Professional Documents
Culture Documents
2023 Institional Investor and ESG Performance
2023 Institional Investor and ESG Performance
institutional investors (i.e., cliques) based on their common block stakes and study how
their coordination shapes the firm’s ESG policy. We document that clique ownership
shows a robust and causal positive impact on the firm’s future ESG performance.
Further tests confirm a positive relationship between ESG and firm value in China.
Additionally, we find that clique members coordinate their trades, and the coordination
enhances their governance via both voice and exit threats. These results suggest that
analyses demonstrate that the positive impact of clique ownership on ESG is greater for
Corporate governance
* Corrsponding author. E-mail address: [email protected] (C.H. Li). Cheng Xiang acknowledges financial
support from the National Social Science Fund of China (22BGL295) and the National Natural Science Foundation
of China (Grant No. 71973018, 72272018).
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
ed
Institutional investor cliques and ESG performance:
iew
Abstract: Using a sample of Chinese firms, we identify groups of coordinated
institutional investors (i.e., cliques) based on their common block stakes and study how
v
their coordination shapes the firm’s ESG policy. We document that clique ownership
re
shows a robust and causal positive impact on the firm’s future ESG performance.
Further tests confirm a positive relationship between ESG and firm value in China.
er
Additionally, we find that clique members coordinate their trades, and the coordination
pe
enhances their governance via both voice and exit threats. These results suggest that
analyses demonstrate that the positive impact of clique ownership on ESG is greater for
Corporate governance
ep
Pr
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
1. Introduction
ed
The Environmental, Social, and Governance (ESG) impact of corporate social
iew
recent decades (Boubakri et al. 2021; Gillan et al. 2021). Rich literature shows that
ESG reputation significantly impacts the firm’s market value (Harjoto & Laksmana
2018), financial performance (Manchiraju & Rajgopal 2017), capital costs (El Ghoul et
v
al. 2011), risk-taking (Albuquerque et al. 2019), social capital (Lins et al. 2017), etc.
re
Given these impacts, there are rapidly growing studies exploring how institutional
investors, who are important stake owners and monitors of public firms, influence their
er
portfolio firms’ ESG policies (Dyck et al. 2019; Chen et al. 2020; Li et al. 2021).
pe
However, these studies reach mixed results (Gillan et al. 2021).
Essentially, institutional investors shape the firm’s ESG performance via their
governance efforts, usually proxied by the firm’s total institutional ownership. However,
ot
each institutional investor usually owns small shares in any given firm and has limited
investors’ governance impact does not simply add up like their ownership but
considerably depends on their coordination (Crane et al. 2019; Doidge et al. 2019;
rin
institutional investors’ impact on ESG. Prior studies offer mixed implications for this
role. Coordination may drive institutional investors to act as one block shareholder and,
Pr
thus, strengthen their governance via voice (Doidge et al. 2019). It may also weaken
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
their governance via exit threats because coordinated institutional investors trade less
ed
aggressively to minimize price impact (Crane et al. 2019). Thus, the net governance
iew
investors’ preferences for ESG depend on its impact on shareholder wealth, on which
prior studies have also reached inconclusive findings (Buchanan et al. 2018; Bartov et
v
mitigate ESG engagement (Borghesi et al. 2014; Dyck et al. 2019). In summary,
re
whether coordination among institutional investors enhances or weakens their portfolio
developed countries (Lin & Fu 2017)1, and Chinese public firms usually have a
controlling shareholder (Jiang & Kim 2020). Given their minority ownership,
ot
institutional investors of Chinese firms need to work together to exert their governance
impact. Second, due to China’s developing legal institutions and its prevalent culture
managers’ business decisions in China (Gu et al. 2019). Thus, China’s unique formal
institutions and social norms may facilitate the formation of coordinated institutional
ep
investor groups, which in turn affects their governance impact (Li & Jiang 2022). Third,
Pr
1 The mean institutional ownership is about 16% for our sample of Chinese firms during 2008-2021. In comparison,
institutional investors own over 65% of American firms by 2010s, according to Borochin and Yang (2017).
3
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
the motivations and consequences of ESG in developing countries, such as China,
ed
significantly differ from those in developed countries and are still under-explored
(Boubakri et al. 2021). Specifically, China may not currently be in a position to make
iew
ESG a high priority (Jiang & Kim 2020), and Chinese consumers are relatively
unwilling to share the firm’s ESG cost (Bartling et al. 2015). These characteristics may
v
Following the methodology of Crane et al. (2019), we identify groups of
re
coordinated institutional investors (i.e., institutional investor cliques) as those
connected via common block stakes (>=5% ownership) in any given firm. The idea is
er
that institutional investor networks formed by common shareholdings motivate and
pe
facilitate their information transfer and coordination (Pareek 2012; Rossi et al. 2018).
