price volatility spillover effect in iran's poultry and livestock market

introduction: the relationship between different market levels is an essential issue in economy. understanding of linkages between different market levels will help to assess the potential impact of agricultural policies. given the importance of the vertical market relationship, the present study examines price volatility spillover in vertical market levels of iranian livestock and poultry market prices. for this, we use monthly returns series of broiler feed, chicken, broiler (as substitutes for broiler vertical market levels), hay, sheep, mutton (as substitutes for mutton vertical market levels), hay, calf and beef (as substitutes for beef vertical market levels) during the period of april 1997 to march 2014. another important aspect that is considered in this study is the volatility regime switching. many variables undergo events that seem their time series’ behavior has changed quite dramatically. such structural breaks are mainly observed in the economic and financial time series data. the regime switching must not be totally considered as a predictable and deterministic event. as if the process has changed in the past, it could obviously change in the future too. therefore, it should be considered as a random variable and hence a full-time series model will include probabilistic inference about switching from one regime to another regime. hence, by doing a present study, we will be able to answer the questions such as: whether a significant regime change happened in livestock and poultry vertical market levels? is there any significant volatility spillover in vertical market levels? is there any difference between the volatility spillovers in high and low volatility regimes? to what extent the price volatility spills over the vertical markets? materials and methods: a multivariate markov switching model, that is best for our study aims, has been introduced by haas and mittnik (2008) to the volatility spillover literature which is a generalization of haas et al. (2004) univariate model. in fact, it is a multi-regime version of bollerslev et al. (1988) vech model. regime depended on variance matrices are defined by equation below: aij, i = 0, …, q, and bij, i = 1, …, p, are parameter matrices. index j is determining the volatility regime. this model does not directly estimable and to ensure positive definite covariance matrices some limitation should be imposed. for this purpose, haas and mittnik (2008) used engle and kroner (1995) proposed approach as: where are lower triangular matrices and aij* and bij* are m  m parameter matrices that must be estimated (hess and mittnik, 2008). the aij* elements are the coefficients that explain the effect of volatility shocks or news and the bij* elements are the coefficients that explain the effect of past volatilities on the current volatility of prices. results and discussion: the results indicate the existence of two volatility regimes in vertical market levels of all three studied markets and also successive switches of volatility regimes, especially for meat (mutton and beef) market. according to evidence obtained in this study, although the durability of low volatility regime is lower than the high volatility regime in all cases and unconditional probability of staying in the high volatility regime is lower, but still the number of months and the length of periods in the high volatility regime are not in acceptable ranges. based on the results, different shock and volatility spillovers between the different levels of the three markets have been occurring in both regimes; although the spillovers in high volatility regimes were more severe. conclusion: price spikes like those we have witnessed for iranian poultry and livestock products in recent years are not just part of a trend of higher prices. they are also part of a different phenomenon, price volatility, as its presence can be proved from our findings of broiler feed, chicken, broiler, hay, calf, sheep, beef and mutton – a combination of the abnormal unpredictability of prices and of unusually large variations, particularly upward. although our findings differ in the magnitude of price volatility in each studied market, we agree that livestock market is more volatile than the poultry market and that volatility will persist in the coming years as past. while higher food prices can be an opportunity for farmers, price volatility hurts both consumers and producers. this extreme range of price volatilities hurts net food consumers and makes their welfare to change time to time. moreover, the unpredictability of prices inhibits planning, makes investment risky and discourages farmers from producing more for the market. this represents a lost opportunity for farmers to raise their incomes, and for the country to develop the potential of programs to contribute to food security.