Firm vertical boundaries, internal capital markets, and firm performance

• Published date: 01 January 2021
• Author: Jaideep Shenoy
• Date: 01 January 2021
• DOI: http://doi.org/10.1111/eufm.12269

Eur Financ Manag. 2021;27:59–97. wileyonlinelibrary.com/journal/eufm © 2020 John Wiley & Sons Ltd.

|

59

DOI: 10.1111/eufm.12269

ORIGINAL ARTICLE

Firm vertical boundaries, internal capital

markets, and firm performance

Jaideep Shenoy

School of Business, University of

Connecticut, Storrs, Connecticut

Correspondence

Jaideep Shenoy, School of Business,

University of Connecticut, Storrs, CT

06269, USA.

Email jaideep.shenoy@uconn.edu

Abstract

We investigate how vertical relatedness between

business segments of firms affects capital allocation

within internal capital markets. Using a battery of

tests including exogenous import tariff reductions, we

show that investments flow toward segments with

better investment opportunities in firms with sig-

nificant vertical relatedness between segments. This

benefit of vertical relatedness is more pronounced in

economic environments prone to information pro-

blems and in imperfectly competitive industries.

Firms with significant inter‐segment vertical relat-

edness also exhibit superior productivity and operat-

ing profitability. Overall, we show that superior

capital allocation is a channel through which vertical

integration impacts real outcomes of firms.

KEYWORDS

internal capital markets, firm boundaries, firm performance,

product market relations, vertical integration

JEL CLASSIFICATION

G30; L25; L22

EUROPEAN

FINANCIAL MANAGEMENT

I am grateful to John Doukas (the Editor) and two anonymous referees whose comments have improved the paper

substantially. I am also grateful for helpful comments from Omesh Kini, Matthew Billett, Gonul Çolak, William Elliott,

Gerald Gay, Charles Hadlock, Jayant Kale, Naveen Khanna, Richmond Mathews, Tom Noe, Chip Ryan, Husayn

Shahrur, Paul Spindt, Venkat Subramaniam, Sheri Tice, as well as from seminar participants at the 2009 Western

Finance Association Meetings, the 2008 Financial Management Association Meetings, Tulane University, California

State University (Fullerton), Iowa State University, Virginia Commonwealth University, and Georgia State University.

All errors and omissions are mine.

1|INTRODUCTION

The literature on the economics of organizations suggests that firms optimally choose their vertical

boundaries to enable value‐enhancing decisions (e.g. Bolton & Scharfstein, 1998; Holmstrom &

Roberts, 1998; Lafontaine & Slade, 2007). Several studies examine the factors that influence the

variation in vertical integration and highlight the costs and benefits of vertical integration

(e.g. Acemoglu, Johnson, & Mitton, 2009; Fan & Lang, 2000). In addition, researchers have shown

that vertical mergers through which firms expand boundaries and vertical divestitures through which

firmsshrinkboundariesaregenerallymotivated by efficiency considerations (e.g. Eckbo, 1983;Fan&

Goyal, 2006; Fan & Lang, 2000; Jain, Kini, & Shenoy, 2011; Kedia, Ravid, & Pons, 2011;Shenoy,2012).

We focus on an issue that is largely underexplored in the literature on vertical integration.

In particular, we examine how vertical relatedness between business segments affects the efficiency of

capital allocations within internal capital markets. We also study the impact of capital allocation

efficiency on firm productivity and operating profitability.

Our research questions are motivated by the theoretical literature which investigates how the type

of integration influences internal capital market allocations. Grant (1996) suggests that firms vertically

integrate into related industries as each stage of production can benefit from the knowledge gained at

the other stages of production. In support of this theory, Atalay, Hortaçsu, and Syverson (2014)find

that vertical integration facilitates the transfer of intangible inputs within firms. Relatedly, Stein (1997)

shows that a corporate headquarters pursuing a related diversification strategy is able to rank‐order

projects more accurately, leading to superior capital allocation within its internal capital market.

Maksimovic and Phillips (2002) show that a vast majority of conglomerate firms grow across

industriesinamannerthatisconsistentwithprofitmaximization and efficiently allocate resources in

response to industry demand shocks. Ozbas (2005) argues that unrelated integration worsens the

quality of information needed by the headquarters to allocate corporate resources. In contrast, the

corporate headquarters in a related diversifier, such as a vertically integrated firm, has better‐quality

information about the prospects of all its segments aiding internal capital allocations. Finally, market

structure theories argue that vertically integrated firms operating in imperfectly competitive

environments acquire information about supplier and customer industries due to network effects

or increased coordination with rival firms (e.g. Dass, Kini, Nanda, Onal, & Wang, 2014;Kedia

et al., 2011; Lafontaine & Slade, 2007), which is helpful in capital allocation decisions.

