Abstract:Purpose This paper aims to investigate the influence of social interactions on mutual fund portfolios from the perspective of alumni network in China. Design/methodology/approach Based on a data set that consists of 162 actively managed equity funds in China during the time period of 2003–2014, this study employs multiple linear regression model to control for organization- and location-based interpersonal connections as well as other confounding factors and clarify the causality relationship between alumni networks of mutual fund managers and their portfolios. Findings After controlling for organization- and location-based interpersonal connections, we find that mutual fund managers who graduated from the same college/university have more similar stock holdings and are more likely to buy or sell the same stocks contemporaneously. As a result, alumni managers exhibit a higher correlation of fund returns. Moreover, the effect of alumni relationship on mutual fund investments becomes weaker when more managers are connected within the network. We also find that valuable information is shared among alumni managers: (1) the average returns for the alumni common holdings portfolios is significantly higher than those for non-alumni holdings portfolios and (2) a long-short strategy composed of stocks purchased minus sold by alumni managers yields positive and significant risk-adjusted returns. Practical implications The findings suggest that information dissemination among connected fund managers could be one of the driving forces for mutual fund herding behavior, and that a portfolio of funds whose managers are educationally connected could be highly exposed to certain stocks and risks. Originality/value This paper contributes to the growing finance literature addressing the influence of personal connections on information dissemination that specifically contributes to price formation. It corresponds more closely to Cohen et al . (2008), who investigate college alumni connections between fund managers and corporate board members. Since the authors simultaneously examine three potentially overlapped social networks, which are based on education, locality and fund family, the authors are able to disentangle their effects on fund managers' investment decisions. Moreover, the findings suggest that institutional investors make investment decisions based on share private information, and therefore, it also contributes to the literature on fund herding behaviors (Grinblatt et al ., 1995; Wermers, 1999).

Inferring Latent Social Networks from Stock Holdings

Count Models of Social Networks in Finance

Network Analysis to Uncover Stock Comovement from a Chinese Financial Portal.

Stock network inference: A framework for market analysis from topology perspective

Inferring financial stock returns correlation from complex network analysis

Fund Manager Educational Networks and Portfolio Performance

Location Choice, Portfolio Choice

Game in another town: Geography of stock watchlists and firm valuation

Social interactions and mutual fund portfolios: the role of alumni networks in China

Statistical Inference on Latent Space Models for Network Data

Estimating Causal Peer Influence in Homophilous Social Networks by Inferring Latent Locations

Friends in Need are Friends Indeed: An Analysis of Social Ties between Financial Analysts and Mutual Fund Managers

Stock portfolio structure of individual investors infers future trading behavior

Hedge fund networks, information dissemination, and stock price comovement: Evidence from China

Multi-Likelihood Methods for Developing Stock Relationship Networks Using Financial Big Data

Estimating Social Influence Using Latent Space Adjusted Approach in R

Alumni Networks and Investment Strategy: Evidence from Chinese Mutual Funds

Extracting Hidden Fluctuation Patterns of Hang Seng Stock Index from Network Topologies

Structural holes and hedge fund return comovement: evidence from network‐connected stock hedge funds in China

Investor network and stock return comovement: Information-seeking through intragroup and intergroup followings

Fund Managers’ Association Networks, Information Sharing and Fund Performance