Diversification Discount and Corporate Governance in … I. Introduction This paper investigates the relationships among industrial diversification, corporate gov-ernance, and firm

427

I. Introduction

This paper investigates the relationships among industrial diversification, corporate gov-ernance, and firm value in Japan, focusing on the period from 2004 to 2012. Berger and Ofek (1995) and Lang and Stulz (1994) find that U.S. firms diversified across industries trade at a substantial discount relative to focused (single-industry) firms operating in the same industries. This phenomenon, commonly known as the diversification (conglomerate) discount, has attracted considerable attention from financial economists. An important theme of this literature is the link between corporate governance and diversification: i.e., the diver-sification discount may arise because managers pursue diversification excessively, especial-ly when they are ill-disciplined due to a weak governance system.

The value destruction by diversified firms can be a serious problem in Japan as well. Since the turn of the last century, a large number of Japanese firms have been engaged in re-structuring under the slogan “choose and focus.” Nevertheless, the question of whether in-vestors value diversified Japanese firms less than comparable focused firms has not been studied sufficiently due to a historical data constraint—segmental reporting data, which are essential to the estimation of diversification discounts, were unavailable for Japanese firms in the last century. However, this constraint was removed by accounting reforms in the late

Diversification Discount and Corporate Governance in Japan＊

Tatsuo UshijimaProfessor, Graduate School of International Management, Aoyama Gakuin University

Abstract

It is widely reported that investors value U.S. firms operating in a diversity of industries less than focused firms in the same industries. Based on segmental reporting data of Japa-nese firms from 2004 to 2012, this paper examines whether such a discount for diversified firms exists and how corporate diversification and governance interact. Several important results are obtained. First, diversified Japanese firms are valued 6 to 7% less highly than fo-cused firms. Second, corporate governance increases firm value, but this effect is attenuated for diversified firms. Third, a stronger governance system increases a firm’s probability of refocusing. These results suggest that managerial agency problems exist in the diversifica-tion of Japanese firms. In particular, managers who are ill-disciplined due to a weak gover-nance system resist value-increasing restructuring actions, such as refocusing.

Keywords: diversification; firm value; agency problems; corporate governanceJEL Classification: G32, G34

* This study was funded by KAKENHI (Grant-in-Aid for Scientific Research), Grant Number 24530421.

Policy Research Institute, Ministry of Finance, Japan, Public Policy Review, Vol.11, No.3, July 2015

1990s. For the present century, the effect of diversification on firm value can be estimated for a large sample of firms based on a standard method developed in U.S. literature.

In this paper, I estimate the effect of industrial diversification on firm value based on a panel of non-financial Japanese firms that were publicly traded during the period 2004 to 2012. In so doing, I pay particular attention to the relationship between diversification and governance. My research sample includes approximately 3,900 firms and 27,300 firm-years. Based on segmental reporting data obtained from the Nikkei NEEDS FinancialQUEST da-tabase, firms that operate multiple industrial segments with distinct 4-digit Japan Standard Industry Classification (JSIC) codes are defined as diversified firms. Following Berger and Ofek (1995), the excess value of a firm is measured by matching each of its segments to a representative focused firm that operates in the same industry.

A corporate governance system consists of a number of factors. To analyze the link be-tween governance and diversification as comprehensively as possible, this study employs the corporate governance index developed by Aman and Nguyen (2008). This index is a composite variable of fifteen factors that are obtained from the Nikkei-Cges (Corporate Governance Evaluation System). In addition to the total index devised by Aman and Nguy-en (2008), I also use sub-indices that measures three distinct aspects of governance: board of directors, equity ownership structure, and information disclosure. As is reported in Sec-tion III, the mean values of the total index and sub-indices are significantly lower for diver-sified firms than for focused firms, suggesting that agency problems can be more serious for the former.

In estimating the value of diversification, regressions are performed to control for vari-ous determinants of firm value, such as size, profitability, and investments in tangible and intangible assets. Estimation results strongly suggest that diversified firms are valued lower by investors than the portfolio of representative focused firms in the same industries. The estimated discount for diversified firms is 6% to 7%, which is smaller than that typically re-ported for U.S. firms (10% to 20%). Nevertheless, the diversification discount in Japan is statistically robust in that it is barely affected by the introduction of firm-fixed effects into regressions.

Regressions also reveal that corporate governance is an important determinant of firm value. An aspect of governance that has a large effect on firm value is board structure. In particular, firms with a higher ratio of independent external directors are valued higher. The equity ownership share of foreign investors is also positively and significantly associated with firm value. An important result is that diversification weakens the value-enhancing ef-fect of governance, perhaps because greater industrial scope increases the complexity and opaqueness of the firm and thereby decreases the effectiveness of governance by external shareholders.

I also investigate the effect of governance on the diversifying and refocusing behaviors of firms. Probit regressions demonstrate that diversified firms with a superior governance system, especially in terms of board and ownership structures, are significantly more likely to refocus than firms with an inferior system. In contrast, the effect of governance on the

428 T Ushijima / Public Policy Review

429

firm’s decision to diversify is not significant. These results suggest that a conflict of interest between managers and shareholders generally does not exist in the initial stage of diversifi-cation, but it emerges as firms continue to expand across industries. For instance, managers may cling to failed diversification projects and, consequently, firms may accumulate ineffi-cient businesses as they diversify. The fact that “choose and focus” has been an overarching theme of restructuring suggests that such a tendency exists in many Japanese firms.

The rest of this paper is organized as follows. The next section reviews the literature on the diversification discount with attention to the association between corporate diversifica-tion and governance. Section III describes the sample and the measurement of industrial di-versification, firm value, and governance system. Regressions are performed in Section IV to estimate the diversification discount and examine its relationship with governance sys-tem. Section V performs probit regressions to explore the effect of governance on the diver-sification behavior of firms. The final section concludes this paper.

II. Corporate Diversification and Governance

II-1. Diversification and firm value

The effect of industrial diversification on firm value is a priori ambiguous. It can be pos-itive if operating and/or financial synergies exist among a firm’s businesses. Operating syn-ergy is the economy of scope in a broad sense. That is, operating synergy exists when a firm’s total costs are less than the sum of the costs its businesses would incur if they were operated as standalone units and/or a firm’s revenue exceeds the sums of the revenues its businesses would have obtained as standalone units. A basic mechanism underlying these effects is the intra-firm sharing of valuable resources. For example, firms may use propri-etary technologies developed in one business to improve the competitiveness of other busi-nesses. Brand equity can also be used to increase the sales of multiple businesses.

Financial synergies are classified into two types: the coinsurance effect and efficient capital allocation due to internal capital markets. The coinsurance effect arises when multi-ple businesses with imperfectly correlated cash flows are combined into a firm. Such combi-nation decreases the variance of total cash flow and thereby increases the firm’s debt capaci-ty by decreasing its bankruptcy risk. On the other hand, the internal capital market is a mechanism through which managers allocate funds among alternative investment opportu-nities within their firm. When information asymmetry is serious between investors and firms, external capital markets can fail to allocate funds to higher value opportunities. As stressed by Williamson (1975) and Stein (1997), the internal capital market can achieve more efficient allocation of funds because managers are better informed about their firms’ businesses than external investors.

