Did corporate boards matter during the Great Depression? Belgian
evidence
This version: July 2015
Marc Deloof
University of Antwerp and Antwerp Management School
E-mail: [email protected]
Veronique Vermoesen
University of Antwerp
E-mail: [email protected]
1
Abstract
We investigate how corporate governance as reflected by board characteristics, was related to
the value of listed Belgian firms in the 1928-1931 period, when investor protection was weak
and firms were hit by the biggest financial crisis of the 20th century. We find that firms
typically had a large board, with directors holding multiple directorships in banks and other
firms. Board size, busy boards and bankers on the board were positively associated with
corporate value, but the value of busy boards strongly decreased in the 1929-1931 period. We
also find a negative but less pronounced crisis effect for board size and bankers on the board.
These negative crisis effects seem to be largely driven by firm risk: riskier firms have busier,
larger boards and experience a larger drop in value.
JEL classification: G21, G30, G34
Keywords: Busy boards, director interlocks, board size, bank affiliation, corporate value, Great
Depression, Belgium
2
1. Introduction
Does corporate governance matter during a financial crisis? A number of studies have
found that corporate governance practices significantly affected firm value during the East
Asian Crisis in 1997-1998 (Baek et al., 2004; Johnson et al., 2000; Lemmon and Lins, 2003;
Mitton, 2002) and the recent 2008-2009 financial crisis (Erkens et al., 2012; Nguyen et al.,
2015). In this paper, we take a unique historical perspective by investigating the value of
corporate governance in Belgium during the Great Depression. This was a period when
investor protection was weak and firms were hit by the worst financial crisis of the twentieth
century. Our focus is on characteristics of the corporate board: we consider board busyness,
i.e. the extent to which directors on the board held multiple directorships, board size, i.e. the
number of directors on the board, and the presence of bankers on the board.
Belgium at the time of the Great Depression provides a very interesting environment to
study the value of corporate boards. Despite weak investor protection and mostly unregulated
corporate governance, Belgium had an active stock market. In 1929, the ratio of stock market
capitalization to the GDP in Belgium (1.31) was similar to that in the United Kingdom (1.38)
and Japan (1.20), and much higher than in Germany (0.35) or the United States (0.75). The
fraction of gross fixed capital formation raised via equity was much higher in Belgium (0.85)
than in the United Kingdom (0.35), the United States (0.38), or Japan (0.13) (Rajan and
Zingales, 2003). Moreover, Belgium combined an active stock market with a powerful banking
industry (e.g. Chlepner, 1943; Durviaux, 1947; van der Valk, 1932; Vanthemsche, 1991).
Rajan and Zingales (2003) find that the ratio of commercial and savings deposits to the gross
domestic product (GDP) in 1929 was higher in Belgium (0.48) than in the United States (0.33),
Germany (0.27), or Japan (0.22). The banking industry was dominated by a limited number of
banks with close ties to the Belgian industry via director interlocks and equity stakes (e.g.
Chlepner, 1943).
3
***Figure 1 about here***
The Great Depression had a strong negative effect on the value of Belgian firms. Figure 1
shows the cumulative stock returns of all firms listed on the Brussels Stock Exchange in the
period 1927–19311. We set the index at 100 on January 1st 1927. During the entire decade
after world war I, the Stock Exchange was almost always in a state of excitement (e.g.
Chlepner, 1943). The index rose from 100 to 215 from January 1927 to January 1929, the
culminating point of the rising market. This is depicted by the vertical line on the chart. After
January 1929, stock prices started decreasing. From January 1929 to December 1931, the
index fell from 215 to 77.
Corporate governance is likely to be especially relevant during a financial crisis, for at
least two reasons (Baek et al., 2004; Johnson et al., 2000; Mitton, 2002). First, a crisis can lead
to greater expropriation by management as their expected return on investments falls. Second,
corporate governance attracts more attention from investors and financial institutions during
crisis periods resulting in a shift of financing from firms that appear to have weak corporate
governance to firms with better corporate governance. Overall, firms with weaker corporate
governance are expected to lose relatively more value during the crisis. In this paper we focus
on corporate boards, because the board is often considered as the most important internal
corporate governance mechanism that monitors and advises management in the interests of the
shareholders (e.g. Francis et al., 2012), and its role should be especially important during crisis
periods (e.g. Mace, 1986; Nguyen et al., 2015).
An additional advantage of focusing on a crisis such as the Great Depression is that we can
avoid the endogeneity problem arising from endogenously composed boards (e.g. Boone et al.,
2007; Linck et al., 2008). Prior papers have tried to address this problem in different ways.
Wintoki at al. (2012) and Forman-Peck and Hannah (2013), for example, use instruments.
1
Indices are Laspeyres market cap weighted return indices calculated by linking monthly returns through a chain
index. The weight of each firm’s return is given by its relative market capitalization.
4
Others, like Black et al. (2006), Black and Kim (2012) and Dahya and Mc Connell (2007) use
a legal shock to governance. Still others use a crisis period: Francis et al. (2012) study the
recent financial crisis for 876 nonfinancial U.S. firms. Following Graham et al. (2011), we use
the Great Depression which was a relatively exogenous shock that negatively affected firm
value. The advantage of this crisis period is the sudden change in economic conditions that
could not have been anticipated by firms by altering the composition or the size of their board.
By considering board characteristics in 1928 prior to the start of the Great Depression, we
ascertain that board characteristics are not driven by firm performance during the Great
Depression.
Today, it is often argued that boards should be independent and small and that directors
should limit their number of directorships. These assumptions are formally written down in
corporate governance codes such as the Sarbanes-Oxley Act in the U.S or the Principles of
Corporate Governance of the OECD. However, there is no conclusive evidence on the effect of
these board characteristics on firm value. Busy directors may enhance the quality of the
board’s advice because they may be a source of valuable experience and they may enlarge a
firm’s network (e.g. Ferris et al., 2003; Miwa and Ramseyer, 2000). However, board busyness
may also hinder the advising and monitoring quality because of time limitations if directors
become overcommitted resulting in lower firm value (e.g. Fich and Shivdasani, 2006; Loderer
and Peyer, 2002). As board size increases, communication and coordination problems (e.g.
Lipton and Lorsch, 1992) can result in less effective advising and monitoring and lower firm
value (e.g. Eisenberg et al., 1998; Yermack, 1996). However, a large board can have a positive
effect on the board’s ability to advise efficiently because each director adds experience and
knowledge to the board and reduces environmental uncertainty (e.g. Dalton et al., 1999). This
may be beneficial for firm value especially when advising needs are large (e.g. Coles et al.,
2008; Faleye et al., 2011). Bankers on the board can provide important industry-specific
5
financial expertise obtained through lending to firms in that industry (e.g. Kroszner and
Strahan, 2001) and they can mitigate information asymmetries between the bank and its client
firm by close and successive monitoring so that the availability of credit is increased and its
cost reduced (e.g. Hoshi et al., 1990; Petersen and Rajan, 1994; Weinstein and Yafeh, 1998).
However, a banker on the board can also abuse its private information to extract rents from the
firm (e.g. Agarwal and Elston, 2001; Rajan, 1992; Weinstein and Yafeh, 1998).
Using a sample of 150 large Belgian firms listed on the Brussels Stock Exchanged in 19281931, we find that Belgian firms in 1928 typically had a large board, with directors holding
multiple directorships in banks and other firms. Board busyness, board size, and bankers on the
board in 1928 were positively associated with corporate value, but the value of busy boards
significantly decreased during the 1929-1931 period. We also find a negative but less
pronounced crisis effect for board size and bankers on the board. If we take into account the
impact of firm risk on the change in value during the crisis, the negative effect becomes less
significant, both economically and statistically. This indicates that the negative relationship
between board characteristics and firm value during the crisis is partly driven by risk. Riskier
firms have larger boards, busier boards and more bank directors on their board, and they
experience a stronger drop in value during the 1929-1931 period.
Our paper provides a unique contribution to the literature by investigating the value of
corporate boards during the most important crisis of the 20th century. It not only contributes to
the body of literature that investigates the value of corporate governance mechanisms during a
crisis (Baek et al., 2004; Barton and Waymire, 2004; Graham et al., 2011; Johnson et al., 2000;
Lemmon and Lins, 2003; Mitton, 2002; Nguyen et al., 2015), but also to a small but growing
literature that investigates the value of corporate boards from a historical perspective.
Foreman-Peck and Hannah (2013) find a positive effect between board size and firm
performance in 1911 UK which they attribute to the directors’ information function and
6
expertise. Van Overfelt et al. (2009) provide evidence that bank directors on the board were
beneficial for firm value in Belgium in the period 1905-1909. Braggion and Moore (2013) find
that politicians on the board increased the value of firms quoted on the London Stock
Exchange between 1895 and 1904 and Graham et al. (2011) find, for a sample of 446 U.S.
firms between 1926 and 1938, that board size is positively related to firm value for complex
firms and negatively related to firm value for simple firms.
The remainder of the paper is organized as follows. Section 2 discusses the historical
background of interwar Belgium with a description of its economy, corporate boards and
banking industry. Section 3 reviews prior literature on seat accumulation by directors, the
value of board size and bank affiliation for firm performance. The construction of the sample,
the empirical model and the variables are discussed in section 4 and section 5 reports empirical
findings. Our conclusions are presented in section 6.
2. Interwar Belgium
2.1. Economy
In Belgium, as in other European countries, the imperatives of war finance during WWI
created unbalanced budgets and rocketing inflation rates. At the start of the war, when the gold
standard was abandoned, the circulation of bank notes was about one billion francs. At the
Armistice in November 1918, Belgian note issue was about 2.7 billion (Chlepner, 1943). In
October 1926, Belgium rejoined the gold standard in a modified form: the gold exchange
standard2. The Belgian franc was devaluated to one-seventh of its prewar value which resulted
in a considerable undervaluation compared to the British pound and American dollar (Aldcroft,
1997). This undervaluation stimulated Belgian exports and strongly boosted the Belgian
economy.
