Causal modeling and inference for electricity markets

How does dynamic price information flow among Northern European electricity spot prices and prices of major electricity generation fuel sources? We use time series models combined with new advances in causal inference to answer these questions. Applying our methods to weekly Nordic and German electricity prices, and oil, gas and coal prices, with German wind power and Nordic water reservoir levels as exogenous variables, we estimate a causal model for the price dynamics, both for contemporaneous and lagged relationships. In contemporaneous time, Nordic and German electricity prices are interlinked through gas prices. In the long run, electricity prices and British gas prices adjust themselves to establish the equilibrium price level, since oil, coal, continental gas and EUR/USD are found to be weakly exogenous.     

What explain the short-term dynamics of the prices of CO2 emissions?

Using a Bayesian Structural VAR (BSVAR), this paper analyzes the short-term dynamics of the prices of CO₂ emissions in response to changes in the prices of oil, coal, natural gas and electricity. The results show that: (i) a positive shock to the crude oil prices has an initial positive effect on the CO₂ allowance prices, which later becomes negative; (ii) an unexpected increase in the natural gas prices reduces the price of CO₂ emissions; (iii) a positive shock to the prices of the fuel of choice, coal, has virtually no significant impact on the CO₂ prices; (iv) there is a clear positive effect of the coal prices on the CO₂ allowance prices when the electricity prices are excluded from the BSVAR system; and (v) a positive shock to the electricity prices has a negative impact on the price of the CO₂ allowances. We also find that the energy price shocks have a persistent impact on the CO₂ allowance prices, with the largest effect occurring 6 months after a shock strikes. The effect is particularly strong in the case of the shocks to the natural gas and crude oil prices. Finally, the empirical findings suggest an important degree of substitution between the three primary sources of energy (i.e., crude oil, natural gas and coal), particularly when electricity prices are excluded from the BSVAR system.     

基于谱分析的国内外能源价格周期关系研究

为了更好地应对经济全球化背景下的能源价格问题,基于谱分析方法对国内外典型市场的石油、煤炭、电力、天然气四种主要能源价格的周期互动关系进行了初步研究与探讨,考虑到我国天然气与电力市场的发展现状,借鉴发展相对成熟的美国市场进行分析。结果表明,煤炭、电力、天然气三种能源市场价格均滞后于国际原油价格波动,其中煤炭价格波动略滞后于石油,而根据美国市场经验,天然气和电力价格波动周期均滞后于煤炭和石油市场价格。     

Electricity prices and fuel costs: Long-run relations and short-run dynamics

The paper examines the long-run relation and short-run dynamics between electricity prices and three fossil fuel prices – coal, natural gas and crude oil – using annual data for the U.S. for 1960–2007. The results suggest (1) a stable long-run relation between real prices for electricity and coal (2) Bi-directional long-run causality between coal and electricity prices. (3) Insignificant long-run relations between electricity and crude oil and/or natural gas prices. And (4) no evidence of asymmetries in the adjustment of electricity prices to deviations from equilibrium. A number of implications are addressed.     

Price discovery in energy markets

In this study, we empirically analyze the price discovery process in the futures and spot markets for crude oil, heating oil and natural gas using daily closing prices. We use two different information share measures that are based on the methods proposed by Gonzalo and Granger (1995) and Lien and Shrestha (2014). Both measures indicate that almost all the price discovery takes place in the futures markets for the heating oil and natural gas. However, for the crude oil, the price discovery takes place both in the futures and spot markets. As a whole, our study indicates that futures markets play an important role in the price discovery process.     

High gas dependence for power generation in Thailand: The vulnerability analysis

Thailand uses 74% of its natural gas supply for power generation and 70% of its power comes from gas-based technology. High dependence on natural gas in power generation raises concerns about security of electricity supply that could affect competitiveness of Thai manufacturing and other industries at the global level. The effect of fuel dependence on security of electricity supply has received less emphasis in the literature. Given this gap, this research examines the economic impact of high dependence on natural gas for power generation in Thailand by analyzing the effect of changes in fuel prices (including fuel oil and natural gas) on electricity tariff in Thailand. At the same time, the research quantifies the vulnerability of the Thai economy due to high gas dependence in power generation. Our research shows that for every 10% change in natural gas price, electricity tariff in Thailand would change by 3.5%. In addition, we found that the gas bill for power generation consumed between 1.94% and 3.05% of gross domestic product (GDP) between 2000 and 2004 and in terms of GDP share per unit of energy, gas dependence in power generation is almost similar to that of crude oil import dependence. We also found that the basic metal industry, being an electricity intensive industry, is the most affected industry. Additionally, we find that volatility of gas price is the main factor behind the vulnerability concern. The research accordingly simulates two mitigation options of the problem, namely reducing gas dependence and increasing efficiency of gas-fired power plants, where the results show that these methods can reduce the vulnerability of the country from high gas dependence in power generation.     

