Business Research Quarterly: Are Religion and Culture Relevant For Corporate Risk-Taking? International Evidence

BRQ Business Research Quarterly (2019) 22, 36---55
Business
BRQ
Research
Quarterly
www.elsevier.es/brq
REGULAR ARTICLE
Are religion and culture relevant for corporate

risk-taking? International evidence
José María Díez-Esteban a , Jorge Bento Farinha b , Conrado Diego García-Gómez c,∗
a
University of Burgos, Department of Economics and Business Administration, Pza. Infanta Elena, 09001 Burgos, Spain
b
University of Porto, Faculty of Economics and Center for Economics and Finance (CEF-UP), Rua Roberto Frias, 4200-464 Porto,
Portugal
c
University of Valladolid --- Duques de Soria Campus, Department of Financial Economics and Accounting, Calle Universidad s/n,
42004 Soria, Spain
Received 19 August 2017; accepted 11 June 2018

Available online 21 July 2018
JEL Abstract Using a large sample of firms from 37 countries over the period of 2007---2015, we
CLASSIFICATION empirically analyse the impact of religion and national culture characteristics on the level of
G15; corporate risk-taking around the world and the channels through which this can take place. First,
G32; we initially observe that different religious backgrounds have different impacts on corporate
G34 risk-taking, these being negative for Catholic and Islamic-based countries and positive for firms
in Protestant nations. Secondly, we observe that companies in countries with high scores of
KEYWORDS power distance, masculinity, individualism and long-term orientation tend to increase risk-
Corporate risk-taking; taking while high levels of uncertainty avoidance moderates corporate risk-taking behavior.
National culture; We also show results that in companies where institutional investors are the most relevant
Religion; reference shareholder the influence of religion on corporate risk-taking is not felt, unlike when
Ownership structure the main shareholder is an individual or a family.
© 2018 ACEDE. Published by Elsevier España, S.L.U. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction and motivation the recent 2007---08 financial crisis (see for instance Erkens
et al., 2012 as well as a survey of senior managers conducted
Excessive corporate risk-taking is a particularly important by KPMG, 2009). Thus, better understanding which factors
topic given that it is usually pointed out as a direct cause of can influence corporate risk-taking is crucial for investors,
regulators and indeed all major remaining stakeholders.
Extant research usually focuses on governance, ownership
∗ structure and incentive systems as the direct causes of cor-
Corresponding author.
E-mail address: [email protected] porate risk taking and is typically more concerned with the
(C.D. García-Gómez). financial sector (ex. Laeven and Levine, 2009; Coles et al.,
2006; Kempf et al., 2009; Conyon et al., 2011; Berger et al.,
https://doi.org/10.1016/j.brq.2018.06.003
2340-9436/© 2018 ACEDE. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
Influence of religion and culture on corporate risk-taking 37
2016). We take the view that, notwithstanding the above Although religion has long been part of the economic
factors, both religion and culture are dimensions that need though (cf. the studies of the Spanish School in the XVI
also to be considered in explaining corporate risk-taking Century or Smith, 1976), research in the field known as
decisions (for both financial and non-financial firms). ‘‘economics of religion’’ is relatively new (Iyer, 2016).
In our paper, we seek to develop further the literature Whereas early works studied the motivations that individuals
that connects corporate risk taking with culture by intro- might have to hold religious beliefs (Ekelund et al., 2002),
ducing a religious dimension and extending the analysis to more recent research is focused quite heavily on the socioe-
include explicit interactions between the national culture conomic consequences of religion, using economic theory
and religion, notably by taking into account the major reli- and more sophisticated statistical tools. In fact, research
gion groups in the world, while using an extensive sample of in accounting and finance increasingly focuses on the link
observations. Although religion and culture have been the between religion and corporate decision-making, empha-
subject of a vast literature in many areas of study, research sizing two aspects: (1) the role of religion as an external
on these subjects in the field of economics is relatively new monitoring mechanism and (2) its relationship to risk aver-
and more limited, especially in the specific field of finan- sion (Kanagaretnam et al., 2015). Thus, and following the
cial economics (see Iyer, 2016 for a survey on Religion and seminal study by Iannaccone (1998), we address the reli-
Economics, and Beugelsdijk and Maseland, 2011, and Reuter, gious issue by studying religion as an independent variable
2011, for literature reviews on Culture and Economics). In on risk attitudes.2
this paper we seek to fill some of that void by using a large Most of previous research on the role of religion in firms
sample of 34,251 observations for the period 2007---2015 and corporate risk-taking, focuses on the religiosity differ-
encompassing 37 different countries to analyse if, and how, ences within a country. For instance, Hilary and Hui (2009)
religion and national culture characteristics impact on the investigate how a firm’s investment decisions are affected by
risk-taking behaviour of listed companies across different the religiosity of its environment. Shu et al. (2011) link local
countries and industries, and try to shed a light on potential religiosity to risk taking by mutual funds. Kumar et al. (2011)
interactions between these two dimensions. see religion as a proxy for gambling propensity and relate
In a recent paper, Minkov and Hofstede (2014) conclude, geographical heterogeneity in religion to differences in cor-
on the basis of a survey on personal values, that national porate decisions and stock returns. Their findings suggest a
culture influence is a much stronger influence than global positive association between religiosity and risk aversion.3
religions. In our research we seek to verify if indeed such Nevertheless, while the literature is quite clear about the
assertion that religion plays a secondary role (or no role at fact that religious people of any faith generally display
all) truly holds in the determination of corporate risk-taking greater risk aversion,4 it is unclear regarding differences
around the world. We also attempt to verify if any interac- across religious groups. Besides, previous research mainly
tions exist between the cultural and religious dimensions for focuses on the differences between Christian denomina-
explaining the observable corporate behaviour towards the tions.
assumption of risks. Traditionally, OECD countries can be divided into
The remainder of this paper is organized as follows. Sec- four groups, depending on the religious background:
tion 2 reviews previous research on the relations between Catholic/Orthodox,5 Protestant, Muslim and Eastern Reli-
national culture, religion and corporate risk-taking and for- gions (including, Buddhist, Hindus and Taoists) (Iannaccone,
mulates our hypotheses. Section 3 describes the sample and 1998).
variables and explains the empirical methodology. Section 4 The distinction between Christian denominations derives
shows the empirical results and assesses the degree to which from the different treatment given to creditors’ and
our initial hypotheses are confirmed or not. The final section investors’ rights that resulted from the Calvinist Reform in
draws our major conclusions and suggests some directions the XVI Century.6 What distinguishes Catholic social thought7
for future research. from the Protestant culture is that it does not regard pri-
vate property and its economic benefits as absolute goods,
Literature review and hypotheses
development 2 By focusing on religion, we overcome the causality issue (from
culture to economics and vice versa), as ‘‘religion is a cultural

Religion and corporate risk-taking dimension inherited by individuals from previous generations rather
than voluntarily accumulated’’ (Guiso et al., 2006).
As Kanagaretnam et al. (2015) state, the culture of a nation 3 The risk preference theory (Miller and Hoffman, 1995) supports
involves many dimensions such as language, education, eth- these findings, as not belonging to a religion is an inherently risky
nic background and religion. In fact, religion should be choice.
4 The current trend in sociology of religion is toward a rational
considered more ancient than other cultural values (Guiso
et al., 2004, 2008). Thus, a country’s principal religion back- choice perspective of religious behavior, i.e., religious believes and
behavior are viewed as the outcome rational calculations of per-
ground becomes a relevant dimension to be considered in
ceived costs and rewards (Iyer, 2016; Stark and Bainbridge, 1987)
the study of a company’s decision-making process.1 5 Catholic and Orthodox are treated equally as their differences
lay more on theological aspects rather than economic ones.

