Forecasting for Time Series Data with Structural Changes : An Empirical Analysis on Prices of Hog

• Main Authors: CHI MING FEN, 紀明芬
• Other Authors: Esher Hsu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/36605837127413506569

碩士 === 國立臺北大學 === 統計學系 === 95 === Hog industry is one of important industries in Taiwan’s agricultural economy, therefore it’s very important to understand pricing volatility of hogs. This study aims to find a suitable forecasting model to forecast future prices of hogs. From historical pattern of prices of hogs, we can see the structural changes clearly, therefore this study tried to employ Markov switching model to conduct an empirical analysis on prices of hogs. The comparisons among Markov switching model and some other models used for forecasting prices of hogs suggested by pervious studies, ARFIMA, ARCH, and GARCH models, are also conducted based upon the RMSE values. This study results show that Markov switching model has a better performance of forecast than other models for the hog prices with structural changes. A Markov switching model, MS(2), is thus suggested as a forecasting model for predicting future prices of hogs with structural changes.