Synergies do not guarantee that diversification generates profit because the success of a business depends on many other factors as well. If firms hang on to unprofitable businesses, the effect of diversification on firm value can be negative. Diversification can cause a de-


cline in firm value also because it increases the information processing load of managers, which negatively affects the quality of their decisions, as well as the complexity of organi-zational structure and processes. Moreover, financial synergy can be negative if corporate socialism prevails in the internal capital market (Scharfstein and Stein, 2000). That is, scarce funds may be allocated more (less) than optimally to unpromising (promising) businesses because divisional managers engage in rent-seeking activities.

To estimate the effect of diversification on firm value, the values of diversified and fo-cused firms must be compared. However, diversified firms differ considerably from each other in the industry-composition of business portfolio. If this heterogeneity is not account-ed for, the value of diversification can be confounded by the effect of industry-level factors. Berger and Ofek (1995) and Lang and Stulz (1994) address this issue by comparing the val-ue of diversified firms and the value of a size-weighted portfolio of representative focused firms in the same industries.

Specifically, the analysis of Berger and Ofek (1995) is based on the excess value defined as follows:

　　Excess Value it = ln VIV

it

it（

（

(1)

　　 IV Sijt Medianjt= × VSj

t

tit （

（ (2)

where V is the enterprise value of the firm defined as the sum of the market value of equity and book value of debt. IV (imputed value) is the value the firm’s segments would collec-tively command if they had operated as standalone firms. The imputed value is estimated as segmental sales (S) times the valuation multiplier (V/S) summed over all of a firm’s seg-ments. The multiplier for a segment is the median multiplier of focused firms in the same industry.1 The excess value is positive if V is greater than IV (premium) and negative if V is smaller than IV (discount).

Berger and Ofek (1995) find that the mean excess value of U.S. diversified firms is sig-nificantly negative during 1986 to 1991. According to their estimations, diversified firms are on average valued 15% lower than focused firms. Similar results hold when excess value is measured based on segmental assets and cash flow instead of sales. Using a slightly differ-ent approach, Lang and Stulz (1994) also demonstrate that diversified firms trade at a dis-count relative to focused firms.

Following Berger and Ofek (1995) and Lang and Stulz (1994), many studies have esti-mated the effect of diversification on the value of U.S. firms. Consistent with the view that conglomerate mergers have destroyed shareholder wealth and served as a seedbed for large-scale restructuring, such as bust-up takeovers and divestitures, these studies almost invari-

1 The imputed value of focused firms is estimated as firm-wide sales multiplied by the median valuation multiplier of focused firms in the same industry.


431

ably find a significant discount for diversified firms. Therefore, the existence of a diversifi-cation discount is a stylized fact. However, whether the discount emanates from diversification itself is an intensely debated issue. See Martin and Sayrak (2003) and Erdorf et al. (2013) for a survey of the literature.

II-2. Diversification and agency problems

A prominent view regarding the cause of the diversification discount is that it is a revela-tion of managerial agency problems. That is, managers excessively pursue diversification to increase private benefits even at the cost of shareholder wealth. Diversification benefits managers through at least three channels.2 The first is the increase of compensation and perks as larger firms generally pay more to their managers. Non-pecuniary benefits such as personal satisfaction and reputation that managers obtain from managing a larger firm can also drive them toward empire- building through diversification.

The second is entrenchment. Diversification increases the cost for shareholders of re-placing incumbent managers because the more diversified a firm, the more difficult to find an external manager who is knowledgeable about all of the firm’s businesses. Forced turn-over is particularly difficult when the firm’s diversification has been guided by the personal knowledge and expertise of incumbent managers (Shleifer and Vishny, 1989).

The third is personal wealth protection. Diversification decreases bankruptcy risks due to the coinsurance effect. Such risk reduction does not generate value in itself for sharehold-ers who can diversify away wealth risks by investing in many assets. In contrast, managers have considerable exposures to the idiosyncratic risk of their firms because their most im-portant asset is firm-specific human capital, which loses value if the firm fails (Amihud and Lev, 1981). Accordingly, managers have a strong incentive to diversify their firms so as to reduce personal risks.

Many studies supply evidence that is consistent with the agency-theoretic view of diver-sification. Denis et al. (1997) find that firms with larger shares of managerial and block-holder ownerships tend to be less diversified. Hyland and Diltz (2002) supply similar evidence, although the link between a firm’s likelihood to diversify and managerial owner-ship is significantly positive according to their estimations. Anderson and Reeb (2003) find that firms owned more by the founding family are less likely to diversify and are valued more by investors.

Berger and Ofek (1996) provide evidence from the market for corporate control. They find that diversified firms with a larger discount are more likely to be the target of a hostile takeover. Moreover, once acquired, these firms are more likely to be broken up. Jiraporn et al. (2006) observe that firms with stronger takeover protection are more likely to diversify and are more deeply discounted by investors.

Jiraporn et al. (2008) investigate the role of boards of directors. They find that diversi-

2 Aggarwal and Samwick (2003) investigate which type of private benefits drives diversification of U.S. firms.


fied firms with a higher ratio of directors who are “busy” due to multiple directorships are more heavily discounted. This pattern suggests that weak monitoring by the board can result in inefficient diversification. Berry et al. (2006) find that poor operating performance signifi-cantly increases CEO turnover. However, this effect is specific to focused firms. Nam et al. (2006) find that stock-based executive compensations increase firm value, especially for di-versified firms. Anderson et al. (2000) note that the governance systems of diversified firms and focused firms are qualitatively different.

Chen and Chen (2012) take a simultaneous look at a number of governance factors. They find that diversified firms with a more effective governance system allocate funds more efficiently across businesses and are valued higher by the market. Hoechle et al. (2012) estimate the diversification discount while controlling for a number of governance factors. They conclude that the diversification discount partly captures the effect of poor gover-nance.

II-3. Diversification of Japanese firms

Managerial agency problems can also be serious in the diversification of Japanese firms. Well-known episodes of corporate failures (e.g., Kanebo) and massive restructuring to avoid bankruptcy (e.g., Daiei) suggest that Japanese managers do have a taste for empire building. Moreover, managers of Japanese firms are normally promoted from within and therefore tend to serve as an agent of employees. This tendency can also lead to diversification that is excessive from shareholders’ viewpoint. This is because in Japan, where the mobility of la-bor across firm is limited, the value of human capital critically depends on the survival of the firm. Accordingly, employees as well as managers have strong demand for the risk re-duction effect of diversification.

Prior studies suggest that the value of diversification is indeed negative for Japanese firms. For the period 1992 to 1994, Lins and Servaes (1999) identify a diversification dis-count of approximately 10%. They also find that keiretsu relationships promote inefficient diversification. Hiramoto (2002) examines a sample of TSE-listed firms. He finds that diver-sified firms are on average valued 10% lower than focused firms. Fukui and Ushijima (2007) examine the diversification of large manufacturing firms during 1973 to 1998. They report that the Tobin’s Q ratio of diversified firms is significantly lower than that of focused firms. They also find that the discount is larger for firms engaged in unrelated diversification.

