Coordination, Control, or Charade? Board Ties Among Members of an Italian Business Group

Coordination, Control, or Charade?Coordination, Control, or Charade?

Board Ties Among Members of an Board Ties Among Members of an Italian Business GroupItalian Business Group

Fabio ZonaBocconi University

Brian BoydArizona State University

Study FocusStudy Focus

Italian business groups are a key feature of the Italian corporate landscape

There are many interlocks, or shared board ties, across members of these groups

No studies have examined what role, if any, is played by these board ties

Corporate Corporate Governance in ItalyGovernance in Italy

Italian system is unique relative to both theAnglo-Saxon model and German system

Classified in Latin subgroup, which is characterized by controlling shareholders and weak capital markets (DeJong, 1997)

Banks and institutional investors are typically not key shareholders

Stock market plays a limited role (Melis, 2000) Two-tier board: Statutory auditors evaluate the

performance of the main board Extensive reforms in the form of commercial law

and governance code, starting in late 1990s

Italian Business Italian Business GroupsGroups

Many flavors of business groups – such as keiretsu and chaebol

Pyramids are legally independent firms, controlled by a unidirectional chain of ownership

Dominant form for Italian industrial firms Majority of firms traded on the Milan Stock Exchange

use pyramidal organization In 2002, there were 52,000 Italian business groups,

covering 4.9 million employees One-third of groups have fewer than 500 employees Pyramids can be found in other countries as well.

Also used by US railroad companies in 1800s.

A Simplified Pyramidal GroupA Simplified Pyramidal Group

PesentiFamily

Source: Volpin (2002) JFE 64: 69

ItalmobiliareItalmobiliareItalmobiliareItalmobiliare

owns45%

(diversified(diversifiedinvestments)investments)

ItalcementiItalcementiItalcementiItalcementi

(cement)(cement)owns54%

Franco TosiFranco TosiFranco TosiFranco Tosi

(steam turbines)(steam turbines)

owns62%

CementerieCementeriede Sardegnade Sardegna

CementerieCementeriede Sardegnade Sardegnaowns

78%(cement)(cement)

CementerieCementerieSicilianeSiciliane

CementerieCementerieSicilianeSiciliane

owns74%

(cement)(cement)

Other holdings include Other holdings include food packaging, banking, food packaging, banking, insulation, othersinsulation, others

Other holdings include Other holdings include food packaging, banking, food packaging, banking, insulation, othersinsulation, others

Pesenti family is able to control Sardegna Pesenti family is able to control Sardegna and Sicliane with lesser investment than if and Sicliane with lesser investment than if held directly.held directly.

The ownership of Italmobilaire is the The ownership of Italmobilaire is the weakest link, as the family does not own or weakest link, as the family does not own or control 51%.control 51%.

Pesenti family is able to control Sardegna Pesenti family is able to control Sardegna and Sicliane with lesser investment than if and Sicliane with lesser investment than if held directly.held directly.

The ownership of Italmobilaire is the The ownership of Italmobilaire is the weakest link, as the family does not own or weakest link, as the family does not own or control 51%.control 51%.

Are Pyramids Beneficial or Are Pyramids Beneficial or Harmful?Harmful?

ProsPros Historically, financial markets in

Italy are underdeveloped – stock exchange is a relatively recent phenomenon. Pyramids allow firms with access to capital.

Pyramids can facilitate access to resources and other transactions across firms

Together, firms may be better able to withstand financial crises

Ability to focus on long-term outcomes

ConsCons Creates a greater imbalance

between ownership and control Potential for firms at lower end

of pyramid to suffer in order to support higher level members

Blockholder may pursue interests other than maximizing returns

‘Tunnelling’ and other unethical or illegal activities

Overall, there is no evidence that a Overall, there is no evidence that a pyramidal form is inherently harmful.pyramidal form is inherently harmful.

Overall, there is no evidence that a Overall, there is no evidence that a pyramidal form is inherently harmful.pyramidal form is inherently harmful.

What Role Do What Role Do Boards Play?Boards Play?

Coordination:Coordination: Boards work as a tool to manage environmental constraint. Key roles include providing access to resources and information.

Control:Control: Boards work as representatives of owners, and provide monitoring and control roles.

Charade:Charade: Boards are window dressing and serve no useful purpose. These exist because of legal requirement and to enhance legitimacy.

Evidence on Evidence on Italian BoardsItalian Boards

2002 code of best practice highlight both agency and resource dependence roles

Monitoring, strategy, and resource roles all contribute to firm performance (Zona & Zattoni, 2007)

Broad range of activity across firms – some craft strategy, others are rubber stamps (Ravasi & Zattoni, 2006)

83 percent of directors act as agents of blockholders, despite reform guidelines

Mixed results on use of interlocks across studies

Related Research:Related Research:Subsidiary BoardsSubsidiary Boards

Key RolesKey Roles External

- Acquire local information and resources from host country

Internal- Coordination with

other subsidiaries, - Conduit for

management oversight

(Leskell & Lindgren, 1982; Kriger, 1988)

FindingsFindings Historically, SB’s have

been exemplars of ‘rubber stamp’ model

SB’s becoming more active over time

Boards of subsidiaries with partial ownership more active than WOS

HypothesesHypotheses

H1:H1: Interlocks between pyramid boards serve as a monitoring mechanism, and will be driven by ownership (agency explanation).

