Accepted Manuscript
Ownership Structure and Market Efficiency
Stockholder/Manager Conflicts at the Dawn of Japanese Capitalism
Masaki Nakabayashi
PII:
DOI:
Reference:
S1042-4431(18)30033-7
https://doi.org/10.1016/j.intfin.2019.03.003
INTFIN 1112
To appear in:
Journal of International Financial Markets, Institutions & Money
Received Date:
Accepted Date:
5 February 2018
17 March 2019
Please cite this article as: M. Nakabayashi, Ownership Structure and Market Efficiency, Journal of International
Financial Markets, Institutions & Money (2019), doi: https://doi.org/10.1016/j.intfin.2019.03.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ownership Structure and Market Efficiency
Stockholder/Manager Conflicts at the Dawn of Japanese Capitalism
Masaki Nakabayashi
Institute of Social Science, The University of Tokyo, Hongo 7–3–1, Bunkyo, 103–0033 Tokyo, Japan. E-mail: mn@iss.u-tokyo.ac.jp
Abstract
We present a framework to analyze the impact of ownership structure on stockholder/manager conflicts. We first
predict that, in an inefficient market, investors motivate managers to pursue a higher return on equity instead of a
higher return on asset and that this focus on short-term performance leads to leverage distortion. Using a sample
of late nineteenth- to early twentieth-century Japanese firms, we show that mediocre performing firms boosted
the return on equity by bond flotation, and a higher president-ownership concentration raised the return on asset
and controlled bond leverage. President-ownership concentration offsets market inefficiency.
Acknowledgments
The author is grateful to William Megginson, Kentaro Asai, Susumu Cato, and other participants in the 2018
Summer Tokyo Conference on Economics of Institutions and Organizations, the International Finance and Banking Society 2018 Oxford Conference, and the 2017 Southwestern Finance Association Annual Conference for
their helpful comments. He is indebted to the editor, Jonathan Batten, for improving the presentation and deepening the analysis of this paper. He is also thankful to Mayo Morimoto for the support in building the dataset for
this study. This research was funded by the JSPS Grant-In-Aid KAKENHI JP17K18558 and The Japan Securities Scholarship Foundation.
Keywords: stockholder/manager conflicts; multitask moral hazard; ownership structure; financial leverage;
self-fulfilling distortion; skewness-adjusted variation coefficient.
JEL Codes: G32; L23; O16; K22
Preprint submitted to Elsevier
March 23, 2019
1. Introduction
On whether ownership structure matters in stockholder/manager conflicts, Smith (1937[1776]) concluded it
does, and was concerned that a diffused ownership structure might cause conflicts between shareholders and
managers due to information asymmetry and lack of monitoring incentives for shareholders. Smith’s capital
market anxiety is another aspect of the same suggestion by Modigliani and Miller (1958) that if the financial
markets were perfect, ownership structure should not matter. Linking the benchmarks, Jensen and Meckling
(1976) and Fama (1980) predicted that discipline by more efficient capital markets and managers markets would
discourage moral hazards of managers and majority shareholders.
We track changes in ownership structure, financial leverage, performance, and market valuation during the
Japanese modernization. To do this, we construct a dataset of all firms listed on the Tokyo Stock Exchange from
1878 to 1910 by collecting financial statement data. Japan in the late nineteenth century was one of the early
cases of non-US nations that succeeded in nurturing a capitalist economy.
We consider the possibility of leverage distortion by non-owner managers to manipulate the return on equity
(ROE) in the short term instead of maximizing the return on asset (ROA). For a focused and practical prediction,
we deploy a multitask moral hazard model tailored for managerial incentives. Shareholders can use two proxies to
measure non-owner managerial performance—ROE and ROA. Shareholders might want to motivate non-owner
managers by a performance-based payment using those proxies. Meanwhile, managers can reduce the ROE
variance by leverage distortion. This is a shareholder maximization problem. Suppose a sufficiently inefficient
and large market where investor information asymmetry is severe and shareholders cannot govern themselves
by relational contracts. Then, it can be optimal for shareholders to reward the ROE instead of ROA and save
the quality premium to be paid to risk-averse non-owner managers. Although the moral hazard of non-owner
managers is predictable, anonymous shareholders encourage it in a self-fulfilling way.
In the inefficient Japanese market from the late nineteenth century to the early twentieth century, market
participants predominantly rewarded the ROE but not the ROA. Responding to that, poorly and mediocre performing firms tended to mechanically raise the return on equity by the leverage of bond flotation. By contrast,
firms whose share concentrated at the president pursued the growth in the ROA, generally contained the leverage
by bond flotation but raised it if it was to accompany the growth in the ROA. As a result, a president-ownership
concentration led to the growth in the ROA but was irrelevant to the ROE and the leverage by bond flotation
contributed to the ROA only in case of the top tier firms. President-ownership concentration offset the market
inefficiency. Those facts seem to explain the significant role of ownership concentration in the non-US markets
1
today.
The rest of the paper is organized as follows. Section 2 briefly describes the historical development of
Japanese corporate finance under the corporate law modeled on German law. Section 3 introduces related literature. Section 4 presents a model to capture the self-fulfilling financial leverage distortion by risk-averse managers
in the context of stockholder/manager conflicts under the separation of ownership and management. Thus, among
possible stockholder/manager conflicts discussed by Jensen (2000), we focus on leverage distortions. We deduce
a few hypotheses to be empirically tested. We predict that non-owner managers have incentives to raise a more
than optimal leverage. Moreover, in an inefficient market with information asymmetry about non-owner manager
actions, investors encourage leverage distortion to save the risk premium to be paid as part of risk-averse nonowner manager compensations. Section 5 describes the dataset we build. Section 6 examines whether ownership
structure affected performance, and how the market rewarded corporate performance. We also test whether the
Commercial Code of 1899 enactment affected the impact of ownership structure on performance. Section 7
focuses on bond flotation distortion. Section 8 concludes and discusses results.
2. Retrospection in German-Japanese resemblance
After the 1868 Meiji Restoration, the Japanese government adopted the civil law from continental Europe.
The Commercial Code of 1899, modeled on German law, completed corporate law. The rising power of Germany inspired Japan’s modernization in the adoption of the German-style constitution from 1890 to 1945 and the
German-style commercial code from 1899 to date. The effort at taking inspiration from Germany included the
establishment of the Industrial Bank of Japan in 1897 to support crucial industries with a government debt guarantee (Lehmbruch (2001) and Vitols (2001)). The shared legal foundation is a basis for establishing corporate
governance features that emphasize not only shareholder value but also stakeholder interests in contemporary
civil law countries; Japan, Germany, and France (Shleifer and Vishny (1997); Tirole (2001); Salazar and Raggiunti (2016)).
However, we cannot characterize the challenges faced by Germany, Japan and other emerging powers from
the late nineteenth century to the early twentieth century only by an effort of the domestic industrialization of
a small closed economy. In the first age of globalization from the 1870s to the 1910s (Mauro et al. (2006), pp.
1–45; Thomadakis et al. (2017); Betrn and Huberman (2016); Varian (2018)), internationalization of financial
markets—the well-integrated international financial markets centered at the London market and efficient crossborder capital flow—as well as the free trade of goods was also of vital interests. Japan embraced advantage of
2
the imposed free trade (Nakabayashi (2014); Kawashima (2018)) and, furthermore, adopted the gold standard
(Mauro et al. (2006), pp. 49–54; Nakabayashi (2012)).
Before the First World War, the financial markets of the industrial world were even more deeply integrated,
and the cross-border capital flow was active (Rajan and Zingales (2003); Mauro et al. (2006)). National financial
markets were well embedded in international financial markets and showed minor differences. The German
economy before the First World War was a competitive market and its civil law characteristics did not influence
German corporate finance and governance (Fohlin (2007) and Burhop and Lübbers (2009)).
Japanese and German corporate governance began to change gradually post-First World War and in earnest
during the Second World War. Cartels gained bargaining power in Weimar Germany when state-direction was
combined with capital property relations and brought about the “Social Market Economy” in the Federal Republic
of Germany. Similarly, state coordination was institutionalized during the Second World War, and post-war Japan
inherited state-guided characteristics through “industrial policies.” The transformations in Japan and Germany
accompanied a rise in the role of the banking sector in corporate finance under the stringent regulations introduced
in the 1920s and 1930s and survived until the deregulation in the 1980s (Okazaki (1999); Jackson (2001); Vitols
(2001); Ferguson and Voth (2008)).
Thus, the distinction between common law countries, represented by the US and the UK, and civil law
countries represented by Japan and Germany, became significant due to structural changes from the 1920s to
the 1940s when the latter formed an axis. While the axis shared legal origins dating back to the late nineteenth
century, the heterogeneity of the industrial world was smaller under British dominance before the First World
War.
Therefore, we begin our study not from La Porta et al. (2008)’s view on the post-Second World War divide
between common law and civil law countries, but from a data-driven approach. As we later demonstrate, the
Japanese market until the early twentieth century was not efficient and hence, ownership structure mattered. The
difference between the common law and civil law distinction is inconsequential to this fact.
The Japanese experience, particularly in the period when the separation of ownership and management according to Berle and Means (1933) and Chandler (1977) was underway, would be a promising case. It transformed itself from the samurai’s nation to a modern capitalist economy without sharing history with the West,
as many emerging economies did. After toppling the Shogunate in 1868, the new imperial government began
its modernization efforts. In 1878, the Tokyo Stock Exchange and the Osaka Stock Exchange were established.
Furthermore, the Commercial Code of 1899 stipulated legal requirements for a joint-stock company and stan-
3
dardized financial statement forms. More information became publicly available and prompted further expansion
of the stock and bond markets.
From the late nineteenth century, Japanese corporate finance and governance experienced two distinctive
phases. First was the entrepreneurial boom of the mid-1880s. The cotton-spinning, railway, and other modern
industries incurred massive initial expenses by taking the form of joint-stock companies. They issued corporate
shares while relying on bank loans. Second phase was a reduction in bank loan reliance and an increase in bond
flotation from the late 1890s (Hoshi and Kashyap (2001), pp. 15–50). From the late 1890s, senior employees
began to climb to management positions and be promoted to board members. Functional diversification of the
board toward professional management meant the shareholders faced a possibility of managerial moral hazard.
Studies on advanced nations’ experiences in the nineteenth and early twentieth centuries such as Borg et al.
(1989), Leeth and Borg (1994, 2000), and Banerjee and Eckard (2001) on the US, Franks et al. (2006) and Kling
(2006) on Germany, Hamano et al. (2009) on Japan also provide us with contemporary policy implications.
Financial markets of advanced economies had been tightly regulated until the 1980s. Most advanced economy
regulations were introduced as a response to the financial markets collapse followed by the Great Depression
in the 1930s. Amid the Great Depression, advanced nations tightened corporate finance regulations, reckoning
severe market distortion due to asymmetric information. For example, the US enacted the Securities Act of 1933
and Securities Exchange Act of 1934. This created the Securities and Exchange Commission and established
the Generally Accepted Accounting Principles in the 1930s. Among advanced nations, regulations in Japan and
Germany were made particularly stringently; the banking sector replaced the stock and bond markets as the
primary source of corporate finance.
In the US, the more stringent banking sector regulations induced households to reallocate financial assets
from bank to brokerage accounts in the 1970s. The banks demanded deregulation, which led to disintermediation
and a brokerage-banking re-convergence from the 1980s to the 1990s. The development of information and
communication technologies that improved financial market efficiency validated the deregulation.
Other advanced nations followed the US experience from the 1980s. In the reform efforts of Japan and
Germany, a cornerstone has been the stock and bond market deregulation. The stock and bond market deregulation and the subsequent disintermediation from the 1980s meant the recovery of the pre-Great Depression direct
finance. Borg et al. (1989), Leeth and Borg (1994, 2000), Banerjee and Eckard (2001), Franks et al. (2006),
Kling (2006), Hamano et al. (2009), Nakabayashi (2017) on the pre-Great Depression stock markets of advanced
nations shared the viewpoint. Likewise, cross-country overviews such as La Porta et al. (2008) give regulatory
4
alternatives. However, one of the most basic questions is not addressed: Did the market discipline work or did
ownership structure complement a potentially imperfect market under lighter regulations in each nation before
the Great Depression?
Most nations have implemented structural reforms to recover vibrant stock and bond markets without being
conscious about how markets worked under lighter regulations, to what extent they were distorted due by asymmetric information, and to what extent the ownership structure complemented the potentially imperfect pre-Great
Depression market. This study attempts to lay a foundation for understanding the origin of the Japanese capital
market alongside previous works on pre-regulated markets. Reflecting on Japan’s century-old experiences of
ownership structure changes would supply meaningful lessons to Japan’s ongoing structural reforms and also
other nations’ reforms.