These studies based on data for the last century share a problem that diversification through subsidiaries is ignored. In other words, diversification is measured at the unconsoli-dated (parent-only) level because consolidated segmental reporting was not present in Japan before accounting reforms in the late 1990s. For the present century, however, this constraint has been removed. Nakano et al. (2002) estimate the value of diversification based on con-solidated segmental data for 1999 to 2002 and identify a diversification discount of approxi-mately 5%.

To summarize, although prior studies suggest that diversification decreases the value of


433

Japanese firms, evidence for recent periods is sparse. Fukui and Ushijima (2007) demon-strate that refocusing activities of diversified firms were a major driver of corporate restruc-turing at the turn of the last century. Accordingly, the relationship between diversification and firm value could have changed in the past decade or so. In what follows, the value of di-versification for Japanese firms is estimated for the period from 2004 to 2012, with particu-lar attention given to the relationship between diversification and the corporate governance system.

III. Sample and Data

III-1. Diversification and firm value

The sample is based on nonfinancial firms that were publicly traded in Japan during the sample period 2004 to 2012. In measuring the diversification of sample firms, I use segmen-tal reporting data obtained from the Nikkei NEEDS FinancialQUEST database, which as-signs JSIC codes to each segment reported by firms. Diversified firms are defined as firms that have two or more segments with distinct 4-digit JSIC codes. Nikkei assigned up to three JSIC codes to a segment. When a segment has multiple codes, the first (primary) code is used to define the industry. When a firm has multiple segments with same primary code, the segments are merged into one.3

Excess Value, as defined by Equations (1) and (2), is measured based on the June-end market value of equity and the book value of debt for the latest fiscal year ending before June. Following Berger and Ofek (1995), segments are matched to an industry at the 4-digit level if no less than five focused firms exist in the industry. If the 4-digit industry includes no more than four focused firms, matching is performed at the finest lower-digit level at which five or more focused firms exist.4 Overall, 27% of segment-industry matching is per-formed at the 4-digit level; 36% at the 3-digit level; 31% at the 2-digit level, and 6% at the 1-digit level. For focused firms, 61% of the firm-industry matching is made at the 4-digit level; 22% at the 3-digit level; 14% at the 2-digit level; and 3% at the 1-digit level. The seg-mental sales of a diversified firm can fail to sum up to the firm’s total sales. When the devia-tion exceeds 5%, the firm is excluded from the sample. When the deviation is within ±5%, the sales of all segments are increased or decreased by an equal percentage such that the sum of adjusted segmental sales matches the firm’s entire sales. Firms with a financial seg-ment are also excluded from the sample.

Section (1) of Table 1 displays the number of sample firms, the ratio of diversified firms, and the mean number of segments of diversified firms. The sample includes approximately 3,000 firms per year. For the entire period, the sample includes 3,917 firms and 27,307 firm-years. The ratio of diversified firms was stable through the 2000s before increasing in 2011. 3 Therefore, firms reporting multiple segments are treated as focused firms if all of their segments share the same primary JSCI code.4 The JSIC codes of focused firms are obtained from the Corporate Fundamental database of the Nikkei NEEDS.


The mean number of segments decreased in 2011 because firms that diversified in that year had a relatively low degree of diversification. The largest number of segments operated by the sample firms during the sample period is eleven.

Section (2) shows the excess values of diversified firms and focused firms. The mean value for focused firms is positive over the sample period. In contrast, the mean value for diversified firms was negative for 2004, 2005, 2006, and 2001. The mean excess value is greater for focused firms than for diversified firms for all years, with the difference being significant at the 1% level for five years. In 2005, when the difference was the largest, diver-sified firms were on average valued 9% lower than focused firms. When all years are pooled, the difference in mean excess value is 5% and significantly different from zero at the 1% level.

III-2. Corporate governance

To capture a firm’s corporate governance system, which consists of many factors, the present study employs the corporate governance index developed by Aman and Nguyen (2008). This index is based on fifteen variables that cover three important areas of gover-nance: board of directors, equity ownership, and information disclosure. Specifically, the in-dex is based on the flowing variables provided by the Nikkei NEEDS-Cges (Corporate Gov-ernance Evaluation System):

Table 1. Diversification, excess value, and governance indices of sample firms

Year

Board Ownershipstructure

Informationdisclosure

2004 2,878 0.512 3.15 -0.044 0.037 -0.081 (0.00) 4.820 1.102 2.073 1.645

2005 3,031 0.510 3.12 -0.060 0.030 -0.090 (0.00) 5.404 1.423 2.265 1.716

2006 3,096 0.506 3.13 -0.031 0.042 -0.073 (0.00) 5.521 1.461 2.474 1.586

2007 3,165 0.508 3.12 0.007 0.064 -0.057 (0.00) 5.543 1.510 2.512 1.522

2008 3,247 0.505 3.12 0.030 0.058 -0.028 (0.11) 5.597 1.630 2.413 1.554

2009 3,171 0.507 3.12 0.030 0.041 -0.011 (0.54) 5.717 1.743 2.206 1.768

2010 3,101 0.508 3.13 0.016 0.040 -0.024 (0.17) 6.006 1.787 2.245 1.973

2011 2,847 0.558 2.93 -0.023 0.041 -0.064 (0.00) 6.168 1.957 2.092 2.118

2012 2,771 0.579 2.88 0.025 0.042 -0.017 (0.34) 6.092 2.010 2.041 2.041

2004-2012 27,307 0.521 3.076 -0.005 0.044 -0.049 (0.00) 5.922 1.777 2.288 1.857

(Source) The author's calculations based on data obtained from Nikkei NEEDS FinancialQuest and NEEDS-Cges databases.

(1) Diversification (2) Excess value (3) Governance system

# firmsRatio of

diversified firms# segments

(mean) Diversified FocusedGovernance

index

Sub-indices

Difference (p-value)

Table 1Diversification, excess value, and governance indices of sample firms


435

Board of directors

(1) The number of directors (–)(2) The number of executive directors (–)(3) The ratio of independent external directors (–)(4) The ratio of auditors to directors (+)(5) The ease of board renewal (+)

Ownership structure

(6) Ownership share of foreign investors (+)(7) Cross-shareholdings ratio (–)(8) Ratio of stable shareholders (–)(9) Mean value of equity owned by a director (+)(10) Stock options dummy (+)

Information disclosure

(11) Number of auditors’ comments in the past three years (–)(12) Number of changes in accounting policies in the past three years (–)(13) Days from the fiscal-year end to the release of the annual accounting report (–)(14) Number of firms that held a shareholders’ meeting on the same date (–)(15) Rating by Nikko IR on the usefulness of the firm’s website for investors (+)

A plus (minus) sign indicates that the quality of a firm’s governance increases (decreas-es) with the variable. The governance index is the number of variables that is better than the median. Here, better means larger for variables with a plus sign and smaller for variables with a negative sign. Accordingly, the index takes an integer value between zero and fifteen and increases with the quality of a firm’s governance system. To capture longitudinal varia-tions as well as cross-sectional variations during the sample period 2004 to 2012, I calculate this index based on the period median value of all firms included in the NEEDS-Cges.5

The above governance index comprehensively captures a firm’s governance system. However, it obscures the contributions of individual components of the system. To over-come this problem, the sub-indices for each of the above three areas of governance are also used in estimations. In addition, the effects of five specific factors—foreign ownership ratio, stable shareholding ratio, managerial ownership ratio, board size (the number of directors), and independent external director ratio—are also examined.6

5 NEEDS-Cges changed the definition of stable shareholders in 2005. To ensure continuity of data, for all firms, the stable shareholdings ratio for 2004 was replaced by the value for 2005 in this study.