H2:H2: Interlocks between pyramid boards serve as a mechanism to procure information and resources, and will be driven by financial transactions between members (resource dependence explanation).

Null:Null: Pyramid boards are rubber stamps. Lack of support for hypotheses can be construed as indirect support for this rationale.

Board Ties: Board Ties: A Network ViewA Network View

Most studies use counts of interlocks to measure board ties across firms

Simple tie counts assume that all interlocks are equally valuable

Network analysis allows assessment of the value of an individual interlock based on its position in a broader system of ties

Consequently, we use network centrality as the Consequently, we use network centrality as the dependent variable for hypothesis testsdependent variable for hypothesis tests

Network Network CharacteristicsCharacteristics

Measures of Centrality:Measures of Centrality:

Power: Taps the range of direct connections Power: Taps the range of direct connections available to an actor (degree centrality)available to an actor (degree centrality)

Betweenness: Role as a unique intermediary Betweenness: Role as a unique intermediary between other actorsbetween other actors

Reach: Taps the range of total connections Reach: Taps the range of total connections available to an actor through intermediariesavailable to an actor through intermediaries

Measures of Centrality:Measures of Centrality:

Power: Taps the range of direct connections Power: Taps the range of direct connections available to an actor (degree centrality)available to an actor (degree centrality)

Betweenness: Role as a unique intermediary Betweenness: Role as a unique intermediary between other actorsbetween other actors

Reach: Taps the range of total connections Reach: Taps the range of total connections available to an actor through intermediariesavailable to an actor through intermediaries

Network Network CharacteristicsCharacteristics

Network Network Characteristics:Characteristics:

BetweennessBetweenness

Network Network Characteristics:Characteristics:

PowerPower

Network Network Characteristics:Characteristics:

ReachReach

FRANCE

SPAIN

ITALY

GERMANY

POLAND

GREAT

BRITAIN

IRELAND

PORTU

GAL

AUSTRIA

DENMARK

CZECH

REP.

UKRAINE

ROMANIA

HUNGARY

SLOVAKIA

Geographic Geographic StructuralStructural

Holes Holes

Hypothesis 3Hypothesis 3

Internationalized firms have greater information and resource requirements

Pyramids have many foreign members Lateral ties in MNCs used to acquire market

information, supplier and customer contacts, access to local capital

Indirect/secondary ties have high information value Geographic diversity of TMTs facilitates learning and

decision-making H3:H3: Boards of foreign members will be more densely

connected to the pyramid network

SampleSample

We obtained proprietary data from a large Italian corporate group. Key characteristics:

2005 revenues of 30 billion euros, assets of 96 billion euros

155 companies operating in five business units 63 percent of companies headquartered outside

Italy, mainly in Europe and South America

VariablesVariables

Board centrality:Board centrality: Used UCINET to creates measures of betweenness, power, and reach centrality for each firm.

Financial dependence:Financial dependence: Analyzed over 8,000 cross-firm transactions to identify resource exchange across group members

Ownership:Ownership: Counted inward, outward and overall ownership links for each firm

Nationality:Nationality: Whether firm is headquartered in Italy or foreign

Firm size:Firm size: Net sales

Descriptive StatisticsDescriptive Statistics Board characteristics:Board characteristics: A typical firm had 4 directors, and

interlocks with 6 – 7 other firms- 26 firms were isolates (no ties to other firms)- One firm was tied to 27 other pyramid members

Financial dependence:Financial dependence: A typical firm reported having financial transactions with 5 other firms- 32 firms reported no transactions with other members- One firm had transactions with over half (57%) of pyramid

members Ownership:Ownership: Firms had between one and five owners, with an

average of slightly over one owner per firm.- One firm had a direct ownership stake in 40 other pyramid

members Nationality:Nationality: 63 percent of firms headquartered outside Italy

Interlocks

FinancialTies

Reach

Between

Power

Ownership

Nationality

Size

Structural Structural Model of Model of

HypothesesHypotheses

Summary Summary StatisticsStatistics

Global Fit:Global Fit:

2 = 36.84

2 /df = 4.6

GFI = 0.94

RMSR = .07

Predictive Power:Predictive Power:

CEDseq = 0.29

R2between = 0.28

R2power = 0.91

R2reach = 0.66

Mean R2dv = 0.62

Interlocks

Financial Ties

Between

Power

Ownership

Nationality

Size

Reach

A single-factor null model A single-factor null model

reported a reported a 2 = 117.17 (14df).117.17 (14df).

A single-factor null model A single-factor null model

reported a reported a 2 = 117.17 (14df).117.17 (14df).

Interlocks

FinancialTies

Reach

Between

Power

Ownership

Nationality

Size

.25 (2.8)

.09 (1.0)

-.27(3.6)

.28(3.4)

.56 (6.7)

.86(10.7)

1.0

Tests of Tests of HypothesesHypotheses

ImplicationsImplications

To date, no studies have addressed the role of board ties among pyramid members. This project is a first step toward resolving that gap.

Evidence points strongly toward the board as a resource role

Our results consistent with positive versus negative arguments for pyramids.

Future ResearchFuture Research Normative models – how do high performing

firms act? Replication – other pyramids in Italy, and

board ties within other types of business groups.

Compare board roles for small and large pyramids

Kriger’s work on subsidiary boards used both archival and survey data. Desirable to study perceived roles of pyramid directors