Based on the above arguments, we posit that the corporate headquarters in vertically integrated

firms will likely have an informational advantage in allocating resources across business segments.

Thus, we expect the internal capital markets to work more efficiently in vertically integrated firms

than in firms that engage in unrelated diversification. Furthermore, if vertically integrated firms

indeed allocate corporate resources to projects with valuable investment opportunities, then this

should be reflected in superior operating performance. Accordingly, we expect vertically integrated

firms to exhibit superior productivity and profitability compared to firms that engage in unrelated

diversification.

Our sample comprises of 46,600 multi‐segment firm–years during the period 1984–2018 for which

we obtain data on the Compustat industry‐annual and segment databases. We use the benchmark

input–output tables of the US economy to find the industry‐level vertical relatedness coefficients for

the different industries in which our sample firmshavebusinesssegments.Wefollowthemetho-

dology in Fan and Lang (2000), weighting these industry coefficients by the segment sales weights

and arriving at a firm‐level measure of vertical relatedness. Based on prior literature, we classify a

multi‐segment firm as vertically integrated if the firm‐level vertical relatedness measure exceeds 5%

(e.g. Fan & Goyal, 2006;Shenoy,2012). We identify firms that meet this criterion by creating a

60

|

EUROPEAN

FINANCIAL MANAGEMENT

SHENOY

dummy variable VrcDum_5pct. As a robustness test, we also create a dummy variable VrcDum_1pct

(VrcDum_10pct)basedonafirm‐level vertical relatedness coefficient of 1% (10%).

To measure the efficiency of internal capital market allocations, we employ the relative

value added by allocation (RVA) measure from Rajan, Servaes, and Zingales (2000) and Peyer

and Shivdasani (2001). Intuitively, this measure captures the association between segment

investment and segment investment opportunities aggregated across all business segments of

the firm. RVA would be positive in firms that allocate more investment to segments with better

investment opportunities.

We start our analyses by comparing the average value of RVA for vertically integrated and

non‐vertically integrated firms. We find that firms with significant vertical relatedness exhibit

positive and statistically significant values of RVA. We also find that the difference in means

between the vertical and non‐vertical sample of firms is positive and significant. This shows that

investment flows more toward segments with better investment opportunities in vertically

integrated firms. Next, we take a peek inside vertically integrated firms and compare the values

of RVA for vertically related and non‐vertically related segments. We find that the average RVA

for vertically related segments is indeed positive and statistically significant, while the average

RVA for non‐vertically related segments is generally positive but statistically insignificant. This

suggests that the superior allocative efficiency in vertically integrated firms is primarily driven

by the vertically related segments. The above results hold for all three measures of vertical

relatedness VrcDum_5pct,VrcDum_1pct, and VrcDum_10pct.

We next conduct cross‐sectional analyses to further validate the relation between vertical

relatedness and internal capital market efficiency. Following Rajan et al. (2000) and Berger and

Ofek (1995), we control for horizontal relatedness or firm focus by including the segment‐sales‐

based Herfindahlindex of the firm in our regressions. The higher the Herfindahl index thegreater

is the concentration of the firm's activities in an industry, which would imply that the segments

are more horizontally related. Since horizontal diversifiers are also able to rank projects more

accurately due to higher correlation in project ranking errors, we expect a positive relation

between the Herfindahl index and the efficiency of internal capital allocations.

1

We also include

firm size and a diversity in investment opportunities measure as additional control variables. We

document a positive and significant relationship between VrcDum_5pct and RVA.Ourinferences

remain unchanged if we use VrcDum_1pct or VrcDum_10pct to identify vertically integrated

firms. The economic impact of vertical relatedness on the efficiency of internal capital

market allocations is meaningful. For example, a change from 0 to 1 in VrcDum_5pct explains

8.46–12.22% of the standard deviation in RVA. Finally, we find that the segment‐sales‐based

Herfindahl indexis positively related to RVA when horizontal overlaps exist between primary and

secondary segments of the firm. This result indicates that firm focus benefits internal capital

allocations when horizontal connections exist between segments within the firm.

We next examine some specific channels through which vertical relatedness benefits internal

capital allocations. First, we investigate how uncertainty in segment investment opportunities affects

the relation between vertical relatedness and internal capital allocations. It is likely that the in-

formation environment is more challenging when firms face higher dispersion in segment investment

opportunities. If the dispersion in segment investment opportunities is high, internal capital markets

are more valuable for vertically related firms since they facilitate knowledge transfer across segments

about input/output markets and superior ranking of projects by headquarters. These informational

1

We conduct additional tests to rule out the possibility that horizontal relations spuriously drive our results.

SHENOY EUROPEAN

FINANCIAL MANAGEMENT

|

61