Section (3) of Table 1 shows the mean values of governance indices. The mean value of the governance index indicates that the quality of governance of sample firms generally im-proved over the sample period. The mean values of the sub-indices show that this improve-ment was mainly caused by changes in two areas, board of directors and information disclo-sure. In contrast, the mean ownership structure index declined after 2008.

VI. Analysis of Firm Value

VI-1. Regression model

Diversified firms differ from focused firms in factors other than industrial scope as well. Regression is performed in this section to estimate the value of diversification while con-trolling for other potential determinants of firm value. Accounting for the effect of gover-nance system can be particularly important. Hoechle et al. (2012) find for U.S. firms that the diversification discount partly captures the effect of the corporate governance system. Their regressions show that the diversification discount survives a control for governance factors. However, the control renders the estimated discount smaller.

To estimate the effects of diversification and governance on firm value based on an un-balanced panel of public firms, the regression model is specified as follows:

　　EVit＝α＋β・Divit＋γ・Govit＋λ・zit＋ηi＋∊it (3)

where the dependent variable (EV) is the excess value as defined by Equations (1) and (2), Div is a measure of diversification, Gov is a vector of governance variables, z is a vector of control variables, and η is a firm-fixed effect. The specifications without Gov and firm-fixed effects are also estimated to examine the stability of the coefficient for diversification (β). Diversification is measured by a dummy variable, which takes a value of one if a firm has multiple 4-digit segments and zero otherwise, and the number of 4-digit segments. The gov-ernance variables include the governance index and sub-indices introduced in the last sec-tion. The control variables include logged total assets, EBITDA/sales, capital expenditure/sales, R&D expenditure/sales, marketing expenditure (advertising and sales promotion ex-penses)/sales, and leverage (book value of debt normalized by the market value of equity). All independent variables are for the latest fiscal year closing before June.

Table 2 tabulates the descriptive statistics of regressions variables for the entire sample as well as diversified-firm and focused-firm subsamples. As shown in Table 1, the excess values of diversified firms and focused firms are significantly different. The table suggests that this difference can be caused by many factors other than industrial scope. For instance, diversified firms are on average larger, less profitable, and more leveraged than focused

6 Although Aman and Nguyen (2008) use the mean value of equities owned by managers to measure the level of managerial ownership, I use the share of equities owned by managers, which is more commonly used in the governance literature.


437

firms.

The governance system also differs between diversified firms and focused firms. The mean values of the governance index and sub-indices are significantly smaller for diversi-fied firms than for focused firms. Individual governance variables are also different between the two groups of firms. Specifically, the ownership shares of foreign investors and stable shareholders are larger for diversified firms, but the managerial ownership ratio is greater for focused firms. The average board size is larger for diversified firms. However, the ratio of independent directors is not significantly different between diversified firms and focused firms.

VI-2. Value of diversification

Table 3 reports the estimation results of regressions that do not control for governance factors. Columns (1) and (2) present the results of pooled cross-sectional regressions with-

Table 2. Descriptive statistics of regression variables

Diversified Focused

Mean Mean

[14,214] [13,093]

Excess value 0.019 0.512 -0.005 0.044 -0.049 (0.000)

Diversification dummy 0.521 0.500 1 0 －－

# segments 2.081 1.321 3.076 1 －－

Governance index 5.922 2.453 5.660 6.207 -0.548 (0.000)

Board of directors 1.777 1.293 1.630 1.936 -0.306 (0.000)

Ownership structure 2.288 1.286 2.260 2.319 -0.059 (0.000)

Information disclosure 1.857 1.072 1.770 1.952 -0.182 (0.000)

Board size 8.017 3.430 8.658 7.320 1.338 (0.000)

Ratio of independent directors 0.059 0.113 0.058 0.059 -0.001 (0.623)

Foreign ownership ratio 0.079 0.108 0.086 0.070 0.016 (0.000)

Stable ownership ratio 0.074 0.083 0.086 0.060 0.026 (0.000)

Managerial ownership ratio 0.091 0.139 0.071 0.114 -0.043 (0.000)

Total assets (logged) 10.35 1.569 10.78 9.887 0.892 (0.000)

EBITDA/Sales 0.082 0.092 0.080 0.083 -0.003 (0.000)

Capital expenditure/Sales 0.043 0.060 0.044 0.042 0.001 (0.056)

R&D expenditure/Sales 0.016 0.040 0.015 0.017 -0.002 (0.000)

Marketing expenditure/Sales 0.015 0.036 0.013 0.017 -0.004 (0.000)

Leverage 0.769 1.132 0.910 0.616 0.294 (0.000)(Note) The number of observations are reported in brackets.(Source) The author's calculations based on data obtained from Nikkei NEEDS FinancialQuest and NEEDS-Cges databases.

Total [27,307]

Mean SDDifference (p-value)

Table 2Descriptive statistics of regression variables


out firm-fixed effects. Diversification is measured by the diversification dummy in Column (1) and by the number of 4-digit segments in Column (2). The reported results strongly sug-gest that diversified Japanese firms trade at a discount. The negative coefficient on the diver-sification dummy indicates that, ceteris paribus, diversified firms are valued 6.9% lower than focused firms. This effect is smaller in magnitude than the diversification discount typically reported for U.S. firms. However, it is significant at a high level.

The significantly negative coefficient on the number of segments in Column (2) implies that more diversified firms are more deeply discounted. However, when focused firms are excluded from estimation in an unreported regression, the effect of this variable is no longer significant. As such, the effect of the number of segments estimated in Column (2) likely picks up the difference between diversified firms and focused firms, not necessarily the ef-fect of diversification degree. Berger and Ofek (1995) find that an increase in the degree of diversification does not always result in a deeper discount.

