Files

Abstract

This article applies recent developments in time-series modeling to analyze the retail prices of beef, pork, and chicken. Specifically, generalized autoregressive conditional heteroscedasticity (GARCH) models were fitted to these data to determine if, unlike more traditional time-series models, the conditional variances of the underlying stochastic processes are nonconstant. The estimation results indicate that the constant conditional variances assumption can be rejected. Furthermore, ex post forecast intervals generated from the GARCH processes indicate that the forecasting accuracy of the estimated models has varied widely over time with substantial volatility occurring during the 1970s and early 1980s.

Details

PDF

Statistics

from
to
Export
Download Full History