2
Monetary authorities tied its currency to gold indirectly: they maintained a fixed exchange rate with a foreign
currency that was convertible into gold at a stable rate of exchange. Those currencies were mostly the U.S. dollar
or the British sterling.
7
By 1929, the international financial and economic situation was deteriorating. This was
mainly due to the rigid international monetary system, with unrealistic parities, budget deficits,
and excess capacity in heavy industry and agriculture (e.g. Bernanke and James, 1990). In
Belgium, industry’s rigidity and low productivity levels led to an economic downturn by the
beginning of 1929. The market crash in the United States in October 1929 seriously aggravated
the emerging international crisis. The British abandonment of the gold exchange standard in
September 1931 was especially harmful for Belgium, since Great Britain was one of the main
customers and principal competitor of Belgian industries in foreign markets (Chlepner, 1943).
After the British abandonment, the Belgian government introduced a deflationary policy to
counterbalance the cheap British pound, but this policy failed because fiscal revenues were
disappointing, and government expenditures rose as unemployment compensation increased
(Buyst, 2005; Chlepner, 1943). In March 1935 the government devalued the Belgian franc by
28% after which the Belgian economy started to recover (Baudhuin, 1944; Chlepner, 1943).
2.2. Corporate boards
The board of directors (“conseil général”) of Belgian firms had a dual structure with
executive directors (“administrateurs”) and supervising directors (“commissaries”). Firms were
not legally obliged to have a formal board of directors with both executive and supervising
directors. Whether or not to have one and other specifications were written down in the articles
of incorporation (Gilis, 1929; Wauwermans, 1927).
The executive directors acted on behalf of and for the account of the company and were
appointed by the articles of incorporation or by the general meeting of shareholders. Their
responsibilities were confined by the company’s articles of incorporation. A firm needed to
have at least three executive directors and their mandate could not exceed six years. However,
they were eligible for re-election and their mandate was revocable at all times (Wauwermans,
1933).
8
Supervising directors were charged with the supervision of the executive directors and had
to approve the company’s annual accounts. They were also appointed by the general meeting
of shareholders. The number of supervising directors was set by the general meeting of
shareholders with a minimum of one and their mandate could not exceed six years but they
were eligible for re-election (Wauwermans, 1933).
Daily management and representation of the firm was delegated to a management
committee that was composed of managers with or without a board seat (“administrateur
délégué” and “directeur” respectively). Only the names of the managers with a board seat had
to be published (e.g. Gilis, 1929; Wauwermans, 1933).
Conflicts of interests between shareholders and executive directors had to be reported and
executive directors with conflicting interest were not allowed to attend board meetings (e.g.
Gilis, 1929; Wauwermans, 1933). Furthermore, names of all directors, both executive and
supervising, had to be made public and all directors were obliged to deposit shares of the firm
(“cautionnement”) to make sure that directors would act according to the articles of
incorporation and to the law. The number of shares to deposit was set by the articles of
incorporation or in absence by the general meeting. However, directors were not obliged to
own those shares themselves (e.g. Wauwermans, 1933).
Although investor protection in Belgium significantly improved since the law of May 25th
1913 (following other European countries such as Britain, Germany and France), corporate
governance was still in its infancy. For example, rules to impose board independence, which is
today generally considered as a requirement for good corporate governance (e.g. SarbanesOxley Act of 2002 in the US and the Principles of Corporate Governance of 1999/2004 of the
OECD), were lacking: the law allowed supervising directors to be employees of the firm and
they were also allowed to be family members or other relatives of the executive directors of
the firm (Centre d'étude des Sociétés, 1956; Wauwermans, 1933). As several studies have
9
shown that board independence is positively related to the effectiveness of monitoring because
of the independent directors’ perceived objectivity (e.g. Adams and Ferreira, 2007; Faleye et
al., 2011; Hwang and Kim, 2009), the lack of rules to enforce board independence may have
hampered monitoring efficiency in the 1920s. This is confirmed by Gilis (1929), a
contemporary accountant, who points to the fact that, de facto, the rules for supervising
directors were not complied with. He describes that supervising directors mostly limited their
monitoring activities to some random checks of the annual accounts. Important steps to
improve corporate governance would only be taken as a response to the crisis in the banking
industry in 1934-1935.
2.3. Banking industry
After WWI, the Belgian industry was in great need of capital to rebuild infrastructure and
resume activities (e.g. Durviaux, 1947). Belgian banks, which had played a crucial role in the
economic development of Belgium before WWI (e.g. Van Nieuwerburgh et al., 2006), were
able to match the enormous demand for capital with an increasing supply, since more people
could save and savings were increasingly invested in stocks (e.g. Chlepner, 1943). Banks were
also able to enlarge their clientele thanks to the development of an extensive network of
branches all over the country (Chlepner, 1943; Durviaux, 1947). During the 1920s, the Société
Générale reinforced its prewar supremacy, while the Banque de Bruxelles emerged as its main
competitor, although it remained significantly smaller. After 1927 the number of acquired or
liquidated banks began to exceed the number of newly created banks, manifesting a strong
trend toward concentration (Chlepner, 1943; Durviaux, 1947; van der Valk, 1932).
By the end of the 1920s, most Belgian banks had a portfolio of equity stakes in industrial
firms which allowed them to control a significant part of Belgian industry (e.g. van der Valk,
1932). Their controlling power was reinforced by the use of holding companies and the
issuance of multiple voting shares. Lamal (1930) calculated that 47% of the votes of limited
10
liability companies were controlled by groups/individuals owning merely 3.5% of the quoted
equity value. A large number of these shares were held by the large banks. Moreover, it was
common for bank directors to have several board seats in other firms in different industries
(e.g. Tilman, 2006; Van Overfelt et al., 2009).
Up until 1932 the Belgian banking industry did not experience severe adverse shocks. For
several reasons it could endure the strain of the first phase of depression without major
problems. First, during the period of high inflation preceding the stabilization of the Belgian
franc in 1926, banks were forced to increase their equity, while many of their assets, mainly
equity stakes, did not lose value. This created hidden reserves (Chlepner, 1930; Durviaux,
1947). Second, between 1927 and 1929 banks used favorable capital market conditions to issue
new equity (Chlepner, 1943). Third, after 1929, when many firms became unable to repay their
bank debts because of the deflationary policy that lowered the value of their assets relative to
nominal bank debts, Belgian banks could convert this bad debt into equity stakes rather than to
write off loans (Durviaux, 1947). Only from 1932 onward did Belgian banks become
financially distressed. In 1934 two mid-sized banks, the Bank van den Arbeid and the
Algemeene Bankvereeniging, had to close down, and in 1935 the Belgian government created
a public institution to rescue the banking industry.
3. Corporate boards and corporate value
3.1. Busy boards
It is not clear whether busy directors, i.e. directors holding multiple directorships, are good
or bad for a firm. It could be argued that busy directors enhance firm performance. According
to the resource dependence theory (e.g. Boyd, 1990; Pfeffer, 1972; Pfeffer and Salancik,
1978), directors holding directorships at other firms and financial institutions provide
connections with the outside world which allow the firm to cope with uncertainty and to
11
facilitate access to resources. Busy directors have better access to information about market
conditions, competitors, customers, suppliers etc. and they help the firm to obtain resources at
lower cost (Coles et al., 2012). Boards with ties to strategically related organizations are able
to provide better advice and counsel: directors use their business connections to acquire
information that is trustable, timely and up-to-date, or that is otherwise unavailable (Carpenter
and Westphal, 2001). Having multiple directorships might also proxy for reputational capital
(e.g. Kaplan and Reishus, 1990) which could be a signal of director quality: higher quality
directors are more frequently asked to serve on other boards (Coles et al., 2012). Consistent
with these arguments, Ferris et al. (2003) find that the prior performance of firms positively
affects the number of directorships that their directors holds. They also find a significant
positive announcement return when the appointment of a busy director to the board is
announced. For a sample of S&P 1500 firms during the period 1996-2007, Coles et al. (2012)
find that firm value increases in advising quality, measured by the connections of outside
directors, as firm complexity increases. Busy boards also contribute positively to the value of
IPO firms, which are likely to rely heavily on their directors for advising (Field et al., 2013).
However, other empirical studies suggest that directors with multiple board seats can become
overcommitted, rendering them unable to effectively perform their monitoring duties (e.g. Fich
and Shivdasani, 2006; Loderer and Peyer, 2002).
It is a priori not clear how the Great Depression affected the value of busy boards. If board
busyness reflected weak monitoring by the board of managers and/or controlling shareholders,
it may have been negatively related to corporate value during the crisis, as a crisis can lead to a
greater expropriation by managers and controlling shareholders, and make investors more
aware of weak corporate governance (Baek et al., 2004; Johnson et al., 2000; Mitton, 2002).
This will especially be the case in an environment where investors are weakly protected by
external corporate governance mechanisms. On the other hand, it could be argued that the
12
advising role of boards becomes especially important during a crisis (Francis et al., 2012). If a
busy board reflects the quality of its directors who can make use of their network, knowledge,
experience and information, we expect a positive impact of board busyness on corporate value
during the crisis.
3.2. Board size
How does board size affect corporate value? Smaller boards can monitor management
more effectively and therefore reduce agency costs (Eisenberg et al., 1998; Jensen, 1993;
Yermack, 1996). During a crisis, smaller boards may be more able to reach consensus and
initiate strategic actions, while larger boards may more easily develop factions and coalitions
that lead to group conflict, which may hinder the process of reaching consensus and taking
action (e.g. Goodstein et al., 1994). A tendency to react slowly or indecisively in a crisis might
jeopardize the very existence of a firm (e.g. Daily and Dalton, 1994).