3. Relevant literature
When residual claimants do not directly perform residual control, a moral hazard such as managerial exploitation of shareholders may arise (Smith (1937[1776]), pp. 699–799). It might be the stockholder/manager
conflict Smith (1937[1776]) feared, which Byrd et al. (1998) and Parrino et al. (2005) revisited. It might also
be stockholder/bondholder conflicts by controlling-shareholders who are often founders as Jensen and Meckling
(1976) highlighted. It is expected to be severe when controlling-shareholders do not invest “real capital” in the
firm (Morck et al. (2005)).
A remedy for moral hazard is an active secondary market for corporate shares (Holmstrom and Tirole (1993)).
The threat of acquisition and replacement of managers is expected to discipline current managers, as argued since
Jensen and Meckling (1976) and Fama (1980).
A question regarding this view is whether ownership structure matters for corporate governance. Demsetz
and Lehn (1985) rejected a possible relationship between ownership concentration and performance for major
US-listed firms, later supported by Himmelberg et al. (1999) and Demsetz and Villalonga (2001). Using US
data, Anderson and Reeb (2003) did not find evidence for minority shareholder exploitation by founding ownermanagers. Morck et al. (1988) using US data found weak evidence that founding family participation by founding
family on boards might deteriorate performance. Helwege et al. (2007) described the evolution of listed firms
using a 1970–2001 US initial public offering dataset and found that better performers have become faster and
more widely-held after being listed and that agency costs do not significantly affect the ownership structure
evolution.
5
As Shleifer and Vishny (1986), Bolton and Scharfstein (1996), Mahrt-Smith (2005), Gorton and Kahl (2008),
Aslan and Kumar (2012), and Dhillon and Rossetto (2015), among others, predict, there is an ownership structure
diversity among US firms, and there must be a rationale for this diversity. Empirical results on the irrelevance
of the difference in ownership structure do not contradict theoretical predictions and the reality of diversity.
Consider an efficient market. A sufficiently efficient market implies market participation by price distortion and
resource reallocation through arbitrage transactions. Thus, on equilibrium, we see multiple ownership structure
types but hardly find statistical differences in performance among them.
Meanwhile, Davies et al. (2005) using British data found a co-deterministic relationship between ownership structure and performance. A characteristic of continental European ownership structure is blockholding
(Enriques and Volpin (2007)). However, the structural implications are mixed. Using European data, Laeven
and Levine (2008) showed that multiple blockholders help prevent managers from exploiting small shareholders, which indicates that ownership structure matters for performance. Ben-Nasr et al. (2015) demonstrated,
using French data, that ownership structure does affect financial leverage; that is, firms with a larger ownershipmanagement divides are prone to extended debt maturity, while multiple blockholder presence curbs such distortions. Although Julian and Mayer (2001), using German data, deny the ownership structure effect on performance, their results t do not reject the hypothesis by Laeven and Levine (2008). Blockholders often control major
German firms and that the banking sector dominates German corporate finance, although this has been gradually
changing ever since the last two or three decades (Ringe (2015)).
Pindado et al. (2014), using Western European data, extracted an inverse-U-shaped relationship between
ownership concentration and performance of family firms; performance increases to a threshold in ownership
concentration and decreases beyond that. In non-family firms, ownership power is more favorable. Hamadi
and Heinen (2015), using Belgian data, found that market valuation of non-family firms tends to monotonically
increase in the degree of ownership concentration while the relationship is inversely U-shaped in family firms.
Abdallah and Ismail (2017), using data from the Gulf Cooperation Council, also showed that a smaller ownership
concentration should be accompanied by better governance to achieve the same performance. Using Ukrainian
data, Mykhayliv and Zauner (2017) found that the state ownership tends to lower the level of investment while
the management ownership has not significant impact on investment. Haider et al. (2018) found that the state
ownership and resulting soft financial constraints improve corporate performance particularly in more corrupted
countries by a dataset from 81 nations.
Japan is no exception among such non-US economies. The institutional backdrop of post-war Japan is rela-
6
tively complicated. After Japan’s surrender, the US attempted to transform Japan’s market into the “widely-held”
market such as the US by procuring conglomerate corporate shares and selling them to small investors and corporate employees. As a result, Japan became a “widely-held” market along with the US and the UK (La Porta
et al. (1999)).
Contrary to the US, however, diffused ownership did not nullify ownership discipline. Lichtenberg and
Pushner (1994) found a positive relationship between insider ownership concentration and performance. Morck
et al. (2000) validated the result by showing that managerial ownership monotonically contributed to corporate
valuation. While post-war Japan-specific factors such as the main bank system made the relationship relatively
ambiguous (Gedajlovic et al. (2005)), the overall tendency is that a more concentrated managerial ownership is
positively correlated with better performance (Gedajlovic and Shapiro (2002)). Using data in the 2000s, Aman
and Nguyen (2013) found that institutional ownership improves corporate credit rating. Sakawa and Watanabel
(2018), using data from the late 2000s to the early 2010s, demonstrated that parent firm’s control contributes to
the growth of subsidiary firm. Thus, despite the US’s experiment to transform the Japanese market in its image,
there exists a positive relationship between ownership concentration and performance-valuation like in non-US
advanced economies.
The differing observations between the US and non-US countries indicate that the significance of ownership
structure is dependent on a condition that the US satisfies but others do not; a sufficiently efficient market. Let us
summarize observations of previous works on two dimensions of market efficiency and ownership concentration.
To close to the Pareto frontier on the plane, firms must be traded in a perfectly efficient market, be exclusively
owned, or be between the extremes. The US is in the northwest of the plane, emerging economies are on the
southeast, and non-US advanced economies are between them. As an economy departs from the west, ownership
matters more in the economy (Figure 1).
[INSERT Figure 1 HERE]
The less efficient the market, the greater ownership concentration must be to offset inefficiency and curb distortion. In particular, we share a common concern about the stockholder/manager conflicts under managerial
risk-aversion with Parrino et al. (2005). While Parrino et al. (2005) evaluated possible distortions in investment
decision with the leverage as given, we focus on possible leverage distortions by risk-averse managers.
7
4. Model
4.1. Model of self-fulfilling leverage distortion
Among the possible stockholder/manager conflicts mentioned by Jensen (2000), we focus on leverage distortions by risk-averse managers. We make predictions by applying Holmstrom and Milgrom (1991)’s multitask
principal-agent model to the context of an undesirable self-fulfilling equilibrium in an imperfect market (Diamond and Dybvig (1983); Goldstein and Pauzner (2004); Kunieda and Shibata (2016)).
For simplicity, we consider an extreme case where managers do not own shares. Assuming a two-dimensional
task for a manager, the first dimension, t1 , is to increase the ROE and the second, t2 , is to increase the ROA.
We standardize managerial human resource endowment as 1 such that t1 + t2 = 1. Let C denote the total
personal cost to be incurred by the manager. We assume that the effort costs to raise the ROE and the ROA are
2
identical. We further assume that C is strictly convex such that C11 C22 − C12
> 0 where C11 ≡ ∂ 2 C/∂t21 ,
C22 ≡ ∂ 2 C/∂t22 , and C12 ≡ ∂ 2 /∂t1 ∂t2 . The identical costs in both dimensions imply that C11 = C22 . Thus,
under the strict convexity assumption, C11 = C22 > C12 . Note that we do not exclude the possibility that efforts
in both dimensions are complements such that C12 < 0.
Let B1 and B2 denote the marginal effort contribution in each dimension such that B1 ≡ ∂ROE/∂t1 and
B2 ≡ ∂ROA/∂t2 . For simplicity, we assume that marginal contribution of the first best efforts for both ROA
and ROE are identical and standardized such that B1 = B2 = 1. The following theoretical predictions also hold
when allowing B1 6= B2 . Given the random market shock, we assume that the ROE and ROA are realized such
that ROE = t1 + ǫ1 and ROA = t2 + ǫ2 , where ǫ1 ∼ N (0, σ12 ), ǫ2 ∼ N (0, σ22 ), and ǫ1 ǫ2 ≡ σ12 .
We further assume that the manager is risk-averse such that his utility function is approximated by an
absolute-constant-risk-averse utility function, u (w − C) = 1 − exp [−r (w − C)], where w is the remuneration
and r is the constant absolute risk-averse coefficient. Conventional wisdom encourages managers to be risktolerant. However, as many empirical works have shown, managerial compensations in contemporary US firms
are largely designed to reduce managerial risk (Blanchard et al. (1994); Murphy (1999); Kraft and Niederprüm
(1999) and Bertrand and Mullainathan (2001)). The most persuasive explanation of the phenomenon is that
managers are risk-averse humans (Murphy (2002)).
Since Knight (1921), the ability to bear risk and uncertainty, which is transformed into subjective risk (Savage
(1954)), has centered on the essential managerial abilities. The argument is consistent with the emphasis on the
risk aversion of managers. Firms that take a higher risk tend to make massive payments to firm executives
notably in the US. Those should be if managerial utility is marginally diminishing over remuneration, that is, if
8
their utility function is concave. The concavity of the utility function is equivalent to risk aversion of the agent.
Curvature of the utility function is the measure of risk aversion.
For simplicity, we temporarily assume that E [ROE] = E [ROA]. In a perfect market under symmetric
information, any financial leverage distortion is impossible. Hence, σ12 = σ22 and σ12 = 1, since random
shock arises only in the current profit—the common numerator. However, in an imperfect market, managers
can mechanically stabilize or increase the ROE by manipulating leverage, withholding the information about the
manipulation.
Suppose that the market evaluates managers by the ROE as well as the ROA, which is not manipulable by the
financial leverage, and that the market is inefficient. Then, risk-averse managers would distort the distribution of
the manipulable ROE such that σ12 < σ22 and σ12 < 1 and its expected value is than those of the ROA. We see
this type of manipulation in emerging markets whose transparency is still yet to be completed (de Wet and du Toit
(2007)). However, this is also an issue in advanced economies (Bergstresser et al. (2006)). Japan’s early-stage
experience should provide practical lessons to contemporary investors.
Note that for shares to be actively traded and for sufficient liquidity to be maintained, the market needs a
sufficient number of “uninformed” investors who know only publicly available information (Kyle (1985); Admati
and Pfleiderer (1988); Collin-Dufresne and Fos (2016)).
For analytical simplicity, we proceed with holding the assumption that E [ROE] = E [ROA], ǫ1 ∼ N 0, σ12 ,
and ǫ2 ∼ N 0, σ22 , ǫ1 ǫ2 ≡ σ12 . Relying on the liquid market’s monitoring power (Holmstrom and Tirole
(1993)), to motivate risk-averse managers, their compensations are designed to reflect stock prices, either directly
by stock options or indirectly by bonuses. We standardize the compensation schedule as
(1)
w = α + STP = α + β1 ROE + β2 ROA = α + β1 (t1 + ǫ1 ) + β2 (t2 + ǫ2 ),
where STP is the firm’s stock price, and α is the minimum transfer that satisfies the individual rationality constraint by equality.
We have E [u (w − C)] = 1−exp [−r (E [w] − C − rV [w] /2)] = 1−exp −r β T t − C(t) − rβ T Σβ/2 ,
T
T
where t = (t1 , t2 ) , β = (β1 , β2 ) , and Σ denotes the covariance matrix whose diagonal elements are σ12
and σ22 and off-diagonal elements are σ12 . The manager then chooses t, given remuneration schedule β, such
that t = arg maxt β T t − C(t) − rβ T Σβ/2. Its first order condition to maximize the managerial payoff is
β T = ∂C(t)/∂t, which is the incentive compatibility constraint of the manager.
9
Given ∂C(t)/∂t = β T , Shareholder j of n total shareholders maximizes the total surplus multiplied by share
owned such that max sj B(t) − C(t) − rβ T Σβ/2 , where sj denotes stock holding ratio of shareholder j, and
Pj=n
j=1 sj = 1, given the incentive compatibility constraint of the manager.
The first order condition of shareholder maximization gives the optimal vector of incentive weights, β ∗ =
−1
(∂B/∂t) [I + rΣ∇C(t)]
, where I is a unit matrix and ∇C(t) is a Hessian matrix of C(t). Therefore, under
the assumptions B1 = B2 = 1 and C11 = C22 , we have optimal incentive vector β ∗ as follows.
1 + r σ22 − σ12 (C11 − C12 )
=
2 ) (C 2 − C 2 ) ,
1 + r [(σ12 + σ22 ) C11 + 2σ12 C12 ] + r2 (σ12 σ22 − σ12
11
12
2
1 + r σ1 − σ12 (C11 − C12 )
β2∗ =
2 ) (C 2 − C 2 ) .