To examine the stability of diversification discount, cross-sectional regressions are per-formed on yearly data. The estimation results are summarized in Columns (3) and (4) in the spirit of Fama and MacBeth (1973). That is, the reported coefficients are the average of

Table 3. Estimation results of diversification discounts

(1) (2) (3) (4) (5) (6)

Diversification dummy -0.069 *** -0.071 *** -0.058 ***(0.006) (0.010) (0.007)

# segments -0.016 *** -0.015 *** -0.027 ***(0.002) (0.004) (0.003)

Total assets (logged) 0.024 *** 0.022 *** 0.024 *** 0.022 *** -0.003 -0.002(0.002) (0.002) (0.006) (0.006) (0.009) (0.009)

EBITDA/Sales 1.306 *** 1.315 *** 1.397 *** 1.407 *** 0.239 *** 0.236 ***(0.034) (0.034) (0.156) (0.156) (0.037) (0.037)

Capital expenditure/Sales 1.657 *** 1.662 *** 1.641 *** 1.646 *** 0.447 *** 0.448 ***(0.050) (0.050) (0.109) (0.108) (0.042) (0.042)

R&D expenditure/Sales 1.735 *** 1.737 *** 1.721 *** 1.723 *** 1.043 *** 1.041 ***(0.072) (0.073) (0.080) (0.078) (0.146) (0.146)

Marketing expenditure/Sales 0.919 *** 0.939 *** 0.916 *** 0.936 *** 0.028 0.028(0.080) (0.080) (0.067) (0.067) (0.152) (0.152)

Debt/Market equity 0.031 *** 0.030 *** 0.027 *** 0.027 *** 0.014 *** 0.014 ***(0.003) (0.003) (0.005) (0.005) (0.003) (0.003)

F-value of firm-fixed effectsR-squared# observations

Fixed effectsFama-MacBethOLS

0.163 0.171 0.783

(Note) In parentheses are standard errors. *** Significant at the 1% level. ** Significant at the 5% level. * Significant at the 10% level.The R-squared of Fama-MacBeth estimations are the mean value of the R-squared of annual cross-sectional regressions.

27,307

17.09***－

0.16027,307

－

0.17427,307

－

27,307 27,307

－

0.78327,307

17.02***

Table 3Estimation results of diversification discounts


439

yearly estimates, while standard errors are based on the variance of yearly coefficients. The results are very similar to the pooled estimation results. The estimated coefficient on the di-versification dummy ranges from -0.110 to -0.059 and is significant at the 0.01 or higher level for eight out of nine years.

Columns (5) and (6) report the estimation results with firm-fixed effects. In estimating the value of diversification, endogeneity is a serious concern because diversification is a re-sult of deliberate decisions of managers. It is possible that the diversification discount ema-nates not from diversification itself but from factors that affected the managerial decision to diversify. Firm-fixed effects cope with such endogeneity due to time-invariant factors. For U.S. firms, Campa and Kedia (2002) report that diversification discounts are halved in size when firm-fixed effects are introduced to regressions. In Column (5) where diversification is measured by the diversification dummy, the effect of diversification is smaller in size than that reported in Column (1), but highly significant. Column (6) shows that the estimated ef-fect of the number of segments becomes larger when firm-fixed effects are included in the regression. These results indicate that the diversification discount in Japan is robust to con-trol for endogeneity.7

VI-3. The effect of corporate governance

Table 4 reports regressions that incorporate governance factors. Pooled cross-sectional regressions are reported in the odd numbered columns, while estimations with firm-fixed ef-fects are reported in the even numbered columns. In all specifications, diversification is measured by the diversification dummy. Across the board, the coefficient on the dummy is negative and significant at the 0.01 level. The estimated size of diversification discount is 6.4–7.2% for pooled cross-sectional estimations and 5.7–5.9% for fixed effects estimations. These estimates are virtually identical to those obtained in the last subsection without a con-trol for governance factors.

In Columns (1) and (2), the effect of the governance system is controlled by the gover-nance index. The coefficients for the index are positive and highly significant with and with-out firm-fixed effects. Therefore, evidence strongly suggests that firm value positively cor-relates with governance quality. However, the coefficient on the governance index in Column (2) is less than half of that in Column (1). This difference suggests that governance is not fully exogenous to firm value. That is, while good governance increases firm value, it is also the case that firms valued higher by investors employ a better governance system.

In Columns (3) and (4), where the governance system is measured by the three sub-indi-ces, the effect of the board index is significantly positive with and without firm-fixed effects. Therefore, firms with a board that is more effective in monitoring and disciplining managers are valued more by investors. In Column (3), the effect of the ownership structure index is 7 As an alternative test for endogeneity, propensity-score matching estimations are performed in a separate paper (Ushijima, 2015). These estimations also robustly show that diversified firms in Japan are discounted relative to focused firms in the same industries.


also positive and significant. However, as reported in Column (4), the effect of ownership structure loses significance (p=0.13) when firm-fixed effects are included in the regression. The effect of information disclosure is negative and significant in Column (3), but not sig-nificant in Column (4).

In Columns (5) and (6), the governance system is captured by five variables: foreign

Table 4. Effects of diversification and governance on firm value

Diversification dummy -0.064 *** -0.059 *** -0.072 *** -0.059 *** -0.064 *** -0.058 ***(0.006) (0.007) (0.006) (0.007) (0.006) (0.007)

Governance index 0.021 *** 0.008 ***(0.001) (0.001)

Board of directors 0.045 *** 0.014 ***(0.002) (0.002)

Ownership structure 0.037 *** 0.005(0.002) (0.003)

Information disclosure -0.035 *** 0.002(0.003) (0.002)

Board size -0.003 *** -0.001(0.001) (0.001)

Ratio of independent directors 0.099 *** 0.078 ***(0.026) (0.024)

Foreign ownership ratio 0.540 *** 0.427 ***(0.033) (0.043)

Stable ownership ratio -0.012 0.053(0.037) (0.063)

Managerial ownership ratio 0.035 0.055 *(0.023) (0.029)

Total assets (logged) 0.027 *** -0.002 0.035 *** 0.000 0.010 *** -0.015 *(0.002) (0.009) (0.002) (0.009) (0.003) (0.009)

EBITDA/Sales 1.225 *** 0.225 *** 1.261 *** 0.230 *** 1.222 *** 0.236 ***(0.034) (0.037) (0.034) (0.037) (0.034) (0.037)

Capital expenditure/Sales 1.624 *** 0.446 *** 1.595 *** 0.447 *** 1.644 *** 0.435 ***(0.050) (0.042) (0.050) (0.042) (0.050) (0.042)

R&D expenditure/Sales 1.578 *** 1.032 *** 1.586 *** 1.046 *** 1.593 *** 1.065 ***(0.073) (0.146) (0.072) (0.146) (0.072) (0.146)

Marketing expenditure/Sales 0.764 *** 0.030 0.717 *** 0.047 0.789 *** 0.060(0.080) (0.151) (0.079) (0.152) (0.080) (0.151)

Debt/Market equity 0.038 *** 0.014 *** 0.033 *** 0.014 *** 0.039 *** 0.017 ***(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)

F-value of firm-fixed effects 0.171 0.783 0.187 0.783 0.174 0.784

R-squared － 16.81 *** － 16.41 *** 16.83 ***

# observations 27,307 27,307 27,307 27,307 27,307 27,307

(Note) In parentheses are standard errors. *** Significant at the 1% level. ** Significant at the 5% level. * Significant at the 10% level.

