Investor heterogeneity and trading

• Date: 01 September 2018
• Published date: 01 September 2018
• DOI: http://doi.org/10.1111/eufm.12169

DOI: 10.1111/eufm.12169

ORIGINAL ARTICLE

Investor heterogeneity and trading

Anzhela Knyazeva

1

|

Diana Knyazeva

1

|

Leonard Kostovetsky

2

1

Securities and Exchange Commission,

Washington, District of Columbia

Emails: diana.knyazeva@gmail.com;

anzhela.knyazeva@gmail.com

2

Boston College, Carroll School of

Management, Chestnut Hill,

Massachusetts

Email: kostovet@bc.edu

Abstract

Institutional investors play a crucial role in the information

environment of firms. We argue that heterogeneity in the

information ability of institutional investors has a signifi-

cant impact on trading around information releases. We

propose novel measures of within-firm investor heteroge-

neity and find that investor heterogeneity increases

abnormal trading volume around news, holding constant

the average levels of investor sophistication. We also find

larger spread reductions around announcements for firms

with greater investor heterogeneity. The effect of investor

heterogeneity on trading continues to hold after accounting

for total institutional ownership, the presence of certain

types of institutional investors, and analyst coverage.

KEYWORDS

information, institutional investors, investor heterogeneity, spreads,

trading volume

JEL CLASSIFICATION

G10, G12, G14

The authors are grateful to the Editor (Lu Zhang), an anonymous referee, and seminar participants at George Washington

University, New York University PhD Alumni Conference, Hui Guo, Jeffrey Harris, and Jonathan Kalodimos for helpful

input, and to Brian Bushee for providing the institutional investor classification information on his website. All errors and

omissions are our own. The Securities and Exchange Commission disclaims responsibility for any private publication or

statement of any SEC employee or Commissioner. This article expresses the authors’views and does not necessarily reflect

those of the Commission, the Commissioners, or other members of the staff.

680

|

© 2018 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/eufm Eur Financ Manag. 2018;24:680–718.

1

|

INTRODUCTION

Institutional ownership has been linked to the information environment of firms, trading patterns, and

corporate decisions. Institutions are traditionally regarded as sophisticated investors with superior

information. However, viewing institutional investors as a single group of sophisticated investors

overlooks crucial differences in their ability to gather and analyze information and their incentives for

doing so. This paper examines the effects of differences in information capabilities of institutional

investors of a given firm on trading of a firm's shares around news releases.

We argue that heterogeneity in the information gathering and processing ability of institutional

investors has a first-order effect on a firm's information environment that is not explained by existing

metrics. Our measures focus on within-firm heterogeneity in the characteristics of institutional

investors that proxy their ability to gather and analyze information.

1

We empirically test the relation

between investor heterogeneity and trading around earnings surprises. Our measures have additional

power to predict trading around news releases after accounting for commonly used information

environment proxies.

We hypothesize that heterogeneity in the information ability within a firm's investor base results in

differences in the precision of investor information about firm performance. As an example, the

precision of a trader's ex ante private information may be a function of the trader's overall or company-

specific investment experience, local knowledge, or the extent of resources that the trader can allocate

to information gathering. The arrival of a corporate announcement causes updating of beliefs about

firm performance among differentially informed investors. Kim and Verrecchia (1991b) show that

differential precision of pre-announcement investor information results in a greater volume reaction to

the announcement. The information contained in the news announcement is more important to traders

with less precise prior information, who revise their beliefs more. In contrast, ex ante better informed

traders place less weight on the public announcement, revising their beliefs less. Thus, even if

the announcement is interpreted identically by all traders, differences in the precision of pre-

announcement information lead to higher trading volume on the announcement. The greater the

informational differences between investors, the larger the abnormal trading volume reaction to

the announcement. In the absence of differences in precision, there is no trading volume on the

announcement.

Kim and Verrecchia (1997) and Wang (1994) extend and generalize the Kim and Verrecchia

(1991b) framework. Kim and Verrecchia (1997) incorporate differences both in the precision of pre-

announcement information and in the interpretation of information in the public announcement. When

news about current earnings is released, heterogeneously informed investors may differ in how well

they interpret the information that is released. For instance, an earnings release may not only lead

investors to revise their beliefs about the previous quarter's performance but also to update their beliefs

about future firm performance. Wang (1994) extends the Kim and Verrecchia prediction to a setting

with both noninformational and information-driven trading.

