Concentrated Shareholders as Substitutes for Outside Analysts

• Published date: 01 November 2008
• Author: Stephen D. Smith,Sanjiv Sabherwal
• DOI: http://doi.org/10.1111/j.1467-8683.2008.00706.x
• Date: 01 November 2008

Concentrated Shareholders as Substitutes for

Outside Analysts

Sanjiv Sabherwal* and Stephen D. Smith**

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: We examine the relationship of concentration of shareholdings with the number of f‌inancial

analysts following a f‌irm to see if concentrated shareholders substitute for the monitoring activities of analysts.

Research Findings/Results: Using a clean ownership dataset with a sample of 3,115 f‌irm-year observations for U.S. f‌irms

and regression techniques that address any potential endogeneity, we f‌ind that analyst following is negatively related to the

concentration of outsider and insider shareholdings. We f‌ind similar relations for changes in analystfollowing and changes

in ownership concentration.

Theoretical Implications: Our results support the argument that an outsider with a larger stake in a f‌irm is more likely to

produce its own in-house information for the monitoring of the f‌irm’s managers and avoid both the cost and moral hazard

problems associated with analysts. The results also support the argument that if senior managers hold large stakes in the

f‌irm, there is a greater likelihood that managerial incentives will be aligned with those of other shareholders.

Practical Implications: We contend that there is a governance substitution effect, with concentrated shareholders substi-

tuting for the monitoring activities of analysts.Our results are consistent with the opinion that regulators need not fear large

shareholders. This is especially applicable to large outside shareholders as we f‌ind that the economic effect of concentrated

outsider shareholdings is quite strong and greater than that of concentrated insider shareholdings.

Keywords: Concentrated Shareholdings, Financial Analysts, Corporate Governance, Agency Theory

INTRODUCTION

Agency theory implies that corporate managers may

engage in activities that are harmful to shareholders.

Therefore, it is in the interest of the shareholders that the

f‌irm’s managers be monitored. Jensen and Meckling (1976)

argue that important monitors of managers include the

f‌inancial analysts employed by investment banks, broker-

age f‌irms, and independent research f‌irms. Financial ana-

lysts provide analyses of the prospects of the f‌irms they

follow considering the industries in which those f‌irms

operate and the overall f‌inancial markets. Analysts tend to

be highly specialized and typically follow a handful of

f‌irms in a few industries. They produce research reports by,

among other things, examining f‌inancial statements,

having discussions with company management, analyzing

industry and market trends, and building f‌inancial models.

The reports include dividend and earnings forecasts;

ratings whether to buy, sell, or hold the stock; stock price

targets; and a detailed analysis. The research produced is

provided to clients such as institutional investors, retail

investors, information vendors (e.g., Bloomberg), and sales

forces of stock brokerage f‌irms. The research is also pub-

licly released, but generally the public release follows the

provision of research to the main client(s) and is not as

elaborate.

The input provided by analysts to the f‌inancial markets

help the markets become more informationally eff‌icient.

Moreover, in the process of providing these services, ana-

lysts collect informationfrom a variety of sources, both inter-

nal and external to the f‌irm, to assess the f‌irm’s business and

f‌inancial prospects and its investment potential. Conse-

quently, they perform an additional role of potentially

important monitoring of management’s actions.

Prior literature has examined the monitoring role of ana-

lysts. For example, Moyer, Chatf‌ield, and Sisneros (1989)

empirically test the hypothesis that the extent of monitoring

activity performed by security analysts is directly related to

the level of potential agency costs in a f‌irm. Their results are

consistent with Jensen and Meckling’s (1976) hypothesis

*Department of Finance and Real Estate,University of Texas at Arlington, Arlington,

TX 76019. Tel: 817-272-5520; Email: sabherwal@uta.edu

**Georgia State University and Federal Reserve Bankof Atlanta, Atlanta, GA 30303

562 CORPORATE GOVERNANCE

Volume 16 Number 6 November 2008 © 2008 TheAuthors

Journal compilation © 2008 BlackwellPublishing Ltd

doi:10.1111/j.1467-8683.2008.00706.x

that analyst activity serves as a monitoring device in the

presence of potential agency problems. Coffee (2002) terms

the f‌inancial analysts as professional “gatekeepers.” He

def‌ines gatekeepers as reputational f‌inancial intermediaries

providing verif‌ication and certif‌ication services to investors.

Analysts are responsible for f‌iltering, verifying, and evaluat-

ing complicated f‌inancial information in order to assess the

f‌irm’s prospects relative to its rivals.

The argument by Jensen and Meckling (1976) that one of

the roles of analysts lies in mitigating moral hazard prob-

lems between insiders and outside claimants suggests that

the demand for f‌inancial analysts could depend on the com-

position of outside shareholders and the distribution of

shares between insiders and disparate groups of outsiders.