1 + r [(σ12 + σ22 ) C11 + 2σ12 C12 ] + r2 (σ12 σ22 − σ12
11
12
β1∗
(2)
We immediately have the following lemma.
Lemma 1. Self-fulfilling distortion:
(i) In an efficient market, the incentive is not distorted.
(ii) In an inefficient market, the incentive is distorted toward an overemphasis on the return on equity.
(iii) Distortion is increasing in the degree of market inefficiency.
Proof.
(i) In an efficient market, σ12 = σ22 . That implies β1 = β2 which is the first best under B1 = B2 .
(ii) In an inefficient market, σ12 < σ22 due to the manipulated financial leverage. That implies β1∗ > β2∗ , which
deviates from the first best under B1 = B2 .
(iii) The more inefficient the market is, the smaller σ12 is. Furthermore,
2
r2 σ22 − σ12 (C11 − C12 )
∂ (β1 /β2 )
=
2 > 0,
∂σ12
[(rσ12 − rσ12 ) (C11 − C12 ) + 1]
in an inefficient market where σ12 < σ22 .
Specifically, if managerial distortion successfully reduces ROE risk, then the variance of the ROE would
become smaller than that of the ROA, which is standardized by the expected value and the skewness (Kraus and
Litzenberger (1976); Scott and Horvath (1980); Adrian and Rosenberg (2008); Conrad et al. (2013)). That is,
10
h
i
3
(σ1 /E [ROE]) |γ1 | < (σ2 /E [ROA])/ |γ2 |, where σ1 and σ2 are standard deviations and γ1 ≡ E (ROE − E [ROE]) /σ13
h
i
3
and γ2 ≡ E (ROA − E [ROA]) /σ23 are the skewness of ROE and ROA respectively.
Thus, our statement is described by variances standardized by the mean and the third-order central moment;
or equivalently, by the skewness adjusted variation coefficients, instead of raw variances as follows: If the market
is perfectly efficient, then
σ12
σ12
i =
× h
E [ROE] E (ROE − E [ROE])3
σ2 /E [ROA]
σ1 /E [ROE]
=
γ1
γ2
(3)
=
σ22
σ22
i ;
× h
E [ROA] E (ROA − E [ROA])3
and, if the market is inefficient, then
σ1 /E [ROE]
σ2 /E [ROA]
<
.
γ1
γ2
(4)
Investors can be aware that the skewness-adjusted variation coefficient of the ROE is smaller than that of the
ROA by a cross-sectional comparison and hence can infer that some managers might have distorted the financial
leverage to smoothen or increase the ROE mechanically. However, given ones’ small share, individual investors
do not have incentives to investigate what a specific firm is doing, and because of this, they rely on the market
price to know monitor firm performance. It encourages free-riding among investors. The resulting financial
leverage distortion implies that the skewness-adjusted variation coefficient of the ROA is higher than that of the
ROE. Given that, investors increase the ROE weight to save risk premium to be paid to risk-averse managers,
which induces managerial overemphasis on the ROE. Financial leverage distortion mechanically attains an ROE
increase.
Risk aversion in an inefficient market where managers can withhold information about their financial leverage
manipulation implies that the distortion is encouraged by investors and arises in a self-fulfilling way. Although
investors reckon that emphasis on short-term ROE would distort the leverage and reduce the long-term value of
the firm, everyone is free-riding each other and the myopic emphasis on the ROE continues.
A way to imperfectly remove the distortion is to have a dominant shareholder who owns long-term shares.
The adverse effects on long-term profitability of distorted leverage intended to mechanically smoothen or increase
11
the ROE become discernible in time when confronting the repayment of more than optimal debt. If an uninformed
shareholder pursues short-term transactions, then he would believe that he can successfully sell at a profit to
another uninformed investor before the distortion is finally revealed rather than make costly efforts to curb the
distortion.
Short-sighted trades by small uninformed investors pursuing a higher ROE is an individually optimal response
to one another, and hence can be an equilibrium strategy. Alternatively, a high ROE might be correctly perceived
as a signal of leverage distortion. If so, large buyers expect that correction of the distortion would lead to a better
long-term performance by block holding and find a reason to buy. If a current shareholder perceive the possibility
of distortion but he does not have an incentive and a claim to correct it, the possible distortion is a reason to
sell for him. In that case, a market that rewards the ROE brings about incremental improvement of resource
allocation through transfer of ownership. Trade transfers ownership between equally uninformed investors but
from a market participant who is more likely to make a wrong decision to one who is more likely to make a right
decision in the sense of Bond and Eraslan (2010). Thus, if there already exists a dominant shareholder to seek
long-term growth in the share price, or if an investor finds an opportunity to become a dominant shareholder and
correct distortion, he has an incentive to refuse or remove distortion.
The return on a commitment to long-term hold can be greater only if he recognizes the short-term divergence
between the share price and the fundamentals unknown to other market participants. This means that he is an
informed investor. The best-informed position is to be on the board. If he manages the firm, he would know
the business fundamentals better than outsiders. A higher manager-ownership concentration would reduce the
agency problem because of having a higher claim and being better informed.
Earlier cases for the advantages of ownership concentration included the privately-owned British cottonspinning firms in the industrial revolution. The firms addressed the concern of moral hazard suggested by Smith
(1937[1776]). A more recent case of public firm ownership concentration by a founding family is an early
generation of rising East Asian family firms (Claessens et al. (2000)). In Japan, from the late nineteenth century
to the early twentieth, a single dominant shareholder implied family ownership.
However, whether they can be long-lived is another question. Founding family exploitation of minority shareholders, highlighted by Jensen and Meckling (1976), is a challenge. Another is successor talent. Consanguineous
descendants of the talented are not necessarily talented. A Japanese choice is adopting a talented adult as a successor of the family business. The system dating back to the late seventeenth century from farmer to samurais has
disciplined Japanese family businesses. On average, family firms perform better than non-family firms, different
12
from other advanced economies (Mehrotra et al. (2013)).
Paraphrasing implications of Lemma 1, our hypotheses to be empirically tested are as follows:
H1 In an inefficient market where the skewness-adjusted variation coefficient of the ROE is smaller than that
of the ROA, the stock prices are more responsive to the ROE than the ROA.
H2 In an inefficient market, a higher president-ownership concentration implies a smaller financial leverage
distortion.
H3 In an inefficient market, a higher president-ownership concentration implies a better performance measured
by the ROA.
5. Data
5.1. Ownership structure
As Berle and Means (1933) and Chandler (1977) observed in US cases; Foreman-Peck and Hannah (2013) in
British cases; and Yui (1979, 1989, 1992), Miyamoto and Abe (1999), and Nakamura (2000, 2007) in Japanese
cases, senior employees were promoted to be managers and independent were hired businessmen as “professional managers” among leading companies from the 1890s to the 1900s. Furthermore, Miwa and Ramseyer
(2002) showed that “prominent” managerial board participation positively contributed to corporate performance
in Japanese cases in the early twentieth century.
These studies, however, did not deal with the possible effects of ownership structure changes within the
board. To differentiate the ownership structure, we introduce two simple measures. First is the president’s stockholding ratio. Second is the product of the president’s stockholding ratio and that of the board member with
the smallest stockholding. The first measure is expected to capture the moral hazard effect that decreases in the
president’s stockholding ratio. In other words, the performance of the case firm is expected to increase in this
measure (H2 and H3). The second measure is to examine how the degree of managerial ownership consolidation
affects performance. When the board structure is closer to shareholder representative, the value of the second
measure is greater. Meanwhile, if an employee is promoted to a board member, the value of the second measure
is expected to become smaller. The measure evaluates how the deviation from the classical form of the board and
the employee promotion to board member could affect performance (H2 and H3).
In the entire Tokyo market, having small shareholders, ownership concentration was considerable. Overall,
the top 1% largest shareholders owned 53% of shares of listed firms as of 1897 (Table 1). We exploit the
13
ownership variance for our estimates.
INSERT Table01 HERE
5.2. Description of dataset
Our sample covers all 95 firms (i) listed on the Tokyo Stock Exchange from the first half of 1878 to the second
half of 1910 (t). The financial statements of the firms are available in the business archives of the Japan Digital
Archives Center delivered by Maruzen-Yushodo.1 Note that firms predominantly owned by conglomerates such
as Mitsubishi and Mitsui were not listed and are not included in our samples. Thus, distortion due to substantial
conglomerate protection by the government is not captured. We manually collected information about financial
status and the stockholding to build a panel dataset of 95 firms.
Financial status variables we use are sales (SALi,t ), total assets (TASi,t ), paid-in stock (STKi,t ),2 outstanding bank loans (LONi,t ), outstanding bond (BNDi,t ), profit in the current term (PRFi,t ), total dividend
(DVDi,t )3 , and balance brought forward (BBFi,t ) for firm i in term t. Discrepancies of the total observation
numbers come from unstandardized financial statements, particularly before the enactment of the Commercial
Code of 1899.
As measures of ownership structure, we calculate the president’s stockholding ratio (SCEOi,t ), the stockholding ratio of the board member with the smallest ratio (SMINi,t ), and their product (CNSLi,t ≡ SCEOi,t ×
SMINi,t ) for firm i in term t.
Regarding the share prices, we use average prices STPi,t for firm i in term t published in Tokyo Stock
Exchange (1928).4 The observations are fewer than that of financial reports because over-the-counter exchanges
were active.
To control for the financial market conditions when estimating determinants of bank loans and bond flotation,
we use average bank interest rates in the prefecture of Tokyo surveyed by the Bank of Japan.5 The interest rates
are available only from the second half of 1886. Descriptive statistics are shown in Table 2. The dataset is
available in Mendeley Data and Data in Brief.
1 https://j-dac.jp/top/eng/index.html
Last accessed: September 12, 2016.
Japanese Commercial Code then, as its counterparts in the West, required a joint stock company to specify the face value of its
share and permitted partial payment at subscription and hence there existed two kinds of “capital” as legal terms; the capital stock registered,
which was the total sum of face value of issued shares, and the paid-in capital, which was the amount really invested. Thus, the paid-in stock
is the capital in an ordinary sense.
3 The sum of ordinary dividend and special dividend in the term.
4 Tokyo Stock Exchange (1928), ”Sho tokei (Statistics),” pp. 125–261.
5 Historical Statistics: Institute for Monetary and Economic Studies, Bank of Japan (http://www.imes.boj.or.jp/hstat/:
Last accessed on September 18, 2016).
2 The
14
INSERT Table02 HERE
We use cross-section fixed effects model as an estimation method to control for invariant variables during
the sample period, such as long-established routines, historical legacy, corporate culture, corporate philosophy,
and other constant factors. We can then identify the effect of ownership structure changes on financial leverage
and performance. When using cross-section fixed effects model, we need to control for exogenous, and often
cyclical, shocks. Thus, we use the growth in the real gross national product (∆GNPt ≡ GNPt − GNPt−1 ) as a
control variable.6
In our estimates below, we stick to fixed effects models because of a concern that error terms and independent
variables might be correlated. With our dataset, the Hausman pretests do not necessarily reject the pretest null
hypothesis that the random effects model is correct. Furthermore, we have confirmed that the random effects
models do not qualitatively change our results. However, given the concern on the Hausman pretest (Guggenberger (2010)), we conservatively adopt the fixed effects model. We want to warrant reproducibility by sharing
our dataset in Menedeley Data and Data in Brief.
6. Structure of ownership and efficiency of the market
6.1. Responsiveness and prediction power of the market
We first evaluate whether the Japanese market from 1878 to 1910 was distorted due to the market inefficiency.
The skewness-adjusted variation coefficient of the ROE (ROEi,t (≡ PRFi,t / (STKi,t + BBFi,t )) and that of the
ROA (ROAi,t ≡ PRFi,t /TASI,t ) are shown in Table 3.
INSERT Table03 HERE
The skewness-adjusted variation coefficient of the ROE becomes smaller as managers mechanically smoothen
or raise the ROE by leverage distortion. Thus, a change in the gap between the skewness-adjusted variation coefficient of the ROE and that of the ROA tracks the evolution of the market distortion. Table 4 shows that the gap
rose over time. As seen in the number of samples, the number of listed firms also increased. With the market
size being fixed, an increase in the number of listings might lower the stock liquidity and suppress efficiency.
6 The GNP series from 1877 to 1884 is from Teranishi (1983), p. 181 and those from 1885 to 1910 are from Ohkawa et al. (1974), p.
