This dissertation focuses on studying the impact that weighting schemes can have on forecasting performance and dynamic analysis in global vector autoregressive (GVAR) models. The first chapter discusses an existing gap in the literature regarding weighting scheme choice and develops a simple, yet powerful method for defining richer spatial linkages in a way that doesn’t sacrifice economic context. The new technique called convex weighting, extends the set of available options for defining spatial linkages in models that handle the curse of dimensionality via compression and offers a justifiable approach to alleviating uncertainty. The second and third chapters apply the newly developed convex weighting method to regional and international level models to show that improvements in forecasting performance are possible and that inferences drawn from dynamic analysis can be highly sensitive to the underlying weighting scheme.
Number of Pages
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
Leach, Garrison, "Essays on Convex Weighting for Global Vector Autoregressive Models" (2019). Economics Theses and Dissertations. 8.