Larger clique ownership indicates that institutional investors have more incentives and
firms during 2008-2021, we document that Chinese firms with higher clique ownership
are associated with significantly better ESG performance in the subsequent year. The
tn
positive impact survives in a battery of robustness tests, including those with alternative
clique ownership or ESG performance measures and those with an alternative algorithm
rin
Our finding on the positive impact of clique ownership on ESG may suffer from
ep
endogeneity issues, such as omitted variables and reverse causality. To address the
concern about omitted variables, we add firm fixed effects in the regression, and the
Pr
finding above still holds. To establish causality, we use the expansion of China’s QFII
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
(Qualified Foreign Institutional Investors) scheme during 2012-2013 as an exogenous
ed
shock that causes variation in Chinese firms’ clique ownership. The Chinese authority
increased QFIIs’ investment quota from 30 billion USD in 2011 to 80 billion in 2012
iew
and 150 billion in 2013. Accordingly, the number of approved QFIIs increased from
165 in 2011 to 277 in 2013. We expect the newly approved QFIIs to cause exogenous
changes in the networks of institutional investors and, thus, the firm’s clique ownership.
v
More importantly, individual firms’ ESG policies can not affect the Chinese authority’s
re
decision to expand its QFII scheme. Based on this assertation, we instrument the firm’s
clique ownership with the treatment effects of the QFII expansion.2 The two-stage least
er
square regression (2SLS) results using this instrument still document a positive impact
pe
of clique ownership on ESG, implying that the impact is causal.
examine their trade correlations. We find that a given institutional investor’s change in
ot
position in a certain firm significantly and positively correlates with that of other
institutional investors within the same clique. These results suggest that the cliques
tn
indeed coordinate their behaviors, which in turn affects their governance impact on
ESG. However, the positive relationship between clique ownership and ESG can be
rin
interpreted in two different ways. Under the premise that ESG benefits Chinese firms
and institutional investors demand ESG accordingly, the positive impact indicates that
ep
2 See Equation (3) and Section 4.4 for details on this instrument.
5
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
also indicate that coordination weakens institutional investors’ governance, assuming
ed
that ESG is initiated by agency problems and detrimental to firm value.
Our further tests support the former interpretation but are against the latter. First,
iew
we document that Chinese firms with better ESG performance have higher market
values and better financial performance in the future. Second, we find that institutional
investors conduct more site visits to firms with higher clique ownership, suggesting that
v
they send more shareholder voices to those firms. Additionally, we employ the short-
re
selling pilot program in China as an exogenous liquidity shock and examine its impact
on the changes in firm value. The results show that the increase in liquidity causes a
er
larger increase in firm value for firms with higher clique ownership, indicating that
pe
institutional investors’ governance impact via exit threats is larger for those firms.3
Altogether, these results support the premise that ESG benefits Chinese firms, and
institutional investors are motivated to promote ESG. More importantly, the results
ot
suggest that coordination strengthens institutional investors’ governance via both voice
and exit threats, consequently improving the ESG performance of their portfolio firms.
tn
In line with this argument, cross-sectional analyses show that the positive impact of
clique ownership on ESG is more pronounced for firms with weaker governance
rin
The contribution of our study to the literature is threefold. First, our study
ep
3This finding is opposite to the study of Crane et al. (2019) on American firms. We discuss this difference and its
potential causes in detail in Section 5.4.
6
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
is found to be positive in some studies (Dyck et al. 2019; Chen et al. 2020), but negative
ed
or nonlinear in others (Borghesi et al. 2014; Harjoto & Laksmana 2018; Cheng et al.
2022). Additionally, most prior studies in this field focus on total institutional
iew
ownership, institution types, or investment horizons and relatively ignore the
interactions between institutional investors (Cheng et al. 2022). Unlike these studies,
we reveal that coordination among institutional investors results in their portfolio firm’s
v
ESG improvement, controlling for the impact of total institutional ownership. In this
re
way, we enrich the understanding of the relationship between institutional investors and
ESG and, more importantly, the plausible underlying mechanism of the relationship.
er
Second, we add to the fast-growing studies on coordination among institutional
pe
investors. While existing studies explore the impact of such coordination on the firm’s
operational and financial outcomes (Crane et al. 2019; Doidge et al. 2019; Gonzalez &
Calluzzo 2019), we examine its impact on social welfare. Additionally, most existing
ot
studies focus on developed markets and find that coordination strengthens institutional
investors’ governance via voice but weakens governance via exit (Edmans & Manso
tn
2011; Bharath et al. 2013; Crane et al. 2019). However, our study on China, the largest
emerging market with low institutional ownership and weak investor protections,
rin
both voice and exit. These findings suggest that the impact of such coordination is
ep
Third, our study is also related to the broad research on ESG. As summarized by
Pr
Boubakri et al. (2021), the motivation and outcomes of ESG in emerging markets
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
notably differ from those in developed ones and are still under-explored. Firms’ ESG
ed
engagement is less expected by external stakeholders and the general public in
developing countries (Ali et al. 2017). Nevertheless, we find that the coordination of
iew
institutional investors improves their portfolio firms’ ESG performance by enhancing
their governance, implying the positive role of corporate governance in promoting ESG
v
firms’ market value and financial performance. Thus, we also contribute to the debate
re
on the relationship between ESG and firm value in developing countries (Zhang et al.
