The Control and Performance of Joint Ventures

• Author: Kyojik Song,Haowen Luo,Tomas Mantecon
• DOI: http://doi.org/10.1111/fima.12100
• Published date: 01 May 2016
• Date: 01 May 2016

The Control and Performance of Joint

Ventures

Tomas Mantecon, Kyojik Song, and Haowen Luo∗

We analyze the control and performance of assets operating in joint ventures (JVs). Control in

JVs is determined by the allocation of voting rights and by the contracts that governthe JVs. This

hybrid allocation of control seeks to reduce the potential for ex post opportunism. The results

suggest that contractual provisions encourage collaboration and improve JV performance when

one of the partners accepts a minority position on the board. The analysis also reveals that with

the exception of JVs with contractual option provisions, assets operating in our sample of JVs

generate higher returns than assets in fully controlled subsidiaries.

Managers have different alternatives for expanding the boundaries of their operations. These

alternatives, which differ bythe degree of control, include acquiring assets, sharing the ownership

of assets in joint ventures (JVs), and contracting for the use of assets in arm’s-length contracts.

The study of these alternatives has been a central topic in financial economics since the pioneering

framework of Coase (1937). Holmstrom and Tirole (1989) and Coase (1994) express concerns

about the very low evidence to theory ratio in this field due to the lack of essential information

needed to conduct empirical research. The proportion of theoretical to empirical work has not

changed significantly since these studies were published. Our goal is to provideempirical evidence

regarding the governance and performance of JVs, prevalent arrangements characterized by

control sharing among several independent entities.

JVs are described as “hybrid” structures as their governance shares characteristics of the hierar-

chical control within a firm and the lack of def initivecontrol in market transactions (Williamson,

1979). Similar to independent firms, JVs are governed by a board of directors, and the partners re-

ceive control and cash flowrights propor tional to their equity investments.Similar to arm’s-length

market transactions, the terms of the collaboration are governed by contracts. Little is known,

however, about how these two institutional arrangements, the board of directors and contractual

provisions, interact to configure the governance of JVs.

The first part of the study provides content to the generalization of JVs as hybrid structures

of control. We analyze the composition of the boards of JVs to ascertain the allocation of voting

rights, as well as contractual clauses that are likely to affect the control of JVs. Specifically,

we focus on supermajority voting and options provisions that condition the use of power in

collaborative arrangements (Demski and Sappington, 1991; Habib and Mella-Barral, 2007).

We also explore cross-equity investments among partners as alternative methods of monitoring

(Shleifer and Vishny, 1986; Grossman and Hart, 1988; Pisano, 1989) and identify the real

We are grateful to an anonymous referee, Raghavendra Rau (Editor), and James Conover for their valuable comments

and suggestions.

∗TomasMantecon is an Associate Professor in the Finance, Insurance,Real Estate, and Law Department of the University

of North Texas in Denton, TX. Kyojik Song is a Professor at Sungkyunkwan University in Seoul, South Korea. Haowen

Luo is a Ph.D. student in the Finance, Insurance,Real Estate, and Law Department of the University of North Texas in

Denton, TX.

Financial Management •Summer 2016 •pages 431 – 465

432 Financial Management rSummer 2016

authority in JVs, the management of day-to-day operations, which, in some cases, differs from

the formalized authority of the board (Aghion and Tirole, 1997).

We analyze the interaction of the board and contracting provisions to identify the allocation

of control in a sample of 109 JVs. We find that in JVs with boards controlled by one of the

partners, contracts are more likely to include supermajority voting and option clauses. Minority

partners are also more likely to acquire blocks of equity in the controlling partner. These findings

suggest that firms are willing to accept a minority position in the board if the contract includes

provisions to avoidtheir par tners behavingopportunistically once assets have been contributed to

the JV. This analysis confirms that control in JVs is multidimensional and is determined by joint

interactions among the allocation of voting rights, contractual clauses, and equity investments

among JV partners.

In the second part of our study, armed with our knowledge about the factors affecting the

allocation of control in JVs, we turn our attention to the association between control and per-

formance. A rich body of literature, which we summarize in the next section, highlights the

costs and benefits of ownership sharing. Some empirical evidence suggests that equal control

sharing is the preferred choice when organizing JVs.1Theoretically, the observed pervasiveness

of the equal distribution of voting rights could be the optimal outcome of competitive forces that

motivate efficiency.The dominance of the equal distribution of ownership in JVs can also be due

to fears of ex post opportunism by the controlling partner. Independent firms would be unwilling

to surrender control of assets unless they are forced by special circumstances. For instance, when

the value of their relative contributions is clearly inferior or they have weaker bargaining power

due to strategic or financial circumstances. In this alternative explanation, there is not a clear

expectation regarding the direction of the association between the control and performance of

JVs.