6 For a deeper analysis on the importance of religion in the eco-
1 Considering that religion is a key component of a system of nomic development of countries, see Stulz and Williamson (2003).
beliefs, it has been previously used as a common proxy for national 7 For a widely explanation of the Catholic social thought, we rec-
culture (Stulz and Williamson, 2003; La Porta et al., 1999). ommend reading the Pope Leo’s XII encyclical Rerum Novarum.
38 J.M. Díez-Esteban et al.
but as those that are subject to the good of society. In fact, Lastly, we must refer to Eastern religions. Although
this view holds that those responsible for business enter- the importance of Eastern countries in global economy is
prises have an obligation to consider the good of persons and growing (Cohn, 2016), most of previous research has just
not only the increase in profits (Bainbridge, 2002).8 On the adapted Western postulates to those countries, forgetting
contrary, Protestant thought is based on the idea that indi- their cultural and religious particularities.
viduals are responsible for their own actions (Luther’s free Whereas in most Western societies a rejection of main-
exam) and, thus, each individual determines on his own what stream religion’s claims clearly represents a form of
is right. In that sense, the existence of such a common good risk-taking (Adhikari and Agrawal, 2016), in many East-
for Catholics prevents competition in contrast to Protestants ern societies, where Buddhism, Hinduism or Shinto are the
(Stulz and Williamson, 2003). dominant religious traditions, the risk of non-participation
Furthermore, Catholic faith emphasizes private charity, is more unclear. Following Iannaccone’s (1998) and Miller
very useful where society lacks formal structures, such as (1992, 1995), individuals in Eastern societies create their
insurance markets and government welfare programs, to own strategies for obtaining religious goals, and this
deal with individual uncertainty (McCleary and Barro, 2006). often comprises selection and choosing from a diversity
In contrast, Protestant value highly economic success, but of religious and secular philosophies. Additionally, Eastern
charitable acts are downplayed, as going against God’s will.9 religious traditions are grounded on the consequences of
Thus, considering these two different points of view a person’s individual behaviour (e.g. the laws of karma).
among the two main Christian denominations, we postulate Hence, while religions offer a personal guidance on appro-
our first hypothesis. priate behaviour, it is ultimately the behaviour itself and not
the religious affiliation that is supposed to be effective.
Hypothesis 1. There is a positive (negative) relation The three main Eastern religions (Buddhism, Hinduism
between a nationś Protestant (Catholic/Orthodox) back- and Shinto) teach that one’s individual actions, not one’s
ground and corporate risk-taking. religious affiliation or the performance of prescribed rituals,
are what ultimately determine the quality of one’s afterlife
Although, like Christianity, Islam emphasizes exclusivity, reward. Consequently, in such a cultural setting, becoming
claiming that there is only one way to heaven, there is a big part of a specific religion becomes irrelevant and, there-
difference between the two major religions in the world: fore, does not necessarily constitute a specific risk-taking
the link between faith and reason. Indeed, Islam is legal- behaviour.
istic, stressing the fulfilment of laws that are communally Accordingly, our third hypothesis is the following:
enforced (Michalopoulos et al., 2016) rather than a faith
that should be subject to thought. Islamic belief has clear Hypothesis 3. There is no relation between a nation’s Bud-
guidelines, based on the five pillars of faith, for improving dhist, Hindu or Shinto background and corporate risk-taking.
a person’s chances of an after-death life (Voll, 1982), and
a good example can be find in the characteristics of Islamic
banking (Alnahas et al., 2017). Two of the fundamental prin-
ciples of Islamic banking are the sharing of profit and loss and Shareholder national heterogeneity and
the prohibition of the collection and payment of interest by monitoring effectiveness
lenders and investors (Iqbal and Molyneus, 2005).
From an Islamic perspective of corporate governance Agency theory states that companies’ assumption of risk may
practices, previous literature has detected the so-called be conditioned by the different shareholders, managers or
‘‘negative Islamic effect’’: a common, systematic and undi- creditors’ attitudes towards risk (Kubicek et al., 2013), as
versifiable risk factor that negatively affects cross-sectional well as the possibility of obtaining private benefits associ-
expected returns in Islamic stock returns in Saudi Arabia ated with the adoption of such risk.
(Merdad et al., 2015). Besides, Chee-Wooi and Ali (2017) In the previous section we have developed a set of
show that firms with Muslim CEOs have weaker performance. hypotheses considering that the dominant religion of a coun-
In such a context, promoting riskier decision would be try is likely to influence the financial decisions of companies
seen as not contributing to the welfare of the society, in that country (Salaber, 2013). Developing deeper that
given that what directs financial decisions is not value cre- statement and following the social identity and social impact
ation, but to follow the guidelines for appropriate religious theory (Hogg and Abrams, 1988), we take the view that
behaviour (Voll, 1982). ‘‘the predominant local religion could influence local cultu-
Hence, our second hypothesis is defined as follows: ral values and norms and consequently affect the financial
and economic decisions of individuals located in that region,
Hypothesis 2. There is a negative relation between a even if they do not personally adhere to the dominant local
nationś Islamic background and corporate risk-taking. faith’’ (Kumar et al., 2011). Furthermore, we also take the
perspective that ‘‘individuals have less control over their
culture than over other social capital. They cannot alter
8 Consistent with this idea, Baxamusa and Jalal (2014) find that their ethnicity, race or family history, and only with diffi-
CEOs who identify themselves as Catholics tend to take less risk culty can they change their country or religion. Because of
than those who consider themselves as Protestant. the difficulty of changing culture and its low depreciation
9 The Calvinist idea of predestination is behind this thought. In rate, culture is largely ‘given’ to individuals throughout
fact, such kind of activities should be condemned for promoting their lifetimes’’ (Becker, 1996) Additionally, religious prac-
idleness. tices, even when they respond to economic circumstances,
can change through time only over centuries or even millen- Hypothesis 4. There is a positive relation between nation-
nia (Botticini and Eckstein, 2005). ality heterogeneity among non-main reference shareholders
The reality is that a given shareholder may hold a dis- and corporate risk-taking.
tinct nationality from its country of residence, but given that
religion is difficult to be changed, the only relevant dimen-
Culture and corporate risk-taking
sion to be considered is what refers to national culture.
Thus, and going further than previous work, we consider that
shareholder nationality becomes a relevant dimension to be Culture has been recently the subject of considerable
considered (Sanchirico, 2015; Lester, 2015), as there may research in Economics even though with just limited atten-
be a divergence between a certain company’s ‘nationality’ tion in the field of Finance (Karolyi, 2016). As individuals
(commonly related to where its headquarters are located) make economic decisions in the presence of incomplete
and that company’s shareholder national heterogeneity. information and with limited or no previous experience, they
Previous literature has shown considerable differences tend to rely on prior beliefs as a basis for their choices,
in terms of the institutional environment and has also which is to a great extent shaped by culture. Hofstede
revealed that numerous companies in our environment are (1980) considers culture to be a collective programming of
characterized by the simultaneous presence of several large the mind that distinguishes different groups of individuals
shareholders. Indeed, it has been observed that agency while Guiso et al. (2006, p. 23) define culture as ‘‘those cus-
relationships and conflicts between different types of share- tomary beliefs and values that ethnic, religious, and social
holders may be more significant than the separation of groups transmit fairly unchanged from generation to gener-
ownership and control in most companies of the OECD coun- ation’’. Culturally transmitted preferences are thus seen to
tries (Bennedsen and Nielsen, 2010; Konjin et al., 2011; Ruiz a large extent persistent (Giavazzi et al., 2014), a feature
and Santana, 2011). that Guiso et al (2006, p. 24) describe as a ‘‘low depreciation
Firms with a higher concentration of ownership present rate’’ and determined early in an individual’s life. As such,
higher levels of profitability (Hu and Izumida, 2008) and cultural values and culturally transmitted preferences are
achieve higher productivity (Claessens and Djankov, 1999). therefore likely to become major influences on individual’s
More specifically, Shleifer and Vishny (1986) argue that large decision-making processes.12
shareholders have the means to direct companies towards Although, similar to Hofstede (1980), other researchers
projects with higher levels of risk and, following Hill and (see for example Parsons and Shils, 1951; Kluckhohn and
Snell (1988), these shareholders may discourage companies Strodtbeck, 1961; Inkeles and Levinson, 1969; Schwartz,
from engaging in unrelated investment strategies. There- 1994, 2004) have also proposed that culture should be rep-
fore, it seems that companies with a reference shareholder resented by more than one dimension, Hostede’s (1980)
assume higher levels of risk and, given the relationship classification has become the most accepted and cited one
between profitability and risk, tend to obtain better results in studies that connect culture with economic phenomena
in the long run (Díez-Esteban et al., 2013; Nguyen, 2011). (Aggarwal et al., 2016, 2014),13 in spite of some criticisms
In this context, the monitoring role10 of other reference (see for instance McSweeney, 2002). Originally proposed on
shareholders becomes relevant and we ask whether that role the basis of the results from an empirical survey under-
is different in terms of national culture.11 Considering that, taken among IBM employees between 1967 and 1973, and
to a greater or lesser extent, other reference shareholders further refined since then, Hofstede (1980) initially found
influence the major shareholder decisions (Attig et al., 2008, systematic differences in national culture that led him to
2009; Laeven and Levine, 2008; Maury and Pajuste, 2005), propose four cultural dimensions, later to be extended to
we analyse whether that monitoring role is more relevant six following a number of additional surveys and refinements
when there is no culture heterogeneity among sharehold- (Hofstede et al, 2010). Of these dimensions, we consider five
ers or, on the contrary, if that role is more difficult to be to be the most relevant for our study, which are described
exercised when other reference shareholders have different as follows14 :
national cultures.
In the presence of heterogeneity between other refe- (i) Power distance: this dimension measures the acceptance
rence shareholders’ nationality, reaching agreements to of hierarchy or power differentials within a society.
monitor the main shareholder may turn out to be more dif- According to Hofstede, high power distance cultures pre-
ficult. In this context, a reference shareholder may have fer strong authority and steep hierarchies in part because
fewer barriers in promoting riskier investments.
Thus, we posit our fourth hypothesis as follows:
12 Following the social legitimation approach discussed by Davidson
(1995), we consider that cultural dimensions should be addressed