225. The GNP series in those sources are the annual basis, and hence we produced bi-annual series by linear supplements.
15
Regarding the efficiency of the Tokyo Stock Exchange, Nakabayashi (2017), using micro data in the 1890s,
presented that the Bank of Japan’s world’s first unconventional monetary policy in the 1890s substantially lowered equity risk premium, as the unconventional monetary policies by major central banks from the late 2000s to
the 2010s held down risk premia in the bond markets. Hamano et al. (2009) pointed out inefficient pricing due
to low liquidity in the Tokyo Stock Exchange in the early twentieth century. Capitalization of the Tokyo Stock
Exchange continued to rise from 50% of the gross domestic product in 1920 and hit the pre-war hight, 122% in
1936 (Hoshi and Kashyap (2001), p. 39). Still, Bassino and Lagoarde-Segot (2015) demonstrated that the price
index of the Tokyo Stock Exchange did not satisfy the weak form efficiency, using data in the 1930s.
Our result shows that distorted incentives of managers remained to be an issue in the entire sample period
in line with our hypothesis H1 on the difference in the skewness-adjusted variation coefficients. The finding is
consistent with those of previous works on the inefficient Tokyo market before the Second World War.
We then test hypothesis H1 on the market responsiveness to the ROE and the ROA. We first regress the growth
in the stock price (∆ log (STPi,t )) on the growth in the ROE (∆ROEi,t ), the growth in the ROA (∆ROAi,t ) in
line with (1), and the growth in real gross national product (∆GNPt ) to control for cyclical shocks common to
all cross sections as follows:
(5)
∆ log (STPi,t ) = β0 + β1 ∆ROEi,t + β2 ∆ROAi,t + β3 ∆GNPt + µi + ǫi,t ,
where µi is the dummy variable for firm i and ǫi,t is the error term.
When separately including ∆ROEi,t and ∆ROAi,t in specifications 4–1 and 4–2, both have a significantly
positive coefficient. However, following (5), once we control for both in specification 4–3, only ∆ROEi,t has a
significantly positive coefficient. The market predominantly responded to the ROE rather than ROA. The result
supports our hypothesis H1 on short-sighted ROE emphasis by an inefficient market.
We also test the market response to the dividend. If the market is sufficiently efficient such that payout reveals
no additional information privately withheld by firms, this term is expected to have a significantly negative
coefficient to keep shareholder value constant as predicted by Miller and Modigliani (1961). If dividend growth
reveals additional information to predict future cash flow increase, the term is expected to have a significantly
positive coefficient, as predicted by Sasson and Huffman (1986). Our estimate specifications thus are
16
(6)
∆ log (STPi,t ) = β0 + β1 ∆ROEi,t + β2 ∆ROAi,t + β3
TODi,t
+ β4 ∆GNPt + µi + ǫi,t ,
TASi,t
When only TODi,t /TASi,t is in specification 4–4, it has a significantly positive coefficient. The result is
robust when (6) is applied in specification 4–5. The market responded to the payout as a positive signal, which
indicates a low level market efficiency.
INSERT Table 4 Here
6.2. Ownership structure and performance
We now analyze the ownership structure and performance relationship. We regress the ROE (ROEi,t ), the
ROA (ROAi,t ), and the return on sales (ROS, ROSi,t ) on two ownership structure indicators: 1) the president’s
stockholding ratio, SCEOi,t , and 2) degree of the ownership consolidation within the board characterized as
CNSLi,t = SCEOi,t × SMINi,t , where SMINi,t denotes the stockholding ratio of the board member with the
smallest stockholding ratio. The ROA (ROAi,t ) captures efficiency in using corporate total asset, and the ROS
(ROSi,t ) measures how large the margin is or how operational costs are saved. Our interest is in whether the
ownership structure affects efficiency in asset usage and operations.
The first indicator directly measures firm controllability by the president who is often the founding owner in
the sample period. The second one measures whether the board functions as the consolidated representative of
shareholders. If the ownership structure diffuses or employees are promoted as board members, then CNSLi,t
decreases. A decrease in CNSLi,t implies that the board becomes less representative of shareholders and, hence,
they might more likely deviate from the maximization of shareholder value. We also insert the sales (SALi,t ) as
a regressor to control for cyclical but heterogeneous changes in business volumes.
Thus, for the ROE (ROEi,t ), we run
(7)
ROEi,t = β0 + β1 SCEOi,t + β2 SALi,t + β3 ∆GNPt + µi + ǫi,t ,
ROEi,t = β0 + β1 CNSLi, t + β2 SALi,t + β3 ∆GNPt + µi + ǫi,t ,
for the ROA (ROAi,t ),
(8)
ROAi,t = β0 + β1 SCEOi,t + β2 SALi,t + β3 ∆GNPt + µi + ǫi,t ,
ROAi,t = β0 + β1 CNSLi,t + β2 SALi,t + β3 ∆GNPt + µi + ǫi,t ,
17
and for the return on sales (ROSi,t ), dropping the sales from regressors,
ROSi,t = β0 + β1 SCEOi,t + β2 ∆GNPt + µi + ǫi,t ,
(9)
ROSi,t = β0 + β1 CNSLi,t + β2 ∆GNPt + µi + ǫi,t .
The results are presented in Table 5. We find that the president-ownership concentration (SCEOi,t ) did not
significantly improve the ROE (ROEi,t , specification 5–1), but substantially improved the ROA (ROAi,t , specification 5–3) and the ROS (ROSi,t , specification 5–5). Furthermore, a higher consolidation of ownership within
the board (CNSLi,t ) improved all of the ROE, ROA, and ROS (specifications 5–2, 5–4, and 5–6).
INSERT Table05 HERE
Thus, we can conclude that the higher president-ownership concentration or higher consolidation of ownership within the board contributed to long-term growth and profitability by raising asset usage and operations
efficiency. The results support our hypothesis H3 on the positive impact of president-ownership concentration on
the ROA.
6.3. Impact of the Commercial Code of 1899 enactment
In 1899, the Commercial Code came into force. It was modeled on German law and introduced German
corporate law for corporate governance. An immediate change was greater transparency in the disclosure of
financial status. It obligated joint-stock companies to disclose their financial status in a detailed and standardized
form. It made more corporate financial status information publicly available and might have reduced distortion
due to asymmetric information. We insert the interaction term between the dummy variable of enactment (d1899)
(which takes a value 1 if the year is 1899 or later and 0 if otherwise) and the ownership structure variables
(d1899 × SCEOi,t , d1899 × CNSLi,t ), and the enactment dummy variable itself (d1899) into specifications (7),
(8), and (9) to examine the effect.
The results are presented in Table 6. All specifications indicate that the Commercial Code did not affect
performance on its own. However, the positive interaction term d1899 × SCEOi,t coefficient in specification
6–5 suggests that the Code enactment improved the operational efficiency of firms with president-ownership
concentration. The Code was intended to make the market more transparent. However, its enactment did not
make ownership discipline less compelling. Ownership concentration and the judiciary system’s development
were not substitutes but complements in that stage.
18
INSERT Table06 HERE
By Tables 5 and 6, contrary to modern US firms, we conclude that ownership structure was relevant. The
results indicate that the Japanese market was not sufficiently efficient and allowed for self-fulfilling distortion
predicted by Lemma 1.
7. Distorted financial leverage
7.1. Financial leverage and performance
We have shown that president-ownership concentration improved asset usage efficiency in the inefficient
market. Our prediction on its cause is that smaller ownership concentration would allow risk-averse managers to
distort leverage and manipulate the ROE as a response to the market inefficiency (H2).
To examine the validity of the hypothesis, we first regress the ROE (ROEi,t ) on two channels of financial
leverage; the bank loans (LONi,t ) and the outstanding bond (BNDi,t ), over the paid-in capital (STKi,t ) and
balance brought forward (BBFi,t ),
ROEi,t = β0 +β1
(10)
BNDi,t
LONi,t
+ β2
STKi,t + BBFi,t
STKi,t + BBFi,t
+β3 SALi,t + β4 ∆GNPt + µi + ǫi,t .
The results are shown in Table 7. Specification 7–1, including the entire sample, does not show a significant
tendency. The result hints at a heterogeneous effect depending on corporate profitability. Thus, specifications
7–2, 7–3, 7–4, 7–5, and 7–6 separate the sample into ROE ranges: less than 0%, 0% to 10%, 10% to 20%, 20% to
30%, and greater than 30%. For the sub-sample where the ROE is less than 0% and less than 10$ (specifications
7–2 and 7–3), we see that the leverage by the outstanding bond significantly contributed to the ROE. For subsample between 20% and 30% of the ROE, the outstanding bond slightly contributed to the ROE (specification
7–5).
INSERT Table07 HERE
To further investigate leverage effects, we next regress the ROA (ROAi,t ) on the financial leverages,
19
ROAi,t = β0 +β1
(11)
LONi,t
BNDi,t
+ β2
STKi,t + BBFi,t
STKi,t + BBFi,t
+β3 SALi,t + β4 ∆GNPt + µi + ǫi,t .
Table 8 shows that for the range of the ROE higher than 30% (specification 8–6), the outstanding bond positively
contributed to the ROA. Thus, excluding the most profitable firms, financial leverages did not improve asset usage efficiency.
INSERT Table08 HERE
Next, we regress the ROS (ROSi,t ) on the leverages,
(12)
ROSi,t = β0 + β1
LONi,t
BNDi,t
+ β2
+ β3 ∆GNPt + µi + ǫi,t ,
STKi,t + BBFi,t
STKi,t + BBFi,t
where we drop SALi,t from the regressors to avoid a mechanical correlation. The results are in Table 9. We
observe that in the ROE ranges between 0% and 10%, and 10% and 20% (specifications 9–3 and 9–4), the
outstanding bond negatively affected. Meanwhile, the bank loans results are mixed, showing a negative impact
in the ROE range 10 to 20% (specification 9–4) and a positive one in the ROE range 0 to 10% (specification 9–3).
INSERT Table09 HERE
Therefore, concerning the most profitable firms whose ROE was higher than 30%, the bond leverage positively contributed to the asset usage efficiency (specification 8–6 in Table 8). The leverage by the outstanding
bond negatively affected the ROS in the ROE range of 0 to 20% (specifications 9–3 and 9–4 in Table 9). By contrast, the impact of the outstanding bond on the ROE was positive in the ROE range less than 10% (specifications
7–2 and 7–3 in Table 7). The results indicate leverage distortion to smoothen or increase the ROE mechanically
among mediocre performing firms.
7.2. Ownership structure and financial leverage
From Lemma 1, we predict that smaller ownership concentration in an inefficient market implies a greater
financial leverage distortion to smoothen or increase the ROE mechanically at the expense of optimal capital
20
structure (H2). To specify a possible distortion, we first regress the financial leverage changes by the bond
flotation (∆ [BNDi,t /(STKi,t + BBFi,t )]) on ownership structure changes (—considering a possible association
between ownership structure changes and changes in the ROA (∆ROAi,t )), with controlling for business volume
changes by the growth in sales (∆SALi,t ) and changes in the Tokyo market interest rate (∆TKRt ), as follows:
∆
∆
BNDi,t
= β0 +β1 ∆SCEOi,t
STKi,t + BBFi,t
+β2 ∆SALi,t + β3 ∆TKRt + β4 ∆GNPt + µi + ǫi,t ,
(13)
BNDi,t
= β0 +β1 SCEOi,t + β2 ∆SCEOi,t × ∆ROAi,t + β3 ∆ROAi,t
STKi,t + BBFi,t
+β4 ∆SALi,t + β5 ∆TKRt + β6 ∆GNPt + µi + ǫi,t ,
and
BNDi,t
∆
STKi,t + BBFi,t
(14)
= β0 + β1 ∆CNSLi,t + β2 ∆SALi,t + β3 ∆TKRt + β4 ∆GNPt + µi + ǫi,t ,
BNDi,t
∆
STKi,t + BBFi,t
= β0 + β1 ∆CNSLi,t + β2 ∆CNSLi,t × ∆ROAi,t
+ β3 ROAi,t + β4 ∆SALi,t + β5 ∆TKRt + β6 ∆GNPt + µi + ǫi,t .
The results are presented in Table 10. First, we observe that president-ownership concentration (SCEOi,t )
tended to lower the financial leverage by the bond flotation (specification 10–1). However, we also observe that it
raised the leverage by the bond flotation when it was accompanied by an increase in the ROA (ROAi,t ) as shown
by the significantly positive coefficient of the interaction term (∆SCEOi,t × ∆ROAi,t ) in specification 10–2.
The president-ownership concentration was likely to control financial leverage unless it was associated with improvement in asset usage efficiency. The result is consistent with our hypothesis H2 on the leverage distortion
reduction by ownership concentration.