Earlier authors (e.g., Bhushan, 1989; O’Brien and Bhushan,

1990) have in fact provided evidence that the number of

analysts following a f‌irm is inversely related to the percent-

age of the f‌irm held by insiders and positively related to

the number of institutional investors. These authors argue

that the insider relation can be attributed to lower moral

hazard costs when insiders collectively hold a higher stake

in the f‌irm. The relationship between analyst following

and the number of institutional shareholders comes from

the idea that sophisticated institutions will demand more

professionally-generated third party informationconcerning

the f‌irm’s prospects than individual investors.

One purpose of this paper is to argue that for a given

distribution of shares between outsiders and insiders, the

demand for analyst services will be inversely related to the

concentration of shareholdings of outsiders. The argument

follows from the fact that the typical outside shareholders

with a small stake in a f‌irm have little incentive, given even

small f‌ixed costs, to produce independent information that

would help them monitor the managers. So, these investors

get most of their information from either management or

analysts. However, an outsider with a large enough stake in

the f‌irm will at some point f‌ind it reasonable to produce its

own in-house information and avoid both the cost and

moral hazard problems associated with analysts (Diamond,

1984; Irvine, 2004; Jackson, 2005). Another purpose of this

paper is to argue that the demand for analyst services will be

inversely related to the concentration of shareholdings of

insiders. The idea is that if senior managers hold large stakes

in the f‌irm, there is a greater chance that managerial incen-

tives will be aligned with those of other shareholders. We

also brief‌ly investigate if the relation between the concentra-

tion of insider holdings and analyst following is a curvilin-

ear one, as suggested by Morck, Shliefer, and Vishny (1988)

and McConnell and Servaes (1990).

We empirically test our arguments using a sample of 3,115

f‌irm-year observations for U.S. f‌irms during 1996 to 2001.

The standard source used for ownership data (Spectrum data

included in Compact Disclosure) suffers from biases and mis-

takes.1Dlugosz, Fahlenbrach, Gompers, and Metrick (2006)

clean the ownership data for the period 1996 to 2001 and

create a blockholders data set that they make publicly avail-

able for research.2We use this data set for our study and

employ Herf‌indahl indices of blockholdings as measures of

ownership concentration.3

If we were to perform yearly regressions of level of analyst

following on levels of ownership concentration and interpret

statistically signif‌icant results as meaning that a change in

concentration can lead to a change in analyst following, the

interpretation could be criticized for ignoring potential

endogeneity. Therefore, we follow two alternative ap-

proaches that do not have this drawback. First, we conduct

level-based regressions on a panel data set and control for

f‌irm-specif‌ic heterogeneity by including f‌irm f‌ixed effects.

We f‌ind that analyst following is signif‌icantly negatively

related with outsider and insider ownership concentration.

In the second approach, we perform regressions based on

year-to-year changes in analyst following and concentration.

We f‌ind the change in analyst following to be signif‌icantly

negatively related with the changes in ownership concentra-

tion. Because any f‌irm f‌ixed effects cancel when changes are

considered, the negative relation observed cannot be due to

endogeneity that could arise from such effects. Therefore,

the change-based results provide further evidence that as

insider or outsider ownership concentration increases,

analyst following decreases.

The rest of the paper is organized as follows. The next

section examines the issues and hypotheses in more detail.

The third section describes the sample. We describe the vari-

ables, specify the empirical models, and discuss the results

in the fourth section. The f‌ifth section examines the robust-

ness of results, multicollinearity, and the economic signif‌i-

cance of results. The f‌inal section provides concluding

comments.

ISSUES AND HYPOTHESES

Our main objective is to address whether the presence of

concentrated ownership is associated with less information

production by third parties; in this case outside f‌inancial

analysts. In this section, we f‌irst present our hypotheses

regarding how concentrated holdings by outsiders and by

insiders affect analyst following. We then discuss the other

variables that are related to analyst following. These vari-

ables are employed as control variables in our tests.

Concentrated Holdings by Outsiders and the

Number of Analysts

As noted in the introduction, we argue that a concentration

of outsider shareholdings inf‌luences the number of f‌inancial

analysts that follow a f‌irm. This conjecture could be justif‌ied

by a number of papers in the theoretical literature, but the

discussion in this section draws on the theory of f‌inancial

intermediation in Diamond (1984). In particular, he builds a

model in which multiple outsiders lend to an entrepreneur

in the presence of asymmetric information. This leads to

incentive problems and a decision by outsiders to either

each pay a f‌ixed cost to produce information to monitor the

entrepreneur themselves or to delegate the monitoring task

to an intermediary. While Diamond discusses the demand

for intermediary services by lenders, the analysis extends to

the case in which the principals are outside security holders

of the f‌irm. Specif‌ically, in the context of this paper, outside

shareholders can be considered as the principals. They are

concerned with whether to produce information to monitor

CONCENTRATED SHAREHOLDERS AS SUBSTITUTES FOR OUTSIDE ANALYSTS 563

Volume 16 Number 6 November 2008© 2008 TheAuthors

Journal compilation © 2008 BlackwellPublishing Ltd