An empirical investigation on different methods of economic growth rate forecast and its behavior from fifteen countries across five continents

• Main Authors: Yip, Chee Yin, Lim, Hock Eam
• Format: Conference or Workshop Item
• Language: English
• Published: 2012
• Subjects: HB Economic Theory
• Online Access: http://repo.uum.edu.my/12413/1/document%2816%29.pdf; http://repo.uum.edu.my/12413/; http://dx.doi.org/10.1063/1.4757500
• Institution: Universiti Utara Malaysia

Our empirical results show that we can predict GDP growth rate more accurately in continent with fewer large economies, compared to smaller economies like Malaysia. This difficulty is very likely positively correlated with subsidy or social security policies.The stage of economic development and level of competiveness also appears to have interactive effects on this forecast stability.These results are generally independent of the forecasting procedures. Countries with high stability in their economic growth, forecasting by model selection is better than model averaging. Overall forecast weight averaging (FWA) is a better forecasting procedure in most countries.FWA also outperforms simple model averaging (SMA) and has the same forecasting ability as Bayesian model averaging (BMA) in almost all countries.