Suresh Bhagavatula NSRCEL - NFCG · Suresh Bhagavatula NSRCEL . Additional Research on Corporate Board Interlocks: Identifying influential Directors through analysis of boards interlocks

Additional Research on Corporate Board Interlocks:

Identifying influential Directors through analysis of boards interlocks during

the period 2010 – 2012

Suresh Bhagavatula

NSRCEL

Additional Research on Corporate Board Interlocks:

Identifying influential Directors through analysis of boards interlocks during the period

2010 – 2012

Executive Summary

It has been found that directors who occupy board positions of both listed and unlisted companies

influence the performance of the firms in two ways (Hillman and Dalziel, 2003). They can do so by

effectively monitoring the activities of the company as representatives of shareholders. In addition, they

bring in scarce intangible resources such as legitimacy, advice, connections to other organizations. From

a corporate governance literature, monitoring function of the board is of greater importance than

bringing in resources. The theoretical lens that is utilized to understand this monitoring function is

agency theory. This aspect focuses on the issues that arise when ownership and control are separated in

organizations (Berle and Means, 1938). So specifically, the board needs to monitor the agents –

managers to protect the interests of the principals – owners (Mizruchi, 1996). Some of the monitoring

activities that the board needs to perform are identifying the CEO, monitoring the CEO and the top

management, chart out the compensation for the top management including the CEO.

Previous research

In our previous study, we found that board interlocks are quite widespread in India. Taking a

(numerically) small but nevertheless (in terms of market capitalization) an important slice of available

corporate data, we observed that in 2010, ‘highly networked’ directors (defined as those on the board

of 5 or more listed NSE companies) who constitute just 6 percent of the overall pool of directors

among NSE100 companies are associated with 486 NSE listed companies which account for a whopping

66 percent of the total market capitalization of all NSE listed companies. Interestingly, there appears to

be a marked increase in market capitalization of these ‘highly networked’ companies, which these ‘highly

networked’ directors are linked to over the last several years. For instance, for the 3 years from 2001

to 2003, the market capitalization of ‘highly networked’ companies ranged between 33 percent to 43

percent; it moved up to peak of 70 percent in 2007 and was at 66 percent in 2010. The substantive rise

in market capitalization of these ‘highly networked’ companies has coincided with only a marginal

increase (from 5% to 6%) in the proportion of ‘highly networked’ directorships. These trends suggest

that despite the well-intentioned regulatory reforms (a) the extent of over-boarding/interlocking among

directors has not come down (there is actually a marginal increase) and (b) there appears to be

increasing concentration of power among key individuals. Given the general view that concentration of

power in a few individuals or entities is not desirable in the larger interests of society, it would appear

that the observed trends in the concentration of power among a handful of the country’s corporate elite

is a matter for substantive public policy concern.

This study

In this study, we took the years 2010-2012 and captured data for all directors of all listed companies

unlike the previous study where we focused only on directors of 100 top companies in NSE. In this

report, we identify the top 25 directors in each of these years. Our research also attempts to

understand the connections of these 25 directors. The relationships between common directors help

companies on which they are board members to be linked to one another. Sometimes this could be

large companies sometimes this could be small companies therefore the network helps us identify

companies that are connected to each other. These connections help companies to understand the best

practices from each other and have many other benefits.

Introduction

Board interlocks occur when a director of one organization is on the board of another

organization. The causes and consequences of such interlocks have been debated in the western

literature but in Indian, comparatively little is known about interlocks corporate boards. Central

to this debate is to understand what the consequences for society would be if economic power is

concentrated within a small set of individuals. As the process of globalisation increases India

would have to adopt internationally accepted corporate governance principles. One of the laws

that has been introduced in India in mid-2000 Clause 49 to increase the number of independent

directors in boards depending on the ownership structure. Although we have not conclusively

shown but the analysis of director interlocks seem to indicate that while the number of

independent director positions has increased but the pool of independent directors has not

increased correspondingly. This means that the number of board positions an independent

director is currently holding has increased but the pool of independent directors has not increased

correspondingly. This means that the board interlocks may have increased after the introduction

of clause 49.

In this study, we took the years 2010-2012 and captured data for all directors of all listed companies

unlike the previous study where we focused only on directors of 100 top companies in NSE. In year

2010, there were 17170 directors in all NSE companies. Of these, 1940 director were on 5 or more

boards and this constitutes about 11% of all directors. On the other hand about 9500 directors or 55%

of all directors are on only one company board. Both In 2011 and 2012, the numbers are quite similar

indicating that the board of directors across all companies seems to be quite stable. In this study our

analysis will be on a subset of these directors. We try to identify the top 25 directors in each of these

years. What we noticed is that although overall number of directors in all NSE companies in each of

these years has not changed much but what has changed over the years is who the top 25 directors are.