present the empirical results showing the impact of clique ownership on ESG. Section
5 verifies the mechanism underlying the impact, and Section 6 reports the cross-
ot
Institutional investors are globally important owners and vital external monitors
of publicly traded firms (Aggarwal et al. 2011; Borochin & Yang 2017). Meanwhile, it
ep
has become a standard strategy for publicly traded firms to incorporate ESG impact into
their operations over the recent decades (Cheng et al. 2022). Accordingly, the question
Pr
of how institutional investors shape the firm’s ESG performance has been intensively
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
explored (Gillan et al. 2021). However, prior studies reached mixed conclusions on this
ed
question. Depending on different samples, methodologies, and ESG measures, existing
studies document that the impact of institutional investors on ESG can be positive
iew
(Dyck et al. 2019; Chen et al. 2020), negative (Borghesi et al. 2014; Cheng et al. 2022),
The main reason for the inclusive relationship is that institutional investors’
v
demand for ESG depends on its impact on shareholder wealth, on which prior studies
re
have not reached conclusive findings, either. On the one hand, the stakeholder theory
states that ESG activities increase firm value by enhancing the firm’s reputation and
er
social capital among stakeholders (Bartov et al. 2021), such as debtors, customers,
pe
suppliers, etc. In line with this view, empirical studies show that ESG performance
decreases the firm’s financing costs (El Ghoul et al. 2011), increases customer loyalty
and the firm’s access to trade credit from suppliers (Zhang et al. 2014), and offers
ot
insurance-like protection against negative events (Lins et al. 2017; Shiu & Yang 2017;
Ding et al. 2021). Given these positive impacts, institutional investors may act as a
tn
On the other hand, opponents of ESG argue that ESG investments are a waste of
corporate resources since such investments deviate from the firm’s objective of
ep
shareholder wealth maximization (Friedman 1970). Based on this assertion, the agency
theory claims that the firm’s ESG engagement is initiated by agency problems between
Pr
shareholders and managers, thus detrimental to firm value (Buchanan et al. 2018). For
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
example, managers may overinvest in ESG for their private benefits, such as improving
ed
their own reputation as good global citizens (Barnea & Rubin 2010). Additionally,
managers may use ESG as a strategic tool to disguise their inefficiency or misconduct
iew
in daily operations (Du 2015; Flugum & Souther 2022). Consistent with agency theory,
Borghesi et al. (2014) demonstrate that institutional investors act as external monitors
v
firm’s institutional ownership and corporate social responsibility activities.
re
2.2 Institutional investor cliques and ESG
Essentially, institutional investors shape the firm’s ESG outcomes via their
er
governance impact on corporate actions. Although global institutional ownership has
pe
risen considerably in the last few decades, it is dispersed among individual firms
(Borochin & Yang 2017), which may impede institutional investors’ governance impact
due to the “free rider” problem. Given their fractional ownership in any given firm,
ot
institutional investors may not have sufficient power to influence corporate actions if
behaviors and how such coordination affects their governance have attracted increasing
attention from investors, policy-makers, and scholars (Azar et al. 2018; Ge et al. 2022).
rin
Most existing studies in this field focus on coordination events that are observable
(Appel et al. 2019; Doidge et al. 2019; Gonzalez & Calluzzo 2019). Unlike these
investors ex-ante through the connections in the network of institutional holdings. They
10
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
label these groups as cliques and define a clique as a group of institutions where each
ed
member is connected via a common block stake (>=5% ownership) in any given firm.
In this way, clique ownership proxies for institutional investors’ incentives and
iew
propensity to coordinate their governance based on their connections. Crane et al. (2019)
investors’ governance via voice but weakens governance via exit threats.
v
Following the methodology of Crane et al. (2019), we plan to examine the impact
re
of institutional investor cliques on the ESG performance of Chinese firms. Our
motivation is that prior studies, including that of Crane et al. (2019), offer mixed
er
implications on the impact. According to Crane et al. (2019), institutional investor
pe
cliques facilitate governance via voice but mitigate governance via exit threats in
America. Therefore, the net effect of clique ownership on governance impact may be
China may differ from that in America since the two markets differ notably in
ownership structures, institution backgrounds, and culture (Gul et al. 2010; Jiang &
tn
Kim 2020). Meanwhile, as discussed earlier, institutional investors may either promote
or mitigate the firm’s ESG activities, depending on whether such activities help align
rin
prior studies on the driving forces of ESG and their impact on firm value in China reach
ep
decreases Chinese firms’ ESG performance, even if the net governance impact of the
11
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
coordination is explicit.
ed
In summary, whether firms with higher clique ownership have better or worse ESG
iew
institutional investor cliques shows a positive net governance impact, an increase in
clique ownership should enhance (weaken) ESG performance if ESG helps increase
(decrease) firm value. On the contrary, Chinese firms with larger clique ownership
v
should be associated with worse (better) ESG performance if ESG activities are
re
beneficial (detrimental) to firm value, given that coordination exhibits a negative net
2008-2021. We start our sample from 2008 since most Chinese public firms had
ep
completed their split share reform by 2007, which exhibited profound influences on
their ownership structure and stock liquidity. Following prior studies, we exclude firms
Pr
with special treatment marks, financial firms, and observations with missing variables.
12
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
The final sample consists of 27,594 firm-year observations from 3,422 firms.
ed
The primary ESG data used in our study is obtained from the Sino-Securities Index
iew
the core of international ESG standards and China’s unique characteristics to build its
ESG rating system, including a four-tier structure with more than 300 underlying
indicators.5 SNSI provides ESG scores ranging from 0 to 100 for all listed firms in
v
China’s A-share market dating back to 2009. As discussed later, we examine the impact
re
of clique ownership on the firm’s ESG performance in the subsequent year.
Accordingly, SNSI ESG scores well fit our sample period. In robustness tests, we also
er
use ESG data from an alternative ESG rating agency: Rankins ESG Ratings (RKS
pe
hereafter). Unless otherwise stated, data other than ESG scores is obtained from
CSMAR, i.e., the China Stock Market & Accounting Research database.
investors (i.e., cliques) within the network formed by institutional investors’ holdings.
tn
Prior studies suggest that common holdings drive institutional investors’ direct
communication and interactions (Pool et al. 2015). Based on this notion, Crane et al.
rin
(2019) state that there is a connection between a pair of institutions if they hold common
block stakes (>=5% ownership) in any given firm. They then define a clique as a group
ep
of institutional investors connected to each other, arguing that the connections proxy
Pr
4Tian and Tian (2022) also use SNSI ESG data in their study and refer it as Hua Zheng ESG ratings.