Correlation between Chinese and international energy prices based on a HP filter and time difference analysis

To establish a reasonable system and mechanism for Chinese energy prices, we use the Granger causality test, Hodrick–Prescott (HP) filter and time difference analysis to research the pricing relationship between Chinese and international energy prices. We find that Chinese and international crude oil prices changed synchronously while Chinese refined oil prices follow the changes of international oil prices with the time difference being about 1 month to 2 months. Further, Australian coal prices Granger causes Chinese coal prices, and there is a high correlation between them. The U.S. electricity price is influenced by the WTI crude oil price, the U.S. gasoline price and the HenryHub gas price. Due to the unreasonable price-setting mechanism and regulation from the central government, China′s terminal market prices for both electricity and natural gas do not reflect the real supply–demand situation. This paper provides quantitative results on the correlation between Chinese and international energy prices to better predict the impact of international energy price fluctuations on China′s domestic energy supply and guide the design of more efficient energy pricing policies. Moreover, it provides references for developing countries to improve their energy market systems and trading, and to coordinate domestic and international energy markets.     

The dynamic linkages between crude oil and natural gas markets

The time varying price spillovers between natural gas and crude oil markets for the period 1994 to 2014 are investigated. Contrary to earlier research, we show that in a large part of our sample the natural gas price leads the price of crude oil with price spillover effects lasting up to two weeks. This result is robust to a battery of tests including out-of-sample forecasting exercises. However, after 2006, we detect little price dependencies between these two energy commodities. These findings arise due to a conjunction of both demand and supply-side shocks arising from both natural and economic events, including Hurricane Katrina, the Tohoku earthquake and the Global Financial Crisis, as well as infrastructure and technological improvements. The increased use of new technologies such as hydraulic fracking for the extraction of gas and oil in particular affected supply in the latter part of the study. We conclude that the long term relation present in the early part of the sample has decoupled, such that price determination of these two energy sources is now independent.     

Energy prices and CO2 emission allowance prices: A quantile regression approach

We use a quantile regression framework to investigate the impact of changes in crude oil prices, natural gas prices, coal prices, and electricity prices on the distribution of the CO₂ emission allowance prices in the United States. We find that: (i) an increase in the crude oil price generates a substantial drop in the carbon prices when the latter is very high; (ii) changes in the natural gas prices have a negative effect on the carbon prices when they are very low but have a positive effect when they are quite high; (iii) the impact of the changes in the electricity prices on the carbon prices can be positive in the right tail of the distribution; and (iv) the coal prices exert a negative effect on the carbon prices.     

Horizontal and vertical transmissions in the US oil supply chain

Oil prices, inventory levels, and utilization rates are influenced by changes that are transmitted horizontally and/or vertically through the energy supply chain. We define horizontal transmissions as changes that are generated by linkages among fuels at a similar stage of processing while vertical transmissions are changes that are generated by upstream/downstream linkages in the oil supply chain. Here, we investigate vertical and horizontal transmissions by estimating vector error correction models (VECMs) that represent relationships among the price of crude oil, US refinery utilization rates, US stocks of crude oil, US stocks of motor gasoline, the US price of motor gasoline, and the US price of a substitute fuel, natural gas. Causal relationships estimated from both weekly and quarterly observations indicate that the price of crude oil is an important gateway for disturbances to the oil supply chain. Impulse response functions indicate that disturbances to crude oil prices ripple down the oil supply chain and affect inventory behaviors, refinery utilization rates, and the price of motor gasoline, and are transmitted laterally to the natural gas market.     

