Authors: Shahnawaz Akhtar, Zeynep Irem Eyibak, Filippo Grittini, Aikaterini Christina Katsimpri, Ashvik Viswanathan
We investigate the connectedness of GDP growth risk over 12 OECD member countries. Understanding the Growth-at-Risk of GDP has been a popular area of discussion in recent years. Even more recently, it has been increasingly imperative to acknowledge GDP downside risk from the lower quantiles of its conditional distribution. Utilizing methods introduced by Adrian, Boyarchenko, and Giannone (2019), we observe the quantile dynamics of these 12 OECD member countries with respect to the vulnerability of GDP growth as a function of relative financial and economic conditions. Further, utilizing network estimation methods from Diebold and Yilmaz (2014), we find that network connectedness is stronger and more volatile at the 5th quantile compared to that at the 50th quantile, and that 5th quantile connectedness increases during the Financial Crisis of 2008. Finally, we decompose the country pairwise connectedness into explanatory channels, and find that along with trade and domestic financial conditions, foreign financial conditions are important in explaining the connectedness between two countries.
Data: Datasets used in analysis can be found in 'data' directory.
Code: Codes used in analysis can be found in 'code' directory.
Results: Outputs from code is available as html files in 'results' directory
Graphs: All the graph and charts from result can be found in "graphs/" directory. Files prefixed with F{h}O{p} indicate output for 'h' period ahead quantile forecast used in 'p' ordered Vector Auto-Regression.
Tables: Table outputs can be found in 'tables' directory.