5 The details on the methodology of SNSI ESG Ratings is available at https://www.chindices.com/files/Sino-
Securities%20Index%20ESG%20Ratings%20Methodology.pdf
13
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
for clique members’ incentives and propensity to coordinate.
ed
To identify the institutional investor cliques defined by Crane et al. (2019), we
construct an N×N matrix each year, representing the relationships among the N
iew
institutional investors in the market. For a pair of institutional investors denoted as
institutions i and j, the off-diagonal element of Xij in the matrix equals 1 if both
institutions own a large stake in common in at least one firm and 0 otherwise. Following
v
Crane et al. (2019), we use no less than 5% ownership as the cutoff of “a large stake”
re
when identifying cliques.6 Although it is easy to judge whether two institutional
investors belong to the same clique, identifying cliques in the massive network of
er
institutional holdings depends on complicated algorithms developed by network
pe
researchers. We use the Louvain algorithm developed by Blondel et al. (2008) to
The Louvain algorithm identifies an average of 6 cliques each year during our
ot
sample period, and the median (mean) amount of members in identified cliques is 8
(9.2). For any given year, about 17.4% of institutional investors in China are members
tn
of the identified cliques, and the cliques hold an average of 5.8% ownership of our
sample firms. Compared to the American market studied by Crane et al. (2019), there
rin
are fewer cliques and lower clique ownership in China’s A-share market. A plausible
reason is that Chinese public firms’ total institutional ownership is much lower than
ep
their American peers. The average institutional ownership is about 16% for our sample
Pr
6 As discussed later in Section 4.3, we use alternative cutoffs (a 3%, 2.5%, or 1% ownership) in robustness tests.
14
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
firms. In contrast, institutional investors own over 65% of American firms by the 2010s,
ed
according to Borochin and Yang (2017).
iew
Based on the identified cliques, we test the impact of coordination among
institutional investors on the firm’s ESG performance with the following model:
v
where ESGi,t+1 is the firm’s one-year lead ESG performance proxied by its SNSI ESG
re
score. The independent variable of most interest is firm i’s total clique ownership,
belongs to at least one clique over year t. Otherwise, CliqueMemberj,t equals 0. i,j,t is
ot
firm i’s shares owned by the jth institutional investor at the end of year t. Accordingly,
CliqueOwni,t is firm i’s stock shares owned by all institutional cliques, which can proxy
tn
control for several firm characteristics in Equation (1). First, we use a set of firm-level
accounting and financial proxies as control variables, including the firms’ book-to-
ep
market ratio (BMi,t), total assets (LnAsseti,t), Tobin’s Q (TobinQi,t), asset tangibility
(Tani,t), profitability (Profiti,t), and financial leverage (Levi,t). Since this study explores
Pr
the impact of the ownership of certain institutional investor groups on ESG, we also
15
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
control for the firms’ total institutional ownership (IOi,t). Similarly, we control for a set
ed
of proxies for the firm’s ownership structure, which matters to its governance structure
and, thus, its ESG performance. Those proxies include the firm’s stake owned by its
iew
largest shareholder (Top1i,t), top managers (ManOwni,t), and the state (SOEi,t). Finally,
we control for the firm’s board size (BSizei,t) and independence (BIndepi,t) since they
v
definitions of all variables are reported in the Appendix.
re
Table 1 reports the summary statistics of the main variables used in our study. All
continuous variables are winsorized at the 1% and 99% percentiles for the concern of
er
outliers. The average SNSI ESG scores of sample firms are 73.360, equivalent to a
pe
rating of “B” (the fifth grade) in SNSI’s nine-grade ESG rating of “AAA-C.” This is
consistent with the notion that, on average, Chinese firms do not make social concerns
a high priority (Jiang & Kim 2020). As discussed earlier, the total clique ownership
ot
(CliqueOwni,t) of the average sample firms is 5.8%, which is significantly lower than
that of American firms since the total institutional ownership in China is much lower.
tn
The standard deviation and the maximum value of CliqueOwni,t are 0.094 and 0.475,
Table 2 reports the regression results of our baseline model shown in Equation (1).
16
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
In all settings, we control for industry and year fixed effects and use standard errors
ed
clustered at the firm level to infer the statistical significance of estimated coefficients.
In Column (1), we use SNSI ESG scores as the proxy of ESG performance and employ
iew
the full sample in the regression. The coefficient of CliqueOwni,t is 2.121 with a 1%
significance level, consistent with Hypothesis H1a that firms with higher clique
ownership are associated with better future ESG performance. This finding implies that
v
coordination among institutional investors promotes ESG for Chinese firms.
re
We then employ alternative settings to explore the robustness of the results in
Column (1). In Column (2), we exclude observations during 2020-2021 to address the
er
potential impact of the COVID-19 pandemic on our regression results, considering that
pe
firms may change their ESG policies to cater to investors’ change in demand for ESG
during the pandemic (Broadstock et al. 2021; Ding et al. 2021). In Column (3), we
exclude the firm’s blockholders with at least 5% ownership from cliques while
ot
calculating the total clique ownership. Prior studies suggest that blockholders play an
important role in the governance of Chinese firms (Jiang & Kim 2020). By excluding
tn
blockholders, we can further verify the role of coordination in the governance impact
of institutional investors with only fractional shares in the firm. In Column (4), we
rin
alternatively measure ESG performance with ESG scores issued by RKS (denoted as
on Chinese firms (Zhang et al. 2020; Xiang et al. 2021). In Columns (2) to (4), the
17
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
The results of control variables are generally consistent with existing studies. The
ed
coefficient of institutional ownership (IOi,t) is statistically insignificant in most columns,
implying that institutional investors of Chinese firms may not influence the firm’s ESG
iew
policies without coordination. Both ManOwni,t and BIndepi,t are significantly and
positively related to the firm’s future ESG performance. The results suggest that
managers more motivated or better monitored are more likely to engage in ESG
v
activities, implying a positive relationship between the firm’s governance quality and
re
ESG engagement.7 Additionally, Table 2 shows that firms with bigger sizes (LnAsseti,t),
higher profitability (Profiti,t), or lower financial leverage (Levi,t) have better future ESG
er
performance. These results are consistent with the idea that firms less financially
pe
constrained make more ESG investments since ESG is costly (Xu & Kim 2022).