On the other hand, the resource dependence theory suggests that larger boards are
associated with a higher corporate value. A larger board may establish a larger network and
may therefore be better able to cope with various external uncertainties and secure external
resources. Furthermore, an outside director may add additional knowledge and experience and
may therefore increase a board’s advising quality. Some studies (e.g. Dalton et al., 1999)
indeed find a positive relation between board size and firm performance. Coles et al. (2008)
and Graham et al. (2011) find that large boards have a positive influence on the corporate
value of complex firms which have typically more extensive advising needs. As argued before,
the advising role of corporate boards might be especially valuable in times of crisis.
So as with board busyness, it is a priori not clear how board size affected the value of
Belgian firms during the Great Depression.
13
3.3. Bankers on the board
Prior to WWII, Belgian firms often had bankers on their board (e.g. Van Overfelt et al.,
2009). Having a banker on the board can have a positive impact on firm value because it
increases credit availability. It can mitigate information asymmetries between the lending bank
and the firm (e.g. Hoshi et al., 1990; Petersen and Rajan, 1994; Weinstein and Yafeh, 1998)
and it can reduce conflicts of interest between shareholders and debt holders which arise if
shareholders prefer riskier projects than lenders (e.g. Diamond, 1984). Furthermore, bankers
on the board can provide valuable financial advice to management (e.g. Fohlin, 1999).
Consistent with this argument, Van Overfelt et al. (2009) find a significant positive effect of
bankers on the board on the value of 129 Belgian firms listed on the Brussels Stock Exchange
in the period 1905-1909.
The impact of bankers on the board should be especially pronounced during a crisis
period. In the absence of a banker on the board, the free rider problem can reduce the incentive
of creditors to grant financial relief or extend credit to financially distressed firms (e.g. Elsas
and Krahnen, 1998 in Germany; Ferri et al., 2001 for the financial crisis of 1997-1998 in
Korea; Hoshi et al., 1990 during the 1980s in Japan). Moreover, bankers can add financial
expertise to the board which is especially valuable during a crisis period (e.g. Francis et al.,
2012).
However, bankers on the board could also have a negative effect on firm value. It allows
the bank to create an information monopoly so that banks can extract rents from the borrowing
firm (e.g. Agarwal and Elston, 2001; Rajan, 1992; Weinstein and Yafeh, 1998). Switching
banks can become too costly, such that banks hold up related firms. Bank affiliations can be
detrimental to firm performance during an economic downturn as a close relationship with a
firm provides a bank with an unfair information advantage compared to the other creditors,
such that in times of financial distress the bank can extract rents, take actions to insulate itself
14
from trouble, and shift the burden to other creditors (e.g. Kroszner and Strahan, 2001). This
effect could also be more pronounced during a crisis.
4. Research design
4.1. Sample
Our study is based on a sample of non-financial Belgian firms listed on the Brussels stock
exchange during the period 1928–1931. Except for the stock data, we hand-collected the data
used for this research from the Recueil Financier, a financial annual containing firm-specific
information on firms listed on the Brussels Stock Exchange, including accounting data and
names and positions of the members of the boards of directors. Stock data were taken from a
database constructed by the StudieCentrum voor Onderneming en Beurs (SCOB) at the
University of Antwerp. This database contains information on all stocks listed on the Brussels
Stock Exchange. Since most of the information needed for this study had to be hand-collected,
we initially restricted our sample to the 220 largest firms based on market capitalization in
1928. For 185 firms we were able to find all the necessary information on the board of
directors and financial statements in the Recueil Financier. To be included in the sample, a
firm also had to be listed on the Brussels Stock Exchange in 1927 and 1928, to allow us to
calculate beta coefficients (discussed further below). After excluding 20 financial firms, 11
firms with missing accounting data, one firm with insufficient stock return data, and three
firms that had listed bonds but no listed stocks, our final sample consisted of 150 firms, for
which we have an unbalanced panel of 574 firm–year observations for the four-year period
1928 to 1931.
15
4.2. Empirical model
We estimate a random effects model with the following specification3:
Ln(MTB)i,t = a + b (Crisis Year Dummies) + c Board measurei + d Board measurei x
(Crisis Year Dummies) + e (Firm Characteristicsi,t) + f (Industry Dummiesi)
The dependent variable is the natural log of the market-to-book ratio (MTB). Using MTB
as a measure of firm value is consistent with related studies such as, for example, Fich and
Shivdasani (2006), Yermack (1996), Coles et al. (2008), Field et al. (2013) and Ferris et al.
(2003). This ratio is calculated as the market value of equity divided by the book value of
equity at the end of the fiscal year. The advantage of this measure is that when a crisis occurs,
it will immediately incorporate the effect of the crisis on profit expectations. Accounting
measures of profitability on the other hand are by nature backward looking, and accounting
profits in Belgium were highly susceptible to manipulation (Van Overfelt et al., 2010).
The main independent variables in our models are (1) three dummy variables equal to one
for the crisis years 1929, 1930 and 1931 to take into account the effect of the crisis, (2) board
variables measuring board busyness, board size and bankers on the board, and (3) the
interaction between the crisis year dummies with our board measures, to investigate how the
relation between corporate value as measured by MTB and board characteristics changed
during the crisis years. Additionally we control for firm characteristics and industry effects.
Board measures, firm characteristics and industry effects are further discussed in the next three
subsections.
3
We are unable to use a fixed effects model as the board characteristic variables are defined only once so they
would drop out of the regressions. However, The Breusch-Pagan Lagrange multiplier test confirmed presence of
random effects. Moreover, the results of the random effect model are qualitatively the same as those of the OLS
regressions. Therefore, we use a random effects model for all regressions in this paper.
16
4.3. Board measures
We consider several measures of board busyness, board size and bankers on the board.
Board busyness is measured in three different ways. To make a maximum use of the
information we have, these measures are based on the larger initial sample of non-financial
firms for which we have sufficient information on the board of directors. First, following Ferris
et al. (2003), we use directorships per director which is the total number of directorships held
across all the directors of a board divided by board size. If there is a dispersion in the number
of board seats held by the different directors of the same firm, the average number of
directorships per director may not be a perfect measure. Therefore, following Fich and
Shivdasani (2006), our second busyness measure is the percentage of busy directors on a
firm’s board. A director is busy if he/she holds three or more directorships. We use three
directorships as the cut-off rate because the mean and median number of directorships per
board in the sample is close to three, resulting in a roughly even split between busy and nonbusy directors. It is also in line with prior work by Fich and Shivdasani (2006) and Ferris et al.
(2003). Third, we use ln(no. of external directorships), which is the natural logarithm of the
total number of external directorships held by all the directors of the board plus one. Board
size is the total number of executive and supervising directors.
To measure bank affiliation, we use interlocking directorships between firms and banks
(e.g. Becht and Ramírez, 2003; Fohlin, 1997; Fohlin, 1998; Van Overfelt et al., 2009). A firm
is assumed to be bank affiliated if it has a bank director on its board. We identify interlocking
directorships by the names of the directors. If a director holds directorships with more than one
bank, we attribute the largest of these banks (in terms of its industrial portfolio) to this director.
We consider the directors of 41 banks based on a list of banks drawn up by Durviaux (1947)
for 1930. We use different affiliation measures. First, bank affiliated (0,1) is a dummy equal to
one if the firm has at least one bank director on its board, and zero otherwise. Second, firms in
17
our sample were often affiliated with more than one bank. Multiple bank affiliations may
reduce the incentive of individual banks to monitor a firm or to grant credit to it because the
costs are borne by just one bank, while the benefits are enjoyed by the other banks as well (e.g.
Foglia et al., 1998). This may have a negative effect on firm value. Therefore, we include the
no. of affiliated banks, which is the number of banks which have at least one director on a
firm’s board. Third, the percentage of bank directors is the number of bank directors on the
board of a firm divided by the total number of directors of the firm.
Although we investigate the effect of corporate board structure on firm value during an
external shock, firms may still react fast to the deteriorating economic conditions by adapting
their board (e.g. Gilson, 1990). They could, for example, appoint new busy directors so that the
firm can take advantage of valuable reputation and experience to better cope with the crisis.
The firm could also increase the number of directors to get more advice on how to tackle the
crisis. In principle, this endogeneity issue is controlled for by defining board structure
variables at the end of 1928, before the start of the crisis. To further rule out the possibility that
board size is dependent on firm performance, we take a closer look at the evolution of board
size and board busyness during our sample period. If a firm’s performance would affect the
number of directors in our sample, we should observe important changes in the number of
directors over time. Therefore, we compare the names of directors in 1928 with the names in
1933 for 128 surviving firms. We find a mean change in the board size of 0.54 and a median
change of 0. Furthermore, 60% of the boards have no change or a change of only one director
over this six year period. The correlation between performance and the change in the board of
directors is very small (-0.050) and statistically insignificant. We also find that directorships
per director decrease from 2.95 in 1928 to 2.73 in 1933. This change is not statistically
significant (p-value is 0.164). The percentage of busy directors decreases from 0.38 to 0.30.
This decrease is statistically significant (p-value is 0.016) and is inconsistent with the idea that
18
badly performing firms would attract busy directors or increase the number of external
directorships to better cope with the crisis.
*** Table 1 about here***
Panel A of Table 1 reports descriptive statistics for the board measures. The average board
size is 15.38. The average number of executive directors is 10.87 and the average number of
supervising directors is 4.51. By way of comparison, Foreman-Peck and Hannah (2013) found
an average board size in the UK in 1911 of 9.6, Frydman and Hilt (2010) found an average
board size of 12.43 for listed U.S. railroad companies for the period 1905-1923 and Graham et
al. (2011) found an average board size of 10.97 for U.S. firms during the Great Depression.
