A Limited Information Bayesian Forecasting Model of the Cattle SubSector

The first step towards forecasting the price and output of the cattle industry is understanding the dynamics of the livestock production process. This study follows up on the Weimar and Stillman (1990) paper by using data from 1970 to 2005 to estimate the parameters that characterizes the cattle output supply. The model is then used to estimate forecast values for the periods 2006 and 2007. Bayesian limited information likelihood method is used to estimate the parameters when endogeneity exists between these variables. The forecasting ability of the model for a two-step ahead forecast for majority of the variables are relatively good and test statistic of the forecast are reported.