not from an aggregate psychological perspective.
10 Although the analysis of the monitoring role of other reference 13 Arosa et al (2014, p. 182) mention that ‘‘despite its pitfalls,
shareholders is not one of our study focus, we have included a vari- Hofstede’s model is still the most accepted and broadly used mea-
able (MONITOR) in the empirical model to test the existence of that sure of culture today and has been validated by a number of recent
role in our sample. studies. No other researcher has been able to develop a model that
11 Obviously, the influence of other reference shareholders in the equals or exceeds his in sample size, methodology, or acceptance
decisions of the main shareholder will be conditioned by the power among academics worldwide’’.
of the latter or, in other words, by their level of participation in 14 Studies in this topic usually consider five or fewer of Hofstede’s
the property. This issue has already been analyzed by Konishi and dimensions. See for example Li et al. (2013), Anderson et al. (2011)
Yasuda (2004). and Ashraf et al. (2016).
they help preserve the existing social order and its greater focus on risks rather than rewards. Companies in col-
related distribution of power. In high power distance cul- lectivistic societies may also give priority to maintaining the
tures, organizations tend to be centralized, with power overall interests of stakeholders (especially large sharehold-
concentrated in a few hands. In low power distance cul- ers and major creditors)15 and thus would not be interested
tures, organizations are more decentralized, there is in encouraging excessive risk-taking.
more consultation in decision making, and independent
action by less-powerful actors is valued and encouraged (iii) Masculinity measures the acceptance of masculine val-
(Li and Harrison, 2008). ues and rigid gender roles in a society and a focus
on work success relative to fostering the wellbeing
One may expect that organizations in high-power dis- of others (Griffin et al., 2017). Masculine cultures
tance societies are more prone to promote corporate favour managerial decisiveness and a performance ori-
risk-taking. In support of this, Anderson and Galinsky (2006) entation, with an emphasis on proactive competitive
suggest and find evidence that possessing power increases behaviour and self-confidence, whereas in feminine
the propensity of individuals to engage in risk. This asser- cultures, a more supportive social orientation prevails,
tion is related to the Approach/Inhibition theory (Gray, accompanied by a strong concern for the preservation
1991; Sutton and Davidson, 1997) which states that when of existing relationships (Li and Harrison, 2008) and a
owning power, individuals have their behavioural approach preference for cooperation instead of competition.
system more active and attend to more reward-laden infor-
mation. As a result, when in the presence of choices to It is likely that societies whose cultures are more femi-
engage in a risky course of action, powerful people will nine are expected to give women a stronger role in society
have a greater focus on the potential payoffs rather than which, given the empirically observable tendency of women
on its risks. In addition, Anderson and Galinsky (2006) also to be more risk-averse than men (see for example Barber and
argue and present evidence that the sense of power can Odean, 2001, in the case of equity investments and Weber
increase optimism in perceiving risks thus conducing to a et al., 2002, for financial risk-taking in general), may also
riskier behaviour. lead to lower risk-taking. Proposed theories to justify such
The perspective that power increases risk-taking may observed relations include biological (Saad and Gill, 2000;
however be at least partly counterbalanced by the fact Buss, 1989) and social-based ones (Anselmi and Law, 1998).16
that powerful individuals may also develop a degree of risk- However, Meier-Pesti and Penz (2008) show that it is mas-
aversion in order to preserve power or to diversify their culinity, rather than gender, that is better able to predict
wealth arising from current power. For instance, Pathan financial risk-taking, while also documenting that masculin-
(2009) shows that a measure of CEO power in banks is associ- ity has a positive impact on risk-taking. Given all the above,
ated with a reduction in corporate risk-taking, a feature the it is reasonable to expect that societies that feature a high
author associates with the desire of these CEOs to reduce score on the masculinity cultural dimension will be more
the risk related to its exposure to un-diversifiable wealth likely to promote corporate risk-taking.
related to the human capital vested in their banks and
dependence on a relatively large fixed salary component.
(iv) Uncertainty avoidance is defined by Hofstede as the
Thus, it is essentially an empirical matter whether the first
‘‘extent to which the members of a culture feel
effect dominates the second or not.
threatened by uncertain or unknown situations’’ or
in other words, as the ‘‘level of a societyś toler-
(ii) Individualism relates to the degree to which people in
ance for ambiguity’’ (Hofstede, 1991). Cultures with
a society are effectively integrated into groups. In soci-
high uncertainty avoidance attempt to mitigate the
eties that are more individualistic, its members tend to
stress associated with uncertainty in favour of safety
have strong ties only to their immediate families and
and security (Hofstede, 2001). Consequently, managers
not to more extended groups and even less so to soci-
in high uncertainty avoidance societies are expected
ety as a whole. The individualism dimension emphasizes
to avoid undertaking excessively innovative or riskier
independence, freedom and individual achievement,
projects.
whereas in contrast collectivism prioritizes the group’s
interests, its preservation and cohesion.
One would anticipate that in environments where a
Breuer et al (2014) observe that overconfidence and over- strong culture of uncertainty avoidance is present, individ-
optimism are related to individualism and find evidence that uals and organizations will have a preference for following
this cultural dimension has a positive impact on risk tak-
ing in household financing decisions. Accordingly, one could
15 Agency theory has traditionally defined two agency problems.
expect that firms in more individualistic countries would
exhibit higher levels of corporate risk-taking due to the Type I issue (shareholder vs. managers) is no longer relevant in
most international companies, where type II problem (major vs.
placement of a higher value to individual accomplishments
minor shareholders) has become more relevant (La Porta et al.,
and to using own judgement more frequently than alterna-
1999, 2000, 2002; Morck et al., 2005). The influence of national
tive group-decision making processes (Kreiser et al., 2010). culture on different corporate shareholder’s structure is analyzed
In contrast, firms in collectivistic countries would be more in a subsequent section of this paper.
focused on preserving the well being of the larger group 16 Notwithstanding, as Anselmi and Law (1998) observe, biological
where individuals belong to and would be more likely to and social factors can interact and so it is difficult in practice to
subject decisions to the consensus of the group, leading to a separate both effects.
established rules and be more comfortable with high lev- et al. (2014) on capital structure decisions, (iii) Ashraf et al.
els of social conformity. In addition, they would be more (2016) on bank risk taking, (iv) Chang and Noorbakhsh (2009)
prone to scrutinize and discourage projects or decisions that and Chen et al (2015) on cash holdings determinants, and (v)
could increase the levels of risk and which are characterised Fidrmuc and Jacob (2010) on the choice of dividend policies.
by highly unpredictable outcomes. In accordance with this, Using a sample of firms from 35 countries, Li et al. (2013)
Mihet (2013) and Li et al. (2013) find that firms in societies show that when analysing the impact of Hofstede’s (1980,
that are seen as highly averse to uncertainty tend to take 2001) and Schwartz’s (1994, 2004) cultural dimensions, they
less risky decisions. Likewise, Graham et al. (2013) show that find that individualism and uncertainty avoidance have a
CEOs with high levels of culturally transmitted uncertainty positive and negative impact, respectively, on the levels of
aversion traits are less likely take riskier decisions like the corporate risk-taking (measured by volatility of earnings and
acquisition of other companies. RandD over assets ratio). In turn, using a sample of US public
companies, Pan et al (2017) conclude that the firm founder‘s
(v) Finally, societies that score low on long-term orientation cultural heritage (proxied by Hofstede’s Uncertainty Avoid-
prefer to maintain time-honoured traditions and norms ance Index) is a driving factor of corporate culture, with
while viewing societal change with suspicion. Those with corresponding risk preferences passed over from generation
a culture which scores high on this dimension, on the to generation through particular selection and promotion
other hand, tend to encourage thrift and efforts in edu- processes.
cation to prepare for the future. In the business context, Antonczyk and Salzmann (2014) look at one specific area
this dimension is often characterised as (short term) of corporate risk-taking, the choice of a capital structure.
normative versus (long-term) pragmatic approaches to The authors look at one particular Hofstede dimension (indi-
corporate behaviour. vidualism) and find evidence that companies in countries
that are seen as being high on the individualism cultural
Lumpkin et al. (2010) define long term orientation as value show a strong level of optimism and overconfidence
‘‘the tendency to prioritize the long-range implications and and this leads to the choice of higher (and riskier) debt ratios
impact of decisions and actions that come to fruition after for their companies. Arosa et al. (2014) utilize a larger set
an extended time period’’ and note that family-run busi- of Hofstede dimensions than Antonczyk and Salzmann (2014)
nesses are likely to be more risk-averse (Naldi et al., 2007) also to analyse capital structure decisions and find that coun-
at the same time that they are also usually seen as more tries with high levels of uncertainty avoidance and also those
long-term oriented. This can be motivated by the family with high power distance tend to have lower debt ratios.
firms’ desire to maintain the status quo and preserve exist- In the context of risk-taking in banking, Ashraf et al.
ing wealth. In addition, as Breton-Miller and Miller’s (2006) (2016) use a sample of banks in 75 countries and, using four
observe, ‘‘long term priorities include good stewardship of Hofstede’s cultural dimensions, find that bank risk taking
aimed at reducing risk’’. Lumpkin et al (2010) further note tends to be higher in countries which possess high individ-
that family firms tend to be more focused on long-term sur- ualism, low uncertainty avoidance, and low power distance
vival rather than on increased profitability or growth, being cultural values.
more likely to refrain from bolder activities that could pro- Regarding the corporate decision to hold cash reserves
vide short-term profitability but with greater risk taking. (which also reflects to a large extent an attitude towards
Adding to this, Gonzalez and André (2014) mention that corporate risk-taking), Chang and Noorbakhsh (2009) show
managerial short-termism is usually considered as one of that, after controlling for governance issues and the level
the main causes for the recent financial crisis and some of of financial development of the country, cultural dimen-
the most relevant recent corporate failures. Aligned with sions also influence the decision to hold cash reserves. The
such view, Hutchinson et al. (2015) show results where authors document that cash holdings tend to be larger in set-
institutional investors may promote short term performance tings of high uncertainty avoidance, high masculinity17 and
over long term value creation when firms are financially longer term orientation. In a later study, Chen et al. (2015)
distressed. Laverty (2004) finds evidence suggesting that report evidence that corporate cash levels are larger when
long term orientation is more related to a firm’s culture individualism is low and uncertainty avoidance is high.
rather than simple managerial myopia. Díez-Esteban et al. Finally, in the context of dividend policy decisions, which
(2016) similarly report evidence consistent with institutional can also contribute to the level of risk-taking by companies,
investors promoting excessive risk-taking by focusing on Fidrmuc and Jacob (2010) show that countries characterised
short-term profitability rather than sustainable long-term by high individualism, low power distance, and low uncer-
value creation, a feature that allegedly helped to foster the tainty avoidance tend to present higher dividend payouts.
2007---08 financial crisis in Europe. The above studies show that, to a large extent, the impli-
Therefore, one can assume that companies in countries cations of Hofstede’s dimensions on corporate risk-taking
with greater long-term orientation will be less interested in decisions are in accordance with the predictions discussed
promoting riskier decisions and will tend to see the future
with a special concern for uncertainty and a stronger com-
mitment to ensure sustainability and long-term survival.
Some of the more relevant exceptions to the rule that 17 Chang and Noorbakhsh (2009) argue that managers in masculine
cultural factors have to a significant extent been ignored societies ‘‘with little oversight from their shareholders, may dis-
in the literature on corporate financial risk-taking (i) are Li perse those cash balances on value-reducing and risky investment
et al. (2013) and Pan et al. (2017) on corporate risk taking projects’’, which is consistent with our hypothesis that masculinity
in general, (ii) Antonczyk and Salzmann (2014) and Arosa traits are associated with higher risk taking.
Table 1 Composition of the sample by countries.

Country # Firms # Observations Country # Firms # Observations
Argentina 19 100 Korea 418 2,544
Australia 175 1,117 Luxembourg 14 84
Austria 25 171 Malaysia 95 545
Belgium 42 266 Mexico 67 400
Brazil 135 752 Netherlands 44 267
Canada 268 1,652 New Zealand 33 202
Chile 72 455 Norway 38 232
Colombia 18 92 Peru 24 146
Denmark 24 144 Poland 42 267
Finland 44 289 Portugal 11 73
France 115 729 Singapore 78 475
Germany 112 706 South Africa 81 499
Greece 17 118 Spain 41 250
India 359 2,091 Sweden 98 617
Indonesia 65 340 Switzerland 59 373
Ireland 27 168 Thailand 97 583
Israel 53 330 United Kingdom 209 1347
Italy 62 404 United States 1463 9,032
Japan 1028 6391 Total 5572 34,251
Source: Thomson One Banker database.
above. This leads us to expect the following null hypothesis Variables description and empirical framework
regarding the influence of culture on risk-taking:
Corporate risk-taking
We use two alternative measures of corporate risk-taking in
Hypothesis 5. Companies strongly influenced by cultures our tests. Firstly, and consistent with previous literature, we
with high levels of (H1a) Power Distance, (H1b) Masculinity, use a measure of risk related to firms’ shares (Ignatowski and
(H1c) Individualism, and/or low levels of (H1d) Long Term Korte, 2014; Huang et al. 2013; Nguyen, 2011; John et al.,
Orientation or (H1e) Uncertainty Avoidance are expected to 2008; Konishi and Yasuda, 2004). Specifically, we assume
engage in more risk-taking activities. that a firm’s risk is associated with the variance of daily
returns.18 Consequently, we define the variable corporate
market risk (MRISK) as the standard deviations of daily stock
However, one can also conclude that in the existing stud- returns of the company for each year.
ies focusing on the relationship between national culture Second, we use the Z-score (ZS), a measure of firm dis-
and corporate risk-taking the religious dimension is notori- tress and distance from insolvency. It consists of an index of
ously absent. accounting measures of profitability, leverage and volatility,
and it is calculated as the sum of the return on assets ratio
plus the capital asset ratio divided by the standard devi-
ation of the return on assets ratio over the entire sample
Data and methodology period. The Z-score indicates the number of standard devia-
tions that a firm’s losses (negative profits) can increase until
Sample and data sources these deplete equity, making the firm insolvent (De Nicolò,
2000). Thus, the higher the Z-score the more stable and the
Our sample consists of 5572 companies from 37 countries more financially healthy the firm will be.
for the period 2007---2015, with a total of 34,251 firm-
year observations. We have obtained accounting data from National culture
financial statements (balance sheet and profit and loss National culture is measured using five of the six
statements), while corporate ownership structure and share Hofstede cultural dimensions: power distance (PD), indi-
prices of the firms were taken from THOMSON ONE BANKER vidualism (IND), masculinity (MAS), uncertainty avoidance
database. Table 1 provides a summary of the sample by (UA) and long-term orientation (LTO). All five variables
country. range from 0 to 100, with higher scores indicating a
The difficulty in obtaining data on the ownership
structure prevents the analysis of all listed companies.
This selection represents companies --- financial firms are 18 We calculate the shareholder returns through the formula
excluded because of their special accounting practices- of Ri = (Pt − Pi )/Pi , Pt being the share price at the end of the day and
all kinds of sizes and from a diverse set of countries, to Pi the initial price. If a share was not listed on any given day, we
encompass different cultural and religion backgrounds. exclude the data from that day to calculate risk.
Table 2 Summary of hypothesis.

