فهرست مطالب masoud taghipou kandsar

The validity of the TAR model with a fixed threshold in a time series variable wich has a trend, is misleading due to the existence of permanent inflation in the general level of prices. Most studies in Iran do not pay attention to this issue. in this paper we apply two innovations; First, the use of a variable (but exogenous) threshold point, second, the use of pannel data that shows the possibility of revealing the heterogeneity of price transfer patterns between different markets. The data includes the monthly retail price of chicken in Tehran, Isfahan, Mazandaran and Golestan during the years 2003 to 2018. A threshold autoregression model was used to show the relationship between prices in different markets. In this model, the threshold is not fixed and is a function of the transportation cost in each period and each two markets. The results showed that the price transmission of chicken meat between the four markets is asymmetric and in the case where the price gap is more than the cost of transportation, the price transmission is fast and unstable. In another case, when the price difference is less than the transportation cost, the price transmission is stable. Hausman test showed that the price transmission has a random component, but it does not affect the parameters and the regime change is based on the transportation cost. Government policies in the chicken meat market have caused a slower price transfer between them. It also helps to explain the market better by using threshold patterns that do not have a fixed threshold point.