(6)

FE OLS FE OLS FE

(5)(1)

OLS

(2) (3) (4)

Table 4Effects of diversification and governance on firm value


441

ownership ratio, stable shareholding ratio, managerial ownership ratio, board size (number of directors), and independent external director ratio. Evidence strongly suggests that board independence increases firm value; the effect of the independent director ratio is positive and highly significant in both regressions. In Column (5), the link between board size and firm value is significantly negative, as reported for non-Japanese firms (e.g., Yermack, 1996). Perhaps larger boards are less flexible and take longer to make decisions, and are, therefore, less effective in policing management. However, Column (6) shows that the effect of board size is not significant when firm-fixed effects are introduced to the regression.

With regard to the ownership variables, the effect of the foreign investor ownership ratio is positive and significant at the 0.01 level. This result is consistent with the view that for-eign institutional investors contribute to the governance of Japanese firms as activist share-holders. The effect of managerial ownership is also positive and significant at the 0.10 level in Column (6). This pattern suggests that managerial equity ownership alleviates agency problems by better aligning the interests of shareholders and managers.8

VI-4. Interactions between diversification and governance

The interaction effect between diversification and governance is examined in this sub-section. In affecting firm value, diversification and governance can interact either positively or negatively. A positive interaction effect arises if good governance improves the quality of managerial decisions and thereby mitigates inefficient diversification. If this effect is im-portant, the value of diversification will be higher for better-governed firms.

A negative interaction effect arises if diversification decreases the effectiveness of the governance system. For instance, if diversification leads to the entrenchment of managers, as suggested by Shleifer and Vishny (1989), the power of external shareholders to discipline managers is weakened for diversified firms. The sheer diversity and complexity of a diversi-fied firm’s activities can be barriers to effective governance by increasing the informational disadvantage of external shareholders vis-à-vis internal managers. Moreover, the organiza-tion of diversified firms is generally characterized by a deeper hierarchy than that of focused firms. As such, if agency problems exist at the middle-management as well as executive lev-els, the effect of governance on firm value can be weaker for diversified firms because the governance system polices only managers at the top of the hierarchy.9

Based on the diversification dummy and governance variables measured as the differ-ence from the sample mean, the regression model is specified as follows:

8 A high level of managerial ownership can entrench managers by sheltering them from external control (Morck et al., 1988). To examine this scenario, I also estimated a specification with a quadratic term of managerial ownership. Consistent with the entrenchment effect, the estimated coefficient on the quadratic term was negative but not significant. 9 For instance, middle managers may distort information they communicate to executives to increase private benefits such as divisional growth and survival (Meyer et al. 1992; Scharfstein and Stein, 2000).


　　EVit＝α＋β・Divit＋γ・(Govit－Gov)＋δ・Divit・(Govit－Gov)＋λ・zit＋ηi＋∊it (4).

In this specification, the coefficient on the diversification dummy estimates the discount for diversified firms with average governance characteristics (Gov). The sign of δ is difficult to predict because, as discussed above, the interaction effect of diversification and governance can be either positive or negative.

Regression results are reported in Table 5. The governance system is measured by the governance index in Columns (1) and (2), while the specifications reported in Columns (3) and (4) employ the sub-indices. The regressions in Columns (1) and (3) are pooled cross-sectional estimations, whereas the regressions in Columns (2) and (4) incorporate firm-fixed effects. The main effect of diversification is negative and highly significant across the board. The estimated diversification discount is in the range of 6% to 7%, which is close to the estimates that were obtained without considering the interaction effect. Consistent with the regressions reported in the last subsection, the main effect of governance is signifi-cantly positive in Columns (1) and (2), confirming that better governance is associated with higher firm value. The coefficient on the interaction term is negative and significant in Col-umns (1) and (2). This result suggests that on average, the negative effect of diversification on the effectiveness of governance system dominates the positive effect of governance on firm value. The estimated coefficients indicate that the effect of governance to increase firm value is halved for diversified firms.10

In Column (3), the main effects of board and ownership structure indices are positive and significant, and all of the interaction effects are significantly negative. These results confirm that diversification decreases the effect of governance to increase firm value. Col-umn (4) suggests that the influence of diversification is particularly large for the effect of ownership structure. In this regression, the interaction effect between ownership structure index and diversification dummy is larger in absolute value than the main effect of the in-dex.

V. Corporate Governance and Diversification Behavior

V-1. Probit models

The preceding analysis shows that diversified firms in Japan trade at a discount relative to a portfolio of representative focused firms in the same industries. It is also found that di-versification mitigates the effect of governance to increase firm value. Is the diversification discount a symptom of managerial agency problems? In other words, do firms engage in in-efficient diversification due to a failure of the governance system? To help answer these

10 In Columns (1) and (2), the sum of the interaction effect and the main effect of governance index is positive and significant-ly different from zero at the 0.01 level. Therefore, corporate governance improves firm value even for diversified firms.


443

Table 5. Regressions with interaction effects between diversification and governance

Diversification dummy -0.065 *** -0.058 *** -0.073 *** -0.059 ***(0.006) (0.007) (0.006) (0.007)

Governance index 0.032 *** 0.011 ***(0.002) (0.002)

Diversification dummy × Governance index -0.023 *** -0.006 ***(0.002) (0.002)

Board of directors 0.052 *** 0.015 ***(0.003) (0.003)

Diversification dummy × Board of directors -0.015 *** -0.002(0.004) (0.004)

Ownership structure 0.056 *** 0.018 ***(0.003) (0.004)

Diversification dummy × Ownership structure -0.037 *** -0.025 ***(0.005) (0.004)

Information disclosure -0.029 *** -0.002(0.004) (0.003)

Diversification dummy × Information disclosure -0.012 ** 0.008 *(0.005) (0.004)

Total assets (logged) 0.028 *** -0.002 0.037 *** 0.003(0.002) (0.009) (0.002) (0.009)

EBITDA/Sales 1.200 *** 0.220 *** 1.234 *** 0.221 ***(0.034) (0.037) (0.034) (0.037)

Capital expenditure/Sales 1.617 *** 0.444 *** 1.586 *** 0.443 ***(0.050) (0.042) (0.050) (0.042)

R&D expenditure/Sales 1.548 *** 1.025 *** 1.554 *** 1.032 ***(0.072) (0.146) (0.072) (0.146)

Marketing expenditure/Sales 0.783 *** 0.031 0.740 *** 0.040(0.080) (0.151) (0.079) (0.151)

Debt/Market equity 0.037 *** 0.014 *** 0.032 *** 0.013 ***(0.003) (0.003) (0.003) (0.003)

R-squared 0.174 0.783 0.190 0.783

F-value of firm-fixed effects － 16.736 *** － 16.347 ***

# observations 27,307 27,307 27,307 27,307

(1) (2) (3) (4)

(Note) The sample mean is subtracted from all governance variables. In parentheses are standard errors. *** Significant at the1% level. ** Significant at the 5% level. * Significant at the 10% level.

OLS FE OLS FE

Table 5Regressions with interaction effects between diversification and governance


questions, the effect of governance on the firm’s choice of industrial scope is investigated in this section.