Hypothesis Description Explanatory variable Expected signs
H1 Christian religions influence corporate PROT/CATH (+) (PROT), (−) (CATH)
risk-taking
H2 Islam reduces corporate risk-taking ISLAM (−)
H3 Eastern religions do not influence corporate EASTERN No significant
risk-taking
H4 Shareholders’ national heterogeneity OWN1-OWN5/NATHET (+) (OWN1-OWN5)/NATHET
reduces monitoring effectiveness,
increasing corporate risk-taking
H5a Power distance reduces corporate PD (+)
risk-taking
H5b Individualism increases corporate IND (+)
risk-taking
H5c Masculinity increases corporate risk-taking MAS (+)
H5d Uncertainty avoidance reduces corporate UAI (−)
risk-taking
H5e Long-term orientation increases corporate LTO (−)
risk-taking
The table shows the summary of the hypothesis (dependent variable: MRISK or ZS) and the expected signs. See Table A1 for variable
definitions.
greater influence of a certain variable in a certain coun- A higher value of this index means that other reference
try. Data for the six dimensions has been obtained from shareholder can play an active monitoring role.21
https://www.hofstede-insights.com. Finally, the variable (NATHET) measures the national-
ity heterogeneity among the second to the fifth reference
Religious background shareholder. It is calculated as the number of different
We have considered four possible groups of religions: nationalities among those shareholders.
Catholicism/Orthodox, Protestantism, Islamic and Eastern
religions (Buddhism, Hindu and Shinto). Following prior Control variables
research on this field, data has been obtained from the 2013 Firstly, following Bruno and Shong Shing (2014), we have
World Fact Book (Stulz and Williamson, 2003). Accordingly, included GDP growth (GDP) which captures each coun-
(CATH) is a dummy variable that takes the value of 1 for try’s overall growth. It is assumed that higher country-level
Catholic countries and 0 otherwise; (PROT) is a dummy vari- growth should be associated with higher earnings volatility,
able that takes the value of 1 for Protestant countries and and thus higher risk. We have obtained GDP data from the
0 otherwise; (EASTERN) is a dummy variable that takes the World Bank database.
value of 1 for Eastern religion countries and 0 otherwise. Second, we determine a country’s market or bank ori-
Finally, (ISLAM) is a dummy variable that takes the value of entated financial markets, to measure to what extent
1 for Islamic countries and 0 otherwise. companies in a country are dependent on financial insti-
tutions financing which may play a moderating role in risk
Ownership and monitoring decisions (Tsai and Luan, 2016). Thus, we use market orien-
The power of the main shareholder is measured through the tation (MO) variable as the ratio between domestic credit
proportion of shares held by that shareholder (OWN1)19 and provided by the financial sector and market capitalization
by the first five shareholders (OWN5). The monitoring role of listed companies for each country and year. The higher
of other reference shareholders different from the main one the MO ratio is, the more important banks are relative to
is measured by the variable (MONITOR), calculated as the capital markets. Data on this is taken from the World Bank
sum of the percentage of shares held by the second to fifth indicators.
shareholder divided by the percentage owned by the first.20 Lastly, we include in the analysis the Market-to-book
value of assets ratio (MB), defined as the sum of the equity
market value plus the debt book value divided by the sum of
19 We consider reference shareholder that owner who owns more
the book values of equity and debt, as it is commonly defined
in current research (Maury and Pajuste, 2005; Villalonga and
than 5% of the capital, as this will allow him to influence the board
of directors, appoint managers and intervene in key strategic deci-
sions. Some databases such as Thomson Financial, Marketguide and
WorldVest also make use of this ratio to identify the reference 21 The monitoring role of other reference shareholders is not con-
shareholders. clusive (Díez-Esteban et al., 2013). We expect a negative sign of

20 This index measures the effectiveness of the monitoring role: (MONITOR) variable, given the post-crisis period of our sample (Hill
PP5 = (P2 + P3 + P4 + P5 )/P1 , being Pi the proportion of shares owned and Snell, 1988; Smith and Watts, 1992; Jensen and Meckling, 1976;
by each of the first five reference shareholders. Campbell et al., 2001).
Table 3 Descriptive statistics.

Variable Mean Std. Dev Median Minimum Maximum
MRISK 0.027 0.026 0.024 0.001 0.979
ZS 16.457 20.761 9.813 −18.451 199.831
PD 3.864 0.333 3.892 2.398 4.605
IND 4.001 0.539 4.204 2.565 4.511
MAS 4.028 0.492 4.127 1.609 4.554
UA 4.049 0.420 3.912 2.079 4.605
LTO 3.830 0.527 3.806 2.565 4.605
CATH 0.197 0.398 0.000 0.000 1.000
PROT 0.414 0.492 0.000 0.000 1.000
EASTERN 0.353 0.477 0.000 0.000 1.000
ISLAM 0.025 0.158 0.000 0.000 1.000
OWN1 0.221 0.186 0.146 0.005 0.946
OWN5 0.427 0.205 0.389 0.006 0.985
MONITOR 1.578 0.987 1.595 0.000 4.000
NATHET 1.812 0.976 2.000 1.000 5.000
GDP 0.018 0.029 0.020 −0.091 0.152
MO 0.024 0.017 0.020 0.000 0.133
MB 2.443 3.289 1.557 0.006 77.928
LEV 0.497 0.198 0.509 0.022 0.950
LNA 21.141 1.686 21.048 14.503 27.405
The table shows the mean, standard deviation, median, minimum, and maximum values of the model variables. See Table A1 for variable
definitions.
Amit, 2006). The rationale is that the higher the market-to- vious section, while showing also the expected signs of the
book ratio is, the lower the value attached to the assets in relationship between our measures of risk-taking and the
place and, in turn, the higher the value related to growth different explanatory variables.
opportunities and also the higher will be corresponding cor-
porate risk levels. We also control for the firms’ capital
structure (LEV), measured as the financial leverage ratio Empirical method
(i.e., debt-to-equity ratio). To account for firm size, we
calculate the log of total assets (LNA). We first report descriptive statistics to show the main char-
Because different industries face different risk levels, acteristics of our sample and to examine the consistency
we have also included appropriate sectorial dummies (see of our data with the results of previous research. This step
Table A2). Thus, our model includes industry dummies and provides preliminary evidence about a possible differential
year dummies (INDUSTRY and YEAR, respectively). All con- impact of cultural variables as well as religious background
trol variables are measured for each firm in each year. on corporate risk-taking.
See the Appendix A for a summary table with the defini- Second, we test our hypotheses through an empiri-
tion of all the variables. cal analysis to validate the relation between corporate
RISKi,t = ˇ0 + ˇ1 PDi,t ˇ2 INDi,t + ˇ3 MASi,t + ˇ4 UAIi,t risk-taking, national culture, religious background and the
effectiveness of the monitoring role among shareholders.
+ˇ5 LTOi,t + ˇ6 CATHi,t Our database combines time series with cross-sectional
+ˇ7 PROTi,t + ˇ8 EASTERNi,t + ˇ9 ISLAMi,t data, allowing the formation of panel data, which we
estimate with the appropriate panel data methodology
+ˇ10 OWN1i,t + ˇ11 OWN5i,t
(1) (Arellano and Bond, 1991; Arellano and Bover, 1990; Bond,
+ˇ12 MONITORi,t + ˇ13 NATHETi,t + ˇ14 GDPi,t 2002). Using this technique has two advantages. First, we
+ˇ15 MOi,t + ˇ16 MBi,t can control the so-called constant unobserved heterogene-
ity, that refers to specific characteristics of each firm
+ˇ17 LEVi,t + ˇ18 LNAi,t + INDUSTRY that remain constant over time as represented by the
+YEAR + i + εi,t fixed-effects term i . Second, we can treat the possible
endogeneity of the variables by using a generalized method
where i denotes the firm, t the time period, i is the fixed- of moments (GMM). We use system estimator, an enhanced
effects term of each firm or unobservable and constant version of the estimator GMM in which variable differences
heterogeneity, and εi ,t is the stochastic error used to intro- are also used as instruments in levels by equations (Blundell
duce possible errors in measurement of the independent and Bond, 2000; Blundell et al., 2000; Bond 2002).
variables and the omission of explanatory variables. The consistency of the GMM estimators depends on the
Considering the above definition of variables, Table 2 absence of a second order serial correlation in the error
presents a summary of the hypothesis explained in the pre- term and the validity of the instruments. For this reason,
Influence of religion and culture on corporate risk-taking
Table 4 Correlation matrix.
Variable MRISK ZS PD IND MAS UA LTO CATH PROT ISLAM EASTERN OWN1 OWN5
ZS −0.081***
PD −0.024*** −0.034***
IND 0.018*** 0.050*** −0.618***
MAS −0.032*** 0.114*** 0.124** 0.128***
UA −0.033*** 0.087*** 0.192*** −0.314*** 0.296***
LTO −0.035*** 0.065*** 0.347*** −0.546*** 0.112*** 0.455***
CATH −0.025*** −0.052*** 0.075*** −0.006 −0.135*** 0.198*** −0.072***
PROT 0.060*** −0.005 −0.472*** 0.377*** −0.225*** −0.332*** −0.422*** −0.479***
ISLAM −0.016*** −0.029*** 0.315*** −0.298*** −0.049*** −0.137*** 0.013** −0.080*** −0.137***
EASTERN −0.020*** 0.033*** 0.508*** −0.606*** 0.255*** 0.348*** 0.732*** −0.366*** −0.621*** −0.120***
OWN1 −0.011** −0.066*** 0.248*** −0.318*** −0.151*** 0.010* 0.091** 0.183*** −0.239*** 0.155*** 0.063***
OWN5 0.006 −0.071*** 0.225*** −0.261*** −0.178*** −0.052*** 0.003 0.133*** −0.138*** 0.131*** 0.014*** 0.860***
MONITOR −0.006 0.069*** −0.219*** 0.345*** 0.153*** −0.081*** −0.186*** −0.192*** 0.253*** −0.134*** −0.116*** −0.781*** −0.544***
NATHET 0.001 −0.042*** 0.058*** −0.152*** −0.141*** 0.008 0.212*** 0.292*** −0.234*** 0.074*** −0.033*** 0.269*** 0.147***
GDP −0.068*** −0.101*** 0.270*** −0.268*** −0.141*** −0.205*** −0.083*** −0.030*** −0.061*** 0.190*** 0.099*** 0.202*** 0.219***
MO −0.008 0.142*** 0.023*** −0.040*** 0.442*** 0.536*** 0.436*** −0.113*** −0.286*** −0.138*** 0.384*** −0.100*** −0.160***
MB 0.004 −0.012** −0.069*** 0.145*** −0.048*** −0.178*** −0.191*** −0.052*** 0.145*** 0.010* −0.149*** 0.015*** 0.041***
LEV 0.004 0.021*** −0.018*** 0.055*** −0.015*** 0.005 0.020*** 0.032*** −0.025*** −0.055*** −0.022*** −0.012** −0.037***
LNA −0.129*** 0.104*** −0.078*** 0.130*** −0.040*** 0.029*** −0.008 0.098*** 0.002 −0.043*** −0.129*** −0.083*** −0.181***
Variable MONITOR NATHET GDP MO MB LEV

NATHET −0.314***
GDP −0.148*** 0.031***
MO 0.055*** −0.022*** −0.414***
MB 0.025*** −0.027*** 0.085*** −0.177***
LEV −0.003 0.005 −0.035*** 0.065*** 0.083***
LNA 0.072*** 0.120*** −0.116*** 0.095*** −0.176*** 0.378***
*Significant at 90% confidence level; **Significant at 95%; ***Significant at 99%. See Table A1 for variable definitions.
45
Table 5 Main descriptive statistics.