with three alternative clique ownership measures. The first is firm i’s shares owned by
its largest institutional investor clique (CliqueTop1i,t), and the second is the ownership
rin
sum of the squares of the fraction of firm i owned by each clique. Both measures are
ep
suggested by Crane et al. (2019) as well and are based on cliques identified with the
Pr
7These results also indicate that coordination increases ESG performance by strengthening institutional investors’
governance, which we formally verify in Section 5.
18
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Louvain algorithm. The third measure is the aggregate ownership of firm i’s cliques
ed
identified with the Bron-Kerbosch algorithm developed by Bron and Kerbosch (1973).
iew
problem, and the Bron-Kerbosch algorithm is one of the classic algorithms to solve the
problem.
Table 3 reports the regression results using the three alternative clique ownership
v
measures. We use SNSI ESG scores to proxy for ESG performance in Panel A and RKS
re
ESG scores in Panel B. The results of control variables are omitted for brevity. The
coefficients of clique ownership measures are positive in all six columns and are
er
statistically significant in five columns. These results suggest a robust positive impact
pe
of clique ownership on ESG.
investors within a clique are connected and motivated to coordinate because they hold
a large stake in common in at least one firm. In previous sections, we set the cutoff of
rin
“a large stake” as no less than 5% ownership, as Crane et al. (2019) do in their study
is much lower than American firms. Accordingly, the common ownership needed to
motivate coordination can also be lower for institutional investors in China, and using
Pr
the 5% ownership cutoff may underestimate their coordination. To address this concern,
19
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
we alternatively set the cutoff as 3%, 2.5%, or 1% ownership and repeat the baseline
ed
model regressions in Table 4. The results show that the aggregate clique ownership
(CliqueOwni,t) is always positively related to the firm’s future ESG performance with
iew
a 1% significance level.8 These results show that our finding about the positive impact
v
re
4.4 Addressing endogeneity issues
So far, we have reached robust results showing that firms with higher clique
er
ownership are associated with better ESG performance. We argue that these results
pe
suggest that coordination among institutional investors leads to an increase in the firm’s
ESG performance. However, this argument may suffer from endogeneity issues such
may simultaneously drive the increase in the firm’s clique ownership and future ESG
holding firms expected to have high ESG scores, but not that their active monitoring
To address the issue of omitted variables, we add firm fixed effects in the
regressions and report the results in Table 5. All four clique ownership measures
ep
previously used are positively related to the firm’s future SNSI ESG score with a 5%
Pr
8In untabulated tests, we reach similar results while measuring clique ownership with CliqueTop1i,t or CliqueHHIi,t.
For brevity, we do not report these results in Table 4.
20
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
or higher significance level. These results demonstrate that unobservable or omitted
ed
firm characteristics do not drive the positive impact of coordination on ESG.
iew
To establish the causality of this impact, we employ the expansion of China’s QFII
scheme during 2012-2013 as an exogenous shock that causes variation in Chinese firms’
v
clique ownership. The Chinese authority launched its QFII scheme in 2002, and the
re
total investment quota of QFIIs was set to 30 billion USD by 2007. The quota was
increased to 80 billion in April 2012 and 150 billion in July 2013. Simultaneously, the
er
number of approved QFIIs increased from 165 in 2011 to 277 by 2013. We expect that
pe
the notable increase in new QFIIs caused the increases in clique members and, therefore,
Chinese firms’ total clique ownership. Specifically, we instrument firm i’s clique
ownership with the treatment of this exogenous shock on firm i, denoted as:
ot
N
1
Treatmenti ,t i , j ,t ( k ) (4)
j |C j | kC j
tn
where λi,j,t is institution j’s ownership in firm i by year t, and N is the number of
common block stakes (>=5% ownership) in any given firm, and | Cj | is its cardinality.
Finally, δk is a dummy that equals 1 if institution k is a new QFII, i.e., a QFII approved
ep
by the Chinese authority later than April 2012. Otherwise, δk equals 0. Using
Treatmenti,t, we instrument the firm’s clique ownership with the change in institutional
Pr
21
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
investor networks resulting from the exogenous increase in QFIIs.9 Since individual
ed
firms’ ESG policies are irrelevant to the Chinese authority’s decision to expand the
QFII scheme, Treatmenti,t meets the exclusion criterion as a valid instrumental variable.
iew
Table 6 reports the 2SLS regressions using Treatmenti,t as the instrument of clique
ownership measures. The Chinese authority further increased the QFII quota to 300
billion USD in January 2019 and removed the quota restriction in September. To isolate
v
the impact of the two events on our tests, we exclude observations during 2019-2021 in
re
Table 6.10 In columns (1), (3), (5), and (7), we regress the four clique ownership
measures on Treatmenti,t and control variables in the baseline model. The coefficients
er
of Treatmenti,t are always positive with a 1% significance level, consistent with the
pe
prediction that the QFII expansion increased the clique ownership of Chinese firms.
