Investigating the Price Transmission Mechanism of the Greek Fresh Tomato Market with a Markov Switching Vector Error Correction model

Abstract

AbstractThe present study investigates the price transmission mechanism between producer and consumer prices in the Greek fresh tomato market, using monthly price data from January 1995 to May 2011. The estimation is carried out by applying a Markov Switching Vector Error Correction model. The results indicate that there are causality and leadership relationships between producer and consumer prices in the short and in the long run. Finally, a multinomial logit model is utilized in order to determine the factors that affect the switching of the price transmission mechanism.Keywords: price transmission, tomato, Greece, Markov, logit.(ProQuest: ... denotes formulae omitted.)1. IntroductionPrice is vital for every market, as it can coordinate efficiently the decisions of producers and consumers. The transactions in a market take place according to the price of the market while producers try to maximize their profit and consumers their utility. Moreover, when conditions of perfect competition exist in a market, the price mechanism leads to optimal allocation of scarce resources. On the contrary, when imperfect competition is observed, the most powerful participants of the market benefit by acquiring part of the surplus of the rest of the participants in the market (Brummer et al, 2009).In this paper, the price transmission mechanism between producer and consumer in the Greek fresh tomato market will be examined. Tomatoes have been selected as they comprise one of the most important agricultural products in Greece. Between 1995 and 2006, fresh tomato production accounted for about 20% to 23% of Greek vegetable production. Moreover, fresh tomatoes are produced throughout Greece. Most tomatoes (60 to 65%) are grown in the open, whereas 35% to 40% are grown in greenhouses. Peloponnesus constitutes the biggest producer of fresh tomatoes grown in the open, whereas Crete is the biggest producer of fresh tomatoes grown in the greenhouse. Dur- ing the period under examination, a decrease in the production of tomatoes in the open can be observed in contrast to an increase in the production of tomatoes in greenhouses. Moreover, in 1997, the greenhouse production of Crete surpassed the production of Peloponnesus tomatoes grown in the open for the first time. For the rest of the period, the production of Crete remained larger than the production of Peloponnesus. At the same time, from 1995 to 2006, exports of fresh tomatoes did not surpass the 1% of the total fresh tomato production and remained stable. On the other hand imports were observed to be increasing, from 1% in 1995 to 3% in 2006. However, it can be said that the Greek tomato market is self sufficient. With regard to consumers, tomatoes are of great importance, as they constitute an essential element of their nutrition. Tomatoes are consumed throughout the year, even during winter. So, their price matters for consumers as much as it matters for producers.A wide range of empirical methods have been developed in order to study price transmission and market integration. Following Brummer et al. (2009) the empirical methods developed so far can be divided into three categories. The first category includes the study of the correlations between the price series. The second category includes the co-integration methods which gave economists the opportunity to distinguish the spurious from the non spurious relationships between the price series. The third category includes non linear models that allow for state specific behavior in the price transmission mechanism according to the state of a transition variable. The non linear models are: 1) The Threshold Vector Error Correction (TVEC) models, in which the state shifts are determined by the size of the error correction term relative to the value of a threshold. 2) The Asymmetric Price Transmission (APT) models where the state shifts depend on whether prices are increasing or decreasing. 3) The Switching Regression models which allow for state specific behavior in the co-integrating vector. …