Estimation of Graphical Models with Shape Restriction
19 Pages Posted: 27 Apr 2017 Last revised: 23 May 2017
Date Written: April 28, 2017
Gaussian graphical models are recently used in economics to obtain networks of dependence among agents. A widely-used estimator is the Graphical Lasso (GLASSO), which amounts to a maximum likelihood estimation regularized using the L1,1 matrix norm on the precision matrix Ω. The L1,1 norm is a lasso penalty that controls for sparsity, or the number of zeros in Ω. We propose a new estimator called Structured Graphical Lasso (SGLASSO) that uses the L1,2 mixed norm. The use of the L1,2 penalty controls for the structure of the sparsity in Ω. We show that SGLASSO is asymptotically equivalent to an infeasible GLASSO problem which prioritizes the sparsity-recovery of high-degree nodes. Monte Carlo simulation shows that SGLASSO outperforms GLASSO in terms of estimating the overall precision matrix and in terms of estimating the structure of the graphical model. In an empirical application to a classic firms’ investment dataset, we obtain a network of firms’ dependence that exhibits the core-periphery structure, with General Motors, General Electric and U.S. Steel forming the core group of firms.
Keywords: Gaussian Graphical Models; Glasso; Inverse Covariance Matrices; Lasso; Precision matrices; Sparsity.
JEL Classification: C55, C10.
Suggested Citation: Suggested Citation