Energy Economics, 2015, 49, 350-358. Liyan Han, Yimin Zhou, Libo Yin*
With the importance of agriculture prices in the economy increasing over the past decades, price links between energy and agriculture futures have attracted considerable attention from global investors (Jebabli et al., 2014; Apergis et al., 2014). A number of empirical studies document that these price links are affected by exogenous factors, such as the financialization of commodities, biofuel policies, food crises and the global financial crisis (see Kristoufek et al., 2014; Nazlioglu et al., 2013; Baffes and Haniotis, 2010). However, there is still no consensus about the consistency of energy and agricultural commodity prices. Because the possibility of multiple states plays an important role in investigating dependence among financial assets (Hamilton, 1988; Gray, 1996), considering dependence in the underlying states is a potential misconception in existing studies of energy and agriculture links.
Therefore, the goal of this paper is to identify the impacts of these exogenous shocks on the dependence structure of energy and agricultural commodity prices. The dependence structure here does more than describe dependencies during different critical periods; instead, it, it emphasizes investigating the correlations among the levels of their underlying states. To achieve this goal, we utilized multivariate normal mixture models, which have been used by Ang and Bekaert (2002), to accommodate the structure of dependence between energy and agriculture commodity prices during different sub-periods of the last decade.
The main contribution of this paper consists of two parts. First, we show that a mixture of multivariate normal distributions provides the underlying structural features of the correlations, whereas other models of dependence, such as copula or multivariate GARCH models, do not. Consistent with Gray (1996) and Ang and Bekaert (2002), the evidence confirms the existence of multiple states and suggests that the state-based structure is likely an important dimension when considering price links between financial assets. In addition, our method of calculating the overall correlation relaxes the linearity inherent in Pearson’s method while remaining tractable and easy to estimate.
Second, we analyze our results from both quantitative and structural perspectives. The empirical results show that abrupt changes in the overall correlations and significant structural heterogeneity in the underlying states coexist only during the financial crisis and suggest that the global financial crisis is the most influential exogenous shock to energy agriculture dependence. In addition, supportive evidence is also found in the dependence features of the tails. For future studies, it is of interest to investigate dependence by introducing underlying states to determine whether investors can benefit from portfolio diversification between the energy and agriculture markets during times of turmoil.
The main conclusions can be summarized as follows. First, we identify the most significant variation in the overall correlations during the financial crisis of 2008. This result demonstrates the worldwide phenomenon of enhanced dependence among financial assets during the financial crisis of 2008.
Second, by investigating the underlying structure, we find the existence of heterogeneity in the price links between energy and agricultural commodities during the biofuel policy and recent financial crisis periods. In addition, the dependence features obtained from the tails provide supportive evidence for the underlying structural analysis. Therefore, the underlying structure is a potential dimension when we consider price dependencies among commodity assets.
Finally, based on quantitative and structural analyses, we conclude that the global financial crisis is the most influential exogenous shock on energy-agricultural price links. Because heterogeneity in dependence is likely to exist during a financial crisis, knowledge of the heterogeneity of relationship in underlying states can assist regulatory authorities in determining suitable effective policies to guard against large financial shocks. Additionally, it can also guide investors in risk management and portfolio diversification across energy and agriculture markets according to circumstances, especially during times of turmoil.
