Price Dependence and Asymmetric Responses between Coffee Varieties

1.IntroductionCoffee is a drink brewed from the seeds of the Coffea genus.1 As a world commodity, coffee is second only to oil. Coffee is mainly produced by developing countries. A large part of the agricultural sector in these economies is involved in the production as well as in the industrial procession of coffee (Talbot, 2004). Economic policies and structural reforms of coffee production and trade are of major importance for many countries in Southern America, Africa and Southern Asia (Russell et al., 2012). Furthermore, political aspects of "coffee policies" are not to be ignored as well (Paige, 1997) since the commodity of coffee is mainly produced by politically unstable countries that suffered dictatorships and political mismanagement for years. On the other hand, coffee production is related to various environmental problems connected with deforestation and land misuse. Sustainable agricultural methods and environmental friendly production and procession might help farmers and workers in the coffee producing countries (Kilian et al., 2006).In 1989, the International Coffee Agreement (ICA) (Akiyama and Varangis, 1990; Ponte, 2002) broke down. As a result, market liberalization policies have allowed producing countries to abandon centralized planning of coffee production levels (Bacon et al., 2008). Hence, coffee production is not regulated by any state or international organization and one can claim that it is a free market regulated mainly by private interests. The market liberalization in several coffee producing countries allowed several researchers to test the Law of One Price (LOP) hypothesis. Under the LOP assumption it is expected that the commodity of coffee (or any other commodity market) is integrated enough to allow price co-movements without asymmetries. However, there is some evidence of the opposite fact. Ghoshray (2009) found evidence of asymmetries in price adjustment between different coffee milds.Coffee quality affects sensory preferences of coffee consumers (Walsh et al., 2011; Yoon and Park, 2012). There are four different coffee varieties: "Colombian Arabicas milds" (CO), 'Brazilian and other natural Arabicas" (BN), 'Other mild Arabicas" (RO) and Robustas (RO).2 The first three, namely Colombian, Brazilian and Others, are of the Arabicas quality. Arabicas coffee beans are considered to be of higher quality than Robustas. Figure 1 displays the main coffee producing countries.Coffee prices are strongly determined by the quality of coffee (Donnet et al., 2008; Wilson and Wilson, 2014) and have shown considerable volatility in the past. Coffee production is relatively sensitive to weather conditions while coffee consumption has relative inelastic demand. A recent research on this topic (Ubilava, 2012) has revealed that El Nino Southern Oscillation (ENSO) influences coffee prices. Moreover, there is strong evidence that price dynamics have non-linear characteristics and there is evidence of asymmetries in price transmission at the farm/wholesale/retail levels (Mehtaa and Chavasb, 2008).In addition, as with many other primary commodities, the global coffee market has been related by high volatility. Several sectors of the food industry are sensitive to prince transmission processes along the production, manufacturing and retail chains (Bakucs et al., 2014; Vavra and Goodwin, 2005) There is cumulative evidence that price transmissions can be asymmetric in many cases and that these asymmetries can be affected by several economic factors. Cointegration and metaregression analysis have been routinely applied in the past to test for asymmetric price transmissions in a variety of commodities (Frey and Manera, 2007). There is strong evidence in the literature that several type of asymmetries exist and various econometric models can capture special sub-cases of asymmetries under certain conditions (Bakucs et al., 2014; Frey and Manera, 2007; Meyer and Cramon-Taubadel, 2004; Rapsomanikis et al. …