman and Doherty (2008), McKillop (2009), Rentschler (2013), Klevnas et al. (2015), Fouquet (2016), and van de Ven and Fouquet (2017) argue that price volatility in fossil fuels may support energy transition because renewables may have much lower price volatility. However, renewable energy prices also have volatility and whether they have sufficiently low volatility to motivate a faster renewable transition is an empirical question (Rentschler, 2013).

   The literature, however, seems to recognize the oil price volatility effect on economic activity as the most relevant uncertainty effect. Based on this consensus, we first examine oil price uncertainty and its implications for the economy, which may hamper transition to less fossil fuel dependent economies. Economic theory suggests that the main channel through which the oil price...