Specifically, probit models are estimated to identify the determinants of two probabili-ties. The first is the probability of a firm focused in the previous year (t-1) to become diver-sified in the current year (t): That is,

　　 1 0 ) )Prob ( (Divit Divit-1 Govit-1, Indit-1, ψtf= = = it-1, dz � (5)

where Div is the diversification dummy, Gov is the vector of governance variables, z is con-trol variables, ind is industry-level factors, and ψ is a year-fixed effect. I control for industry factors because firms in different industries are likely to have significantly different diversi-fication opportunities. The ratio of diversified firms among firms sharing the same 3-digit JSIC code is used to account for such heterogeneity.

The second is the probability of a firm diversified in t-1 becoming focused in t:

　　 0 1 ) )Prob ( (Divit Divit-1 Govit-1, Indit-1, ψtɡ= = = it-1, rz (6)

The refocusing probability model includes two firm-level control variables that are not shared by the diversifying probability model. One is the number of 4-digit industrial seg-ments. The more diversified a firm is, the less likely it is to refocus. The expected effect of this variable on the refocusing probability is therefore positive. The other is a dummy for firms reporting a negative net income. The expected effect of this variable is negative be-cause drastic restructuring actions such as refocusing are likely to be undertaken by finan-cially distressed firms.

V-2. Estimation results

Table 6 shows the estimation results of the diversifying probability model. In Column (1), the coefficient on the governance index is negative but not significant. In Column (2) where governance is measured by sub-indices, the effect of board index is not significantly different from zero, even though it is an important determinant of firm value. The coefficient on the ownership structure index is also insignificant. However, the effect of information disclosure is significantly negative, suggesting that firms that are more willing to share in-formation with investors are less likely to diversify. The specification reported in Column (3) is based on five governance variables. Only the coefficient for the stable shareholder ratio is significant. The positive effect of this variable implies that managers who are more tightly protected by cross-shareholding are more likely to diversify their firms.

Table 7 reports the estimation results for the refocusing probability. The results strongly suggest that firms with a stronger governance system are more likely to refocus. In Column (1), the coefficient on the governance index is positive and highly significant. In Column (2), where the three aspects of governance system are separately measured, the coefficients on


445

Table 6Probit estimation results of the diversifying probability of focused firmsTable 6. Probit estimation results of the diversifying probability of focused firms

Governance index -0.001(0.009)

Board of directors 0.019(0.018)

Ownership structure 0.019(0.018)

Information disclosure -0.060 ***(0.022)

Board size -0.001(0.009)

Ratio of independent directors 0.203(0.190)

Foreign ownership ratio -0.264(0.264)

Stable ownership ratio 0.784 ***(0.290)

Managerial ownership ratio -0.138(0.164)

Total assets (logged) -0.018 -0.012 -0.024(0.016) (0.018) (0.022)

EBITDA/Sales -0.297 * -0.273 -0.244(0.169) (0.171) (0.174)

Capital expenditure/Sales 0.095 0.065 0.158(0.317) (0.321) (0.316)

R&D expenditure/Sales -1.337 ** -1.301 ** -1.270 **(0.640) (0.636) (0.639)

Marketing expenditure/Sales 0.931 * 0.882 * 1.061 **(0.486) (0.489) (0.485)

Debt/Market equity 0.038 ** 0.031 0.036 *(0.019) (0.020) (0.020)

Share of diversified firms in industry 0.300 ** 0.291 ** 0.249 **(0.125) (0.125) (0.125)

Log likelihood -1958.9 -1954.4 -1952.9

Pseudo R-squared 0.141 0.143 0.144

# observations 11,234 11,234 11,234

(1) (2) (3)

(Note) In parentheses are standard errors. *** Significant at the 1% level. ** Significant at the 5% level. *Significant at the 10% level.


Table 7. Probit estimation results of the refocusing probability of diversified firms

Governance index 0.037 ***(0.010)

Board of directors 0.063 ***(0.020)

Ownership structure 0.043 ***(0.020)

Information disclosure -0.007(0.024)

Board size -0.027 ***(0.009)

Ratio of independent directors 0.324(0.206)

Foreign ownership ratio 0.482 *(0.283)

Stable ownership ratio -0.361(0.314)

Managerial ownership ratio -0.125(0.194)

# segments -0.338 *** -0.339 *** -0.336 ***(0.032) (0.032) (0.032)

Total assets (logged) -0.094 *** -0.082 *** -0.085 ***(0.018) (0.019) (0.025)

EBITDA/Sales -0.097 -0.090 -0.075(0.257) (0.256) (0.259)

Negative net income dummy 0.187 *** 0.173 *** 0.165 ***(0.061) (0.062) (0.062)

Capital expenditure/Sales -0.309 -0.332 -0.297(0.383) (0.385) (0.379)

R&D expenditure/Sales 0.081 0.072 -0.012(0.648) (0.649) (0.662)

Marketing expenditure/Sales 2.298 *** 2.270 *** 2.423 ***(0.549) (0.552) (0.552)

Debt/Market equity 0.028 0.024 0.021(0.018) (0.018) (0.018)

Share of diversified firms in industry -0.513 *** -0.526 *** -0.557 ***(0.134) (0.135) (0.135)

Log likelihood -1601.6 -1599.1 -1597.8

Pseudo R-squared 0.137 0.138 0.138

# observations 12,320 12,320 12,320

(Note) In parentheses are standard errors. *** Significant at the 1% level. ** Significant at the 5% level. *Significant at the 10% level.

(1) (2) (3)

Table 7Probit estimation results of the refocusing probability of diversified firms


447

the board and ownership structure indices are positive and highly significant. In Column (3), the effect of board size is significantly negative, suggesting that smaller boards are more ef-fective in pressuring managers to undertake drastic restructuring actions such as refocusing. The effect of foreign investor ownership is also positive and significant at the 0.10 level.

Concerning the effects of control variables, the coefficient on the negative net income dummy is significantly positive in Table 7. This result confirms that refocusing tends to be undertaken by firms that are under strong pressure for restructuring. The effect of the share of diversified firms is significantly positive for the diversifying probability (Table 6) and significantly negative for the refocusing probability (Table 7). As expected, the coefficient for the number of segments is negative in Table 7.

Overall, the above results suggest that the corporate governance system has differential implications for the increase and decrease of a firm’s industrial scope. Even though the ef-fect of governance on the firm’s decision to diversify is limited if any exists, it critically af-fects a firm’s decision to refocus. These findings suggest that the conflict of interest between managers and shareholders, which is not large at the onset of diversification, grows serious as firms continue to operate in multiple industries. One possible scenario for such a pattern is that diversified firms accumulate unprofitable businesses over time, perhaps because man-agers hang on to failed diversification projects. That “choose and focus” has been an overar-ching theme of restructuring suggests that such a tendency exists in many Japanese firms. The corporate governance system contributes to overcoming this problem by pressuring managers to undertake necessary remedial actions.