Panel 5A
Catholic Non-Catholic Test t Protestant Non-Protestant Test t

MRISK 0.026 0.027 4.49*** 0.028 0.026 −8.06***
ZS 14.220 17.004 9.86*** 17.016 16.061 −4.18***
PD 3.918 3.850 −15.11*** 3.632 4.027 133.14***
IND 3.993 4.011 2.46 4.453 3.692 −100.08***
MAS 3.895 4.061 24.98*** 3.960 4.076 21.74***
UA 4.226 4.005 −39.44*** 3.810 4.218 100.65***
LTO 3.753 3.848 13.31*** 3.422 4.118 158.25***
OWN1 0.292 0.204 −35.45*** 0.158 0.265 54.63***
OWN5 0.483 0.413 −25.62*** 0.382 0.458 34.42***
MONITOR 1.184 1.675 37.34*** 1.967 1.303 −64.91***
NATHET 2.401 1.668 −57.94*** 1.535 2.009 45.56***
GDP 0.015 0.019 9.66*** 0.014 0.021 23.85***
MO 0.021 0.025 17.10*** 0.019 0.027 46.04***
MB 2.082 2.532 10.09*** 3.165 1.932 −34.74***
LEV 0.510 0.493 −6.39*** 0.498 0.495 −1.54
LNA 21.482 21.056 −18.68*** 21.270 21.049 −11.96***
Panel 5B
Islamic Non-Islamic Test t Eastern Non-Eastern Test t

MRISK 0.024 0.027 2.96*** 0.026 0.027 3.82***
ZS 12.748 16.556 5.38*** 17.404 15.940 −6.23***
PD 4.509 3.846 −61.60*** 4.093 3.738 −100.01***
IND 3.020 4.034 57.88*** 3.565 4.249 141.09***
MAS 3.879 4.032 9.09*** 4.198 3.935 −48.86***
UA 3.694 4.058 25.71*** 4.247 3.941 −68.72***
LTO 3.872 3.828 −2.42** 4.353 3.544 −200.00***
OWN1 0.399 0.216 −29.17*** 0.237 0.212 −11.760***
OWN5 0.592 0.422 −24.56*** 0.431 0.424 −2.73***
MONITOR 0.764 1.599 25.05*** 1.423 1.662 21.61***
NATHET 2.276 1.800 −14.37*** 1.767 1.837 6.27***
GDP 0.053 0.017 −35.80*** 0.022 0.016 −18.49***
MO 0.009 0.024 25.86*** 0.033 0.019 −77.17***
MB 2.657 2.437 −1.95* 1.779 2.805 27.88***
LEV 0.429 0.498 10.32*** 0.490 0.500 4.17***
LNA 20.694 21.152 7.97*** 20.844 21.302 24.16***
Mean values by religious and legal background. The t-value test is the maximum level of significance to reject the null hypothesis of
equality of means between both subsamples. *Significant at 90% confidence level; **Significant at 95%; ***Significant at 99%. See Table A1
for variable definitions.
in Tables 6---9 we present the model specification tests. The used. The data shows that there is a large variability
validity of the instruments is assessed through the Hansen of values for our risk-taking variables (either for MRISK
test of over-identifying restrictions that evaluates the joint or ZS). Also, the percentage of observations categorised
validity of the selected instruments. We also perform a as being from a Catholic environment is 19.7%, while
test (AR2) to verify that the error terms in the regressions Protestant beliefs encompass 41.4% of the sample, East-
do not present a second-order serial correlation, since the ern religions total 35.3% and Islam corresponds to 2.5%.
definition of the model makes the existence of first-order Regarding ownership variables, the main shareholders own
correlation very likely. on average 22.2% of each firm, while the other refe-
rence shareholders maintain, on average, 42.7% of the
Results shares, what makes the study of their monitoring role
relevant.
Table 4 shows the correlations between the variables for
Descriptive statistics
the whole sample. Both risk measurements (MRISK and ZS)
reveal statically significant relationships with the cultural,
To characterize the sample under analysis, we present
religious and ownership variables considered. Nevertheless,
in Table 3 the descriptive statistics of the variables
Table 6 Results of the estimation of model 1.

Corporate Market Risk (MRISK) Z-score (ZS)
Cultural Religious Ownership Cultural Religious Ownership

dimen- back- structure dimen- back- structure
sions ground sions ground
MRISK/ZS (t−1) 0.419*** 0.275* 0.344** 0.434*** 0.621*** 0.546***
(0.134) (0.147) (0.142) (0.193) (0.150) (0.134)
PD 0.003* 17.682***
(0.002) (6.146)
IND 0.008*** 35.399***
(0.002) (8.566)
MAS 0.002** 12.819***
(0.001) (3.577)
UAI −0.003* −48.936***
(0.002) (13.327)
LTO 0.002* 4.157*
(0.002) (6.216)
CATH −0.001* −0.864*
(0.005) (3.940)
PROT 0.012** 6.991*
(0.005) (3.777)
ISLAM −0.006* −2.232*
(0.007) (4.320)
EASTERN 0.003 4.408
(0.005) (4.035)
OWN1 0.015* 123.723*
(0.008) (68.097)
OWN5 0.022*** 91.085*
(0.007) (54.263)
MONITOR −0.002*** −5.901*
(0.001) (3.434)
NATHET 0.001*** 0.535**
(0.001) (0.662)
GDP −0.061* −0.010* −0.010* −116.393*** −118.375*** −100.713*
(0.092) (0.099) (0.101) (25.743) (25.865) (60.862)
MO −0.058* −0.272** −0.259** −112.012* 61.909 −1.800**
(0.063) (0.133) (0.128) (133.845) (109.682) (64.466)
MB −0.006*** −0.007*** −0.007*** 0.114 −0.300** 0.550
(0.001) (0.001) (0.001) (0.446) (0.349) (0.412)
LEV −0.005 0.001* −0.002 11.372** 32.035* −44.067
(0.011) (0.013) (0.011) (11.015) (19.291) (22.329)
LNA −0.009*** −0.013*** −0.012*** −5.507* −4.046* −1.651*
(0.001) (0.002) (0.002) (2.752) (2.084) (1.333)
INDUSTRY YES YES YES YES YES YES
YEAR YES YES YES YES YES YES
Constant 0.242** 0.315*** 0.301*** −187.391*** 76.919 −11.605
(0.078) (0.101) (0.051) (63.394) (50.954) (24.214)
Wald test (g.l.) 1,161.59*** (21) 1,056.55*** (20) 1,072.88*** (20) 505.14*** (21) 720.19*** (20) 389.37*** (20)
m1 −4.86*** −4.66*** −4.84*** −3.28*** −5.14*** −5.37***
m2 1.02 0.14 0.30 −1.38 −0.91 −1.25
Hansen test (g.l.) 33.18 (27) 27.53 (26) 33.65 (27) 32.72 (27) 36.38 (27) 29.42 (27)
the high correlation among them suggests further analysis icant differences in terms of market risk (MRISK) between
separately. Protestant countries (0.028) and the rest (0.026 and 0.024
In Table 5, we present the mean values of all con- for Catholic and Eastern ones and Islamic, respectively),
sidered variables in our study depending on the religious whereas regarding Z-score values, Eastern religion countries
background. Firstly, it is remarkable that there are signif- are those that present the highest values.
Table 7 Results of the estimation of model 1 for religious background and cultural dimensions (MRISK).
CATHOLIC PROTESTANT ISLAMIC EASTERN

MRISK (t−1) 0.100** 0.516*** MRISK (t−1) 0.083** 0.239***
(0.154) (0.099) (0.140) (0.049)
PD 0.063*** 0.017* PD (ORT) −0.244 0.273**
(0.021) (0.010) (0.277) (0.135)
IND 0.057** 0.223 IND (ORT) −0.174 0.091**
(0.023) (0.048) (0.135) (0.044)
MAS 0.019** −0.033 MAS (ORT) −0.074 0.035*
(0.007) (0.006) (0.080) (0.019)
UAI −0.172*** 0.180 UAI (ORT) −0.264** −0.014*
(0.061) (0.037) (0.116) (0.011)
LTO 0.073* 0.042*** LTO (ORT) −0.068 0.018*
(0.026) (0.010) (0.088) (0.014)
GDP −0.106* −0.125* GDP −0.070** −0.294***
(0.058) (0.178) (0.299) (0.157
MO −0.378*** −0.948*** MO −3.750* −0.454**
(0.113) (0.207) (2.165) (0.232)
MB −0.001** −0.001* MB −0.001*** −0.001**
(0.001) (0.001) (0.001) (0.001)
LEV −0.025 0.009** LEV 0.007 0.019**
(0.017) (0.007) (0.018) (0.008)
LNA −0.005* −0.002** LNA −0.003 −0.002**
(0.002) (0.001) (0.001) (0.002)
INDUSTRY YES YES INDUSTRY YES YES
YEAR YES YES YEAR YES YES
Constant 0.499*** −1.563*** Constant 3.267 1.319**
(0.180) (0.351) (2.506) (0.624)
Wald test (g.l.) 879.78*** (21) 2263.94*** (21) Wald test (g.l.) 341.58*** (21) 828.64*** (21)
m1 −2.17** −3.83*** m1 −3.03*** −2.57
m2 0.01 1.48 m2 1.58 0.72
Hansen test (g.l.) 36.87 (27) 33.58 (27) Hansen test (g.l.) 30.32 (27) 28.35 (27)
We must also note that cultural variables also present Regarding results reported in columns 1 and 4 (cultural
statistically significant differences, specially between dimensions), we observe that PD, IND, MAS and LTO coeffi-
Protestant countries and the rest. This different pattern cients are positive and statistically significant. In high power
anticipates a possible different influence of cultural varia- distance countries, those where individualism and masculin-
bles on corporate risk-taking, depending on the religious ity are characteristics of such societies, and those which are
background more focused on the long-term, companies tend to promote
Regarding ownership variables, there is a higher owner- corporate risk-taking. On the contrary, UAI coefficients sug-
ship concentration (OWN1) in Catholic, Islamic and Eastern gest a negative and statistically significant relation between
countries than in the Protestant ones, and these values UAI and risk. This result confirms that firms in countries with
are observable when the first five shareholders (OWN5) a higher UAI score tend to take less risk. These results to a
are considered. Accordingly, the monitoring role (MONI- large extent confirm our fifth set of hypotheses and pre-
TOR) shows also differences. Lastly, national heterogeneity vious research except for the long-term orientation (LTO)
among the reference shareholders (NATHET) is more rele- trait which shows an opposite sign. A potential explanation
vant in Catholic countries than in the other ones. for this last result can be the mediating role of innovation.
In other words, if long term orientation is associated with
greater a tendency for fostering innovation (Lumpkin et al,
2010), this effect can potentially compensate the risk-taking
Multivariate analysis moderating impact of LTO.
Columns 2 and 5 present results for a country’s reli-
Our major empirical analysis draws on the results of the gious background and its influence on corporate risk-taking.
descriptive analysis. Tables 6---9 report the results from the We observe a significant positive relationship between risk
estimation of Eq. (1). As reported in the previous section, and a Protestant background, a significant negative one for
the high correlation among the variables that describe the Catholic and Islamic but not significant for Eastern ones. This
three considered dimensions (culture, religion and owner- finding, in line with our set of hypotheses related to religion
ship structure) leads us to analyse them separately. (H1 to H3), is consistent with the view that the Protes-
Table 8 Results of the estimation of model 1 for religious background and ownership structure (MRISK).
CATHOLIC PROTESTANT ISLAMIC EASTERN