INSERT Table10 HERE
By running the same regressions for the changes in the leverage by an increase in bank loans, we find no sig-
21
nificant impact of the ownership structure as shown in Table 11. Leverage distortion due to a diffused ownership
structure was severe in the bond market but not with bank loans.
INSERT Table11 HERE
7.3. Bond flotation as the channel of distortion
A higher president-ownership concentration led to an increase in the ROA (Table 5). Higher leverage through
the bond flotation increased the ROA for top firms with higher than 30% ROE (Table 8). A higher presidentownership concentration lowered the financial leverage through the bond flotation but raised it if a rise in the
ROA accompanied it (Table 10). The president-ownership concentration of ownership never affected the ROE
(Table 5). Meanwhile, greater leverage through the bond flotation raised the ROE in the range of less than 10%
(Table 7).
Given the results, we conclude that mediocre performing firms whose ownership structure was more diffused
were more prone to the distortion of financial leverage through over-reliance on the bond flotation. Mediocre
performing firms in the ROE range of less than 10% deceived the market when raising the leverage to smoothen
or increase the short-term ROE mechanically. These results are mutually consistent and support our hypotheses
H2 on a reduction in the leverage distortion by a higher ownership concentration.
Mediocre performing firms were particularly inclined to distort the financial leverage in the inefficient Japanese
market from 1878 to 1910. A higher president-ownership concentration controlled the bond flotation but raised
the bond flotation in the case where it contributed to an increase in the ROA. Thus, in the inefficient market, a
higher ownership concentration contributed to better leverage, as predicted by our hypothesis H2.
8. Conclusion
The inconsequential ownership structure in modern US firms is possibly explained by market discipline
(Demsetz and Lehn (1985); Himmelberg et al. (1999) and Demsetz and Villalonga (2001)). This seems to be the
case because of the sufficient degree of efficiency of the modern US market.
By contrast, our results show that ownership structure substantially affected corporate performance in the
Tokyo market from the late nineteenth century to the early twentieth century. Pricing in the Tokyo market predominantly rewarded the ROE but not the ROA. Contrary to the ROA, the ROE is mechanically manipulable by
leverage distortion. Investors can decern it. However, suppose that the market is so inefficient that it is considerably costly to specify the leverage distortion by each firm and that the discount rate of investors is substantially
22
high. Then, it is optimal for investors to reward the ROE instead of the ROA. The ROE is manipulable by managerial leverage distortion, hence, investors can save the risk premium to be paid to risk-averse managers by
rewarding the ROE rather than the ROA. It sacrifices the long-term performance of the investment. However, it
is beyond the scope of uninformed and short-sighted investors.
The way to offset the market weakness is an ownership concentration centered on the president who is
often the founder. A long-sighted and informed president pursue long-term growth in the firm’s valuation. Our
results show that a higher president-ownership concentration led to a higher ROA, but it did not affect the ROE.
Enactment of the Commercial Code of 1899 augmented the impact of the president-ownership concentration.
Market transparency and the ownership discipline were complements rather than substitutes.
A higher president-ownership concentration tended to hold down the leverage by bond flotation unless it
was to boost the ROA. Higher bond leverage raised the ROA only for the top tier firms. Meanwhile, mediocre
performing firms were tempted to raise the bond leverage to boost the ROE.
By contrast, we do not find evidence of bank loan distortion. The different results for corporate bonds
and bank loans are consistent with what we saw in Japan in the after the deregulation. Agency problem is
more significant in the bond market as banks faced by disintermediation improved their efficiency in screening
(Anderson and Makhija (1999); Uchida and Satake (2009); Nakagawa and Uchida (2011); Uchida and Udell
(forthcoming)).
In summary, the inefficient Tokyo market from the late nineteenth century to the early twentieth century
allowed managers to manipulate the ROE by bond flotation. A higher president-ownership concentration suppressed the adverse effect of the market inefficiency. Thus, our work, along the lines of Morck et al. (2000),
Gedajlovic and Shapiro (2002), Pindado et al. (2014), and Hamadi and Heinen (2015), provides more evidence
that ownership matters in non-US markets.
We observed that the management-ownership concentration enabled the firm to pursue long-term growth. A
remaining question is whether ownership concentration on its own, which was not necessarily at the management, helped. Case studies support the possibility. Railway industry in the late nineteenth was one of the most
technology intensive one. Thus, the tendency was promotion of experts into management (Nakamura (2000)).
Then, the other way of shareholder/manager conflicts arose. Small- and medium-sized shareholders who tended
to sell shares in short order preferred payout to investment for long-term growth. Typically, large shareholders
who tended to be “buy-and-hold” type helped management avoid under-investment (Nakamura (2014)). In case
of another leading industry in the age, the cotton spinning (Nakamura (2015); Dong et al. (2015)), large “buy-
23
and-hold” type shareholders tended to persuade small- and medium-sized shareholders to approve investment
suggested by managers rather than demand payout in shareholders annual meetings (Yuki (2011)). These case
studies indicate a possibility that ownership concentration itself improved management in an emerging Japan, as
Abdallah and Ismail (2017) showed for Gulf Cooperative Council region.
Our results also have a policy implication. Prevalence of family firms in non-US nations often attracts attention because it might accompany a divide between management and control—typically as stockholder/bondholder
conflicts—and hurt efficiency (Claessens et al. (2000) and Hamadi and Heinen (2015)). The exploitation of other
stakeholders by the founding owner is precisely the issue on which Jensen and Meckling (1976) focused. However, we should also admit the virtue of blockholding that includes founding families. If the market is not sufficiently efficient to contain stockholder/manager conflicts, something else must cancel it out. Next to the efficient
market, concentrated ownership is among the second-best alternatives. That is why family firms still prosper
in non-US nations. Desirable reforms in those nations make the market more transparent without restricting
blockholding.
Abdallah, A. A., Ismail, A. K., Jan 2017. Corporate governance practices, ownership structure, and corporate
performance in the GCC countries. Journal of International Financial Markets, Institutions & Money 46, 98–
115.
Admati, A., Pfleiderer, P., Spring 1988. A theory of intraday patterns: Volume and price variability. The Review
of Financial Studies 1 (1), 3–40.
Adrian, T., Rosenberg, J., Dec 2008. Stock returns and volatility: Pricing the short-run and long-run components
of market risk. The Journal of Finance 63 (6), 2997–3030.
Aman, H., Nguyen, P., Aug 2013. Does good governance matter to debtholders? Evidence from the credit ratings
of Japanese firms. Research in International Business and Finance 29, 14 –34.
Anderson, C., Makhija, A., Feb 1999. Deregulation, disintermediation, and agency costs of debt: evidence from
Japan. Journal of Financial Economics 51 (2), 309–339.
Anderson, R. C., Reeb, D. M., Oct 2003. Founding-family ownership, corporate diversification, and firm leverage. The Journal of Law and Economics 46 (2), 653–684.
24
Aslan, H., Kumar, P., Jul 2012. Strategic ownership structure and the cost of debt. The Review of Financial
Studies 25 (7), 2257–2299.
Banerjee, A., Eckard, E. W., Dec 2001. Why regulate insider trading? Evidence from the first great merger wave
(1897–1903). The American Economic Review 91 (5), 1329–1349.
Bassino, J.-P., Lagoarde-Segot, T., Nov 2015. Informational efficiency in the Tokyo Stock Exchange, 1931-40.
The Economic History Review 68 (4), 1226–1249.
Ben-Nasr, H., Boubaker, S., Rouatbi, W., Dec 2015. Ownership structure, control contestability, and corporate
debt maturity. Journal of Corporate Finance 35, 265–285.
Bergstresser, D., Desai, M., Rauh, J., 2006. Earnings manipulation, pension assumptions, and managerial investment decisions. The Quarterly Journal of Economics 121 (1), 157–195.
Berle, A. A., Means, G. C., 1933. The Modern Corporation and Private Property. Macmillan, New York.
Bertrand, M., Mullainathan, S., Aug 2001. Are CEOs rewarded for luck? The ones without principals are. The
Quarterly Journal of Economics 116 (3), 901–932.
Betrn, C., Huberman, M., Feb 2016. International competition in the first wave of globalization: new evidence
on the margins of trade. The Economic History Review 69 (1), 258–287.
Blanchard, O. J., Lopez-de Silanes, F., Shleifer, A., Dec 1994. What do firms do with cash windfalls? Journal of
Financial Economics 36 (3), 337–360.
Bolton, P., Scharfstein, D. S., Feb 1996. Optimal debt structure and the number of creditors. Journal of Political
Economy 104 (1), 1–25.
Bond, P., Eraslan, H., Sep 2010. Information-based trade. Journal of Economic Theory 145 (5), 1675–1703.
Borg, J. R., O., B. M., Leeth, J. D., Mar 1989. The success of mergers in the 1920s: A stock market appraisal of
the second merger wave. International Journal of Industrial Organization 7 (1), 117–131.
Burhop, C., Lübbers, T., 2009. Cartels, managerial incentives, and productive efficiency in German coal mining,
1881–1913. The Journal of Economic History 69 (2), 500–527.
Byrd, J., Parrino, R., Pritch, G., May/Jun 1998. Stockholder-manager conflicts and firm value. Financial Analysts
Journal 54 (3), 14–30.
25
Chandler, A. D., 1977. The Visible Hand: The Managerial Revolution in American Business. Belknap Press,
Cambridge: MA.
Claessens, S., Djankov, S., Lang, L., 2000. The separation of ownership and control in East Asia corporations.
Journal of Financial Economics 58 (1–2), 81–112.
Collin-Dufresne, P., Fos, V., Jul 2016. Insider trading, stochastic liquidity, and equilibrium prices. Econometrica
84 (4), 1441–1475.
Conrad, J., Dittmar, R., Ghysels, E., Feb 2013. Ex ante skewness and expected stock returns. The Journal of
Finance 68 (1), 85–124.
Davies, J. R., Hillier, D., McClogan, P., Sep 2005. Ownership structure, managerial behavior and corporate value.
Journal of Corporate Finance 11 (4), 645–660.
de Wet, J., du Toit, E., Mar 2007. Return on equity: A popular, but flawed measure of corporate financial
performance. South African Journal of Business Management 38 (1), 59–69.
Demsetz, H., Lehn, K., Dec 1985. The structure of corporate ownership: Causes and consequences. The Journal
of Political Economy 93 (6), 11155–1177.
Demsetz, H., Villalonga, B., Sep 2001. Ownership structure and corporate performance. Journal of Corporate
Finance 7 (3), 209–233.
Dhillon, A., Rossetto, S., Feb 2015. Ownership structure, voting, and risk. The Review of Financial Studies
28 (2), 521–560.
Diamond, D. W., Dybvig, P. H., Jun 1983. Bank runs, deposit insurance, and liquidity. The Journal of Political
Economy 91 (3), 401–419.
Dong, B., Gong, J., Peng, K., Zhao, Z., Nov 2015. Little divergence: Evidence from cotton textiles in Japan and
China 1868–1930. Review of Development Economics 19 (4), 776–796.
Enriques, L., Volpin, P., 2007. Corporate governance reforms in Continental Europe. The Journal of Economic
Perspectives 21 (1), 117–140.
Fama, E. F., Jan-Mar 1980. Agency problems and the theory of the firm. The Journal of Political Economy 88 (2),
288–307.
26
Ferguson, T., Voth, H.-J., 2008. Betting on Hitler: The value of political connections in Nazi Germany. The
Quarterly Journal of Economics 123 (1), 101–137.
Fohlin, C., 2007. Does civil law tradition and universal banking crowd out securities markets? Pre-World War I
Germany as counter-example. Enterprise & Society 8 (3), 602–641.
Foreman-Peck, Hannah, L., 2013. Some consequences of the early twentieth-century British divorce of ownership from control. Business History 55 (4), 543–564.
Franks, J., Mayer, C., Wagner, H. F., 2006. The origins of the German Corporation–finance, ownership and
control. Review of Finance 10 (4), 537–585.
Gedajlovic, E., Shapiro, D., Jun 2002. Ownership structure and firm profitability in Japan. Academy of Management Journal 45 (2), 565–575.
Gedajlovic, E., Yoshikawa, T., Hashimoto, M., 2005. Ownership structure, investment behaviour and firm performance in Japanese manufacturing industries. Organization Studies 26 (1), 7–35.
Goldstein, I., Pauzner, A., Nov 2004. Contagion of self-fulfilling financial crises due to diversification of investment portfolios. Journal of Economic Theory 119 (1), 151–183.
