Journal of Energy Markets

Welcome to the second issue of Volume 11 of The Journal of Energy Markets.

In the issue’s first paper, “An analysis of intraday market response to crude oil inventory shocks”, Ziyuan Li and Helyette Geman investigate the intraday market activity of West Texas Intermediate (WTI) crude oil futures around the release of information from the US Energy Intelligence Agency. This indicates how prices respond to inventory shocks. The paper also examines the impacts of belief dispersion, calendar effects and the movement of prices between the releases of American Petroleum Institute and Energy Intelligence Agency reports. Market activity, in terms of price returns, volatility and trading volume, responds to inventory shocks very quickly, with the effects lasting for about twenty-five minutes. Li and Geman’s findings suggest that a positive (negative) inventory shock will result in an immediate price decline (rise), while both kinds of shock increase volatility and trading volume; however, the price reverts quickly after the initial reaction. Moreover, wider belief dispersion is associated with a larger market response to inventory shock. From this basis, the authors discuss various intraday trading strategies on Energy Intelligence Agency report days, documenting the benefits of high-frequency advance information in particular.

Our next paper, “Does the impact of exchange-traded funds flows on commodities prices involve stockpiling as a signature? An empirical investigation” by Steve Ohana and Xiaoying Huang, is also related to the dynamics of price formation. It analyzes the proposition that any abnormal impact of speculators on commodities prices should involve stockpiling. This is in contrast to an alternative perspective that, due to the price inelasticity of supply and demand in commodity markets, speculation could distort commodity prices without causing any change in inventories. Motivated by this debate, the paper examines the relationship between the investment flows into the three main commodity index exchange-traded funds and the prices, inventories and term structures of four US-traded energy commodities. Using weekly inventory data from the Energy Intelligence Agency and futures prices from NYMEX energy contracts, the authors do not find any significant relation between commodity index flows and inventory or term structure. By contrast, they retrieve the short-term impacts of index flows on energy commodities’ futures prices that have already been evidenced in the literature. An extension of this approach to twelve US-traded agricultural contracts confirms these conclusions. Hence, the authors’ results suggest that stockpiling is not necessarily a “signature” of an abnormal impact of speculators on commodities prices.

Looking specifically at electricity price formation, in our third paper, “The Iberian electricity market: analysis of the risk premium in an illiquid market”, Márcio Ferreira and Helder Sebastião analyze the risk premium in the base-load monthly futures contracts traded on the Iberian electricity market (MIBEL) between July 1, 2006 and March 31, 2017. During this period, the ex post risk premium on the last trading day presented on average a value of -5.77%, with negative skewness, excess kurtosis and some persistence. The risk premium depended on the season of the year, with the absolute value for the winter futures being more than five times higher than for the summer futures. The absolute risk premium and its volatility decreased nonlinearly throughout the remaining trading days until maturity. The sequence of futures prices approaching maturity showed some predictive power as regards the risk premium. The futures price path between seven and three days prior to delivery explained around 28% of the variability in the risk premium, and there is some evidence that this information can be used to successfully forecast the risk premium signal.

The issue’s final paper, “Improving the Brazilian electricity market: how to replace the centralized dispatch by decentralized market-based bidding” by Felipe A. Calabria, João Tomé Saraiva and A. P. Rocha, looks at market redesign aspects for Brazilian wholesale electricity. This paper proposes the introduction of a virtual reservoir model. To simulate the behavior of this proposed innovation, the authors implement an agent-based model using a reinforcement learning algorithm, simulated annealing and linear programming. These simulations are based on real data from the Brazilian power system, encompassing more than 98% of the total hydro installed capacity and three years’ worth of market data. The authors’ results indicate that responsibility for the management of (virtual) reservoirs might be assigned to each hydro company, which could then be allowed to operate according to their own risk preferences, while maintaining current levels of efficiency and security as well as reducing prices. Overall, this issue provides four important papers related to behavioral aspects of energy markets, as informed by empirical analysis and simulation. Behavioral complexity is one of the distinguishing features of this sector and, as a consequence, determines why detailed quantitative research remains essential for both the advancement of our insights and the improvement of performance in practice.

Derek W. Bunn
London Business School