A Comparison of Forecasting Models of Live Swine Price

Abstract

The objective of this study is to compare the four forecasting models of live swine price; namely Box-Jenkins method, Holt’s exponential smoothing method, Brown’sexponential smoothing method, and damped trend exponential smoothing method. The live swine average price per month is obtained from the website of Office of Agricultural Economics during January 2005 to January 2018. There are 157 observations and the data are divided into 2 sets. The first set including 150 observations during January 2005 to June 2017 is used to construct the forecasting models. The second set comprising 7 observations during July 2017 to January 2018 is employed to investigate the most suitable forecasting method via the criteria of the lowest mean absolute percentage error (MAPE) and root mean squared error (RMSE). The findings indicate that Box-Jenkins method is the most appropriate and the forecasting model is

with MAPE is 8.4164 and RMSE is 5.1719.