Gorton, G., Kahl, M., Dec 2008. Blockholder scarcity, takeovers, and ownership structures. The Journal of
Financial and Quantitative Analysis 43 (4), 937–974.
Guggenberger, P., Jun 2010. The impact of a Hausman pretest on the size of a hypothesis test: The panel data
case. Journal of Econometrics 156 (2), 337–343.
Haider, Z., Liu, M., Wang, Y., Zhang, Y., Mar 2018. Government ownership, financial constraint, corruption, and
corporate performance: International evidence. Journal of International Financial Markets, Institutions and
Money 53, 76–93.
Hamadi, M., Heinen, A., Dec 2015. Firm performance when ownership is very concentrated: Evidence from a
semiparametric panel. Journal of Empirical Finance 34, 172–194.
Hamano, Y., Hoshi, T., Okazaki, T., 2009. Listing policy and development of the Tokyo Stock Exchange in the
prewar period. In: Ito, T., Rose, A. K. (Eds.), Financial Sector Development in the Pacific Rim, East Asia
Seminar on Economics. University of Chicago Press, Chicago, pp. 51–87.
27
Helwege, J., Pirinsky, C., Stulz, R. M., Jun 2007. Why do firms become widely held? An analysis of the dynamics
of corporate ownership. The Journal of Finance 62 (3), 995–1028.
Himmelberg, C. P., Hubbard, G. R., Palia, D., Sep 1999. Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53 (3), 353–384.
Holmstrom, B., Milgrom, P., Jan 1991. Multitask principal-agent analyses: Incentive contracts, asset ownership,
and job design. Journal of Law, Economics, & Organization 7 (Special Issue), 24–52.
Holmstrom, B., Tirole, J., Aug 1993. Market liquidity and performance monitoring. Journal of Political Economy
101 (4), 678–709.
Hoshi, T., Kashyap, A., 2001. Corporate Financing and Governance in Japan: The Road to the Future. The MIT
Press, Cambridge, MA.
Jackson, G., 2001. The origins of nonliberal corporate governance in Germany and Japan. In: Streeck, W.,
Yamamura, K. (Eds.), The Origins of Nonliberal Capitalism: Germany and Japan in Comparison. Cornell
University Press, Ithaca, pp. 121–170.
Jensen, M., 2000. A Theory of the Firm: Governance, Residual Claims, and Organizational Forms. Harvard
University Press, Cambridge, MA.
Jensen, M. C., Meckling, W., Oct 1976. Theory of the firm: Managerial behavior, agency costs and ownership
structure. Journal of Financial Economics 3 (4), 305–360.
Julian, F., Mayer, C., 2001. Ownership and control of German corporations. The Review of Financial Studies
14 (4), 943–977.
Kawashima, T., 2018. A European political-economic space that embraced Japan: the international context of the
conventional-tariff network, ca.1892-1914. European Review of History: Revue europenne d’histoire 25 (3–4),
612–635.
Kling, G., Dec 2006. Does the merger paradox exist even without any regulations? Evidence from Germany in
the pre-1914 period. Empirica 33 (5), 315–328.
Knight, F., 1921. Risk, Uncertainty and Profit. Houghton Mifflin, Boston.
28
Kraft, K., Niederprüm, A., Sep 1999. Determinants of management compensation with risk-averse agents and
dispersed ownership of the firm. Journal of Economic Behavior & Organization 40 (1), 17–27.
Kraus, A., Litzenberger, R., Sep 1976. Skewness preference and the valuation of risk assets. The Journal of
Finance 31 (4), 1085–1100.
Kunieda, T., Shibata, A., Sep 2016. Asset bubbles, economic growth, and a self-fulfilling financial crisis. Journal
of Monetary Economics 82, 70–84.
Kyle, A. S., Nov 1985. Continuous auctions and insider trading. Econometrica 53 (6), 1315–1335.
La Porta, R., Lopez-de-Silanes, Shleifer, A., Apr 1999. Corporate ownership around the world. The Journal of
Finance 54 (2), 471–517.
La Porta, R., Lopez-de Silanes, F., Shleifer, A., Jun 2008. The economic consequences of legal origins. Journal
of Economic Literature 46 (2), 285–332.
Laeven, L., Levine, R., Mar 2008. Complex corporate ownership structures and corporate valuations. The Review
of Financial Studies 21 (2), 5791–604.
Leeth, J. D., Borg, J. R., Jun 1994. The impact of mergers on acquiring firm shareholder wealth: The 1905–1930
experience. Empirica 21 (1), 221–244.
Leeth, J. D., Borg, J. R., Jun 2000. The impact of takeovers on shareholder wealth during the 1920s merger wave.
Journal of Financial and Quantitative Analysis 35 (2), 217–238.
Lehmbruch, G., 2001. The institutional embedding of market economies: The German “Model” and its impact
on Japan. In: Streeck, W., Yamamura, K. (Eds.), The Origins of Nonliberal Capitalism: Germany and Japan
in Comparison. Cornell University Press, Ithaca, pp. 39–93.
Lichtenberg, F., Pushner, G., Oct 1994. Ownership structure and corporate performance in Japan. Japan and the
World Economy 6 (3), 239–261.
Mahrt-Smith, J., May 2005. The interaction of capital structure and ownership structure. The Journal of Business
78 (3), 787–816.
Mauro, P., Sussman, N., Yafeh, Y., 2006. Emerging Markets and Financial Globalization: Sovereign Bond
Spreads in 1870–1913 and Today. Oxford University Press, Oxford.
29
Mehrotra, V., Morck, R., Shim, J., Wiwattanakantang, Y., Jun 2013. Adoptive expectations: Rising sons in
Japanese family firms. Journal of Financial Economics 108 (3), 840–854.
Miller, M. H., Modigliani, F., Oct 1961. Dividend policy, growth, and the valuation of shares. The Journal of
Business 34 (4), 411–433.
Miwa, Y., Ramseyer, J. M., Jun 2002. The value of prominent directors: corporate governance and bank access
in transitional Japan. Journal of Legal Studies 31 (2), 273–301.
Miyamoto, M., Abe, T., 1999. Kogyoka shoki ni okeru Nihon kigyo no corporate governance: Osaka Boseki
Kaisha to Nihon Seimei Hoken Kaisha no jirei (Corporate governance of Japanese firm in the early age of
industrialization: Cases from Osaka Cotton Spinning Company and Nippon Life Insurance Company). Osaka
Digaku Keizaigaku (Osaka Economic Papers) 48 (3–4), 176–197.
Modigliani, F., Miller, M. H., Jun 1958. The cost of capital, corporation finance and the theory of investment.
The American Economic Review 48 (3), 261–297.
Morck, R., Nakamura, M., Shivdasani, A., Oct 2000. Banks, ownership structure, and firm value in Japan. The
Journal of Business 73 (4), 549–567.
Morck, R., Shleifer, A., Vishny, R. W., Jan-Mar 1988. Management ownership and market valuation. Journal of
Financial Economics 20, 293–315.
Morck, R., Wolfenzon, D., Yeung, B., 2005. Corporate governance, economic entrenchment, and growth. Journal
of Economic Literature 43 (3), 655–720.
Murphy, K. J., 1999. Executive compensation. In: Ashenfelter, O., Card, D. (Eds.), Handbook of Labor Economics, 3B. Elsevier, Amsterdam, pp. 2485–2563.
Murphy, K. J., Summer 2002. Explaining executive compensation: Managerial power versus the perceived cost
of stock options. University of Chicago Law Review 69, 847–869.
Mykhayliv, D., Zauner, K., Aug 2017. The impact of equity ownership groups on investment: Evidence from
Ukraine. Economic Modelling 64, 20–25.
Nakabayashi, M., 2012. The rise of a Japanese fiscal state. In: Yun-Casalilla, B., O’Brien, P. (Eds.), The Rise of
Fiscal States: A Global History, 1500–1914. Cambridge University Press, Cambridge, pp. 378–409.
30
Nakabayashi, M., May 2014. Imposed efficiency of treaty ports: Japanese industrialization and Western imperialist institutions. Review of Development Economics 18 (2), 254–271.
Nakabayashi, M., Apr 2017. Contained crisis and socialized risk: Unconventional monetary policy by the Bank
of Japan in the 1890s. Research in International Business and Finance 40, 231–241.
Nakagawa, R., Uchida, H., 2011. Herd behaviour by Japanese banks after financial deregulation. Economica
78 (312), 618–636.
Nakamura, N., Oct 2000. Meiji-era industrialization and provincial vitality: the significance of the first-enterprise
boom of the 1880s. Social Science Japan Journal 3 (2), 187–205.
Nakamura, N., 2007. Personnel management and the formation of modern business organization: The railway
industry in Japan before the First World War. In: Okazaki, T. (Ed.), Production Organizations in Japanese
Economic Development. Routledge, London, pp. 75–109.
Nakamura, N., 2014. Shisan tokushu sei to kigyo no kyokai: Meijiki Chikuho chiiki ni okeru tanko to tetsudo
(Asset specificity and the boundary of firm: Coal mining and railway industries in the Chikuho area in the
Meiji Era). In: Nakabayashi, M., Ishiguro, S. (Eds.), Kigyo no Keizaigaku: Kozo to Seicho (Economics of the
Firm: Organization and Growth). Yuhikaku, Tokyo, pp. 101–119.
Nakamura, N., Jan 2015. Reconsidering the Japanese industrial revolution: Local entrepreneurs in the cotton
textile industry during the Meiji Era. Social Science Japan Journal 18 (1), 23–44.
Ohkawa, K., Takamatsu, N., Yamamoto, Y., 1974. Estimates of Long-term Economic Statistics of Japan since
1868: 1 National Income. Toyo Keizai Shinposha, Tokyo.
Okazaki, T., 1999. Corporate governance. In: Okazaki, T., Okuno-Fujiwara, M. (Eds.), The Japanese Economic
System and Its Historical Origins. Oxford University Press, New York, pp. 97–144, translated by Susan Herbert.
Parrino, R., Poteshman, A., Weisbach, M., Mar 2005. Measuring investment distortions when risk-averse managers decide whether to undertake risky projects. Financial Management 34 (1), 21–60.
Pindado, J., Requejo, I., de la Torre, C., Jun 2014. Family control, expropriation, and investor protection: A panel
data analysis of Western European corporations. Journal of Empirical Finance 27, 58–74.
31
Rajan, R., Zingales, L., 2003. The great reversals: The politics of financial development in the twentieth century.
Journal of Financial Economics 69 (1), 5–50.
Ringe, W.-G., 2015. Changing law and ownership patterns in Germany: Corporate governance and the erosion
of Deutschland AG. The American Journal of Comparative Law 63 (2), 493–538.
Sakawa, H., Watanabel, N., Oct 2018. Parent control and ownership monitoring in publicly listed subsidiaries in
Japan. Research in International Business and Finance 45, 7–14.
Salazar, A. R., Raggiunti, J., 2016. Why does executive greed prevail in the United States and Canada but not in
Japan? The pattern of low CEO pay and high worker welfare in Japanese corporations. The American Journal
of Comparative Law 64 (3), 721–744.
Sasson, B.-Y., Huffman, L., Mar 1986. The information content of dividends: A signalling approach. The Journal
of Financial Quantitative Analysis 21 (1), 47–58.
Savage, L., 1954. The Foundations of Statistics. Wiley, New York.
Scott, R., Horvath, P., Sep 1980. On the direction of preference for moments of higher order than the variance.
The Journal of Finance 35, 915–919.
Shleifer, A., Vishny, R., Jun 1986. Large shareholders and corporate control. Journal of Political Economy 94 (3,
Part 1), 461–488.
Shleifer, A., Vishny, R., 1997. A survey of corporate governance. The Journal of Finance 52 (2), 737–783.
Smith, A., 1937[1776]. An Inquiry into the Nature and Causes of the Wealth of Nations. The Modern Library,
New York, first published in 1776.
Teranishi, J., 1983. Matsukata-Defure no makuro keizaigakuteki bunseki: Kaiteiban (A macroeconomic analysis
of the Matsukata Deflation: Revised). In: Umemura, M., Nakamura, T. (Eds.), Matsukata-Defure to Shokusan
Kogyo (The Matsukata-Deflation and the Encouragement of New Industries). United Nations University Press,
Tokyo, pp. 157–185.
Thomadakis, S., Gounopoulos, D., Nounis, C., Riginos, M., Aug 2017. Innovation and upheaval: early growth
in Greek capital market listings and IPOs from 1880 to the Second World War in the Athens Stock Exchange.
The Economic History Review 70 (3), 859–892.
32
Tirole, J., 2001. Corporate governance. Econometrica 69 (1), 1–35.