Table 1 also shows that boards tend to be busy: the average number of directorships per
director is 2.95. This is in line with Fich and Shivdasani (2006) and Ferris et al. (2003) for
contemporary US data, but the frequency of busy boards is high (38%) compared to other
studies (e.g. Ferris et al., 2003). We find strong board connections between the firms in our
sample and the banking industry: 87% of the firms in our sample have a bank director on their
board and the median firm is affiliated with two banks. This is remarkable, since Van Overfelt
et al. (2009) found that in 1905, only 36% of Belgian listed firms in capital-intensive industries
had a bank director on their board.
4.4. Firm characteristics
As potential determinants of MTB we consider the following firm characteristics: firm size,
age, leverage, intangible assets, corporate diversification and beta (e.g. Coles et al., 2008; Fich
and Shivdasani, 2006; Yermack, 1996). Firm size is measured by the natural logarithm of total
assets at the beginning of the fiscal year. Age is the difference between the beginning of the
year and the year the firm was first listed4. As a measure of leverage, we use the book value of
4
If a firm had different kinds of stock outstanding with a different period of listing, we use the age of the oldest
stock.
19
total debt divided by the book value of total assets measured at the beginning of the fiscal year.
Investment opportunities are often measured by scaling research and development or capital
expenditures by sales. However, we do not have this information for our sample. Therefore,
we include intangible assets as a proxy for investment opportunities. This variable is defined
as the book value of intangible assets divided by book value of total assets measured at the
beginning of the fiscal year. We account for corporate diversification by including the number
of business segments (e.g. Faleye et al., 2011; Francis et al., 2012; Yermack, 1996). Our risk
measure is beta, which is computed by regressing a firm’s monthly stock return in the precrisis period on the corresponding market return obtained from SCOB. We require at least 24
months of return prior to January 1929 to compute beta and we use a maximum of 63 months’
worth of data. Risky firms generally have a high default risk and are therefore more vulnerable
to external shocks (e.g. Baek et al., 2004).
***Figure 2 about here***
Descriptive statistics on the firm characteristics are reported in panel B of Table 1. The
firms in our sample tend to be mature firms: they have on average been listed for 26 years
(median is 20 years). The average (median) debt ratio is 0.37 (0.34), which is substantially
higher than the debt ratio’s found by Deloof and Van Overfelt (2008) for capital intensive
Belgian firms in 1905-1909. Figure 2 shows the mean and median values of the MTB in each
of the four years. Not surprisingly, the effect of the crisis from 1928 to 1931 is very
pronounced. The mean (median) MTB decreases from 7.50 (2.98) in 1928 to 2.29 (0.84) in
1931.
***Table 2 about here***
20
4.5. Industry effects
To control for industry effects, we include seven industry dummies, a dummy for firms
with their main activities abroad and a dummy for firms with mainly colonial activities. The
industry classification is taken from the SCOB database and is based on the most important
industry in which the firm is active. The industries were double-checked with a description of
the firm’s activities in the Recueil Financier.
Table 2 describes board characteristics and connections between firms across industries.
While firms in all industries tend to have large boards, busy board and bank directors on the
board in all industries, these characteristics are especially pronounced for firms in electricity,
colonial firms and firms operating abroad. For example, 55% of the directors of colonial firms
are “busy”, i.e. they hold three or more directorships. All these firms are affiliated with at least
one bank, and on average they are connected with three banks.
The last two columns show the number of firms connected with at least one firm operating
in the same industry and the percentage of external directorships held in firms operating in the
same industry. These figures show many interlocking directorships within an industry. For
example, in the mining industry, 20 out of 23 firms in the sample are connected to other firms
in the mining industry via interlocking directors. An explanation for the magnitude of these
connections might be cartel formation. Heyman (1928), a former minister of industry,
mentions the creation of international cartels. He provides a list of cartels known at the time
with the countries involved. In the Recueil Financier, we found that ‘Société anonyme
Métallurgique de Prayon’, a metallurgic firm from our sample, mentions restrictions of
production because of international agreements (“entente international”). At the same time,
firms are also strongly connected with firms in other industries. For example, of all the
external directorships held by directors in the mining industry, only 33% is within the mining
industry, other board seats are in other industries. Figure 3 which shows all connections
21
between our sample firms via interlocking directorships illustrates that the firms in our sample
are heavily connected. Only four firms have no interlocking directors with other firms from
our sample.
***Figure 3 about here***
***Table 3 about here***
5. Results
5.1. Correlations
Table 3 shows Pearson correlation coefficients for the variables used in this study. Not
surprisingly, firms with larger boards have more external directorships. Larger boards also tend
to have more busy directors and bank directors. There is also a substantial positive correlation
between the board busyness measures and bank affiliation measures. While larger firms tend to
have larger boards (r = 0.33), older firms tend to have smaller boards (r = -0.32). It is also
interesting to note that each of the three busyness measures are strongly positively correlated
with beta, our measure of firm risk. This indicates that risky firms use board connections to
cope with uncertainty, as predicted by the resource dependency theory.
*** Table 4 about here***
5.2. Busy boards
Table 4 shows the regression results for our busy board measures. The natural log of the
MTB is the dependent variable in all regressions. As a benchmark, in regression (1) we first
look at the effect of the crisis on MTB without taking into account board characteristics. As
expected, the results show a significant decrease of MTB in the crisis years 1929, 1930 and
1931. The negative effect of size and the positive effect of debt ratio on MTB is inconsistent
with findings of Baek et al. (2004) for Korean firms in the crisis of 1997-1998, who argue that
22
larger firms are less vulnerable to external shocks and highly leveraged firms have more
difficulty obtaining external financing during a crisis. However, they confirm findings of Van
Overfelt et al. (2009) for Belgian firms in the period 1905–1909 and Graham et al. (2011) for
US firms in the Depression years. Foreman-Peck and Hannah (2013) also found a negative
effect of firm size on share premia for British firms in 1911. The negative effect of corporate
diversification is in line with previous research (e.g. Baek et al., 2004; Fich and Shivdasani,
2006) but it is not statistically significant. The negative coefficient on intangible assets, our
measure for investment opportunities, is inconsistent with expectations (e.g. Fich and
Shivdasani, 2006) but consistent with Coles et al. (2012). The sign of the coefficient of beta is
statistically insignificant, which suggests that systematic risk was incorporated in stock prices.
In regressions (2), (3) and (4), we include the busy measures directorships per director,
percentage of busy directors and ln(no. of external directorships). We find that MTB is
positively related to all three measures before the crisis, and that this relationship is statistically
and economically significant. For example, one extra board seat is associated with a 20%
increase in the MTB (=100*(exp(0.178)-1)). However, the market value associated with busy
directors significantly decreased in the crisis years 1929, 1930 and 1931. This result is found
for each crisis year and for each of the three busy measures.
The positive association between board busyness and market value before the crisis can
be interpreted in two ways: either busy directors increased firm value because external
directorships are valuable for the firm, or the directors of firms with a higher market value are
more likely to get external directorships because market value signals director quality. As
board composition may be endogenously chosen, we cannot prove the direction of causality
before the crisis. However, the crisis in 1929-1931 was an unexpected negative shock and we
measure board busyness before the crisis. Therefore the observed negative association between
board busyness and market value during the crisis in 1929-1931 suggests that board busyness
23
affected firm value and not vice versa. If firm value is negatively affected by busy directors’
over commitment, we expect the decline in firm value during a crisis to be stronger for firms
with a busy board, which is exactly what we find for 1929-1931. On the other hand, if board
busyness reflects the quality of the directors, we would expect the value of busy board to go up
during a crisis when the knowledge, experience and information of connected directors
arguably will become more important. But this is not what we find.
*** Table 5 about here***
5.3. Board size
Table 5 reports regression results for board size. Board size was positively related to MTB
before the crisis: the board size coefficient is positive and highly statistically significant in
regression (5). Economically, significance is limited: a one percent change in the board size is
associated with a 0.56 percent change in MTB.
The positive relationship weakens in 1929, but the decrease is less pronounced as in the
case of board busyness. The effect in 1930 and 1931 is also negative but the decrease is not
statistically significant. Since boards were composed of both executive and supervising
directors, our results may be driven by either executive or supervising directors. To test this,
we differentiate between the number of executive and the number of supervising directors in
regressions (6) and (7). We again find a significant positive relationship before the crisis, but
the negative crisis effect is now statistically insignificant in all crisis years. In sum, our results
indicate a positive relationship between board size and firm value before the crisis that
becomes somewhat weaker during the crisis, but the negative influence of board size during
the crisis seems to be less pronounced than in the case of board busyness.
*** Table 6 about here***
24
5.4. Bank affiliation
In table 6, we investigate the relation between corporate value and bank affiliation using
three measures of bank affiliation: bank affiliated (0,1) (regression (8)), the no. of affiliated
banks (regression (9)) and the percentage of bank directors on the board (regression (10)).
Bank affiliation and MTB are not statistically significantly related before the crisis (regression
(8)). However, the coefficient of the interaction between bank affiliation and the 1929 dummy
is significant and negative, which suggests that the decline in corporate value in 1929 was
more pronounced for firms which had a bank director on their board. We also find negative
coefficients for the interactions with the 1930 and 1931 dummies, but these are not statistically
significant. In regression (9), we find a positive association between the number of affiliated
banks and MTB which is statistically and economically significant: being affiliated with one
more bank is associated with a 10% increase in MTB (=100*(exp(0.097)-1)). However, the
coefficients of the interaction between bank affiliation and the 1929 and 1930 dummies are
significant and negative which means that being affiliated with more banks puts extra pressure
on firm value in 1929 and 1930. The coefficient of the interaction variable with the 1931
dummy is also negative but not statistically significant. Results are similar when we consider
the percentage of bank directors on the board. These results indicates that the perceived costs
of bank affiliation in Belgium increased during the first years of the Great Depression. The
negative effect of the number of affiliated banks in 1929 and 1930 is in line with literature
arguing that multiple bank affiliations may reduce the incentive of individual banks to
intervene because the costs are borne by one bank while the benefits are enjoyed by the other
banks as well.
