Testing of breakdates in agricultural prices of selected representatives of animal production

Abstract

This paper deals with an investigation of breakdates in agricultural prices. A structural break has occurred if at least one of the model parameters has changed at some date. This date is a breakdate. Ignoring structural breaks in time series can lead to serious problems with economic models of time series. The aim is to determine the number and date of the breakdates in individual time series and connect them with changes in the market and economic environment. The time series of agricultural price relating to animal production, namely the prices of pork, beef, chicken, milk and eggs, are analyzed for the period from January 1996 to December 2011. The autoregressive model (AR) model of Box-Jenkins methodology and stability testing according to Quandt or Wald statistics are used for the purposes of this paper. Multiple breakdates are found in the case of eggs (September 1998, May 2004), milk (October 1999, December 2007) and chicken (October 2002, February 2005) prices. One breakdate was detected in the prices of beef (April 2002) and none in the case of pork prices. The results show the importance of multiple breakdate testing. The Quandt statistic provides one possible way of applying a multiple approach. All breakdates which were confirmed using these statistics can be associated with changes in the agri-food market and economic environment. Information about the date of changes in the time series can be used for other unbiased modelling in more complex models.