Regional and global yield curve dynamics and interactions

In this paper I examine whether regional and/or global yield factors exist, what their dynamics and interactions are, and how they perform in forecasting the country yield factors and curves. For this purpose I extend the single-country dynamized Nelson-Siegel model to a multi-country context; regionally and globally. I use monthly government bond yield data for different maturities of countries in the regions America, Europe and Asia/Pacific. I find important regional and global yield factors, where the regional factor is more important than the global factor for at least 67% of the countries. This is also confirmed by an analysis of the prediction performance, where the regional factor outperforms the global factor.