There are few individuals who come in and go out. Our research also attempts to understand the

connections of these 25 directors. The relationships between common directors help companies on

which they are board members to be linked to one another. Sometimes this could be large companies

sometimes this could be small companies therefore the network helps us identify companies that are

connected to each other. These connections help companies to understand the best practices from each

other and have many other benefits.

Methodology of Network Analysis

Humans intuitively understand that certain actors are more powerful and influential than others because

of the networks they are embedded in. However the first sociogram or a picture that depicted

networks was drawn by Moreno. In early, 1930s, he depicted relationships of individuals by drawing

nodes to represent individuals and lines to represent relationships. He used these diagrams to study

structural properties of groups. Over years, network analysis has become to be known as a disciplined

enquiry into patterns of social relationships between and among actors. Freeman (2004, pg 10) defines

social network analysis as a method of social research that displays four features:

a structural intuition, which recognizes the importance of ties that link social actors;

systematic relational data, which can help generate reports of patterned social interactions (ibid.

pg 16);

graphic images that depict the patterned social interactions

mathematical or computational models to clarity the concepts and spell out the consequences

(pg. 25)

Accordingly, this study uses network analysis to analyze board interlocks in the chosen sample

companies and directors. Following is a brief introduction to the network methodologies used here.

Relational data between various actors is usually one mode in nature, which are straightforward actor to

actor networks. However, sometimes actors are affiliated to one another through common events they

participate in or attend. Networks arising from individuals or entities interacting with each other while

being engaged in a single forum (such as a board of directors or even assembly of companies such as an

industry association) are referred to as ‘category membership networks’ or ‘two mode networks’.

Wasserman and Faust (1994, p. 291) describe “affiliation networks are two mode networks consisting of

set of actors and set of events…and the connections among members of one of the modes are based on

the linkages established through the second mode”. In line with the most popular method of analysis of

two mode networks, we separated them into single mode networks before they can be analyzed. This

method was introduced by Brieger in his paper The Duality of Person and Group (1974). Most two

mode networks are collected from archival data (Valente, 2010). The original data is in a tabular form in

which rows are individuals and columns are organizations they are part of. As Valente (pg 48) explains

that the “table is a matrix, which can then be transposed and post-multiplied to yield an individuals-by-

individuals matrix representing the number of joint memberships. The transposed matrix can be pre-

muliplied (placed first) to yield an organization-by-organization matrix.”

Figure 1 depicts a classic affiliation network where Directors are occupying board position in companies.

For instance, directors A and B are on the board of companies 1 and 3 but B is also on the board of 2,

which he shares with C. Hence, B is connected to A and C. D on the other hand is on the boards of 1

and 4. D is connected to A and B through the company 1 and to C and E through the company 4. This

figure can be separated into a two one mode networks - one which depicts the network of Directors

and another that depicts the network of the Companies.

Once the networks are separated, they were analysed using the Gephi software. One of the primary

role of network analysis is to identify core or powerful individuals, which is done through identifying

central actors.

Fig1. Affiliation network

The two mode network data of directors and companies was separated was done using Ucinet 6.0

software (Borgatti, et al. 1999).

In our study, data on directors from 2010 to 2012 has been analysed using the above mentioned

process.

Sample data description

The sample consists of the NSE listed companies from 20010 to 2012. The process of sample

construction is discussed below:

The data for the study was collected from multiple sources. These include

National Stock Exchange

CMIE, Prowess

Annual Reports from Insight , Money Control, Report Junction

Director’s database, Prime and BSE and

Individual company reports

CMIE’s Prowess was the primary data source. We extracted firm level information from 2009 to 2012

on various characteristics such as financial and performance variables, information relating to ownership

groups, composition of boards of directors, industry classification, etc. The CMIE – Prowess database

3 4

2

1

D A

B

C

D

E

1

2

3

4

Director Companies

A B

C E

Director to Director

Company to Company

was queried to get the Directors list. Missing data was obtained manually by sifting through copies of

company’s annual reports. A great deal of cleaning up was necessary to overcome problems associated

with First Names, Last Names, Middle Names, and so on. After this cleaning CMIE-Prowess database

was queried to get data on multiple directorships.

