Better to Stay Apart: Asset Commonality, Bipartite Network Centrality, and Investment Strategies

37 Pages Posted: 15 Nov 2018

See all articles by Andrea Flori

Andrea Flori

Politecnico di Milano

Fabrizio Lillo

Università di Bologna

Fabio Pammolli

Polytechnic University of Milan - Department of Management, Economics and Industrial Engineering; CERM Foundation

Alessandro Spelta

Human Technopole Foundation - Center for Analysis, Decisions, and Society

Date Written: October 30, 2018

Abstract

By exploiting a bipartite network representation of the relationships between mutual funds and portfolio holdings, we propose an indicator that we derive from the analysis of the network, labelled the Average Commonality Coefficient (ACC), which measures how frequently the assets in the fund portfolio are present in the portfolios of the other funds of the market. This indicator reflects the investment behavior of funds' managers as a function of the popularity of the assets they held. We show that ACC provides useful information to discriminate between funds investing in niche markets and those investing in more popular assets. More importantly, we find that ACC is able to provide indication on the performance of the funds. In particular, we find that funds investing in less popular assets generally outperform those investing in more popular financial instruments, even when correcting for standard factors. Moreover, funds with a low ACC have been less affected by the 2007-08 global financial crisis, likely because less exposed to fire sales spillovers.

Suggested Citation

Flori, Andrea and Lillo, Fabrizio and Pammolli, Fabio and Spelta, Alessandro, Better to Stay Apart: Asset Commonality, Bipartite Network Centrality, and Investment Strategies (October 30, 2018). Available at SSRN: https://ssrn.com/abstract=3275178 or http://dx.doi.org/10.2139/ssrn.3275178

Andrea Flori

Politecnico di Milano ( email )

via Lambruschini 4/B
Milano, Italy 20156
Italy

Fabrizio Lillo

Università di Bologna ( email )

Via Zamboni, 33
Bologna, 40126
Italy

Fabio Pammolli

Polytechnic University of Milan - Department of Management, Economics and Industrial Engineering ( email )

Via Lambruschini 4C - building 26/A
Milano, 20156
Italy

CERM Foundation ( email )

Via Fiorentina, 1
Siena, Siena 53100
Italy

HOME PAGE: http://www.cermlab.it

Alessandro Spelta (Contact Author)

Human Technopole Foundation - Center for Analysis, Decisions, and Society ( email )

Italy

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
47
Abstract Views
442
PlumX Metrics