Tokyo Stock Exchange, 1928. Tokyo Kabushiki Torihikijo Goju Nenshi (Fifty Years of the Tokyo Stock Exchange). Tokyo Stock Exchange, Tokyo.
Uchida, H., Satake, M., Dec 2009. Market discipline and bank efficiency. Journal of International Financial
Markets, Institutions and Money 19 (5), 792– 802.
Uchida, H., Udell, G., forthcoming. Banking in japan. In: Berger, A., Molyneux, P., Wilson, J. (Eds.), The
Oxford Handbook of Banking, 3rd Edition. Oxford University Press, Oxford.
Varian, B. D., Feb 2018. Anglo-American trade costs during the first era of globalization: the contribution of a
bilateral tariff series. The Economic History Review 71 (1), 190–212.
Vitols, S., 2001. The origins of bank-based and market-based financial systems: Germany, Japan, and the United
States. In: Streeck, W., Yamamura, K. (Eds.), The Origins of Nonliberal Capitalism: Germany and Japan in
Comparison. Cornell University Press, Ithaca, pp. 171–199.
Yui, T., 1979. Meiji jidai ni okeru juyaku soshiki no keisei (The formative process of the top management
organization of Mitsubishi during Meiji era). Keiei Shigaku (Japan Business History Review) 14 (1), 1–27.
Yui, T., 1989. Development, organization, and business strategy of industrial enterprises in Japan (1915–1935).
Japanese Yearbook on Business History 5, 56–87.
Yui, T., 1992. The enterprise system in Japan: Preliminary considerations on internal and external structural
relations. Japanese Yearbook on Business History 8, 49–66.
Yuki, T., 2011. Kigyo tochi ni okeru kabunushi sokai no yakuwari: Osaka boseki kaisha no jirei (the role of
shareholders meetings in corporate governance). Keiei Shigaku (Japan Business History Review) 46 (3), 56–
77.
33
Table 1 Distribution of ownership, 1897.
Individuals
Shares owned
Number of
Number of
share
shareholders
shares
1-99
467
24.4%
22,043
100-499
962
50.3%
209,432
500-999
207
10.8%
140,402
1,000-1,999
151
7.9%
201,586
2,000-2,999
50
2.6%
120,260
3,000-3,999
22
1.2%
73,142
4,000-4,999
9
0.5%
40,137
5,000-5,999
13
0.7%
68,844
6,000-6,999
4
0.2%
25,318
7,000-7,999
3
0.2%
22,524
8,000-8,999
2
0.1%
16,766
9,000-9,999
2
0.1%
18,625
10,00020
1.0%
482,785
Total
1,912 100.0% 1,441,864
share
1.5%
14.5%
9.7%
14.0%
8.3%
5.1%
2.8%
4.8%
1.8%
1.6%
1.2%
1.3%
33.5%
100.0%
Number of
shareholders
2
10
7
6
4
2
1
1
0
2
3
1
5
44
Institutions
Number of
share
shares
4.5%
189
22.7%
2,571
15.9%
3,917
13.6%
9,808
9.1%
8,182
4.5%
7,286
2.3%
4,000
2.3%
5,250
0.0%
0
4.5%
14,221
6.8%
25,694
2.3%
9,575
11.4%
691,558
100.0%
782,251
share
0.0%
0.3%
0.5%
1.3%
1.0%
0.9%
0.5%
0.7%
0.0%
1.8%
3.3%
1.2%
88.4%
100.0%
Number of
shareholders
469
972
214
157
54
24
10
14
4
5
5
3
25
1,956
Total
Number of
share
shares
24.0%
22,232
49.7%
212,003
10.9%
144,319
8.0%
211,394
2.8%
128,442
1.2%
80,428
0.5%
44,137
0.7%
74,094
0.2%
25,318
0.3%
36,745
0.3%
42,460
0.2%
28,200
1.3%
1,174,343
100.0%
2,224,115
share
1.0%
9.5%
6.5%
9.5%
5.8%
3.6%
2.0%
3.3%
1.1%
1.7%
1.9%
1.3%
52.8%
100.0%
Table 2 Descriptive statistics of firms listed at the Tokyo Stock Exchange, from the first half of 1878 to the second half half of 1910.
Number of individual firms (cross
sections)
95
variables
Number of
total
observations
Unit
Mean
Median
Maximum
Standard
deviation
Minimum
Skewness
Kurtosis
Amount of sales in the current term
SAL
1,101
Yen
1,673,988
524,863
19,305,644
600
2,818,222.923
2.791
11.617
Total assets as of the current term
TAS
1,119
Yen
15,717,824
3,651,671
301,457,885
52,168
35,907,995.178
5.014
31.721
Paid-in stock as of the current term
STK
1,077
Yen
6,111,014
1,600,000
102,000,000
25,000
12,036,180.793
4.681
32.877
Bank loans as of the current term
LON
1,119
Yen
375,992
0
13,146,042
0
1,083,525.975
4.911
36.511
Outstanding bond as of the current term BND
1,119
Yen
1,471,965
0
93,568,012
0
8,000,257.208
7.924
70.028
Profit in the current term
PRF
1,081
Yen
388,115
97,992
18,084,554
-1,318,361
837,159.095
9.887
188.851
Total dividends in the current term
DVD
979
Yen
283,615
75,000
3,648,813
0
480,640.730
2.721
12.120
Balance brought forward as of the end of
BBF
the current term
1,113
Yen
88,170
10,195
2,316,513
-1,065,271
241,258.728
3.750
26.400
Yen
89.1015
68.3000
425.5000
6.2400
76.721
1.818
6.290
Average share price in the current term
STP
323
Return on equity: =PRF/(STP+BBF)
ROE
1,040
percent
8.1012%
6.2254%
104.6430%
-104.7981%
0.108
1.598
33.371
Return on asset: =PRF/TAS
ROA
1,080
percent
3.3989%
2.9467%
34.6725%
-33.8713%
0.038
0.875
23.463
Stock holding ratio of the president as of
the current term: =[Shares owned by
SCEO
President]/[Total Share]
610
percent
5.1408%
2.7633%
70.0000%
0.0000%
0.071
4.248
30.622
Stock holding ratio of the director whose
stock holding ratio is the smallest in the
board as of the current term: [Share
SMIN
owned by the board member]/[Total
Share]
610
percent
1.1309%
0.7000%
10.0000%
0.0000%
0.015
3.360
17.700
Measure of ownership consolidation in
the board: =SCEO×SMIN
610
per ten
thousand
8.7819‱
2.0000‱
130.0000‱
0.0000‱
0.002
3.630
16.941
CNSL
Notes : All values are nominal terms. Japan had adopted the silver standard until September 1897 and hence its exchange rate against th US dollar and the
Sterling pound had been voltile. From October 1897 to the First World War, Japan adopted the gold standard. The fixed exchange rate stiputed by the Coinage
Act of 1897 was JPY100=USD49.875 and the rate was sustained by the monetary policy of the Bank of Japan until the breakout of the First Word War.
Table 3 Skewness-adjusted variation coefficient from the first half of 1878 to the second half of 1910.
peirod
ROE
ROA
Degree of distortion
b –a
a
b
1878–1888
0.3185
0.3425
0.0239
Number of observations
59
72
1889–1899
0.3891
0.4877
0.0986
Number of observations
245
250
1900–1910
24.8754 119.7480
94.8726
Number of observations
736
758
1878–1910
0.8355
1.2643
0.4288
Number of observations
1,040
1,080
Notes: ROE: return on equity. ROA: return on asset.
Table 4 Determinants of the stock prices (STP), from the first half of 1879 to the second half of 1910.
Δlog(STPi , t )
Δlog(STPi , t )
Δlog(STPi , t )
Δlog(STPi , t )
Dependent variable
4 1
4 2
4 3
4 4
panel least squares
panel least squares
panel least squares
panel least squares
estimation method
Cross section fixed effect fixed
fixed
fixed
fixed
Independent variables
t statistic
t statistic
t statistic
t statistic
Constant
-0.0046
-0.22
-0.0085
-0.40
-0.0042
-0.20
0.0021
0.11
ΔROEi , t
1.5407
3.66 ***
2.2891
3.24 ***
ΔROAi , t
2.4061
2.12 ** -2.4499
-1.31
Δ(TODi , t /TASi , t )
7.4979
4.30
ΔGNPt
0.0000
0.22
0.0001
0.37
0.0000
0.16
adjusted R2
Log likelihood
F statistic
Number of individual firms
(cross sections)
0.04
-0.01
0.04
0.07
-3.38
1.33
-7.75
0.92
-2.40
1.35
9.39
1.69
24
25
24
23
***
Δlog(STPi , t )
4 5
panel least squares
fixed
t statistic
0.0022
0.11
-0.0242
-0.38
-1.5132
-0.74
9.3439
3.22
0.0000
0.34
***
0.08
**
12.66
1.69
22
Number of total
217
218
217
209
201
observations
Notes : STP: Stock price offirm i in semiannual period. ROE: returon on equity. ROA: return on asset. TOD: total payouf of dividend. TAS: total
assets. GNP: Groth National Product. ***, **, and * denote significance of 1, 5, and 10 percent levels respectively.
**
Table 5 The return on equity (ROE), return on asset (ROA), and return on sales (ROS) and the stock ownership structure, from the second half of 1878 to the second half
of 1910.
Dependent variables
ROEi , t
51
Panel least squares
estimation method
Cross section fixed
fixed
effect
Independent variables
Constant
0.0678
SCEOi , t
0.0945
CNSLi , t
SALi , t
0.0000
ΔGNPt
-0.0001
2
adjusted R
Log likelihood
F statistic
Number of individual
firms (cross sections)
t statistic
8.37
0.83
8.41
-3.52
67
ROAi , t
53
Panel least squares
ROAi , t
54
Panel least squares
ROSi , t
55
Panel least squares
ROSi , t
56
Panel least squares
fixed
fixed
fixed
fixed
fixed
***
0.0688
t statistic
11.44
***
7.0553
0.0000
-0.0001
2.29
8.56
-3.55
***
0.47
599.53
8.21
ROEi , t
52
Panel least squares
***
***
***
591.10
8.36 ***
67
t statistic
10.44
2.45
***
0.0000
0.0000
6.77
-2.97
***
***
0.42
1,225.36
6.86
70
0.0317
t statistic
14.96
2.8618
0.0000
0.0000
2.76
6.76
-3.09
***
**
**
0.48
***
0.0292
0.0943
1,226.28
6.90
70
t statistic
-3.90
7.90
***
0.0351
t statistic
0.47
223.2661
5.32
-0.0004
-0.96
***
***
***
***
0.42
***
-0.3955
11.9172
-0.0003
-0.64
0.18
***
-909.40
2.76
70
0.12
***
-927.25
2.17
70
Number of total
560
582
582
582
582
560
observations
Notes : ROE: returon on equity. ROA: returon on asset. ROS: returon on sales. SCEO: ownership share of the CEO. CNSL:Ownership consolidation within the board =
SCEO×SMIN, where SIMIN: ownership shaer of the board member whose onwership was smallest withiint the board. SAL: sales. GNP: gross national product. ***, **,
and * denote significance of 1, 5, and 10 percent levels respectively.
***
Table 6 Impacts of the Commercial Code on the asset and operation efficiency, from the second half of 1878 to the second half of 1910.
ROEi , t
ROEi , t
ROAi , t
ROAi , t
ROSi , t
Dependent variables
63
64
65
61
62
estimation method
Panel least squares
Panel least squares
Panel least squares
Panel least squares
Panel least squares
ROSi , t
66
Panel least squares
Cross section fixed effect fixed
fixed
Independent variables
Constant
SCEOi , t
d1899×SCEOi , t
CNSLi , t
d1899×CNSLi , t
d1899
SALi , t
ΔGNPt
0.0829
0.0555
0.0397
-0.0220
0.0000
-0.0001
2
fixed
t statistic
5.53
0.24
0.17
-1.31
8.57
-3.51
***
***
***
fixed
0.0817
t statistic
7.15
8.2257
-1.2086
-0.0198
0.0000
-0.0001
1.39
-0.20
-1.39
8.74
-3.54
***
***
***
0.0313
0.1086
-0.0186
-0.0034
0.0000
0.0000
fixed
t statistic
6.23
1.47
-0.25
-0.58
6.83
-2.96
***
***
***
fixed
0.0345
t statistic
8.77
3.1851
-0.3774
-0.0045
0.0000
0.0000
1.65
-0.19
-0.89
6.85
-3.09
***
-0.1119
1.0395
13.0829
t statistic
-0.59
0.37
4.59
-0.1238
t statistic
-0.78
2.48
0.46
1.17
-0.98
***
-0.3508
-1.58
192.4261
37.1312
0.2326
-0.0003
-0.62
-0.0004
*
***
***
0.47
0.48
0.42
0.42
0.21
0.13
adjusted R
601.08
592.83
1,225.98
1,227.04
-895.13
-925.62
Log likelihood
8.03 ***
8.19 ***
6.67 ***]
6.72 ***
3.16
2.15
F statistic
Number of individual
67
66
70
70
70
70
firms (cross sections)
Number of total
560
552
582
582
582
582
observations
Notes : ROE: return on equity. ROA: return on asset. ROS: return on sales. SCEO: Share of ownership of the CEO. d1899: Dummy variable of enactment of the Civil Code of
1899, takes 1 if year is 1899 or later and 0 otherwise. CNSL: Ownersship consolidation within the board = SCEO×SMIN, where SMIN: the ownership share of the board member
whose ownership was smallest within the board. ***, **, and * denote significance of 1, 5, and 10 percent levels respectively.
