Building Large Economic and Financial Networks

“The study of high dimensional networks has increased dramatically. Group
wise information from large datasets can be used to build networks where
nodes represent variables and edges represent the conditional dependency
between two variables. Economic policy makers can use these networks to
measure impulse responses and determine how an economy will react over
time. Networks can be used to illustrate the trade and exchange of goods in
non-centralized markets, the provision of mutual insurance in developing
countries, alliances among corporations, and trading agreements. Two
approaches for determining grouping information to build networks are large
covariance matrix estimation and regularization. The purpose of this research
is to investigate whether large covariance matrix estimation or nodewise l1-
regularization will give more interpretable results when building a high
dimensional network through groupwise information.”