5.5. Board characteristics and firm risk
We have found that board busyness, board size and bankers on the board negatively
affected firm performance during the crisis. However, it cannot be ruled out that an
25
unobserved factor correlated with our board measures affected firm performance during the
crisis. Specifically, board busyness, board size and bankers on the board might be related to
firm risk. The resource dependence theory proposes that corporate boards are a mechanism for
reducing environmental uncertainty (Pfeffer, 1972). Hillman and Dalziel (2003) argue that
boards may not only provide resources, but may also “provide channels of communication and
conduits of information between the firm and external organizations”. Boards “provide the
firm with timely and valuable information and serve to reduce the transaction costs of dealing
with uncertainties in the environment” (Hillman and Dalziel, 2003, p387). Therefore, the
higher a firm’s environmental uncertainty or risk, the more a firm may try to connect with the
external environment by having a large board with busy directors and bank directors.
Furthermore, firms with higher risk will typically suffer more during a crisis (e.g. Baek et al.,
2004). The negative crisis effects of our board variables may therefore be driven by firm risk.
Firms with a higher systematic risk enlarge their network by having a large board with busy
directors and bank directors, and they are the ones that suffer most during the crisis because of
their higher systematic risk. Consistent with this argument, Table 3 shows a strong and
statistically significant correlation between our board busyness measures and beta, which
measures systematic risk (correlations between 0.42 to 0.53). Table 7 shows that, on average,
beta differs strongly per industry. For example, mining industry has the lowest beta (0.30) and
the chemicals industry, electricity, tram- and railways and colonial firms have highest beta’s,
respectively 1.12, 1.10, 1.10 and 1.21.
*** Table 7 about here***
*** Table 8 about here***
Our base regressions included beta as a control variable, but this does not take into
account that the effect of beta might change during a crisis. While beta generally tends to be
positively related to stock returns, one would expect firms with a high beta to perform poorly
26
during a crisis period, when the overall market performs poorly (e.g. Baek et al., 2004). To
take into account the changing effect of beta during the crisis, we interact beta with the three
crisis years. Results are shown in table 8. Regression (11) shows a significant positive relation
between beta and MTB before the crisis. However, this relation became negative during the
crisis. This shows that the higher the systematic risk, measured by beta, the higher firm value
losses are during the crisis. When we add the board measures directorships per director, board
size and no. of affiliated banks together with their interactions with the crisis years dummies in
regressions (12), (13) and (14), we still find a significant positive relation for each measure
before the crisis. However the crisis effects become less pronounced. The number of
directorships per director (regression (12)) and the no. of affiliated banks (regression (14)) still
have a statistically significant negative effect in 1929 but not anymore in 1930 and 1931. The
coefficients for 1930 and 1931 are still negative but not statistically significant anymore. In
Table 5 board size had a statistically significant negative effect in 1929, but this effect is not
significant anymore when we take into account the changing effect of beta. This means that,
when we account for systematic firm risk, the negative effect during the crisis becomes less
strong economically and statistically. It seems that firms subject to more systematic risk had a
more extended network reflected by busy boards, large boards and bankers on the board and
suffered more during the crisis.
5.6. Robustness checks
As we only have 574 observations for 150 firms and we include interaction variables in
our model, we have a potential multicollinearity problem. Indeed, calculation of variance
inflation factors shows that there is multicollinearity for the interaction variables. As a
robustness check, we consider a dummy variable crisis which combines the three crisis years,
instead of three crisis year dummies. Table A.1 in the appendix shows results for directorships
per director (regression (15)), ln(Board size) (regression (16)) and the no. of affiliated banks
27
(regression (17)). The results are in line with our previous results: we find a significant positive
relation before the crisis that is reduced during the crisis. In the case of board size, the crisis
effect is negative but not statistically significant.
To measure total firm risk instead of systematic risk, we included stock return volatility
instead of beta. Stock return volatility is defined as the standard deviation of monthly stock
returns over the previous year. We reproduce results from table 8 with stock return volatility
instead of beta in table A.2 in the appendix. All results from this additional test are
qualitatively the same as before.
6. Conclusion
This study investigates the impact of corporate board structure on the value of Belgian firms in
the period 1928-1931. In this period which was characterized by weak corporate governance
and high information asymmetries, firms had large, busy boards and were strongly connected
with other firms and with banks. Board busyness, board size and bankers on the board were
positively related to firm value before the crisis, but this positive effect tended to become
smaller during the crisis in 1929-1931. It seems that these connections provided little
protection during the crisis. Consistent with resource dependency theory which considers
boards as provider of resources, our results suggest that board busyness, board size and
bankers on the board are correlated with firm risk. When we take into account that the effect of
firm risk on market value changed during the crisis, we find that the negative effect of our
board characteristics on firm value become smaller, suggesting that our findings for board
characteristics are partly driven by firm risk. High risk firms had a larger director network in
order to cope with environmental uncertainties, but they suffered more during the crisis.
References
Adams, Renée B., Ferreira, Daniel, 2007. A Theory of Friendly Boards. The Journal of Finance 62, 217-250.
28
Agarwal, Fajshree, Elston, Julie Ann, 2001. Bank-firm relationships, financing and firm performance in Germany.
Economics Letters 72, 225-232.
Aldcroft, Derek H., 1997. Studies in the Interwar European Economy. Ashgate Publishing Limited.
Baek, Jae-Seung, Kang, Jun-Koo, Park, Kyung Suh, 2004. Corporate governance and firm value: evidence from
the Korean financial crisis. Journal of Financial Economics 71, 265-313.
Barton, Jan, Waymire, Gregory, 2004. Investor protection under unregulated financial reporting. Journal of
Accounting & Economics 38, 65-116.
Baudhuin, Fernand, 1944. Histoire Economique de la Belgique 1914-1939. Etablissements Emile Bruylant.
Becht, Marco, Ramírez, Carlos D., 2003. Does Bank Affiliation Mitigate Liquidity Constraints? Evidence from
Germany's Universal Banks in the Pre-World War I Period. Southern Economic Journal 70, 254-272.
Bernanke, B.S., James, Harold, 1990. The Gold Standard, Deflation, and financial crisis in the Great Depression:
an International Comparison. NBER Working Papers Series No. 3488.
Black, Bernard, Hasung, Jang, Woochan, Kim, 2006. Does Corporate Governance Predict Firms' Market Values?
Evidence from Korea. Journal of Law, Economics, & Organization 22, 366-413.
Black, Bernard, Kim, Woochan, 2012. The effect of board structure on firm value: A multiple identification
strategies approach using Korean data. Journal of Financial Economics 104, 203-226.
Boone, Audra L., Field, Laura Casares, Karpoff, Jonathan M., Raheja, Charu G., 2007. The determinants of
corporate board size and composition: An empirical analysis. The Journal of Financial Economics 85,
66-101.
Boyd, B., 1990. Corporate linkages and organizational environment: a test of the resource dependence model.
Strategic Management Journal 11, 419-430.
Braggion, Fabio, Moore, Lyndon, 2013. How insiders traded before rules. Business History 55, 562-581.
Buyst, E., 2005. Beschavingsgeschiedenis en inleiding tot de economische geschiedenis. Acco, Leuven.
Carpenter, Mason A., Westphal, James D., 2001. The strategic context of external network ties: examining the
impact of director appointments on board involvement in strategic decision making. Academy of
Management Journal 4, 639-660.
Centre d'étude des Sociétés, 1956. Le Statut des Commissaires et des Commissaires-Reviseurs. Établissements
Émile Bruylant, Bruxelles.
Chlepner, B. S., 1930. Le marché financier belge depuis cent ans. Falk, Bruxelles.
Chlepner, B. S., 1943. Belgian Banking and Banking Theory. The Brookings Institution.
Coles, Jeffrey L., Daniel, Naveen D., Naveen, Lalitha, 2008. Boards: Does one size fit all? Journal of Financial
Economics 87, 329-356.
Coles, Jeffrey L., Daniel, Naveen D., Naveen, Lalitha, 2012. Board Advising. Working paper.
Dahya, Jay, McConnell, John J., 2007. Dominant shareholders, corporate boards, and corporate value: a crosscountry analysis. Journal of Financial Economics 87, 73-100.
Daily, C.M., Dalton, D.R., 1994. Bankruptcy and corporate governnace: The impact of board composition and
structure. Academy of Management Journal 37, 1603-1617.
Dalton, Dan R., Daily, Catherine M., Johnson, Jonathan L., Ellstrand, Alan E., 1999. Number of directors and
financial performance: a meta-analysis. Academy of Management Journal 42, 674-686.
Deloof, Marc, Van Overfelt, Wouter, 2008. Were modern capital structure theories valid in Belgium before World
War I? Journal of Business Finance and Accounting 35, 491-515.
Diamond, Douglas W., 1984. Financial intermediation and Delegated Monitoring. Review of Economic Studies
51, 393-414.
Durviaux, Roland, 1947. La Banque Mixte: origine et soutien de l'expansion économique de la Belgique.
Etablissements Emile Bruylant, Bruxelles.
Eisenberg, Theodore, Sundgren, Stefan, Wells, Martin T., 1998. Larger board size and decreasing firm value in
small firms. Journal of Financial Economics 48, 35-54.
Elsas, Ralf, Krahnen, Jan Pieter, 1998. Is relationship lending special? Evidence from credit-file data in Germany.
Journal of Banking & Finance 22, 1283-1316.
Erkens, David H., Hung, Mingyi, Matos, Pedro, 2012. Corporate governance in the 2007-2008 financial crisis:
Evidence from financial institutions worldwide. Journal of Corporate Finance 18, 389-411.