Natural gas development in Mexico

Mexico, with a very large manufacturing sector, has a well-developed domestic industrial market for natural gas. Gas accounts for 45% of all energy consumed in Mexican industry. Pemex, the national oil company, is the largest natural gas consumer in the country, using gas as a fuel for oil field operations, pipelines and refineries, and as a fuel and feedstock for the production of fertilizer and petrochemicals. Gas also is heavily used in electricity generation and in manufacturing steel, cement, paper, and other products. Continued rapid growth is anticipated in all of these sectors. In the short run, no additional gas can be made available for export, since gas processing facilities in the southern pan of the country, as well as the trunkline to northern Mexico, are already operating at capacity. Longer run availability of gas for export will depend on export price in comparison with the value of gas in domestic uses. An export price tied to residual fuel oil prices may be inadequate to stimulate additional Mexican gas exports. Current gas development priorities for Mexico include the elimination of flaring, the enhancement of gas-conservation efforts, the restriction of boiler fuel consumption, and the rationalization of domestic gas prices so that the differential over heavy fuel oil prices is increased.     

How competitive are EU electricity markets? An assessment of ETS Phase II

This paper studies the interactions between electricity and carbon allowance prices in the year-ahead energy markets of France, Germany, United Kingdom and the Nordic countries, during Phase II of the EU ETS. VAR and Granger-causality methods are used to analyze causal interfaces, whereas the volatility of electricity prices is studied with basic and asymmetric AR-GARCH models. Among the main results, the marginal rate at which carbon prices feed into electricity prices is shown to be ca. 135% in the EEX and Nord Pool markets, where electricity and carbon prices display bidirectional causality, and 109% in the UK. Therefore, generators in these markets internalized the cost of freely allotted emission allowances into their electricity prices considerably more than the proportionate increase in costs justified by effective carbon intensity. Moreover, electricity prices in France are found to Granger-cause the carbon price. This study also shows how European electricity prices are deeply linked to coal prices among other factors, both in terms of levels and volatility, regardless of the underlying fuel mix, and that coal was marginally more profitable than gas for electricity generation. EU policies aimed at increasing the carbon price are likely to be crucial in limiting the externalities involved in the transition to a low-carbon system.     

Quantifying multiscale inefficiency in electricity markets

One of the basic features of efficient markets is the absence of correlations between price increments over any time scale leading to random walk-type behavior of prices. In this paper, we propose a new approach for measuring deviations from the efficient market state based on an analysis of scale-dependent fractal exponent characterizing correlations at different time scales. The approach is applied to two electricity markets, Alberta and Mid Columbia (Mid-C), as well as to the AECO Alberta natural gas market (for purposes of providing a comparison between storable and non-storable commodities). We show that price fluctuations in all studied markets are not efficient, with electricity prices exhibiting complex multiscale correlated behavior not captured by monofractal methods used in previous studies.     

The production of crude oil and natural gas

This paper is directed at examining the impact of changing prices on the level of production of crude oil and natural gas in the United States. By using a cross-correlation test for unidirectional causality it is clearly demonstrated that, for both crude oil and natural gas, domestic production is affected by changing prices. The implications are clear. The decontrol of the price of crude oil and the deregulation of natural gas prices will lead to additional production in the near term.     

Estimating the value of electricity storage in PJM: Arbitrage and some welfare effects

Significant increases in prices and price volatility of natural gas and electricity have raised interest in the potential economic opportunities for electricity storage. In this paper, we analyze the arbitrage value of a price-taking storage device in PJM during the six-year period from 2002 to 2007, to understand the impact of fuel prices, transmission constraints, efficiency, storage capacity, and fuel mix. The impact of load-shifting for larger amounts of storage, where reductions in arbitrage are offset by shifts in consumer and producer surplus as well as increases in social welfare from a variety of sources, is also considered.     

Hard coal for energetic purposes: price–quality relationships; international coal market observations and Polish practice

Coal is the most abundant and commonly used energy carrier in the world. In coal-producing countries, coal is often the cheapest fuel for electricity and heat production. Prices of steam coals offered by exporters on international markets reflect current economic and market conditions and are also related to the prices of other fossil fuels like crude oil and natural gas. International coal-market observations and analyses lead to the conclusion that steam-coal prices depend only on heating value. In Polish practice, steam-coal prices are calculated using a price formula in which coal price is a function of three quality parameters: net calorific value, ash content and sulphur content, and a price of ‘basic’ or ‘reference’ coal (which means: coal of defined quality). This paper presents the results of international coal-market analyses of relationships between coal price and quality and describes the Polish coal-pricing system. A new solution, relevant to domestic coal mines and power plants, is presented to improve and simplify the conditions of bilateral settlements of coal deliveries.     

