Energy and Agricultural Commodity Markets Interaction: An Analysis of Crude Oil, Natural Gas, Corn, Soybean, and Ethanol Prices.

• Author: Chiou-Wei, Song-Zan

1. INTRODUCTION

   This paper empirically studies the connections between agricultural and energy commodity markets with a focus on five commodities--crude oil, natural gas, corn, soybean, and ethanol, hoping to shed light on how these markets are related in terms of price level and volatility in a more recent period. The study intends to provide more updated empirical evidence on whether or to what degree these commodity markets are connected, and whether these connections, if there are any, make any sense based on economic explanation of the linkages among the markets.

   We believe the study is timely as there is a growing literature on the interaction of energy and agricultural commodity markets. These new and renewed interests have been triggered by at least two recent events in the commodity markets. One such event is the phenomenal increase in the volume of trading in these markets during the period from 2000 to 2008. Cevik and Sedik (2011) suggested that such events as the global financial crisis and the commodity price volatility during the period have heightened interest in the dynamic relationships among the commodity markets. A common theme of the explanation of this growth in virtually all commodity markets during the period is financialization of the commodity markets (Henderson et al., 2015; etc.). Frankel (2014) presented some evidence that supported the argument that both speculation and easy monetary policy had contributed to commodity price changes and their volatilities. Gozgor et al (2016) found some evidence of financialization in the corn and soybean markets. Sockin and Xiong (2015) suggested that noises in futures trading of the commodities would feed back into the demand of commodity producers and the producers had difficulties to tell whether the change was due to financial trading or global demand change due to information frictions. Cheng and Xiong (2014) found evidence to support the claim that financialization had fundamentally changed the functions of the risk sharing and price discovery of the commodity markets. While these studies confirmed to a degree that financialization in general and speculation in particular had affected the commodity markets as a whole, not all studies agreed that financialization or speculation had a leading effect on sharp increases in commodity prices (for example, the case of oil in Knittel and Pindyck (2016)). Even though Knittel and Pindyck (2016) did not directly address the issue of connections between oil and commodity markets, their results implied that speculation could not be the main factor that drove the co-movement of the commodity prices.

   The other factor behind the renewed interest in the connections between energy and agricultural commodity markets is the worldwide ethanol policy, especially in the U.S., which called for cleaner and lower-cost ethanol to replace traditional hydrocarbons such as oil. The Renewable Fuels Standard (RFS2) as established by the 2007 Energy Independence and Security Act requires an increasingly larger number of renewable sources of fuel including biofuels into the fuel mix. Other countries especially Brazil and European Union also ramped up the production of biofuels to supplement gasoline in especially transportation. Serra and Zilberman (2013) had a nice review of the literature on the bio-fuel related price transmission for the energy and agricultural markets with focuses on mainly three commodities--crude oil, ethanol, and corn. Serra and Zilberman (2013) pointed out that even though the results are mixed, most of price transmissions went from oil market to ethanol market and then to corn market. In addition, many studies focused on price transmissions and there was a need for more studies on volatility transmissions. Furthermore, most of the studies only examined the transmission mechanism in the context of time series models without incorporating economic variables. To address this last point, we include several relevant economic variables to control for any possible effects of these variables on returns and volatilities of the commodity prices to better understand how shock from one market affects the others. By directly controlling for these variables, we can directly measure how these economic variables have impacted the commodity prices and better capture the shocks in the commodity prices to study the interactions. In addition, we expand the number of energy and agricultural commodity variables.

   The energy and agricultural markets can be interrelated due to the same set of economic forces that influence them and to unavoidable cross-market arbitrage activities (de Gorter et al., 2008). Agricultural production processes use energy products such as oil and natural gas; thus, energy prices directly and indirectly affect the input and transportation cost of the agricultural products. The increase in oil and gas prices generates an incentive to use biofuels and other alternative energy sources; thus, an increase in the price of biofuels such as ethanol would also increase the prices of some agricultural products, namely food prices, directly. In addition, economic policies concerning biofuels can directly or indirectly strengthen or weaken these relationships. This discussion suggests that relationships among the markets are time-varying and dynamic.

   Economic forces have also been cited in the literature to have caused the commodity markets to move together. These factors include economic activities, speculation (inventories have been used to bear on the role of speculations despite of some unsettlements on this issue), and monetary policy (Frankel (2014)). In addition to some of these macroeconomic variables, Gilbert (2010) also suggested that dollar exchange rate would be a factor that could influence commodity prices as originally analyzed by Ridler and Yandle (1972). Gozgor and Kablamaci (2014) also suggested that speculation and financialization could be the driving factors behind the co-movement of the commodity prices. Even though this paper does not directly address the role of speculation in the movement of the commodity prices, by incorporating these exogenous economic variables directly, our study helps to understand the possible co-movement of the commodity prices and to what degree these factors could directly influence the commodities.

   To thoroughly understand the price level and volatility connections among these markets is important in several aspects. Most motivations for these studies have focused on the importance of food markets in an economy. This is an especially important issue for developing countries, as food makes up a greater share in their spending than in more developed countries, so problems in the food sector can threaten the food security of those countries and undermine their food price stabilization policies. Due to the substitutability of foodstuffs, if there is a close relationship between the price level and volatility of the agricultural and energy commodities, then the higher price and volatility in the corn and soybean markets due to energy market price volatility could spill over to other agricultural sectors such as wheat and so on. In addition, the co-movement of energy and commodity prices can make financial diversification programs less effective. This could be an increasingly important issue, as increasingly agricultural commodities have been included in financial investment portfolios. Furthermore, the increase in price volatility of agricultural commodities could increase the cost of risk management programs and thus the cost of foodstuffs in general. Consequently, it is extremely important to understand the extent of energy and agricultural product price interactions and, in particular, the transmission of price volatility between the markets.

   We study five energy and agricultural commodities: oil, natural gas, ethanol, corn, and soybean and these commodities are selected based on the linkages amount them. It is generally accepted that oil and gas are highly connected as they are both substitutable to a certain degree in technology. Both oil and gas are also associated in the extraction and production process. However, despite of the association in production, substitutability in consumption may be limited to a certain degree. In addition, oil and gas are driven by different sets of variables. Therefore, the price and volatility connection between oil and gas may be time dependent. Among the five variables we consider, ethanol is the variable that connects the energy and commodity markets explicitly as ethanol is used as an alternative to traditional energy. We include corn and soybean in the system as ethanol can be extracted from corn as well as soybeans even though the extraction of ethanol from soybean is to a much lesser degree. The agricultural commodity markets can also be connected to the energy market through other links as oil and natural gas can be the energy used in the agricultural commodity production process.

   We study these interactions for separate markets--spot and futures. Even though we will not be able to directly measure the feedback effect of the futures market on the spot market, the existence/lack of possible difference in the interaction patterns in the spot and futures markets may provide additional information about the energy and agricultural commodity market connections. As spot market is more influenced by physical supply of and demand for the commodity, the differentiation of the impacts from both the spot and futures markets may be an indication that the financial (futures) markets may be impacted differently by the financialization than the spot market.

   This paper utilizes the multivariate GARCH (MGARCH) approach to study the dynamics and cross-dynamics of price and price volatility in oil, natural gas, ethanol, corn, and soybean markets for the period from 2005 to 2017. The price interactions are captured by the...