VI. Conclusion

This paper has examined the relationships among industrial diversification, firm value, and corporate governance based on a large sample of Japanese firms from 2004 to 2012. Ev-idence strongly suggests that diversified firms trade at a discount. That is, other things con-stant, diversified firms are valued 6% to 7% lower than representative focused firms in the same industries. The estimation results also show that corporate governance is an important determinant of the market value as well as industrial scope of Japanese firms. The results also suggest that managers of diversified firms resist the restructuring of inefficient busi-nesses when they are ill-disciplined by a weak governance system.

The discount for U.S. diversified firms has been debated intensely from various view-points. Similar investigations are necessary for understanding the diversification discount in Japan. If diversified firms hang on to unprofitable businesses as suggested above, corporate socialism (i.e., suboptimal allocation of scarce funds to unprofitable businesses at the cost of more successful ones) can be a pervasive problem in Japan. Therefore, the efficiency/ineffi-ciency of internal capital markets is a particularly important issue for future research.


References

Hiramoto, T. (2002), “Industrial Diversification and Firm Value,” Gendai Finance, No.12, pp.31–55 (in Japanese).

Nakano, M., N. Kubo and Y. Yoshimura (2002), “Financial Structure and Valuation of Di-versified Firms,” Shoken Analyst Journal, Vol.40. No.12, pp.76–91 (in Japanese).

Aggarwal, R.K. and A.A. Samwick (2003), “Why Do Managers Diversify Their Firms? Agency Reconsidered,” Journal of Finance, Vol.58, pp.71–118.

Aman, H. and P. Nguyen (2008), “Do Stock Prices Reflect the Corporate Governance Quali-ty of Japanese Firms?” Journal of the Japanese and International Economies, Volume 22, pp.647–662.

Amihud, Y. and B. Lev (1981), “Risk Reduction as a Managerial Motive for Conglomerate Mergers,” Bell Journal of Economics, Vol.12, pp.605–617.

Anderson, R.C., T.W. Bates, J.M. Bizjak, and M.L. Lemmon (2000), “Corporate Gover-nance and Firm Diversification,” Financial Management, Vol.29, pp.5–22.

Anderson, R.C. and D.M. Reeb (2003), “Founding-Family Ownership, Corporate Diversifi-cation, and Firm Leverage,” Journal of Law and Economics, Vol.46, pp.653–683.

Berger, P.G. and E. Ofek (1995), “Diversification’s Effect on Firm Value,” Journal of Finan-cial Economics, Vol.37, pp.39–65.

Berger, P.G. and E. Ofek (1996), “Bustup Takeovers of Value-Destroying Diversified Firms,” Journal of Finance, Vol.51, pp.1175–1200.

Berry, T.K., J.M. Bizjak, M.L. Lemmon, and L. Naveen (2006), “Organizational Complexity and CEO Labor Markets: Evidence from Diversified Firms,” Journal of Corporate Fi-nance, Volume 12, pp.797–817.

Campa, J.M., and S. Kedia (2002), “Explaining the Diversification Discount,” Journal of Finance, Vol.57, pp.1731–1762.

Chen, S.S. and Chen, I.J. (2012), “Corporate Governance and Capital Allocations of Diver-sified Firms,” Journal of Banking & Finance, Volume 36, pp.395–409.

Denis, D.J., D.K. Denis and A. Sarin (1997), “Agency Problems, Equity Ownership and Corporate Diversification,” Journal of Finance, Vol.52, pp.135–160.

Erdorf, S., T. Hartmann-Wendels, N. Heinrichs, and M. Matz (2013), “Corporate Diversifi-cation and Firm Value: A Survey of Recent Literature,” Financial Markets and Portfo-lio Management, Vol.27, pp.187–215.

Fama, E.F., and J.D. MacBeth (1973), “Risk, Return, and Equilibrium: Empirical Tests,” Journal of Political Economy, Vol.81, pp.607–636.

Fukui, Y. and T. Ushijima (2007), “Corporate Diversification, Performance, and Restructur-ing in the Largest Japanese Manufacturers,” Journal of the Japanese and International Economies, Volume 21, pp.303–323.

Hyland, D.C. and J.D. Diltz (2002), “Why Firms Diversify: An Empirical Examination,” Fi-nancial Management, Vol.31, pp.5–35.


449

Hoechle, D., M. Schmid, I. Walter, and D. Yermack (2012), “How Much of the Diversifica-tion Discount can be Explained by Poor Corporate Governance?” Journal of Financial Economics, Volume 103, pp.41–60.

Jiraporn, P., Y.S. Kim, W.N. Davidson, and M. Singh (2006), “Corporate Governance, Shareholder Rights and Firm Diversification: An Empirical Analysis,” Journal of Bank-ing & Finance, Volume 30, pp.947–963.

Jiraporn, P., Y.S. Kim, and W.N. Davidson III (2008), “Multiple directorships and corporate diversification,” Journal of Empirical Finance, Volume 15, pp. 418–435.

Lang, L. and R. Stulz (1994), “Tobin’s Q, Corporate Diversification, and Firm Perfor-mance,” Journal of Political Economy, Vol.102, pp.1248–1280.

Lins, K. and H. Servaes (1999), “International Evidence on the Value of Corporate Diversi-fication,” Journal of Finance, Vol.54, pp.2215–2239.

Martin, J.D. and A. Sayrak (2003), “Corporate Diversification and Shareholder Value: A Survey of Recent Literature,” Journal of Corporate Finance, Volume 9, pp.37–57.

Meyer, M., P. Milgrom and J. Roberts (1992), “Organizational Prospects, Influence Costs, and Ownership Changes,” Journal of Economics and Management Strategy, Vol.1, pp.9–35.

Morck, R., A. Shleifer, and R. W. Vishny (1988), “Management Ownership and Market Val-uation,” Journal of Financial Economics, Vol.20, pp.293–315.

Nam, J., C. Tang, J.H. Thornton Jr., and K. Wynne (2006), “The Effect of Agency Costs on the Value of Single-Segment and Multi-Segment Firms,” Journal of Corporate Fi-nance, Volume 12, pp.761–782.

Scharfstein, D.S. and J.C. Stein (2000), “The Dark Side of Internal Capital Markets: Divi-sional Rent-Seeking and Inefficient Investment,” Journal of Finance, Vol.55, pp.2537–2564.

Shleifer, A. and R.W. Vishny (1989), “Management Entrenchment: The Case of Manag-er-Specific Investments, Journal of Financial Economics, Volume 25, pp.123–139.

Stein, J.C. (1997), “Internal Capital Markets and the Competition for Corporate Resources,” Journal of Finance, Vol.52, pp.111–34.

Ushijima, T. (2015), “Diversification, Organization, and Value of the Firm,” RIETI Discus-sion Paper Series 15-E-019.

Williamson, O.E. (1975), Markets and Hierarchies: Analysis and Antitrust Implications, Free Press.

Yermack, D. (1996), “Higher Market Valuation of Companies with a Small Board of Direc-tors,” Journal of Financial Economics, Volume 40, pp.185–211.


Diversification Discount and Corporate Governance in … I. Introduction This paper investigates the relationships among industrial diversification, corporate gov-ernance, and firm

Documents