**
Table 7 The return on equity (ROE) and the financial leverage, from the second half of 1878 to the second half of 1910.
ROEi , t
ROEi , t
ROEi , t
ROEi , t
Dependent variables
73
74
71
72
estimation method
Panel least squares
Panel least squares
Panel least squares
Panel least squares
fixed
fixed
fixed
Cross section fixed effect fixed
Independent variables
t statistic
t statistic
t statistic
t statistic
-1.92
0.0469
39.50 *** 0.1448
30.77 ***
Constant
0.0639
14.69 *** -0.1969
LONi , t /(STKi , t +BBFi , t )
0.0097
1.24
-0.0137
-0.04
0.0013
0.74
-0.0106
-0.49
**
***
BNDi , t /(STKi , t +BBFi , t )
0.0012
0.31
2.7259
2.21
0.0041
4.10
-0.0023
-0.69
SALi , t
0.0000
7.31 *** 0.0000
-2.32 ** 0.0000
6.10 *** 0.0000
-2.83 ***
ΔGNPt
-0.0001
-3.70 *** 0.0003
1.36 *
0.0000
-1.75 *
0.0000
0.54
2
adjusted R
Log likelihood
F statistic
number of individual firms
(cross sections)
Restriction of observation
by ROE
ROEi , t
75
Panel least squares
fixed
t statistic
0.2266
18.02
-0.0168
-0.17
0.0870
1.74
0.0000
0.55
0.0000
1.24
**
*
ROEi , t
76
Panel least squares
fixed
t statistic
0.2806
2.69
-0.0169
-0.06
1.1332
1.39
0.0000
2.14
0.0000
0.08
0.63
0.54
0.49
0.22
0.24
0.33
1,184.50
10.18 ***
57.67
3.18
1,991.91
9.59
364.04
2.05
132.96
1.91
21.81
2.24
***
***
**
**
89
24
82
37
15
9
no restriction
ROE≤0%
0%<ROE≤10%
10%<ROE≤20%
20%<ROE≤30%
30%<ROE
**
**
*
Number of total
1,031
52
746
148
54
31
observations
Notes: ROE: return on equity. LON: bank borrowing that did not include outstanding bond. STK+BBF=own capital, wheree STK=paid in stock and BBF=Balance broght
forward (retained earnings). ***, **, and * denote significance of 1, 5, and 10 percent levels respectively.
Table 8 The return on asset (ROA) and the financial leverage, from the second half of 1878 to the second half of 1910.
ROAi , t
ROAi , t
ROAi , t
ROAi , t
Dependent variables
83
84
81
82
estimation method
Panel least squares
Panel least squares
Panel least squares
Panel least squares
fixed
fixed
fixed
Cross section fixed effect fixed
Independent variables
t statistic
t statistic
t statistic
t statistic
28.94
Constant
0.0334
19.94
-0.0707
-1.37
0.0241
24.63 *** 0.0754
LONi , t /(STKi , t +BBFi , t )
-0.0018
-0.61
-0.0279
-0.16
-0.0022
-1.48
-0.0292
-2.44
BNDi , t /(STKi , t +BBFi , t )
-0.0007
-0.45
1.0002
1.62
-0.0006
-0.72
-0.0008
-0.44
SALi , t
0.0000
1.79 *
0.0000
-1.65
0.0000
2.61 *** 0.0000
-8.35
ΔGNPt
0.0000
-2.61 *** 0.0001
1.26
0.0000
-0.69
0.0000
-1.04
adjusted R2
Log liklehood
F statistic
number of individual firms
(cross sections)
Restriction of observation
by ROE
***
**
***
ROAi , t
85
Panel least squares
fixed
t statistic
0.1146
12.88
-0.0134
-0.20
-0.0413
-1.17
0.0000
-3.11
0.0000
1.29
***
***
ROAi , t
86
Panel least squares
fixed
t statistic
0.1013
4.22
-0.0127
-0.18
0.5041
2.68
0.0000
-0.29
0.0001
0.75
0.32
-0.08
0.32
0.76
0.78
0.73
2,169.01
6.36
93.48
0.86
2,135.60
5.15
451.61
12.77
151.63
11.70
67.34
7.86
89
24
82
37
15
9
no restriction
ROE≤0%
0%<ROE≤10%
10%<ROE≤20%
20%<ROE≤30%
30%<ROE
***
***
***
***
Number of total
1,031
52
746
148
54
31
observations
Notes: ROA: return on asset. LON: outstanding bank borrowing. STK+BBF=own capital, where STK= paid in capital and BBF=Balance brought forward (retained
earnings). BND: outstanding corporate bond liability. SAL: sales. GNP: gross national product. ***, **, and * denote significance of 1, 5, and 10 percent levels
respectively.
**
***
Table 9 The return on sales (ROS) and the financial leverage, from the second half of 1878 to the second half of 1910.
ROSi , t
ROSi , t
ROSi , t
ROSi , t
Dependent variables
93
94
91
92
estimation method
Panel least squares
Panel least squares
Panel least squares
Panel least squares
fixed
fixed
fixed
Cross section fixed effect fixed
Independent variables
t statistic
t statistic
t statistic
t statistic
-0.31
0.3050
35.39 *** 0.3528
29.63
Constant
0.2522
6.21 *** -1.2166
LONi , t /(STKi , t +BBFi , t )
-4.43
0.1123
1.16
0.8750
0.06
0.0675
3.80 *** -0.3112
*
BNDi , t /(STKi , t +BBFi , t )
-0.0164
-0.34
0.6668
0.03
-0.0177
-1.83
-0.0289
-2.69
ΔGNPt
-0.0003
-1.00
-0.0041
-0.43
0.0000
-0.03
-0.0001
-1.78
adjusted R2
Log likelihood
F statistic
number of individual firms
(cross sections)
Restriction of observation
by ROE
0.10
***
***
***
*
ROSi , t
95
Panel least squares
fixed
t statistic
0.4509
15.35
-0.9404
-2.16
-0.0347
-0.15
0.0003
2.08
ROSi , t
96
Panel least squares
fixed
t statistic
0.6997
2.82
-0.2063
-0.13
2.2648
0.75
-0.0006
-0.26
-0.44
0.59
0.84
0.65
0.39
-137.61
0.40
283.71
13.76
187.88
20.35
50.67
6.75
-33.89
2.78
89
24
82
37
33
9
no restriction
ROE≤0%
0%<ROE≤10%
10%<ROE≤20%
20%<ROE≤30%
30%<ROE
-1,401.04
2.28 ***
***
***
***
Number of total
1,031
52
746
148
54
31
observations
Notes : ROS: return on sales. LON: outtanding bank borrowing. STK+BBF=own capital, where STK= paid in capital and BBF = balance broaght forward (retained
earnings). BND: Outstanding corporate bond liability. GNP: gross national product. ***, **, and * denote significance of 1, 5, and 10 percent levels respectively.
**
**
Table 10 Determinants of the changes in outstanding bond (BND), from the first half of 1887 to the second half of 1910.
Δ[BNDi , t /(STKi , t +BBFi , t )] Δ[BNDi , t /(STKi , t +BBFi , t )] Δ[BNDi , t /(STKi , t +BBFi , t )]
Dependent variables
101
102
103
estimation method
Panel least squares
Panel least squares
Panel least squares
fixed
fixed
fixed
Cross section fixed effect
Independent variables
t statistic
t statistic
t statistic
Constant
0.0011
0.56
0.0018
0.90
0.0013
0.63
ΔSCEOi , t
-0.1238
-1.88 *
-0.0475
-0.63
ΔSCEOi , t ×ΔROAi , t
14.2044
2.03 **
ΔCNSLi , t
-0.0302
-0.01
ΔCNSLi , t ×ΔROAi , t
ΔROAi , t
-0.1619
-2.59 **
ΔSALi , t
0.0000
-0.01
0.0000
1.23
0.0000
-0.06
ΔTKRt
0.0007
0.34
0.0011
0.54
0.0002
0.08
ΔGNPt
0.0000
0.97
0.0000
0.52
0.0000
0.95
adjusted R2
Log likelihood
F statistic
Number of individual firms
(cross sections)
Number of total observations
Δ[BNDi , t /(STKi , t +BBFi , t )]
104
Panel least squares
fixed
t statistic
0.0019
-0.5787
-45.9281
-0.1795
0.0000
0.0002
0.0000
-0.22
-0.56
-2.78
1.50
0.10
0.48
0.02
0.07
0.01
0.05
808.57
1.20
818.37
1.61
806.58
1.11
813.64
1.40
43
42
43
42
397
390
397
390
***
Notes : BND: Outstanding corporate bond liability. STK+BBF=own capital, where STK=paid in capital, BBF=balance brought forward. SCEO: Share of
ownership of the CEO. ROA: return on asset. CNSL: ownership consolidation withint the borad=SCEO×SNIN, where SMIN=share of the board member whose
ownership share was smallest within the board. TAR: market interest rate in Tokyo prefecture. GNP: gross national product. ***, **, and * denote significance
of 1, 5, and 10 percent levels respectively.
Table 11 Determinants of the changes in bank loans (LON), from the first half of 1887 to the second half of 1910.
Δ[LONi , t /(STKi , t +BBFi , t )] Δ[LONi , t /(STKi , t +BBFi , t )] Δ[LONi , t /(STKi , t +BBFi , t )]
Dependent variables
111
112
113
estimation method
Panel least squares
Panel least squares
Panel least squares
Cross section fixed effect
fixed
fixed
fixed
Independent variables
t statistic
t statistic
t statistic
Constant
-0.0185
-0.54
-0.0185
-0.53
-0.0190
-0.56
ΔSCEOi , t
-0.2910
-0.27
0.0342
0.03
ΔSCEOi , t ×ΔROAi , t
55.1128
0.44
ΔCNSLi , t
-1.6403
-0.32
ΔCNSLi , t ×ΔROAi , t
ΔROAi , t
-0.6215
-0.56
ΔSALi , t
0.0000
-0.05
0.0000
0.22
0.0000
-0.06
***
***
ΔTKRt
0.1136
3.17
0.1147
3.15
0.1129
3.17 ***
ΔGNPt
0.0004
1.60
0.0004
1.52
0.0004
1.61
adjusted R2
Log likelihood
F statistic
Number of individual firms (cross
sections)
Number of total observations
Δ[LONi , t /(STKi , t +BBFi , t )]
114
Panel least squares
fixed
t statistic
-0.0190
-0.55
-10.2520
-2.5717
-0.6353
0.0000
0.1130
0.0004
-0.22
0.00
-0.56
0.26
3.13
1.51
-0.04
-0.05
-0.04
-0.05
-307.90
0.64
-305.52
0.63
-307.88
0.64
-305.63
0.62
43
42
43
42
397
390
397
390
***
Notes : LON: outstanding banking borrowing. STK+BBF=own capital, where STK=paid in capital, BBF=balance brought forward. SCEO: ownership share of the CEO.
ROA: return on asset. CNSL: ownership consolidation withint the borad =SCEO×SMIN, where SMIN=ownership share of the board member whose ownership share was
the smallest within the board. SAL: sales: TKR: market interest rate in Tokyo prefecture. GNP: gross national proeuct. ***, **, and * denote significance of 1, 5, and 10
percent levels respectively.
⊲ We analyze effects of ownership structure on stockholder/manager conflicts.
⊲ Inefficient markets motivate non-owner managers to distort leverage to smoothen ROE.
⊲ We construct a dataset of all firms listed on the Tokyo market from 1878 to 1910.
⊲ President-ownership concentration led to better leverage and a higher ROA.
⊲ Ownership concentration offsets market inefficiency to seek long-term performance.