Executive Network Centrality and Stock Liquidity

• Published date: 01 September 2019
• Author: William R. McCumber,Jared F. Egginton
• Date: 01 September 2019
• DOI: http://doi.org/10.1111/fima.12237

Executive Network Centrality and Stock

Liquidity

Jared F. Egginton and William R. McCumber∗

We examine the relationship between stock market liquidity and the network centrality of firm

executives. Wefind that firms whose executive officers are more central in the networkof executives

have narrowerbid-ask spreads. Weuse an exogenous network centralityshock of executive turnover

and report that liquidity improves after firms hire executives with greater centrality. We present

evidence that improvedliquidity is attributable to efficient information flows around executives in

more advantageous network positions.

A voluminous literature addresses information asymmetries in markets wherein heterogeneous

information sets enable sophisticated or informed traders to outperform relatively uninformed

traders (Grossman and Stiglitz, 1980; Hellwig, 1980; Kyle, 1985). Thus, the informational envi-

ronment of a stock is an important factor in determining its liquidity (Healy and Palepu, 2001;

Ravi and Hong, 2014). More opaque firms, where the information environment is not as transpar-

ent, have larger bid-ask spreads. Conversely,g reater transparency and more efficient information

flows reduce information asymmetries and adverse selection costs, thereby reducing spreads.

Recent literature explores the importance of investor networks in trading behavior and asset

pricing (Colla and Mele, 2009; Ozsoylev and Walden, 2011; Han and Yang, 2013; Ozsoylev

et al., 2014; Walden, 2016). These studies suggest that trading behavior, investor profits, and,

by extension, asset prices are partially determined by information dissemination through the

network of market participants. For example, Ozsoylev et al. (2014) construct an Empirical

Investor Network comprised of investors making similar trades on the Istanbul Stock Exchange

and find that traders more central within the network earn higher profits than those on the

periphery. The authors argue that investors whose positions in the network are more central enjoy

an informational advantage in that they receive information more quickly than other investors.

In this study, we examine whether executive network centrality provides a mechanism of

information flows that affects the trading environment of a stock as measured by the bid-ask

spread. This relationship is important as it sheds light on the interplay between the networks of

firm executives and market participants, information asymmetries, adverse selection, and liquidity

costs. A network is comprised of individuals and the connections between them. Executive

network centrality variablesmeasure the size of an individual’snetwork represented by the number

of direct connections to other participants, the “importance” of one’s connections, and the spatial

position of a participant relative to all other participants. We are interested in the diffusion of

We are indebted to Bing Han (Editor) and an anonymous referee for many insightful suggestions. We would like to

thank Marc Lipson, Ethan Watson, Garrett McBrayer, seminar participants at Boise State University, and conference

participants at the Southern Finance Association and the Eastern Finance Association Annual Meetings for helpful

comments. Portions of this research were conducted with high performance computational resources provided by the

Louisiana Optical Network Initiative (http://www.loni.org).

∗Jared F.Egginton is an Assistant Professor in the College of Business at Boise State University in Boise, ID. William R.

McCumber is an Assistant Professor and the JPJ Investments EndowedProfessor of Finance in the College of Business

at Louisiana TechUniversity in Ruston, LA.

Financial Management •Fall 2019 •pages 849 – 871

850 Financial Management rFall 2019

information through a network of direct and indirect connections and the possible subsequent

effect on stock liquidity provided bythis diffusion over a network of professional relationships. If

networks represent the infrastructure through which information flows, the network centrality of

firm executives should play a role in information dissemination and, as such, the liquidity of the

firm’s securities. Increased stock market liquidity benefits firms directly by lowering the cost of

raising equity capital in secondary offerings (Butler, Grulton, and Weston,2005) and, indirectly if

increased liquidity lowers the underpricing of secondary offerings (Corwin, 2003) and/or lowers

the risk premium investors require to hold the stock.

To study the relationship between executive network centrality and liquidity, we calculate four

direct measures of chief executive officer (CEO) and chief financial officer (CFO) network

position as opposed to an indirect measure of networks (Ozsoylev et al., 2014). Utilizing past

and present board positions of North American executives, we calculate degree, eigenvector,

betweenness, and closeness centralities for all of the executives in the network. We find that

firms whose officers are more central in the network of executives have narrower spreads and

reduced stock liquidity costs for all four measures of network centrality. The estimated decrease

in liquidity costs is economically significant. An increase in any centrality measure from the 25th

percentile to the 75th percentile in the sample corresponds to a 2% to 6% reduction in a median

firm’s average annual percentage bid-ask spread. For example, a within-sample change from

the 25th to the 75th percentile in CEO closeness centrality corresponds to a $6 million annual

reduction in stock liquidity costs for a typical firm. Our results suggest that the network position

of a firm’s executives improves the speed and quality of information flows thereby reducing

information asymmetries between firms and investors.

This study contributes to the literature on investor networks by confirming that the network

centrality of a firm’s executives affects the liquidity of a firm’s shares. This paper also contributes

to the literature on equity market liquidity by documenting that the personal characteristics of a

firm’s executives affect the informational environment around a stock.

The remainder of the paper is structured as follows. Section I discusses the related literature and

motivates our hypothesis. Section II details the definition of network centrality and the liquidity

variables and describes the summary statistics. Section III presents our empirical methodology

and primary results. Section IV considers changes in liquidity around executive replacements,

Section V examines the industry effects and alternative measures of liquidity, while Section VI

provides our conclusions.

I. Related Literature and Hypothesis Development

The extant literature considers how information may move through a network of market partic-

ipants. Stock price movements are, in and of themselves, informative. As one trader buys a stock,

other traders observe the trade and update their information sets to include this information.

Ozsoylev et al. (2014) consider traders “linked” if they display similar trading patterns and find

that more central market participants trade earlier and reap higher profits than those on the

periphery of the network of all traders. Market participants also share information independent

of trading decisions, and investor portfolios are shown to reflect information shared by and

with peers for institutional traders (Shiller and Pound, 1989), fund managers (Hong, Kubik, and

Stein, 2004), and households (Ivkovic and Weisbenner, 2007). Walden (2016) introduces a model

wherein agents in a network of “replica networks” (local subnetworks) share information with

their neighbors. Individual agent centrality is shown to drive agent profitability, while network

centrality (i.e., the network structure) drives information dissemination. More importantly, the