Count Models of Social Networks in Finance
Harrison G. Hong
Princeton University - Department of Economics; National Bureau of Economic Research (NBER)
November 29, 2013
We use overdispersed Poisson regression models to study social networks in finance. We count an investor's social connections in different groups, such as cities or industries, as proportional to the number of stocks held by this investor that are headquartered in those cities or part of those industries. When connections are formed randomly, the count of such connections in any group follows a Poisson distribution. Using data from institutional and retail investors' holdings, we find instead overdispersion for some groups. It implies that investors have distinct propensities to form ties because they are part of networks. We relate these propensities to investor demographics, gauge the prevalence of city versus industry networks, and measure their value for investor performance. These models can be utilized to study any financial network where investment data are available.
Number of Pages in PDF File: 47
Keywords: Social Networks, Poisson Regressions, Investor Behavior
JEL Classification: G1, G2, G3working papers series
Date posted: January 8, 2014
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo3 in 0.453 seconds