MRISK (t−1) 0.131** 0.135* 0.090** 0.253***
(0.224) (0.156) (0.144) (0.056)
OWN1 0.167** 0.122** 0.042* 0.248**
(0.081) (0.054) (0.025) (0.115)
OWN5 0.067* 0.079*** 0.034* 0.128**
(0.036) (0.025) (0.018) (0.063)
MONITOR 0.029 −0.006** −0.004 −0.019**
(0.014) (0.004) (0.003) (0.009)
NATHET 0.001 −0.001 0.004*** 0.006***
(0.001) (0.007) (0.001) (0.001)
GDP −0.108** −0.194* 0.279*** −0.063***
(0.079) (0.100) (0.096) (0.023)
MO −0.186** −0.589*** −0.092** −0.078*
(0.184) (0.107) (0.420) (0.061)
MB −0.001* −0.001** −0.001* −0.005***
(0.001) (0.001) (0.001) (0.001)
LEV −0.003 0.001** 0.034** 0.054**
(0.027) (0.003) (0.022) (0.018)
LNA −0.007* −0.005*** −0.003* −0.015***
(0.004) (0.001) (0.001) (0.002)
INDUSTRY YES YES YES YES
YEAR YES YES YES YES
Constant 0.274** 0.105*** 0.034 0.303***
(0.130) (0.027) (0.042) (0.073)
Wald test (g.l.) 359.92*** (20) 757.78*** (20) 315.01***(20) 508.68*** (20)
m1 −2.11** −3.42*** −3.77*** −2.62***
m2 −0.30 −0.60 1.08 0.21
Hansen test (g.l.) 28.66 (27) 24.53 (27) 29.63 (27) 31.01 (27)
tant view of business promotes competition and, hence, interaction between both dimensions may be interpreted in
corporate risk-taking, whereas the theological view of busi- the context of corporate risk-taking.
ness and the common good in Catholic and Islamic countries We must note that, in the case of Islamic countries and
makes companies less risk seekers. those with an Eastern religious background, cultural varia-
Ownership structure results are reported in columns 3 bles are totally correlated among them, which causes a
and 6. As expected, ownership concentration (OWN1 and modelling specification problem (collinearity) when using
OWN5) coefficients are positively related to corporate risk- GMM. To overcome this issue, and following prior litera-
taking, confirming the findings of previous studies. Although ture we have orthogonalized22 cultural variables (Salmerón
it is not part of our hypothesis development, we have et al., 2016; Novales et al., 2015).
included in the model the variable MONITOR to measure It is important to note that not all cultural variables are
the monitoring role exercised by the non-main reference relevant when considering different religious backgrounds.
shareholders over the main one. The coefficient of this Only in Catholic and Eastern countries all five considered
variable is negative and statistically significant, consistent cultural variables become statistically significant for cor-
with a monitoring role that reduces corporate risk-taking. porate risk-taking. In Protestant ones, only Power distance
As expected, NATHET is positively related to corporate PD and long-term orientation LTO are significant whereas
risk-taking, because when there is heterogeneity between in Islamic countries the only relevant cultural variable is
the non-main reference shareholders’ nationality, reach- uncertainty avoidance UAI. These different results are in
ing agreements turns out to be more difficult and, hence, line with the fundamentals of our hypothesis. On the one
the main reference shareholder will not have any barriers hand, we find that in Islamic countries what really deter-
in promoting riskier investments. These results confirm the mines risk-taking behaviour is religion, whereas in Catholic
expectations of our fourth hypothesis. ones there is a joint effect of both religion and culture.
One of the key issues addressed by our study is to better On the other hand, in Eastern countries there is a relevant
understand the relationship between national culture and a impact of cultural variables independently of the religious
country’s religious background. Thus, in Table 7 we report
the results of the estimation of our model for the cultural
variables by dividing our sample into four groups, according 22 By using the residuals of the OLS regression of each variable as
to the religious groups considered. Proceeding this way, the independent variables in model 1.
Table 9 Results of the estimation of model 1 by reference shareholder nature.

1st shareholder individual 1st shareholder institutional investor
Cultural Religious Ownership Cultural Religious Ownership

dimen- back- struc- dimen- back- struc-
sion ground ture sions ground ture
MRISK (t−1) 0.267*** 0.157** 0.041** 0.018** 0.167** 0.085**
(0.097) (0.075) (0.077) (0.061) (0.074 (0.094)
PD 0.004* −0.001
(0.002) (0.002)
IND 0.007*** 0.014***
(0.002) (0.004)
MAS −0.001 0.005**
(0.002) (0.002)
UAI −0.003* −0.010**
(0.002) (0.004)
LTO 0.003 0.012***
(0.002) (0.004)
CATH −0.666** 0.168
(0.304) (0.291)
PROT 0.671** 0.098
(0.310) (0.293)
ISLAM −0.678** −0.067
(0.309) (0.547)
EASTERN −0.666** 0.091
(0.305) (0.292)
OWN1 0.055** 0.122*
(0.021) (0.066)
OWN5 0.028* 0.065**
(0.014) (0.032)
MONITOR 0.002 0.009**
(0.002) (0.007)
NATHET 0.009* 0.011***
(0.004) (0.004)
GDP −0.182*** 0.088 −0.179*** −0.279*** −0.240*** −0.194***
(0.066) (0.103) (0.040) (0.073) (0.073) (0.062)
MO −0.130** −0.219** −0.024** −0.602** −0.354** −0.126**
(0.051) (0.104) (0.067) (0.249) (0.140) (0.051)
MB −0.001** 0.001 −0.002** 0.001 −0.001 −0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
LEV 0.035** 0.079*** 0.036** −0.015 −0.008 −0.005
(0.014) (0.026) (0.018) (0.014) (0.015) (0.028)
LNA −0.007*** −0.006*** −0.012*** −0.002*** −0.005** −0.005
(0.002) (0.002) (0.003) (0.001) (0.002) (0.001)
INDUSTRY YES YES YES YES YES YES
YEAR YES YES YES YES YES YES
Constant 0.114** 0.921*** 0.287*** −0.087** 0.023 0.128***
(0.048) (0.351) (0.079) (0.044) (0.267) (0.040)
Wald test (g.l.) 517.03*** (21) 507.83*** (20) 252.82*** (19) 1174.60*** (21) 1599,91*** (20) 686.82*** (20)
m1 −2.55** −2.45** −2,48** −3.20*** −3.43*** −3.50***
m2 0.49 0.44 1.07 −0.65 0.26 −0.56
Hansen test (g.l.) 35.57 (27) 30.82 (27) 35.93 (27) 30.99 (27) 33.85 (27) 35.49 (27)
background. Lastly, in Protestant ones, corporate risk-taking Although ownership variables present the same pattern
is clearly influenced by two cultural variables. in all environments the effect of the monitoring role is
A similar analysis has been conducted to better explain only relevant in Protestant and Eastern countries. Moreover,
the effect of national heterogeneity among the reference the effect of national heterogeneity among the non-main
shareholders. Results are reported in Table 8. reference shareholders is only significant in Islamic and East-
ern countries, precisely due to the joint influence of both The influence of these cultural dimensions is not relevant
cultural and religious dimensions that may make reaching for all countries but are observed to be important espe-
agreements more difficult. cially in those countries where the religious understanding
Lastly, we try to address a specific issue: Are all kinds of business focuses more on the common good.
of shareholders equally influenced by national culture and To try shedding some light on whether religion and culture
religion? To analyse this question, we identify the identity are separable influences or not on risk taking, we anal-
of the reference shareholder and split the sample into two ysed the impact of cultural dimensions for companies in
groups: one with companies whose first main shareholder is each religion background. While for countries with Catholic
an individual or a family and another with firms whose first and Eastern religions all the cultural traits have a relevant
main shareholder is an institutional investor. Results of the impact over and above the religion factor, in Protestant
estimation are reported in Table 9. countries only the power distance and long-term orientation
Results are very relevant, as we find that the religious dimensions have an additional influence and in Islam-based
background is only relevant for individual shareholders, countries the marginal relevance of cultural traits over the
while cultural dimensions are more relevant for institutional religion factor is only observable for the uncertainty avoid-
investors, except for individualism IND and uncertainty ance trait.
avoidance UAI. In fact, institutional investors, due to their Finally, we analyse if nationality heterogeneity among
characteristics, usually do not ascribe to a certain religion the non-main reference shareholders modifies the monitor-
(i.e., such investors in a sense are ‘‘agnostic’’) but, when ing role under a type II agency perspective. We find that
investing, may be influenced by the cultural environment in such heterogeneity is relevant to corporate risk-taking, but
which was ‘‘created’’. especially so in Islamic and Eastern countries.
We must mention that, as our sample includes countries Our research may have promising implications for prac-
of different sizes, as a robustness check we have redone all titioners, policy makers and academia. Our results are
the analysis by excluding USA companies that represent 26% informative for practitioners about the way national culture
of the total. Moreover, we have also replaced the estimation and religion modulate corporate risk-taking. Considering this
in Tables 7---9 by the Z-score (ZS) risk measurement. The double perspective, policy makers may influence the adop-
results are essentially analogous to those discussed above tion of less risky investment decisions by encouraging the
and are not presented for parsimony reasons. formation of balanced ownership structures in countries
where personal believes are relevant. In the perspective
Conclusions of investors, these should bear in mind that the pres-
ence of large reference shareholders will have an impact
We analyse the relation between national culture, religion, on risk-taking according to whether these are institutional
ownership structure and corporate risk-taking for a sample shareholders and individuals/families and that in the lat-
of large quoted companies from 37 countries for the period ter case their type of religiosity will impact on the riskiness
2007---2015. We specifically examine whether it is valid to of the firm’s activities. Finally, our paper adds to the fertile
consider national culture and religion as determinants of field of academic research on the factors affecting corporate
corporate decisions or, instead, if these are relevant factors risk-taking, especially those which stress the importance of
only in certain environments. qualitative characteristics in explaining corporate decisions.
Leaving cultural dimensions aside, and consistent with
our initial hypothesis, our evidence suggests that certain Acknowledgments
religious environments seem to have either a negative
(Catholic and Islamic) or positive (Protestant) impact on This research has been financed by Portuguese Public Funds
corporate-risk taking. through FCT (Fundação para a Ciência e a Tecnologia) in
On the one hand, we find that national culture, prox- the framework of the project UID/ECO/04105/2013. The
ied by Hofstede cultural dimensions, clearly also influences authors are grateful to Xosé H. Vazquez, Yama Temouri
corporate risk-taking. While power distance, masculinity, and two anonymous referees for their comments on previ-
individualism and long-term orientation exercise a positive ous versions of the paper. All the remaining errors are our
effect, uncertainty avoidance reduces corporate risk-taking. responsibility.
Appendix A.
See Tables A1 and A2 .
Table A1 Definition of variables.

Abbreviations Variable Definition
MRISK Corporate market risk Standard deviation of return on assets
ZS Z Score ROA plus the capital asset ratio to the standard deviation of
the ROA
PD Power distance Power distance index proposed by Hofstede (2001)
IND Individualism Individualism index proposed by Hofstede (2001)
MAS Masculinity Masculinity index proposed by Hofstede (2001)
UA Uncertainty avoidance Uncertainty avoidance index proposed by Hofstede (2001)
LTO Long-term orientation Long-term orientation index proposed by Hofstede (2001)
CATH Catholic Dummy variable that takes 1 for Catholic countries
PROT Protestant Dummy variable that takes 1 for Protestant countries
EASTERN Eastern religion Dummy variable that takes 1 for Eastern religion countries
ISLAM Islamic Dummy variable that takes 1 for Islamic countries
OWN1 First shareholder Proportion of shares owned by the first shareholder
OWN5 First five shareholders Proportion of shares owned the top five shareholders
MONITOR Monitoring role Sum of the percentage of shares held by the second to fifth
shareholder divided by the percentage owned by the first
NATHET Nationality heterogeneity Number of different nationalities among the first five
shareholders
GDP GDP growth GDP data from the World Bank database
MO Market orientation Ratio between domestic credit provided by financial sector
and market capitalization
MB Market to book Equity market value/equity book value
LEV Leverage Total liabilities divided by total equity plus total liabilities
LNA Size of the firm Logarithm of total assets
Table A2 Industry dummy variables.