The F-statistics in those columns are considerably higher than 10. According to Staiger
and Stock (1997), we reject the hypothesis that Treatmenti,t is a weak instrument.
ot
Columns (2), (4), (6), and (8) report that instrumented clique ownership measures are
always positively related to ESGi,t+1 at the 1% significance level. These results support
tn
9 Note that new QFIIs are only used to establish connections in the networks of institutional investors. For any given
firm, its change in QFII ownership may be trivial after new QFIIs become a clique member. Accordingly, our 2SLS
regression results are free from the concern that Treatmenti,t proxies for the firm’s total QFII ownership.
10 We reach similar results when including these observations.
22
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
So far, we have revealed a casual and robust positive impact of clique ownership
ed
on ESG. As discussed in Section 2.2, we can interpret the positive impact in two
different ways. Assuming that ESG benefits Chinese firms, the positive impact implies
iew
that coordination among institutional investors increases their governance impact. On
the opposite, it may suggest that such coordination decreases institutional investors’
v
value. In this section, we empirically verify which interpretation is more plausible. First,
re
we examine the trade similarities among clique members to verify that they coordinate
their behaviors as assumed. We then study whether ESG benefits or harms Chinese
er
firms to indicate whether coordination increases or decreases institutional investors’
pe
governance impact. Finally, we explore how coordination affects institutional investors’
The key assumption of this study is that institutional investors coordinate their
behaviors. Although prior studies provide evidence for this assumption in markets
tn
outside China (Appel et al. 2019; Crane et al. 2019; Doidge et al. 2019), no similar
examine whether the members of institutional investor cliques in China coordinate their
11 Institutional investors exert their governance via either voice (e.g., voting) or exit (e.g., trading). We examine their
coordination in trading rather than in voting since the voting data of Chinese public firms is unavailable to us.
23
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
dSharei, j,h 1dCliqSharei,h 2dNonCliqSharei,h FirmFE InstFE HalfYearFE i,h (5)
ed
where dSharei,j,h is the change in firm i’s shares owned by institution j over half-year h.
The independent variable of interest is the change in firm i’s shares owned by
iew
institutional investors connected to institution j via any clique (dCliqSharei,h). We
collect half-year observations since institutional investors in China only disclose their
detailed mid-year and year-end portfolios. A significant positive β1 indicates that clique
v
members simultaneously increase or decrease their holdings in any given firm. That is,
re
they make coordinated trades. For comparison, we control for the change in firm i’s
Similarly, we include half-year fixed effects in the model to control for the change in
(4), the coefficient of dCliqSharei,h is 0.542 when controlling for dNonCliqSharei,h and
the fixed effects. It implies that a 1% ownership change in any given firm by
rin
(0.542%) in institution j’s ownership in the firm. In comparison, Column (4) shows that
ep
change their holdings in any given firm, i.e., make coordinated behaviors as assumed.
24
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
<Insert Table 7 here>
ed
5.2 Impact of ESG on firm value for Chinese firms
iew
To provide implications on whether institutional investors in China demand ESG
or not, we examine the relationship between Chinese firms’ ESG performance and their
future market value and financial performance in Table 8. Specifically, we measure the
v
firm’s market value with Tobin’s Q (TobinQi,t) and use industry-adjusted ROE
re
(AdjROEi,t) or ROA (AdjROAi,t) as financial performance proxies. Following existing
studies, we control for the firm’s book-to-market ratio (BMi,t), size (LnAsseti,t), financial
er
leverage (Levi,t), ownership of the largest shareholder (Top1i,t), institutional ownership
pe
(IOi,t), state ownership (SOEi,t), and capital expenditures (CapExi,t) in Table 8.
In Column (1) of Table 8, the firm’s lagged SNSI ESG performance (ESGi,t-1) is
positively related to the firm’s industry-adjusted ROE (ROA) with a 10% (5%)
significance level. These results demonstrate that institutional investors benefit from
tn
Chinese firms’ ESG investments, implying that they are motivated to promote ESG via
25
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Since the previous section confirms that ESG benefits Chinese firms, the
ed
documented positive impact of clique ownership on ESG indicates that coordination
iew
explore whether coordination enhances their governance via voice, exit threats, or both.
Dyck et al. (2019) state that governance via voice takes the form of public shareholder
proposals or private engagements between investors and managers. Since the data on
v
Chinese public firms’ shareholder proposals is unavailable to us, we focus on the latter
re
form. Specifically, we examine the impact of clique ownership on institutional
investors’ site visits. Paying site visits allows institutional investors to communicate
er
with managers face-to-face and has been documented as an effective channel of
pe
institutional investors’ monitoring (Jiang & Yuan 2018; Cao et al. 2022; Guo et al.
they conduct more site visits to firms with higher clique ownership.
ot
In Table 9, we regress the proxies for the frequency of institutional investors’ site
visits on measures of the firm’s clique ownership and control variables. DVisiti,t is a
tn
binary variable that equals 1 if firm i is visited by institutional investors at least once
over year t and 0 otherwise. We run logistic regressions when using DVisiti,t as the
rin
dependent variable. LnNVisiti,t is the natural logarithm of one plus the number of times
for which institutional investors visit firm i over year t. Table 9 shows that all four
ep
clique ownership measures used in this study are positively related to DVisiti,t or
investors exert more monitoring efforts on firms with higher clique ownership. That is,
26
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
coordination increases institutional investors’ governance via voice.
ed
<Insert Table 9 here>
iew
5.4 Clique ownership and governance via exit
governance via exit threats. Like Crane et al. (2019), we examine this based on the link
v
between stock liquidity and exit threats since actual threats are unobservable. High
re
stock liquidity decreases the price impact of institutional investors’ trades. Accordingly,
an increase in stock liquidity should facilitate their exit threats and increase firm value
er
(Edmans 2009; Bharath et al. 2013). Based on this idea, Crane et al. (2019) conclude
pe
that a positive liquidity shock should increase more (less) value for firms with larger
threats.
ot
exogenous liquidity shock and investigate whether the sensitivity of Chinese firms’
tn
value to this shock varies with their clique ownership. The Chinese authority announced
the pilot program in March 2010 and gradually expanded it to about 50% of Chinese
rin
public firms by 2021. Prior studies have confirmed that the pilot program enhances the
market (Li et al. 2018). Based on this finding, we examine whether the pilot program’s
impact on designated firms’ value correlates with the firms’ clique ownership.