Finally, after the relevant data regarding the companies listed in NSE and the multiple directors were

obtained, we constructed the pivot table in spreadsheet as shown in Figure 2 where the rows represent

the directors and the columns represent the companies. The value 1 will be entered in a cell if the

corresponding director sits on that particular company else it will be empty. This pivot table is given as

input to the gephi software which is then used to convert two mode network into one mode (director-

director mode). Finally, the gephi network based on the centrality measures (Betweenness, Degree and

Eigen vector) are constructed and analysed. We termed directors who are in the board of 5 and more

companies as highly boarded directors. Multiplicity indicates the number of common directors any two

companies share. The Distance dimension indicates the number of directors who can be reached within,

1, 2 or more than 2 degrees of separation.

Fig. 2. Pivot table construction for 2011 data

Data analysis using Gephi

Gephi is an open-source network analysis and visualization software package. Gephi has been

used in a number of research projects in the university, journalism and elsewhere, for instance in

visualizing the global connectivity of New York Times content and examining Twitter network traffic

during social unrest along with more traditional network analysis topics. Gephi is an interactive

visualization and exploration platform for all kinds of networks and complex systems, dynamic and

hierarchical graphs.

The inputed pivot table data is modelled as a two mode network as shown in Figure 3 using gephi. The

two mode network was then converted into one mode (director mode) was shown in Figure 1. The

nodes represents the directors and an edge between any two directors indicates that they have at least

one company in common between them.

Figure 3 Two Mode Network for 2011 data

● Directors

● Companies

Definition of network terms

One of the primary reason for conducting a social network analysis is to identify central nodes or

vertices, which in this study, would be directors and companies. These central nodes are said to be

powerful and influential due to these positions: ability to reach out to other in the network, the ability

to connect unconnected people, etc. However, there are many forms of centrality that have been used

in network analysis but we would be using three important measures viz. degree centrality, betweenness

centrality and eigen vector centrality.

Degree: Historically first and conceptually simplest is degree centrality, which is defined as the number of

links incident upon a node (i.e., the number of ties that a node has). The degree can be interpreted in

terms of the immediate risk of a node for catching whatever is flowing through the network. In the case

of a directed network (where ties have direction), we usually define two separate measures of degree

centrality, namely indegree and outdegree. (Ref: Wikipedia, Centrality, accessed 24 July, 2014)

In-Degree: The number of head endpoints adjacent to a node is called the indegree of the node.

Out-Degree: The number of tail endpoints is its outdegree of the node.

Betweenness: Betweenness centrality quantifies the number of times a node acts as a bridge along the

shortest path between two other nodes. It was introduced as a measure for quantifying the control of a

human on the communication between other humans in a social network by Linton Freeman. In his

conception, vertices that have a high probability to occur on a randomly chosen shortest path between

two randomly chosen vertices have a high betweenness. (Ref: Wikipedia, Centrality, accessed 24 July,

2014)

Eigenvector: Eigenvector centrality is a measure of the influence of a node in a network. It assigns relative

scores to all nodes in the network based on the concept that connections to high-scoring nodes

contribute more to the score of the node in question than equal connections to low-scoring nodes. (ref:

Wikipedia, Centrality, accessed 24 July, 2014)

Results

Tables 1, 2 and 3 give the details of the two 25 directs in the years 2010, 2011, 2012 respectively. There

is a slight variance in the number of companies each of the directors are in. While the top director in all

the three years is Mr. RA Shah, one or two new directors can be found in the years 2011 and 2012.

The betwenness gephi network of the top 25 highly boarded directors of the year 2010, 2011, 2012 are

shown in Figure 4a, 4b and 4c respectively. Degree centrality of the same set of directors is given in in

figures 5a, 5b and 5c respectively. Finally figures 6a, 5b and 6c show the Eigen vector centrality for the

same set of directors. The size of the line indicates the number of companies the directors are in

together: from 1 to 4 companies. The networks of the companies these directors are board members of

are given in the subsequent set of figures. In all the three years, 25 top directors were on boards of 328

companies. Figures 7a, 7b and 7c are degree centrality figures for these companies in 2010, 2011 and

2012 respectively. Figures 8a, 8b and 8c are betweenness centrality figures for these companies in 2010,

2011 and 2012 respectively. Figures 9a, 9b and 9c are eigen vector centrality figures for these companies

in 2010, 2011 and 201 respectively.

Conclusion and Future work

To conclude, the work on interlocks of director is very nascent in India. Research from developed

economies seems to indicate that interlocks are both good and bad. Through this research, we have

identified directors who are on the large number of companies. While in the western setting, having

directors who are on boards of more than four companies was found to be detrimental to companies.

However, in our previous research we found out that in India it was not so. Furthermore, we also found

out that compared to having these top directors it is better to have more central directors. Directors

who are central in the network seem to be able to provide companies with much needed information

and advice. Through this research we wanted to identify the most boarded directors and the companies

they are board members of. In the process we also identified the companies these directors are board

members of. Not all companies are large companies. In the next round of activity we would like to

compare the performance of companies that have these highly boarded members with companies that

do not have such members. This would give insights into whether or not it is beneficial to have top

directors in a smaller company’s board of directors.