The impact of price variation on the production of crude oil and natural gas

This paper is directed at examining the impact of changing prices on the level of production of crude oil and natural gas in the United States. By using a cross correlation test for unidirectional causality, it is clearly demonstrated that for both crude oil and natural gas, domestic production is affected by changing prices. The implications are clear. The decontrol of the price of crude oil and the deregulation of natural gas prices will lead to additional production in the near term.     

Revisiting long-run relations in power markets with high RES penetration

Electricity generation from renewable energy resources (RES) has become increasingly significant to reach EU and emissions reduction targets. At the same time, one of the main EU policy goals has been the creation of a common internal energy market for Europe. In this paper, we focus on these two issues previously studied separately, considering their possible interactions. We first analyze the long-run relationship between day-ahead electricity prices and fuel prices (natural gas and coal) looking at two samples of years characterized by low and high RES penetration, then we explore the integration of EU markets.We show that the electricity–fuel nexus found over 2006–2008 changed dramatically over 2010–2014 for the majority of countries considered. In particular, the long-run dependence of electricity from gas and coal prices is much lower in recent years. Furthermore, our results confirm that the considered EU countries are becoming less integrated as RES-E increases. Our findings suggest that nationally implemented policies to support renewables are successful in increasing RES penetration, but they have lessened the linkage among EU markets, then making integration more difficult to obtain.     

A long-term view of worldwide fossil fuel prices

This paper reviews a long-term trend of worldwide fossil fuel prices in the future by introducing a new method to forecast oil, natural gas and coal prices. The first section of this study analyses the global fossil fuel market and the historical trend of real and nominal fossil fuel prices from 1950 to 2008. Historical fossil fuel price analysis shows that coal prices are decreasing, while natural gas prices are increasing. The second section reviews previously available price modelling techniques and proposes a new comprehensive version of the long-term trend reverting jump and dip diffusion model. The third section uses the new model to forecast fossil fuel prices in nominal and real terms from 2009 to 2018. The new model follows the extrapolation of the historical sinusoidal trend of nominal and real fossil fuel prices. The historical trends show an increase in nominal/real oil and natural gas prices plus nominal coal prices, as well as a decrease in real coal prices. Furthermore, the new model forecasts that oil, natural gas and coal will stay in jump for the next couple of years and after that they will revert back to the long-term trend until 2018.     

Price dynamics of natural gas and the regional methanol markets

A ‘methanol economy’ based mainly on natural gas as a feedstock has a lot of potential to cope with the current and ongoing concerns for energy security along with the reduction of CO-2 emissions. It is, therefore, important to examine the price dynamics of methanol in order to ascertain whether the price of methanol is mainly natural-gas-cost driven or demand driven in the context of different regions. This paper is the first attempt to investigate the following: (i) is the natural gas price significantly related to the regional methanol prices in the Far East, United States and Europe? (ii) who drives the regional methanol prices? The paper is motivated by the recent and growing debate on the lead-lag relationship between natural gas and methanol prices. Our findings, based on the most recently developed ‘long-run structural modelling’ and subject to the limitations of the study, tend to suggest: (i) natural gas price is cointegrated with the regional methanol prices, (ii) our within-sample error-correction model results tend to indicate that natural gas was driving the methanol prices in Europe and the United States but not in the Far East. These results are consistent, during most of the period under review (1998.5–2007.3), with the surge in demand for methanol throughout the Far East, particularly in China, Taiwan and South Korea, which appears to have played a relatively more dominant role in the Far East compared to that in Europe and the United States within the framework of the dynamic interactions of input and product prices. However, during the post-sample forecast period as evidenced in our variance decompositions analysis, the emergence of natural gas as the main driver of methanol prices in all three continents is consistent with the recent surge in natural gas price fueled mainly, among others, by the strong hedging activities in the natural gas futures/options as well as refining tightness (similar to those that were happening in the crude oil markets).