CV - Job Market Paper  
 

Gudmundsson, Stefan G.

Job market candidate

Contact information

Tel. +34 93 542 1766

stefan.gudmundsson@upf.edu

Available for Interviews at :

Simposio de la Asociación Española de Economía (SAEe), December 14-16, Barcelona, Spain

Allied Social Science Associations (ASSA), January 5-7, Philadelphia, US

 

Research interests

Network Analysis. Time Series. Econometrics.

Placement officer

Filippo Ippolito
filippo.ippolito@upf.edu
 

References

Gabor Lugosi (Advisor)
gabor.lugosi@upf.edu

Christian Brownlees
christian.brownlees@upf.edu

Michael Wolf
michael.wolf@uzh.ch

Piotr Zwiernik
piotr.zwiernik@upf.edu

Majid Al Sadoon
majid.alsadoon@upf.edu

 

 

Research

"Community Detection in Large Vector Autoregressions" (Job Market Paper)
In this work we introduce a class of vector autoregressive (VAR) models in which the series are partitioned into communities, such that spillovers are higher between series within the same community than otherwise. A question of interest in this setting is how to detect the communities from data when the communities are unknown. We propose an algorithm that uses the eigenvectors of a function of the estimated autoregressive matrices to recover the communities and establish its consistency. We also introduce a regularised block VAR estimator inspired by the presence of community structure in data. The methodology is applied to study clustering in industrial production among a group of major economies. An out-of-sample forecasting exercise shows that the block VAR estimator performs favourably relative to a number of alternatives.

“Community Detection in Partial Correlation Network Models", with Christian Brownlees and Gabor Lugosi.
Many real-world networks exhibit a community structure: The vertices of the network are partitioned into groups such that the concentration of linkages is high among vertices in the same group and low otherwise. This motivates us to introduce a class of partial correlation network models with a community structure that replicates this empirical regularity. A natural question that arises in this framework is how to detect the communities from a random sample of observations. We introduce an algorithm called Blockbuster that recovers the communities using the eigenvectors of the sample covariance matrix. We study the properties of the procedure and establish consistency. The methodology is used to study real activity clustering in the U.S. and Europe.

Research in Progress

"Community-Based Block Estimation of Large Vector Autoregressions" (with Christian Brownlees)