Division Standard industrial classification (SIC) description
A Agriculture, Forestry and Fishing
B Mining
C Construction
D Manufacturing
E Transportation, Communication, Electric, Gas and Sanitary services
F Wholesale Trade
G Retail trade
H Finance, Insurance and Real estate
I Services
J Public Administration
References Anderson, C.W., Fedenia, M., Hirschey, M., Skiba, H., 2011. Cultu-
ral influences on home bias and international diversification by
Adhikari, B.K., Agrawal, A., 2016. Does local religiosity matter for institutional investors. J. Bank. Finance 35 (4), 916---934.
bank risk-taking? J. Corp. Finance 38, 272---293. Anselmi, D.L., Law, A.L., 1998. Questions of Gender. McGraw-Hill,
Aggarwal, R., Faccio, M., Guedhami, O., Kwok, C.C., 2016. Culture Boston, United States.
and finance: an introduction. J. Corp. Finance (41), 466---474. Antonczyk, R.C., Salzmann, A.J., 2014. Overconfidence and opti-
Alnahas, A.M., Kabir Hassan, M., Ismail, G.M., 2017. Religion and mism: the effect of national culture on capital structure. Res.
mergers and acquisitions contracting: the case of earnout agree- Int. Bus. Finance 31, 132---151.
ments. J. Corp. Finance 42, 221---246. Arellano, M., Bond, S., 1991. Some tests of specification for panel
Anderson, C., Galinsky, A.D., 2006. Power, optimism, and risk- data: Monte Carlo evidence and an application to employment
taking. Eur. J. Soc. Psychol. 36 (4), 511---536. equations. Rev. Econ. Stud. 58 (194), 277---297.
Arellano, M., Bover, O., 1990. La econometría de datos de panel. Claessens, S., Djankov, S., 1999. Ownership concentration and cor-
Investig. Econ. 14 (1), 3---45. porate performance in the Czech Republic. J. Comp. Econ. 27,
Arosa, C.M., Richie, N., Schuhmann, P.W., 2014. The impact of 498---513.
culture on market timing in capital structure choices. Res. Int. Cohn, T.H., 2016. Global Political Economy: Theory and Practice.
Bus. Finance 31, 178---192. Routledge.
Ashraf, B.N., Zheng, G., Arshad, S., 2016. Effects of national culture Coles, J.L., Daniel, N.D., Naveen, L., 2006. Managerial incentives
on bank risk-taking behavior. Res. Int. Bus. Finance 37, 309---326. and risk taking. J. Financ. Econ. 79 (2), 431---468.
Attig, N., El Ghoul, S., Guedhami, O., 2009. Do multiple large share- Conyon, M., Judge, W.Q., Useem, M., 2011. Corporate governance
holders play a corporate governance role? Evidence from East and the 2008---09 financial crisis. Corp. Govern.: Int. Rev. 19 (5),
Asia. J. Financ. Res. 32 (4), 395---442. 399---404.
Attig, N., Guedhami, O., Mishra, D., 2008. Multiple large sharehold- Davidson, P., 1995. Culture, structure and regional levels of
ers, control contests, and implied cost of equity. J. Corp. Finance entrepreneurship. Entrep. Reg. Dev. 7, 41---62.
14, 721---737. De Nicolò, G., 2000. Size, Charter Value and Risk in Banking: An
Bainbridge, S.M., 2002. The bishops and the corporate stakeholder International Perspective. IFC Discussion Paper No. 689. Board
debate. Villanova J. Law Invest. Finance, 3---37. of Governors of the Federal Reserve, Washington, DC.
Barber, B.M., Odean, T., 2001. Boys will be boys: gender, overcon- Díez-Esteban, J.M., Farinha, J.B., García-Gómez, C.D., 2016. The
fidence, and common stock investment. Q. J. Econ. 1616 (1), role of institutional investors in propagating the 2007 financial
261---292. crisis in Southern Europe. Res. Int. Bus. Finance 38, 439---454.
Baxamusa, M., Jalal, A., 2014. Does religion affect capital structure? Díez-Esteban, J.M., García-Gómez, C.D., López-Iturriaga, F.J.,
Res. Int. Bus. Finance 31, 112---131. 2013. International evidence about the influence of large share-
Becker, G., 1996. Preferences and values. In: Becker, G. (Ed.), holders on corporate risk taking. Span. J. Finance Account. 13
Accounting for Taste. Harvard University Press, Cambridge. (31), 126---144.
Bennedsen, M., Nielsen, K.M., 2010. Incentive and entrench- Ekelund, R.B., Hebert, R.F., Tollison, R.D., 2002. An economic
ment effects in European ownership. J. Bank. Finance 34, analysis of the Protestant Reformation. J. Polit. Econ. 110 (3),
2.212---2.229. 646---671.
Berger, A.N., Imbierowicz, B., Rauch, C., 2016. The roles of cor- Erkens, D.H., Hung, M., Matos, P., 2012. Corporate governance in
porate governance in bank failures during the recent financial the 2007---2008 financial crisis: Evidence from financial institu-
crisis. J. Money Credit Bank. 48 (4), 729---770. tions worldwide. J. Corp. Finance 18 (2), 389---411.
Beugelsdijk, S., Maseland, R., 2011. Culture in Economics. Cam- Fidrmuc, J.P., Jacob, M., 2010. Culture, agency costs, and divi-
bridge University Press, Cambridge, United Kingdom. dends. J. Comp. Econ. 38 (3), 321---339.
Blundell, R., Bond, S., 2000. GMM estimation with persistent panel Giavazzi, F., Petkov, I., Schiantarelli, F., 2014. Culture: Persis-
data: an application to production functions. Econom. Rev. 19 tence and evolution (No. w20174). National Bureau of Economic
(3), 321---340. Research.
Blundell, R., Bond, S., Windmeijer, F., 2000. Estimation in dynamic Gonzalez, A., André, P., 2014. Board effectiveness and short ter-
panel data models: improving on the performance of the stan- mism. J. Bus. Finance Account. 41 (1---2), 185---209.
dard GMM estimator. In: Baltagi, T, Fomby, B, Carter Hill, B Graham, J.R., Harvey, C.R., Puri, M., 2013. Managerial attitudes
(Eds.), Nonstationary Panels, Panel Cointegration, and Dynamic and corporate actions. J. Financ. Econ. 109 (1), 103---121.
Panels,. Emerald, Bingley, pp. 53---92. Gray, J.A., 1991. Neural systems, emotion, and personality. In: Mad-
Bond, S., Working Paper 9/02, The Institute for Fiscal Studies 2002. den IV, J. (Ed.), Neurobiology of Learning, Emotion, and Affect.
Dynamic Panel Data Models: A Guide to Micro Data Methods and Raven Press, New York, pp. 273---396.
Practice. Griffin, D.W., Guedhami, O., Kwok, C.C.Y., Li, K., Shao, L., 2017.
Botticini, M., Eckstein, A., 2005. Jewish occupational selection: National culture: the missing country-level determinant of cor-
education, restrictions, or minorities? J. Econ. Hist. 65 (4), porate governance. J. Int. Bus. Stud. 48 (6), 740---762.
922---948. Guiso, L., Sapienza, P., Zingales, L., 2006. Does culture affect eco-
Breton-Miller, L., Miller, D., 2006. Why do some family businesses nomic outcomes? J. Econ. Perspect. 20 (2), 23---48.
out-compete? Governance, long-term orientations, and sustain- Guiso, L., Sapienza, P., Zingales, L., 2004. Cultural Biases in Eco-
able capability. Entrep. Theory Pract. 30 (6), 731---746. nomic Exchange. National Bureau of Economic Research.
Breuer, W., Riesener, M., Salzmann, A.J., 2014. Risk aversion vs. Guiso, L., Sapienza, P., Zingales, L., 2008. Trusting the stock mar-
individualism: what drives risk taking in household finance? Eur. ket. J. Finance 63, 2557---2600.
J. Finance 20 (5), 446---462. Hilary, G., Hui, J.K., 2009. Does religion matter in corporate deci-
Bruno, V., Shong Shing, H., 2014. Globalization of corporate risk- sion making in America? J. Financ. Econ. 93, 455---473.
taking. J. Int. Bus. Stud. (45), 800---820. Hill, C., Snell, S., 1988. Effects of ownership structure and control
Buss, D., 1989. Sex differences in human mate preferences: evolu- on corporate productivity. Acad. Manage. J. (32), 25---46.
tionary hypotheses tested in 37 cultures. Behav. Brain Sci. 12, Hofstede, G., Hofstede, G.J., Minkov, M., 2010. Cultures and Orga-
1---49. nizations: Software of the Mind, Rev. 3rd ed. McGraw-Hill, New
Campbell, J., Lettau, M., Malkiel, B., Xu, Y., 2001. Have individ- York, United States.
ual stocks become more volatile? An empirical exploration of Hofstede, G.H., 1991. Cultures and Organizations: Software of the
idiosyncratic risk. J. Finance 56 (1), 1---43. Mind. McGraw-Hill, Berkshire, England.
Chang, K., Noorbakhsh, A., 2009. Does national culture affect inter- Hofstede, G.H., 1980. Culture’s Consequences: International Dif-
national corporate cash holdings? J. Multinatl. Financ. Manage. ferences in Work-Related Values. SAGE Publications, Thousand
19 (5), 323---342. Oaks, United States.
Chee-Wooi, H., Ali, R., 2017. Does a Muslim CEO matter in Hofstede, G.H., 2001. Culture’s Consequences: Comparing Values,
Shariah-compliant companies? Evidence from Malaysia. Pac. Behaviors, Institutions, and Organizations across Nations, second
Basin Financ. J. 42, 126---141. edition. SAGE Publications, Thousand Oaks, United States.
Chen, Y., Dou, P.Y., Rhee, S.G., Truong, C., Veeraraghavan, M., 2015. Hogg, M.A., Abrams, D., 1988. Social Identifications: A Social Psy-
National culture and corporate cash holdings around the world. chology of Intergroup Relations and Group Processes. Psychology
J. Bank. Finance 50, 1---18. Press, London.
Hu, Y., Izumida, S., 2008. Ownership concentration and corporate Laverty, K.J., 2004. Managerial myopia or systemic short-termism?
performance: a causal analysis with Japanese panel data. Corp. The importance of managerial systems in valuing the long term.
Gov.: Int. Rev. 16 (4), 342---358. Manage. Decis. 42 (8), 949---962.
Huang, B., Wald, J., Martell, R., 2013. Financial market liberaliza- Lester, S., 2015. Rethinking the international investment law sys-
tion and the pricing of idiosyncratic risk. Emerg. Mark. Rev. 17, tem. J. World Trade 49 (2), 211---222.
44---59. Li, J., Harrison, R.J., 2008. National culture and the composition
Hutchinson, M., Seamer, M., Chapple, L.E., 2015. Institutional and leadership structure of boards of directors. Corp. Gov.: Int.
investors, risk/performance and corporate governance. Int. J. Rev. 16 (5), 375---385.
Account. 50 (1), 31---52. Li, K., Griffin, D.W., Yue, H., Zhao, L., 2013. How does culture
Iannaccone, L.R., 1998. Introduction to the economics of religion. influence corporate risk-taking? J. Corp. Finance 23, 1---22.
J. Econ. Lit. 36 (3), 1465---1495. Lumpkin, G.T., Brigham, K.H., Moss, T.W., 2010. Long-term ori-
Ignatowski, M., Korte, J., 2014. Wishful thinking or effective threat? entation: Implications for the entrepreneurial orientation and
Tightening bank resolution regimes and bank risk-taking. J. performance of family businesses. Entrep. Reg. Dev. 22 (3---4),
Financ. Stab. 15, 264---281. 241---264.
Inkeles, A., Levinson, D.J., 1969. National character: the study Maury, B., Pajuste, A., 2005. Multiple large shareholders and firm
of modal personality and sociocultural systems. In: Lindzey, value. J. Bank. Finance 29 (7), 1.813-1.834.
G., Aronson, E. (Eds.), The Handbook of Social Psychology IV. McCleary, R.M., Barro, R.J., 2006. Religion and economy. J. Econ.
McGraw-Hill, New York, United States, pp. 418---506 (first pub- Perspect. 20 (2), 49---72.
lished 1954). McSweeney, B., 2002. Hofstede’s model of national cultural differ-
Iqbal, M., Molyneus, P., 2005. Thirty Years of Islamic Banking. Pal- ences and their consequences: a triumph of faith --- a failure of
grave MacMillan, London, United Kingdom. analysis. Hum. Relat. 55 (1), 89---117.
Iyer, S., 2016. The new economics of religion. J. Econ. Lit. 54 (2), Meier-Pesti, K., Penz, E., 2008. Sex or gender? Expanding the
395---441. sex-based view by introducing masculinity and femininity as
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial predictors of financial risk taking. J. Econ. Psychol. 29 (2),
behaviour, agency costs and ownership structure. J. Financ. 180---196.
Econ. 3 (4), 305---360. Merdad, H., Hassan, M.K., Hippler, W.J., 2015. The Islamic risk fac-
John, K., Litov, L., Yeung, B., 2008. Corporate governance and risk- tor in expected stock returns: an empirical study in Saudi Arabia.
taking. J. Finance 63 (4), 1.679---1.728. Pac. Basin Financ. J. (34), 293---414.
Kanagaretnam, K., Lobo, G.J., Wang, C., Whalen, D.J., 2015. Reli- Michalopoulos, S., Naghavi, A., Prarolo, G., 2016. Islam, inequality
giosity and risk-taking in international banking. J. Behav. Exp. and pre-industrial comparative development. J. Dev. Econ. 120,
Finance 7, 42---59. 86---98.
Karolyi, G.A., 2016. The gravity of culture for finance. J. Corp. Mihet, R., 2013. Effects of culture on firm risk-taking: a cross-
Finance 41, 610---625. country and cross-industry analysis. J. Cult. Econ. 37 (1),
Kempf, A., Ruenzi, S., Thiele, T., 2009. Employment risk, compen- 109---151.
sation incentives, and managerial risk taking: evidence from the Miller, A., Hoffman, J., 1995. Risk and religion, an explanation of
mutual fund industry. J. Financ. Econ. 92 (1), 92---108. gender differences in religiosity. J. Sci. Study Relig. 34, 63---75.
Kluckhohn, F.R., Strodtbeck, F.L., 1961. Variations in Value Orien- Miller, A.S., 1992. Conventional religious behaviour in modern
tations. Greenwood Press, Westport (CT), United States. Japan: a service industry perspective. J. Sci. Study Relig. 31,
Konishi, M., Yasuda, Y., 2004. Factors affecting bank risk-taking: 207---214.
evidence from Japan. Journal of Banking and Finance 28 (1), Miller, A.S., 1995. A rational choice model for religious behaviour in
215---232. Japan. J. Sci. Study Relig. 34, 234---244.
Konjin, S.J.J., Kräussl, R., Lucas, A., 2011. Blockholder dispersion Minkov, M., Hofstede, G., 2014. Nations versus religions: which has
and firm value. J. Corp. Finance 17, 1.330---1.339. a stronger effect on societal values? Manage. Int. Rev. 54 (6),
KPMG, 2009. Never Again? Risk Management in Banking beyond the 801---824.
Credit Crisis. Morck, R., Wolfenzon, D., Yeung, B., 2005. Corporate governance,
Kreiser, P.M., Marino, L.D., Dickson, P., Weaver, K.M., 2010. Cultural economic entrenchment, and growth. J. Econ. Lit. 43 (3),
influences on entrepreneurial orientation: the impact of national 655---720.
culture on risk taking and proactiveness in SMEs. Entrep. Theory Naldi, L., Nordquist, M., Sjoberg, K., Wiklund, J., 2007.
Pract. 34 (5), 959---983. Entrepreneurial orientation, risk taking, and performance in
Kubicek, A., Bhanugopan, R., Fish, A., 2013. Perceiving safety and family firms. Fam. Bus. Rev. 20, 33---58.
risk in culturally diverse organizations: toward a conceptual Nguyen, P., 2011. Corporate governance and risk taking: evidence
model. Risk Manage. 15 (3), 199---223. from Japanese firms. Pac. Basin Finance J. 19 (3), 278---297.
Kumar, A., Page, J.K., Spalt, O.G., 2011. Religious beliefs, gambling Novales, A., Salmerón, R., García, C.B., García, J., López, M.M.,
attitudes, and financial market outcomes. J. Financ. Econ. 102, 2015. Tratamiento de la multicolinealidad aproximada mediante
671---708. variables ortogonales. An. Econ. Apl., 1212---1227.
La Porta, R., López de Silanes, F., Shleifer, A., 1999. Corporate Pan, Y., Siegel, S., Wang, T.Y., 2017. Corporate risk culture. J.
ownership around the world. J. Finance 54 (2), 471---517. Financ. Quant. Anal. 52 (6), 2327---2367.
La Porta, R., López de Silanes, F., Shleifer, A., Vishny, R.W., 2000. Parsons, T., Shils, E.A., 1951. Toward a General Theory of Action.
Investor protection and corporate governance. J. Financ. Econ. Harvard University Press, Cambridge (MA), United States.
58 (1----2), 3---27. Pathan, S., 2009. Strong boards, CEO power and bank risk-taking.
La Porta, R., López de Silanes, F., Shleifer, A., Vishny, R.W., 2002. J. Bank. Finance 33 (7), 1340---1350.
Investor protection and corporate valuation. J. Finance 57 (3), Reuter, C.H., 2011. A survey of ‘culture and finance’. Finance 32
1.147---1.170. (1), 75---152.
Laeven, L., Levine, R., 2008. Complex ownership structures and Ruiz, M.V., Santana, D.J., 2011. Dominant institutional owners and
corporate valuations. Rev. Financ. Stud. 21 (2), 579---604. firm value. J. Bank. Finance 35, 118---129.
Laeven, L., Levine, R., 2009. Bank governance, regulation, and risk Saad, G., Gill, T., 2000. Applications of evolutionary psychology in
taking. J. Financ. Econ. 93 (2), 259---275. Marketing. Psychol. Mark. 17 (12), 1005---1034.
Salaber, J., 2013. Religion and returns in Europe. Eur. J. Polit. Econ. Smith, C., Watts, R., 1992. The investment opportunity set and
32, 149---160. corporate financing, dividend, and compensation policies. J.
Salmerón, R., García, J., García, C.B., García, C., 2016. Treat- Financ. Econ. 32 (3), 263---292.
ment of collinearity through orthogonal regression: an economic Stark, R., Bainbridge, W.S., 1987. A Theory of Religion. Peter Lang,
application. Bol. Estad. Investig. Oper. 32 (3), 184---202. New York.
Sanchirico, C.W., 2015. As American as Apple Inc.: international tax Stulz, R., Williamson, R., 2003. Culture, openness, and finance. J.
and ownership nationality. Tax Law Rev. 68, 207---274. Financ. Econ. 70, 313---349.
Schwartz, S.H., 1994. Beyond Individualism/Collectivism: New Sutton, S.K., Davidson, R.J., 1997. Prefrontal brain asymmetry: a
Cultural Dimensions of Values. Sage Publications, Inc. biological substrate of the behavioral approach and inhibition
Schwartz, S.H., 2004. Mapping and interpreting cultural differences systems. Psychol. Sci. 8 (3), 204---210.
around the world. In: Vinkins, H, Soeters, J, Esters, P (Eds.), Tsai, H.F., Luan, C.J., 2016. What makes firms embrace risks? A
Comparing Cultures: Dimensions of Culture in a Comparative risk-taking capability perspective. Bus. Res. Q. 19, 219---231.
Perspective. Brill Academic Publishers, Boston, MA. Villalonga, B., Amit, R., 2006. How do family ownership, control,
Shleifer, A., Vishny, R., 1986. Large shareholders and corporate con- and management affect firm value? J. Financ. Econ. 80 (2),
trol. J. Polit. Econ. 94, 461---488. 385---417.
Shu, T., Sulaeman, J., Yeung, P.E., University of Georgia Working Voll, J.O., 1982. Islam: Continuity and Change in the Modern World.
Paper 2011. Local religious beliefs and organizational risk-taking Westview, Boulder (CO), United States.
behaviors. Weber, E.U., Blais, A.R., Betz, N.E., 2002. A domain specific risk-
Smith, A., 1976. In: Raphael, D.D., Macfi, A.L. (Eds.), The theory of attitude scale: measuring risk perceptions and risk behaviors. J.
moral sentiments. Clarendon, Oxford. Behav. Decis. Mak. 15, 263---290.