Table 1: Top 25 Directors on NSE boards for the year 2010

Director's Name

No Of

Boards

Rajendra Ambalal Shah 17

BANSIDHAR SUNDERLAL MEHTA 12

SHAILESH VISHNUBHAI HARIBHAKTI 12

SURESH NARSAPPA TALWAR 11

DEEPAK SHANTILAL PAREKH 11

BALAJI RAO JAGANNATH RAO DOVETON 9

BIDHUBHUSAN NABAGHAN SAMAL 9

Brij Behari Tandon 9

DARIUS ERACH UDWADIA 9

NASSER MUKHTAR MUNJEE 9

PIYUSH GUNWANTRAI MANKAD 9

PRADIP PANALAL SHAH 9

Dharmendar Nath Davar 8

DILIP JAYANTILAL THAKKAR 8

HAIGREVE KHAITAN 8

N. SRINIVASAN 8

PRADIP KUMAR KHAITAN 8

RAKESH RADHEYSHYAM

JHUNJHUNWALA 8

RAMADORAI VENKATRAMAN

SUBRAMANIAN 8

SUBHASH CHANDRA BHARGAVA 8

SURENDER KUMAR TUTEJA 8

Vijay Mallya 8

AMAL GANGULI 7

ARUN KUMAR PURWAR 7


Director's Name No Of Directors

RAJENDRA AMBALAL SHAH 13



NARESH RAMA KANT CHANDRA 11

BRIJ BEHARI TANDON 10


BALAJI RAO JAGANNATH RAO DOVETON 9



DHARMENDAR NATH DAVAR 9





AMAL GANGULI 8

DARIUS ERACH UDWADIA 8

HAIGREVE KHAITAN 8

N. SRINIVASAN 8

NASSER MUKHTAR MUNJEE 8

OMKAR MIHIR GOSWAMI 8

PRATHIPATI ABRAHAM 8

RAKESH RADHEYSHYAM JHUNJHUNWALA 8


SUBRAMANIAN 8

SUBHASH CHANDRA BHARGAVA 8

VIJAY MALLYA 8


Director's Name No Of Directors

RAJENDRA AMBALAL SHAH 16




NARESH RAMA KANT CHANDRA 11


Nasser Mukhtar Munjee 11

BRIJ BEHARI TANDON 10

Darius Erach Udwadia 10


BALAJI RAO JAGANNATH RAO

DOVETON 9


DHARMENDAR NATH DAVAR 9


N. SRINIVASAN 9



AMAL GANGULI 8

JAYANT NARAYAN GODBOLE 8

Jitender Balakrishnan 8

OMKAR MIHIR GOSWAMI 8

PRATHIPATI ABRAHAM 8

RAKESH RADHEYSHYAM

JHUNJHUNWALA 8


SUBRAMANIAN 8

RATAN NAVAL TATA 8

Fig. 3a: Betweenness centrality of 25 top directors in year 2010

Figure 4b: Betweenness centrality of top 25 directors in year 2011

Figure 4c: Betweenness centrality of top 25 directors in year2012

Figure 5a: Degree centrality of top 25 directors for the year 2010

Figure 5b: Degree centrality of top 35 directors for the year2011

Figure5c: Degree centrality of top 25 directors for the year2012

Figure 6a: Eigenvector of top 25 directors for the year 2010



Figure 7a: Degree centrality of the companies top 25 directors are members of for the year 2010

Figure 7b: Betweenness centrality of the companies top 25 directors are members of for the year 2010

Figure 7c: Eigen centrality of the companies top 25 directors are members of for the year 2010



Figure 8c: Eigenvector centrality of the companies top 25 directors are members of for the year 2011



Figure 9c: Eigenvector centrality of the companies top 25 directors are members of for the year 2012

References

Berle, A.A., Jr. and G.C. Means, 1932. The modern corporation and private property (Macmillan, New

York).

Hillman, A. J., & Dalziel, T. (2003). Boards of directors and firm performance: Integrating agency and

resource dependence perspectives. Academy of Management review, 28(3), 383-396.

Mizruchi, M. S. (1996). What do interlocks do? An analysis, critique, and assessment of research on

interlocking directorates. Annual review of sociology, 271-298.

Suresh Bhagavatula NSRCEL - NFCG · Suresh Bhagavatula NSRCEL . Additional Research on Corporate Board Interlocks: Identifying influential Directors through analysis of boards interlocks

Documents