2

1

Investor pre-announcement information can stem from acquiring private information and processing public information

about firms prior to the announcement. Thus, precision of an investor's pre-announcement information depends on the

ability to both gather and process information prior to the announcement. Conceptually and empirically, our main

hypothesis does not distinguish between the two aspects, collectively referred to as the information ability or pre-

announcement informedness of investors.

2

A number of papers explore other theoretical settings in which trading volume arises from disagreement and differences

among investors in a broader context. See Bamber et al. (2011) for an in-depth discussion of prior work.

KNYAZEVA ET AL.

|

681

Empirically, all else equal, these effects predict a positive relation between heterogeneity in

investor informedness and trading volume around earnings releases. Moreover, heterogeneity in

investor informedness may affect the behavior of bid–ask spreads around the announcement.

Heterogeneity in investor informedness prior to the announcement can result in an adverse selection

problem (e.g., less informed investors recognizing that they are less informed and being unwilling to

trade against better informed investors), leading to larger bid–ask spreads prior to the announcement.

The arrival of information in the form of a public announcement makes less informed investors

relatively more informed. The resulting convergence in the level of informedness of investors can

reduce information asymmetries between investors. By reducing information asymmetries between

investors and mitigating the adverse selection problem, the public release of earnings information can

reduce bid–ask spreads. Thus, our empirical prediction is that the reduction in bid–ask spreads after the

announcement would be more pronounced if there were greater differences in investor informedness

prior to the announcement.

While our main hypothesis has re ceived more support in prior th eoretical work, a competing

effect may emerge. If differen ces in the information ability of investors translate into di fferences in

investor ability to interpret th e announcement information, th e announcement may in some cases

increase the information asy mmetry between less informed and better informed inves tors and result

in adverse selection. In Kim and Verrec chia (1994), heterogeneity in inv estor ability to interpret the

announcement information in creases information asymmetr y and widens bid–ask spreads after the

announcement. Notably, t heir model predicts higher bid–ask spreads as well as higher tr ading

volume for such firms. At the sam e time, George, Kaul, and Nimalen dran (1994) show that under

some conditions heterogene ity in the information ability of i nvestors can increase spread s and

decrease trading volume.

3

In their model, differences in the informatio n ability of investors imply

differential ability to interpr et announcement information, whi ch results in adverse selection and

increases transaction costs . Thus, the relationship between d ifferential precision and volu me

becomes ‘both indirect and amb iguous’once transaction co sts are endogenized. Heter ogeneity in

the information ability of investors has two competing eff ects: increased information-driven trading

and reduced liquidity trading (caused by increased transaction cos ts as a result of adverse selection).

Whether the two effects offset e ach other depends on the impact of in creased transaction costs on

liquidity trading. General ly, public announcements of inf ormation tend to reduce informa tion

asymmetries between investo rs. However, if heterogeneit y in information ability affects th e

processing of announcement infor mation in a manner that increases adverse s election, investor

heterogeneity could incr ease abnormal spreads around the announcement and reduc e abnormal

volume.

Ultimately, the effect of differences in the information ability of investors on volume and spreads

around earnings announcements is an empirical question. We are able to disentangle our main

hypothesis (investor heterogeneity increases the volume of trading and reduces spreads around

announcements) from competing effects (investor heterogeneity reduces volume and increases spread)

as they have contrasting empirical predictions.

Our empirical setting relies on earnings announcements for two reasons. First, the release of

earnings news is a well-documented source of new information carefully watched by the market,

so it is relevant for investors. Second, unlike other disclosures, the content of an earnings release

3

Adverse selection caused by the presence of less informed and better informed investors has also been modeled in the

context of share repurchases in Brennan and Thakor (1990) and Lucas and McDonald (1998). More generally, the argument

that uninformed liquidity traders that have discretion with respect to the timing of their trades may trade less when

information asymmetry is high is presented in Admati and Pfleiderer (1988) and Foster and Viswanathan (1990).

682

|

KNYAZEVA ET AL.