Business Research Quarterly: Are Religion and Culture Relevant For Corporate Risk-Taking? International Evidence

Uploaded by

Copyright:

Available Formats

Business Research Quarterly: Are Religion and Culture Relevant For Corporate Risk-Taking? International Evidence

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Business Research Quarterly: Are Religion and Culture Relevant For Corporate Risk-Taking? International Evidence

Uploaded by

Copyright:

Available Formats

BRQ Business Research Quarterly (2019) 22, 36---55

Are religion and culture relevant for corporate

Received 19 August 2017; accepted 11 June 2018

culture to economics and vice versa), as ‘‘religion is a cultural

lay more on theological aspects rather than economic ones.

(1995), we consider that cultural dimensions should be addressed

Table 1 Composition of the sample by countries.

Table 2 Summary of hypothesis.

shareholders. clusive (Díez-Esteban et al., 2013). We expect a negative sign of

Table 3 Descriptive statistics.

Variable MONITOR NATHET GDP MO MB LEV

Table 5 Main descriptive statistics.

Catholic Non-Catholic Test t Protestant Non-Protestant Test t

Islamic Non-Islamic Test t Eastern Non-Eastern Test t

Table 6 Results of the estimation of model 1.

Cultural Religious Ownership Cultural Religious Ownership

CATHOLIC PROTESTANT ISLAMIC EASTERN

CATHOLIC PROTESTANT ISLAMIC EASTERN

Table 9 Results of the estimation of model 1 by reference shareholder nature.

Cultural Religious Ownership Cultural Religious Ownership

See Tables A1 and A2 .

Table A1 Deﬁnition of variables.

Table A2 Industry dummy variables.

You might also like