The analysis indicates that JVs perform better when their boards are controlled by one of

the partners, but this association is restricted to the subsample of JVs that include options

and supermajority clauses in their contracts. Robustness checks reveal that these results are

not explained by the potential for endogeneity between the allocation of control in JVs and their

performance. The contractual provisions analyzedin this study are asymmetrically associated with

performance depending on the allocation of voting rights. Supermajority and option provisions

and the effective control of JVs are associated with better performance in JVs controlled by one

partner than in JVs with split boards. One explanation for these findings is that these contractual

provisions reduce the chance of ex post opportunism, encouraging collaboration and fostering

the performance of JVs controlled by one partner. However, partners in equally controlled boards

have effective veto power that they can use to protect their interests. We conjecture that partners

bargain for the inclusion of these restrictiveprovisions in JVs with equally split boards to facilitate

the dissolution of the JVs when they attach a lower probability of success to the collaboration.

In the third section, we contrast the performance of JVs and fully controlled operations. Both

organizational arrangements involve benefits and costs that should determine their relative per-

formance. Some empirical facts, described in detail in the next section, denote a clear preference

for the control of operations. This preference for control may be the outcome of competitive pres-

sures that yield the most efficient organizational arrangement. The existence of control sharing

in JVs could be explained as a “second best” alternative, only undertaken when full control is

1Forinstance, 65% of the approximately 15,000 JVs in the Thomson Reuters Securities Data Company(SDC) announced

in the period 1994-2011 with at least one US partner reported as having equal ownership. Ownership information

provided bySDC often refers to the allocation of cash flow rights instead of voting rights, but we find a similar ownership

breakdown in our sample: 66% of the JVs in our sample have equallysplit boards.

Mantecon, Song, & Luo rThe Control and Performance of Joint Ventures 433

more costly.2Analternative explanation for the observed preference for control, based on agency

arguments, is that managers can realize private benefits from the control of operations carried

out inside the firm that cannot be obtained when control is shared with other partners in JVs.

We find that assets operating in JVs generate higher returns than assets in fully controlled

subsidiaries, except those in JVs with option provisions, which perform similarly to assets op-

erating inside firms. Consistent with an agency explanation of the preference for control, JVs

perform better than partners with high levels of the Gompers, Ishii, and Metrick (2003) index of

governance.

We investigate the prospect that managers may choose to share control in JVs in the presence

of high levels of risk (Johnson and Houston, 2000; Robinson, 2008). If this is the case, the

riskiness of projects could affect both the choice of control overoperations and their perfor mance.

Our results confirm that JVs are chosen in the presence of risk and uncertainty, but superior

performance in our sample of JVs also is obtained when we consider the potential consequences

of the endogeneity between the choice and the performance of JVs. This finding, however, has

to be taken with caution as, by construction, our sample is not random, but it consists of large

JVs, the performance of which may not be representative of the average JV.

The analysis of the ownership and performance of JVs contributes to several strands of re-

search in financial economics. Groundbreaking theoretical work analyzing the integration and

nonintegration of operations provides a rich framework for the analysisof the optimal allocation

of control in business organizations (Grossman and Hart, 1986; Hart and Moore, 1990; Rajan

and Zingales, 1998). Theorists have extended these models to the allocation of ownershipin par t-

nerships (Cramton, Gibbons, and Klemperer, 1987; Darrough and Stoughton, 1989; Aghion and

Tirole, 1994; Halonen, 2002; Schmitz, 2006; Andreoni and Bernheim, 2009). Weprovide content

to the generalization that JVs are “hybrid” structures of control by describing the interaction of

voting rights and contractual provisions to determine the allocation of control rights in JVs. We

contend that theoretical models that fail to consider the different dimensions of control in JVs

offer an incomplete description of the institutional structure of these entities.

In the absence of transaction costs, Coase’s (1960) irrelevance theorem of organizational design

predicts that the allocation of control inside JVs should have no implications on performance.

Prior empirical studies concerning the association between the control and performance of JVs are

limited and yield contradictory results (Hauswald and Hege, 2005; Slovin,Sushka, and Mantecon,

2007). We use the control structure and accounting information provided in our sample to extend

the empirical work in this area. We also use our sample to test the relative performance of JVs and

fully controlled operations in an attempt to provide empirical evidenceregarding the link between

performance and control over operations, adding to extant work comparing the performances of

strategic alliances and projects carried out inside firms (Gomes-Casseres, Hagedoorn, and Jaffe,

2006; Palia, Ravid, and Reisel, 2007).

The remainder of the paper is organized as follows. Section I summarizes prior literature on

the governance and performance of JVs. Section II explains the sample construction. Section

III analyzes the hybrid structure of control in JVs. Section IV studies the association between

the control and performance of JVs. Section V contrasts the accounting performance of assets

operating in a JV with that of assets under the control of the firm. We present our conclusions in

Section VI.

2For instance, a firm would rather acquire other firms’ assets, but asymmetries of information may prevent partners from

agreeing on a transaction price or the buyer may not have sufficient resources to complete the transaction.