Faleye, Olubunmi, Hoitash, Rani, Hoitash, Udi, 2011. The costs of intense board monitoring. Journal of Financial
Economics 101, 160-181.
Ferri, Giovanni, Kang, Tae Soo, Kim, In-June, 2001. The value of relationship banking during financial crises:
evidence from the Republic of Korea Policy research working paper. - Washington, D.C.
Ferris, Stephen P., Jagannathan, Murali, Pritchard, A.C., 2003. Too Busy to Mind the Business? Monitoring by
Directors with Multiple Board Appointments. The Journal of Finance 58, 1087-1112.
Fich, Eliezer M., Shivdasani, Anil, 2006. Are Busy Boards Effective Monitors? The Journal of Finance 61, 689724.
29
Field, Laura, Lowry, Michelle, Mkrtchyan, Anahit, 2013. Are busy boards detrimental? Journal of Financial
Economics 109, 63-82.
Foglia, A., Laviola, S., Reedtz, P. Marullo, 1998. Multiple banking relationships and the fragility of corporate
borrowers. Journal of Banking & Finance 22, 1441-1456.
Fohlin, Caroline, 1997. Universal banking networks in pre-war Germany: new evidence from company financial
data. Research in Economics 51, 201-225.
Fohlin, Caroline, 1998. Relationship Banking, Liquidity, and Investment in the German Industrialization. Journal
of Finance 53, 587-612.
Fohlin, Caroline, 1999. Universal banking in Pre-World War I Germany: Model or Myth? Explorations in
Economic History 36, 305-343.
Foreman-Peck, James, Hannah, Leslie, 2013. Some consequences of the early twentieth-century British divorce of
ownership from control. Business History 55, 540-561.
Francis, Bill B., Hasan, Iftekhar, Wu, Qiang, 2012. Do corporate boards matter during the current financial crisis?
Review of Financial Economics 21, 39-52.
Frydman, Carola, Hilt, Eric, 2010. Predators or Watchdogs? Bankers on Corporate Boards in the Age of Finance
Capitalism.
Gilis, H., 1929. La Société Anonyme au point de vue Juridique, Administratif, Financier, Comptable & Fiscal. J.
Leherte-Courtin & Fils, Renaix.
Gilson, Stuart C., 1990. Bankruptcy, boards, banks, and blockholders. Journal of Financial Economics 27, 355387.
Goodstein, Jerry, Gautam, Kanak, Boeker, Warren, 1994. The effects of board size and diversity on strategic
change. Strategic Management Journal 15, 241-250.
Graham, John R., Hazarika, Sonali, Narasimhan, Krishnamoorthy, 2011. Corporate Governance, Debt, and
Investment Policy During the Great Depression. Management Science 57, 2083-2100.
Heyman, Henri, 1928. Centralisation et Rationalisation des Entreprises Industrielles en Belgique. Editions de la
société d'études morales, sociales et juridiques, Louvain.
Hillman, Amy J., Dalziel, Thomas, 2003. Boards of directors and firm performance: integrating agency and
resource dependence perspectives. Academy of Management Journal 28, 383-396.
Hoshi, Takeo, Kashyap, Anil, Scharfstein, David, 1990. The role of banks in reducing the costs of financial
distress in Japan. Journal of Financial Economics 27, 67-88.
Hwang, Byoung-Hyoun, Kim, Seoyoung, 2009. It pays to have friends. Journal of Financial Economics 93, 138158.
Jensen, Michael C., 1993. The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems.
The journal of Finance 48, 831-880.
Johnson, Simon, Boone, Peter, Breach, Alasdair, Friedman, Eric, 2000. Corporate governance in the Asian
financial crisis. Journal of Financial Economics 58, 141-186.
Kaplan, Steven N., Reishus, David, 1990. Outside directorships and corporate performance. Journal of Financial
Economics 27, 389-410.
Kroszner, Randall, Strahan, Philip E., 2001. Bankers on boards: monitoring, conflicts of interest, and lender
liability. Journal of Financial Economics 62, 415-452.
Lamal, Emile, 1930. Une Enquête sur les actions a vote privilégié en Belgique. Bruxelles: Maurice Lamertin.
Lemmon, Michael L., Lins, Karl V., 2003. Ownership Structure, Corporate Governance, and Firm Value:
Evidence from the East Asian Financial Crisis. Journal of Finance 108, 1445-1468.
Linck, James S., Netter, Jeffry M., Yang, Tina, 2008. The determinants of board structure. Journal of Financial
Economics 87, 308-328.
Lipton, Martin, Lorsch, Jay W., 1992. A modest proposal for improved corporate governance. Business Lawyer
48, 59-77.
Loderer, Claudio, Peyer, Urs, 2002. Board Overlap, Seat Accumulation and Share Prices. European Financial
Management 8, 165-192.
Mace, Myles L., 1986. Directors: myth and reality. Harvard business school, Boston.
Mitton, Todd, 2002. A cross-firm analysis of the impact of corporate governance on the East Asian financial
crisis. Journal of Financial Economics 64, 215-241.
Miwa, Yoshiro, Ramseyer, J. Mark, 2000. Corporate Governance in Transitional Economies: lessions from the
Prewar Japanese Cotton Textile Industry. Journal of Legal Studies 29, 171-204.
Nguyen, Tu, Nguyen, H.G., Yin, Xiangkang, 2015. Corporate Governance and Corporate Financing and
Investment during the 2007-2008 Financial Crisis. Financial Management 44, 115-146.
Petersen, Mitchell A., Rajan, Raghuram G., 1994. The Benefits of Lending Relationships: Evidence from Small
Business Data. The Journal of Finance 49, 3-37.
Pfeffer, J., 1972. Size and composition of corporate boards of directors: the organization and its environment.
Administrative Science Quarterly 17, 218-228.
30
Pfeffer, J., Salancik, G., 1978. The external control of organizations: a resource-dependence perspective. Harper
& Row, New York.
Rajan, Raghuram G., 1992. Insiders and Outsiders: The Choice between Informed and Arm's-Length Debt.
Journal of Finance 47, 1367-1400.
Rajan, Raghuram G., Zingales, Luigi, 2003. The great reversals: the politics of financial development in the
twentieth century. Journal of Financial Economics 69, 5-50.
Tilman, Samuel, 2006. Les grands banquiers belges (1830-1935). Portrait collectif d'une élite. Académie royale
de Belgique, Louvain-la-Neuve.
van der Valk, H.M.M.A., 1932. De betrekkingen tusschen banken en industrie in België. Haarlem: Bohn.
Van Nieuwerburgh, Stijn, Buelens, Frans, Cuyvers, Ludo, 2006. Stock market development and economic growth
in Belgium. Explorations in Economic History 43, 13-38.
Van Overfelt, Wouter, Annaert, Jan, De Ceuster, Marc, Deloof, Marc, 2009. Do universal banks create value?
Universal bank affiliation and company performance in Belgium, 1905-1909. Explorations in Economic
History 46, 253-265.
Van Overfelt, Wouter, Deloof, Marc, Vanstraelen, Ann, 2010. Determinants of Corporate Financial Disclosure in
an Unregulated Environment: Evidence from the Early 20th Century. European Accounting Review 19,
7-34.
Vanthemsche, Guy, 1991. State, banks and industry in Belgium and The Netherlands, 1919-1939. In: James, H.,
H. Lindgren and A. Teichova, (Ed., The Role of Banks in the Interwar Economy. Cambridge University
Press, Cambridge.
Wauwermans, Paul, 1927. Manuel pratique des sociétés anonymes. Etablissements Emile Bruylant, Brussels.
Wauwermans, Paul, 1933. Manuel Pratique des Sociétés Anonymes. Etablissements Emile Bruylant, Bruxelles.
Weinstein, David E., Yafeh, Yishay, 1998. On the Costs of a Bank-Centered Financial System: Evidence from the
Changing Main Bank Relations in Japan. Journal of Finance 53, 635-672.
Wintoki, M. Babajide, Linck, James S., Netter, Jeffry M., 2012. Endogeneity and the dynamics of internal
corporate governance. Journal of Financial Economics 105, 581-606.
Yermack, David, 1996. Higher market valuation of companies with a small board of directors. Journal of
Financial Economics 40, 185-211.
31
Table 1
Descriptive statistics (period 1928-1931, 150 firms and 574 observations)
Mean
Median
St. Dev.
25th percentile
75th percentile
Panel A: Board characteristics in 1928 (150 observations)
Board size
No. of executive directors
No. of supervising directors
15.38
10.87
4.51
14.00
10.00
4.00
6.35
5.20
1.82
11.00
7.00
3.00
19.00
14.00
5.00
Directorships per director
Percentage of busy directors
No. of external directorships
2.95
0.38
31.95
2.84
0.38
27.00
1.41
0.24
27.68
1.70
0.18
8.00
4.00
0.56
49.00
Bank affiliated (0,1)
No. of affiliated banks
Percentage of bank directors
0.87
2.21
0.22
2.00
0.20
1.62
0.15
1.00
0.09
3.00
0.33
0.88
53
10
0.23
0.16
1.00
0.62
3.59
200
36
0.48
0.63
3.00
1.17
Panel B: Firm characteristics in 1928-1931 (574 observations)
MTB
Size (in millions of Belgian franks)
Age
Debt ratio
Intangible assets
Corporate diversification
Betaa
4.52
200
26.07
0.37
0.40
1.88
0.94
1.59
100
20.00
0.34
0.38
2.00
0.89
16.57
310
20.40
0.19
0.27
1.04
0.45
Notes: board size is the number of board seats of a firm; no. of executive directors is the total number of directors
at the board of a firm minus the number of supervising directors; no. of supervising directors is the number of
supervising directors at the board of a firm; directorships per director is the total number of directorships held
across all the directors of a board divided by board size; percentage of busy directors is the percentage of busy
directors on a firm’s board (a director is defined as busy if he/she holds three or more directorships); no. of
external directorships is the total number of directorships held across all the directors of a board minus board
size; bank affiliated is a dummy equal to one if the firm has at least one bank director on its board, and zero
otherwise; no. of affiliated banks is the number of banks with which a firm is affiliated; percentage of bank
directors is the number of bank directors on the board of a firm divided by the total number of directors of that
firm; MTB is the market value of equity divided by the book value of the equity at the end of the fiscal year; size
is total assets; age is the number of years the firm has been listed; debt ratio is the book value of total debt
divided by total assets; intangible assets are the intangible assets divided by total assets; and corporate
diversification is the number of industries in which the firm is active; beta is computed by regressing a firm’s
monthly stock return in the pre-crisis period on the corresponding market return obtained from SCOB. We require
at least 24 months of return prior to January 1929 to compute beta and we use a maximum of 63 months’ worth of
data. Board variables and industry variables are defined at the start of 1929. Size, age, debt ratio and intangible
assets are defined every year of the four-year sample period 1928-1931.