Pr
Table 10 reports the results of regressing the firm’s value proxied by Tobin’s Q
27
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
on lagged clique ownership, MTi,t, their interaction term, and control variables. MTi,t is
ed
an indicator that equals 1 for pilot firms allowed for short-selling and 0 otherwise. The
coefficients of MTi,t are positive with a 5% or higher significance level, consistent with
iew
the notion that short-selling increases firm value by increasing liquidity. The interaction
terms of the four clique ownership measures and MTi,t are always positively related to
Tobin’s Q with a 1% significance level. These results show that the value effect of the
v
positive liquidity shock caused by the removal of short-selling constraints is larger for
re
firms with higher clique ownership. As discussed earlier, this finding implies that
that institutional investors of cliques with larger ownership have to trade less
aggressively due to the concern about the price impact of their trades. However, the
ot
total institutional ownership and the clique ownership of Chinese firms are much lower
than American firms. For example, the mean clique ownership is 5.8% in our study and
tn
29% in Crane et al. (2019). Thus, institutional investor cliques in China are relatively
less concerned about their trades’ price impact. In contrast, they need to trade
rin
6. Cross-sectional analyses
28
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Previous sections reveal that ESG benefits Chinese firms, implying that
ed
institutional investors in China are motivated to exert their governance impact to
iew
via voice and exit threats, resulting in the documented positive relationship between
clique ownership and ESG. As final tests of these findings, we run subsample
v
6.1 Subsamples based on governance mechanisms
re
First, we split our sample into subsamples based on the firm’s governance
mechanisms. We use the firm’s internal control quality to proxy for its internal
governance quality and employ analyst coverage to measure its external governance
ot
quality. Rich literature documents that internal control is an effective governance tool
(Chen et al. 2017), and analysts serve as external monitors by collecting, disseminating,
tn
In Panel A of Table 11, we divide observations into the High (Low) subsample if
rin
the firm’s internal control quality is above (below) the sample median. Similarly, we
divided observations into the High or Low subsample in Panel B depending on whether
ep
their analyst coverage is above the sample median. We proxy for internal control quality
with the DIB internal control index obtained from DIB internal control and risk
Pr
29
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
management database,12 and analyst coverage is the number of analysts following the
ed
firm over the year. We then estimate the baseline model using each subsample.
iew
and positively related to ESGi,t+1 in all columns, its coefficient is larger in magnitude
for the Low subsamples. For example, the coefficient of CliqueOwni,t is 3.920 for the
subsample of Low internal control quality and 2.168 for the High sample. The Wald
v
Chi2 of Seemingly Unrelated Regressions (SUR) suggests that the coefficient difference
re
is statistically significant at the 10% level. Similarly, the coefficient of CliqueOwni,t in
the subsample of Low analyst coverage is larger than that in the High subsample, and
er
the difference is statistically significant at the 5% level. In summary, Panels A and B
pe
of Table 11 support our prediction that the positive impact of clique ownership on ESG
Previous sections suggest that institutional investor cliques in China promote ESG
since it benefits Chinese firms. If so, the cliques should be more motivated to promote
tn
ESG in firms with larger potential ESG benefits. Accordingly, the positive impact of
clique ownership on ESG should be greater for those firms. Based on prior studies, we
rin
expect that firms in heavily polluting industries and firms that are industry followers
benefit more from ESG investments. First, firms in heavily polluting industries are
ep
12 See http://www.dibdata.cn/.
30
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Accordingly, ESG investments by heavily polluting firms should show larger effects in
ed
mitigating these risks than other firms (Huang et al. 2021; Brown et al. 2022). Second,
compared to industry leaders with high market shares, ESG investments may provide
iew
industry followers with more benefits in increasing product differentiation and sales
We identify heavily polluting industries according to the Industry List for the
v
Category Management of Public Firms’ Environmental Verification issued in 2008 by
re
the Ministry of Ecology and Environment, China. Examples of those industries include
the thermal power industry, the cement industry, the mining industry, the petrochemical
er
industry, etc.13 Additionally, we define firms as industry followers (leaders) if their
pe
gross profit margins are below (above) the industry median. We then divide our sample
into subsamples depending on whether the firm is a heavily polluting firm (an industry
follower) and report the subsample regression results in Panel C (D) of Table 11.
ot
columns. Nevertheless, Panel C shows that the coefficient of CliqueOwni,t for heavily
tn
polluting firms is about twice the magnitude of that for other firms. Similarly, Panel D
reports that the coefficient is larger for industry followers than for industry leaders. The
rin
Wald Chi2 values reported at the bottom of the panels suggest that the coefficient
differences are statistically significant. Thus, these results confirm that the positive
ep
impact of clique ownership on ESG is more pronounced for firms potentially benefiting
Pr
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
more from their ESG investments.
ed
<Insert Table 11 here>
iew
7. Conclusion
connections in their common block holdings and examine how coordination affects
v
their governance impact on the ESG performance of Chinese firms. Our empirical
re
results suggest a robust and positive impact of clique ownership on ESG. Using the
document that coordination increases institutional investors’ governance via both voice
ot
and exit threats. Altogether, our findings suggest that institutional investors in China
coordinate their governance to improve their portfolio firms’ ESG performance, given
tn
that ESG benefits Chinese firms. Consistent with this argument, we show that the
positive impact of clique ownership on ESG is more pronounced for firms with weaker
rin
governance mechanisms or firms that potentially benefit more from ESG investments.