a
Based on 150 observations
32
Table 2
Board characteristics and firm connections across industries (mean figures)
5
6
7
8
9
No. of
Bank
No. of
% external
Connected
external
affiliated
affiliated
directorships
firms within
directorships
(0,1)
banks
within industry
industry
Mining
23
13.70
2.97
30%
27.74
0.91
1.87
33%
20
Metal
17
14.95
2.74
33%
26.94
0.88
2.41
18%
15
Electricity
14
20.29
3.67
57%
54.64
1.00
2.93
27%
14
Stone
9
10.78
2.03
20%
12.56
0.78
1.67
19%
4
Tram/railways
9
17.78
2.98
43%
34.22
0.78
1.78
14%
8
Chemicals
8
14.13
2.15
25%
17.63
0.88
2.13
6%
4
Textile
5
10.80
1.48
16%
5.80
0.60
1.80
23%
3
Foreign
32
15.94
3.00
40%
34.88
0.88
2.34
36%
28
Colonial
23
16.39
3.79
55%
45.61
1.00
3.00
49%
21
Rest (*)
10
13.82
2.22
24%
17.82
0.55
1.00
39%
0
(*)
Notes:
The rest category includes firms from the following industries: automobile industry, catering, construction, food industry, other manufacturing, other
utilities, shipping, weapon industry and wholesale
Industry
1
No of
firms
2
Board size
3
Directorships
per director
4
% busy
directors
33
Table 3
Pearson correlations
1
Variables
Board size
1
1
2
3
4
5
6
7
8
9
2
No. of executive directors
0.97
1
3
No. of supervising directors
0.70
0.49
1
4
Directorships per director
0.25
0.24
0.18
1
5
Percentage of busy directors
0.22
0.21
0.15
0.86
1
6
Ln(no. of external directorships) 0.65
0.63
0.45
0.85
0.72
1
7
Bank affiliated (0,1)
0.27
0.26
0.20
0.36
0.39
0.34
1
8
No. of affiliated banks
0.49
0.51
0.23
0.39
0.43
0.53
0.54
1
9
Percentage of bank directors
0.21
0.24
0.04
0.45
0.48
0.42
0.57
0.75
10 MTB
-0.06
-0.03 -0.13
0.03
0.01
0.02
-0.16 -0.04 -0.05
11 Size
0.33
0.32
0.05
0.01
0.15
0.11
12 Age
-0.32
-0.37 -0.04
13 Debt ratio
-0.12
14 Intangible assets
10
11
12
13
14
15
1
1
0.09
-0.11
1
-0.11 -0.22 -0.21 -0.06
-0.17 -0.01
0.01
0.02
1
-0.13 -0.03
-0.06 -0.05 -0.08 -0.01
-0.07 -0.05 -0.01
0.17
0.17
1
0.06
0.05
0.07
0.04
0.01
0.05
0.02
0.07
0.09
-0.10
0.06
0.02
-0.32
1
15 Corporate diversification
0.15
0.16
0.07
-0.01
0.08
0.03
0.04
0.06
0.06
0.11
-0.03
-0.06
0.06
-0.29
16 Beta
0.22
0.21
0.18
0.42
0.53
0.43
0.13
0.20
0.16
0.07
-0.01
-0.24 -0.09 -0.09 0.21
0.23
0.19
1
This table shows Pearson correlation coefficients. All variables are defined as before. Bold denotes significance at the 5% level.
34
Table 4
Corporate value and busy boards
(1)
(2)
(3)
(4)
Directorships per
Percentage of
ln(no. of external
director
busy directors
directorships)
1929
-0.275***
0.023
-0.039
0.131
(0.000)
(0.761)
(0.533)
(0.115)
1930
-0.772***
-0.553***
-0.578***
-0.392***
(0.000)
(0.000)
(0.000)
(0.004)
1931
-1.101***
-0.851***
-0.873***
-0.705***
(0.000)
(0.000)
(0.000)
(0.000)
Busy measure
0.178***
0.595*
0.230***
(0.001)
(0.091)
(0.001)
Busy measure*1929
-0.101***
-0.628***
-0.134***
(0.000)
(0.000)
(0.000)
Busy measure*1930
-0.074**
-0.524**
-0.126***
(0.030)
(0.010)
(0.002)
Busy measure*1931
-0.084**
-0.621***
-0.132***
(0.016)
(0.003)
(0.002)
ln (Size)
-0.499***
-0.503***
-0.492***
-0.511***
(0.000)
(0.000)
(0.000)
(0.000)
ln (Age)
0.116
0.124
0.141*
0.155*
(0.145)
(0.120)
(0.081)
(0.054)
Debt ratio
1.569***
1.592***
1.592***
1.607***
(0.000)
(0.000)
(0.000)
(0.000)
Intangible assets
-0.569***
-0.582***
-0.596***
-0.593***
(0.001)
(0.000)
(0.000)
(0.000)
Corporate diversification
-0.079
-0.055
-0.079
-0.065
(0.259)
(0.455)
(0.292)
(0.372)
Beta
0.015
-0.118
-0.017
-0.114
(0.920)
(0.437)
(0.923)
(0.481)
Industry dummies
Yes
Yes
Yes
Yes
Observations
574
574
574
574
Notes: this table displays the regression coefficients and robust p-values for random effects regressions. The
results are based on clustered standard errors at the firm level. The dependent variable is the natural logarithm of
the MTB. 1929, 1930 and 1931 are year dummies; directorships per director is the total number of directorships
held across all the directors of a board divided by board size; percentage of busy directors the percentage of busy
directors in a firm’s board; ln(no. of external directorships) is the natural logarithm of the total number of
directorships held across all the directors of a board minus board size plus one; all other variables are defined as
before. Industry dummies are included. Intercept is included but not reported. Here ***, **, and * indicate
significance at the 1%, 5%, and 10% levels, respectively.
Busy measure
35
Table 5
Corporate value and board size
(6)
(7)
ln(No. of executive ln(No. of supervising
directors)
directors)
1929
0.151
-0.014
-0.190
(0.460)
(0.940)
(0.131)
1930
-0.423
-0.585**
-0.718***
(0.220)
(0.032)
(0.001)
1931
-0.629*
-0.858***
-1.038***
(0.077)
(0.009)
(0.000)
Board size measure
0.561***
0.356*
0.346**
(0.003)
(0.083)
(0.038)
Board size measure * 1929
-0.160**
-0.114
-0.055
(0.033)
(0.139)
(0.491)
Board size measure * 1930
-0.130
-0.082
-0.030
(0.297)
(0.476)
(0.829)
Board size measure * 1931
-0.176
-0.107
-0.034
(0.168)
(0.447)
(0.798)
ln (Size)
-0.538***
-0.522***
-0.530***
(0.000)
(0.000)
(0.000)
ln (Age)
0.179**
0.165*
0.114
(0.031)
(0.075)
(0.145)
Debt ratio
1.600***
1.603***
1.577***
(0.000)
(0.000)
(0.000)
Intangible assets
-0.580***
-0.580***
-0.566***
(0.000)
(0.001)
(0.001)
Corporate diversification
-0.095
-0.093
-0.082
(0.152)
(0.172)
(0.229)
Beta
-0.031
-0.009
-0.041
(0.832)
(0.949)
(0.789)
Industry dummies
Yes
Yes
Yes
Observations
574
574
574
Notes: this table displays the regression coefficients and robust p-values for random effects regressions.
The results are based on clustered standard errors at the firm level. The dependent variable is the natural
logarithm of the MTB. Ln(Board size) is the natural logarithm of the number of a firm’s board seats;
ln(no. of executive directors) is the natural logarithm of the total number of directors at the board of a
firm minus the number of supervising directors; ln(no. of supervising directors) is het number of
supervising directors at the board of a firm; all other variables are defined as before. Industry dummies
are included. Intercept is included but not reported. Here ***, **, and * indicate significance at the 1%,
5%, and 10% levels, respectively.
Board size measure:
(5)
Ln(Board size)
36
Table 6
Corporate value and bank affiliation
(8)
(9)
(10)
Bank affiliated No. of affiliated
Percentage of
(0,1)
banks
bank directors
1929
-0.117
-0.118**
-0.115*
(0.162)
(0.032)
(0.059)
1930
-0.654***
-0.659***
-0.685***
(0.000)
(0.000)
(0.000)
1931
-0.966***
-1.029***
-1.023***
(0.000)
(0.000)
(0.000)
Bank affiliation measure
0.127
0.097**
0.717*
(0.448)
(0.033)
(0.097)
Bank affiliation measure * 1929
-0.184**
-0.071***
-0.728***
(0.043)
(0.000)
(0.001)
Bank affiliation measure * 1930
-0.138
-0.051*
-0.397
(0.434)
(0.078)
(0.215)
Bank affiliation measure * 1931
-0.159
-0.032
-0.355
(0.356)
(0.312)
(0.240)
ln (Size)
-0.495***
-0.512***
-0.503***
(0.000)
(0.000)
(0.000)
ln (Age)
0.121
0.141*
0.121
(0.132)
(0.084)
(0.127)
Debt ratio
1.576***
1.545***
1.534***
(0.000)
(0.000)
(0.000)
Intangible assets
-0.560***
-0.593***
-0.592***
(0.001)
(0.000)
(0.000)
Corporate diversification
-0.079
-0.079
-0.082
(0.261)
(0.261)
(0.235)
Beta
0.017
-0.007
-0.004
(0.912)
(0.964)
(0.979)
Industry dummies
Yes
Yes
Yes
Observations
574
574
574
Notes: this table displays the regression coefficients and robust p-values for random effects
regressions. The results are based on clustered standard errors at the firm level. The
dependent variable is the natural logarithm of the MTB. Bank affiliated (0,1) is a dummy
that equals one if a firm has at least one banker on its board and zero otherwise; no. of
affiliated banks is the number of banks of which the firm has a director on its board;
percentage of bank directors is the number of bank directors on the board of a firm divided
by the total number of directors of that firm; and all other variables are defined as before.
Industry dummies are included. Intercept is included but not reported. Here ***, **, and *
indicate significance at the 1%, 5%, and 10% levels, respectively.
Bank affiliation measured by:
37
Table 7
Average beta per sector in 1929 (mean figures)
Industry
No of firms Beta
Mining
23
0.304
Metal
17
0.838
Electricity
14
1.101
Stone
9
0.291
Tram/railways 9
1.100
Chemicals
8
1.119
Textile
5
0.517
Foreign
32
0.647
Colonial
23
1.210
Rest (*)
10
0.791
(*)
Notes:
The rest category includes
firms from the following industries:
automobile
industry,
catering,
construction, food industry, other
manufacturing, other utilities, shipping,
weapon industry and wholesale
38
Table 8
Corporate value, board characteristics and firm risk during the crisis
(11)
(12)
(13)
(14)
Directorships per
ln (Board size)
No. of affiliated
director
banks
1929
0.045
0.159*
0.208
0.129*
(0.582)
(0.057)
(0.284)
(0.093)
1930
-0.283**
-0.275**
-0.306
-0.250**
(0.021)
(0.034)
(0.331)
(0.038)
1931
-0.696***
-0.642***
-0.534
-0.683***
(0.000)
(0.000)
(0.116)
(0.000)
Beta
0.340**
0.173
0.285*
0.300**
(0.025)
(0.255)
(0.050)
(0.046)
Beta * 1929
-0.347***
-0.262***
-0.332***
-0.307***
(0.000)
(0.001)
(0.000)
(0.000)
Beta * 1930
-0.542***
-0.536***
-0.541***
-0.529***
(0.000)
(0.000)
(0.000)
(0.000)
Beta * 1931
-0.456***
-0.412***
-0.436***
-0.453***
(0.000)
(0.003)
(0.000)
(0.000)
Board measure
0.136***
0.449**
0.076*
(0.007)
(0.019)
(0.090)
Board measure * 1929
-0.065***
-0.065
-0.055***
(0.006)
(0.379)
(0.001)
Board measure * 1930
-0.003
0.011
-0.020
(0.930)
(0.924)
(0.499)
Board measure * 1931
-0.031
-0.065
-0.007
(0.445)
(0.601)
(0.829)
ln (Size)
-0.464***
-0.476***
-0.506***
-0.480***
(0.000)
(0.000)
(0.000)
(0.000)
ln (Age)
0.157**
0.153**
0.203**
0.171**
(0.041)
(0.045)
(0.011)
(0.030)
Debt ratio
1.548***
1.542***
1.564***
1.515***
(0.000)
(0.000)
(0.000)
(0.000)
Intangible assets
-0.649***
-0.650***
-0.653***
-0.669***
(0.000)
(0.000)
(0.000)
(0.000)
No of business segments
-0.094
-0.069
-0.107
-0.093
(0.171)
(0.342)
(0.101)
(0.178)
Constant
8.696***
8.704***
8.261***
8.902***
(0.000)
(0.000)
(0.000)
(0.000)
Industry dummies
Yes
Yes
Yes
Yes
Observations
574
574
574
574
Notes: this table displays the regression coefficients and robust p-values for random effects regressions. The
results are based on clustered standard errors at the firm level. The dependent variable is the natural logarithm of
the MTB. All variables are defined as before. Intercept is included but not reported. ***, **, and * indicate
significance at the 1%, 5%, and 10% levels, respectively.
Board composition variable:
39
Figure 1
Cumulative stock return of all firms on the Brussels Stock Exchange
230%
210%
190%
170%
150%
130%
110%
90%
70%
50%
10 1931
7 1931
4 1931
1 1931
10 1930
7 1930
4 1930
1 1930
10 1929
7 1929
4 1929
1 1929
10 1928
7 1928
4 1928
1 1928
10 1927
7 1927
4 1927
1 1927
Source: Own calculations based on SCOB database
40
Figure 2
Market-to-book ratio 1928-1931
8
7
6
5
4
3
2
1
0
1928
1929
Mean
1930
1931
Median
Source: Own calculations based on SCOB database and data from the Recueil Financier
41
Figure 3
Connected firms via interlocking directorships
Note: This graph shows connections between 146 firms in our sample based on interlocking directors. Four firms
share no directors with other firms in our sample.
42
Appendix
Table A.1
Corporate value and bank affiliation with one dummy for the crisis
(15)
(16)
(17)
Directorships
ln (Board size)
No. of affiliated
per director
banks
Crisis
-0.347***
-0.534***
-0.487***
(0.000)
(0.000)
(0.000)
Average no. of mandates per firm
0.194***
0.607***
0.114**
(0.001)
(0.004)
(0.018)
Average no. of mandates per firm * crisis
-0.083***
-0.003
-0.047**
(0.001)
(0.613)
(0.047)
ln (Size)
-0.653***
-0.709***
-0.671***
(0.000)
(0.000)
(0.000)
ln (Age)
-0.102
-0.046
-0.083
(0.231)
(0.599)
(0.342)
Debt ratio
1.854***
1.857***
1.845***
(0.000)
(0.000)
(0.000)
Intangible assets
-0.655***
-0.657***
-0.667***
(0.000)
(0.000)
(0.000)
Number of business segments
-0.023
-0.069
-0.048
(0.789)
(0.366)
(0.556)
Beta
-0.167
-0.075
-0.041
(0.352)
(0.650)
(0.814)
Constant
12.572***
12.387***
13.106***
(0.000)
(0.000)
(0.000)
Industry dummies
Yes
Yes
Yes
Observations
574
574
574
Notes: this table displays the regression coefficients and robust p-values for random effects
regressions. The results are based on clustered standard errors at the firm level. The dependent
variable is the natural logarithm of the MTB. Crisis is a dummy variable that equals one if year is
1929, 1930 or 1931 and zero otherwise. All variables are defined as before. The control variables
ln(Size), ln(Age), Debt ratio, no of business segments and beta and the intercept are included but not
reported. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
Board measure:
43
Table A.2
Corporate value, board characteristics and firm risk during the crisis
(18)
(21)
No. of affiliated
banks
1929
0.016
0.188*
0.437**
0.148*
(0.8691)
(0.0527)
(0.0256)
(0.0970)
1930
-0.278**
-0.166
0.274
-0.138
(0.0126)
(0.1849)
(0.3929)
(0.2338)
1931
-0.575***
-0.492***
-0.190
-0.547***
(0.0000)
(0.0001)
(0.5250)
(0.0000)
Stock return volatility
2.773***
2.378***
2.781***
2.650***
(0.0009)
(0.0027)
(0.0005)
(0.0008)
Stock return volatility * 1929
-2.811***
-2.299***
-2.846***
-2.711***
(0.0009)
(0.0046)
(0.0004)
(0.0007)
Stock return volatility * 1930
-6.171***
-5.827***
-6.304***
-6.311***
(0.0001)
(0.0003)
(0.0001)
(0.0001)
Stock return volatility * 1931
-6.429***
-6.084***
-6.426***
-6.512***
(0.0000)
(0.0000)
(0.0000)
(0.0000)
Board measure
0.137***
0.531***
0.086*
(0.0037)
(0.0049)
(0.0578)
Board measure * 1929
-0.076***
-0.157**
-0.065***
(0.0005)
(0.0239)
(0.0000)
Board measure * 1930
-0.050
-0.204*
-0.061***
(0.1175)
(0.0679)
(0.0085)
Board measure * 1931
-0.040
-0.144
-0.012
(0.2088)
(0.1765)
(0.6601)
ln (Size)
-0.469***
-0.476***
-0.504***
-0.482***
(0.0000)
(0.0000)
(0.0000)
(0.0000)
ln (Age)
0.152**
0.157**
0.212***
0.171**
(0.0476)
(0.0385)
(0.0062)
(0.0277)
Debt ratio
1.555***
1.565***
1.582***
1.517***
(0.0000)
(0.0000)
(0.0000)
(0.0000)
Intangible assets
-0.643***
-0.634***
-0.654***
-0.670***
(0.0003)
(0.0003)
(0.0003)
(0.0002)
Number of business segments
-0.078
-0.054
-0.092
-0.077
(0.2546)
(0.4454)
(0.1573)
(0.2628)
Constant
8.777***
8.539***
7.895***
8.864***
(0.0000)
(0.0000)
(0.0000)
(0.0000)
Industry dummies
Yes
Yes
Yes
Yes
Observations
574
574
574
574
Note: this table is a reproduction of table 8 with stock return volatility instead of beta.
Board measure:
(19)
Directorships
per director
(20)
ln (Board size)
44