Based on the unique Chinese setting, our findings provide additional empirical
ep
Unlike prior studies in developed markets, we find that coordination in China’s A-share
Pr
32
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
explanation is that institutional investors in China are less concerned about the price
ed
impact of their simultaneous trades since their ownership is relatively low. Also, our
study contributes to the literature on institutional investors’ role in shaping the firm’s
iew
ESG policy and the driving forces of ESG investments in emerging markets.
References
v
Aggarwal, R., Erel, I., Ferreira, M., Matos, P., 2011. Does governance travel around the world? Evidence
from institutional investors. Journal Of Financial Economics 100, 154-181
re
Albuquerque, R., Koskinen, Y., Zhang, C.D., 2019. Corporate Social Responsibility and Firm Risk:
Theory and Empirical Evidence. Management Science 65, 4451-4469
Ali, W., Frynas, J.G., Mahmood, Z., 2017. Determinants of Corporate Social Responsibility (CSR)
Disclosure in Developed and Developing Countries: A Literature Review. Corporate Social
Responsibility and Environmental Management 24, 273-294
er
Appel, I.R., Gormley, T.A., Keim, D.B., 2019. Standing on the Shoulders of Giants: The Effect of Passive
Investors on Activism. Review of Financial Studies 32, 2720-2774
Azar, J., Schmalz, M.C., Tecu, I., 2018. Anticompetitive Effects of Common Ownership. Journal of
pe
Finance 73, 1513-1565
Barnea, A., Rubin, A., 2010. Corporate Social Responsibility as a Conflict Between Shareholders.
Journal of Business Ethics 97, 71-86
Bartling, B., Weber, R.A., Yao, L., 2015. Do Markets Erode Social Responsibility? Quarterly Journal of
Economics 130, 219-266
ot
Bartov, E., Marra, A., Momente, F., 2021. Corporate Social Responsibility and the Market Reaction to
Negative Events: Evidence from Inadvertent and Fraudulent Restatement Announcements.
Accounting Review 96, 81-106
tn
Bharath, S.T., Jayaraman, S., Nagar, V., 2013. Exit as Governance: An Empirical Analysis. Journal Of
Finance 68, 2515-2547
Blondel, V.D., Guillaume, J.L., Lambiotte, R., Lefebvre, E., 2008. Fast unfolding of communities in
large networks. Journal of Statistical Mechanics-Theory and Experiment, P10008
rin
Borghesi, R., Houston, J.F., Naranjo, A., 2014. Corporate socially responsible investments: CEO
altruism, reputation, and shareholder interests. Journal of Corporate Finance 26, 164-181
Borochin, P., Yang, J., 2017. The effects of institutional investor objectives on firm valuation and
governance. Journal Of Financial Economics 126, 171-199
ep
Boubakri, N., El Ghoul, S., Guedhami, O., Wang, H., 2021. Corporate social responsibility in emerging
market economies: Determinants, consequences, and future research directions. Emerging
Markets Review 46, 100758
Broadstock, D.C., Chan, K.L., Cheng, L.T.W., Wang, X.W., 2021. The role of ESG performance during
Pr
times of financial crisis: Evidence from COVID-19 in China. Finance Research Letters 38,
101716
Bron, C., Kerbosch, J., 1973. Algorithm 457: finding all cliques of an undirected graph. Communications
33
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
of the ACM 16, 575-577
Brown, J.R., Martinsson, G., Thomann, C., 2022. Can Environmental Policy Encourage Technical
ed
Change? Emissions Taxes and R&D Investment in Polluting Firms. Review of Financial Studies
35, 4518-4560
Buchanan, B., Cao, C.X.Y., Chen, C.Y., 2018. Corporate social responsibility, firm value, and influential
institutional ownership. Journal of Corporate Finance 52, 73-95
iew
Cao, J., Liang, H., Zhan, X.T., 2019. Peer Effects of Corporate Social Responsibility. Management
Science 65, 5487-5503
Cao, J., Wang, H.Y., Zhou, S.L., 2022. Soft activism and corporate dividend policy: Evidence from
institutional investors site visits. Journal of Corporate Finance 75, 102221
Chen, J., Chan, K.C., Dong, W., Zhang, F., 2017. Internal Control and Stock Price Crash Risk: Evidence
v
from China. European Accounting Review 26, 125-152
Chen, T., Dong, H., Lin, C., 2020. Institutional shareholders and corporate social responsibility. Journal
re
of Financial Economics 135, 483-504
Chen, T., Harford, J., Lin, C., 2015. Do analysts matter for governance? Evidence from natural
experiments. Journal Of Financial Economics 115, 383-410
Cheng, X., Wang, H., Wang, X.J., 2022. Common institutional ownership and corporate social
responsibility. Journal of Banking & Finance 136, 106218
er
Crane, A.D., Koch, A., Michenaud, S., 2019. Institutional investor cliques and governance. Journal of
Financial Economics 133, 175-197
Ding, W.Z., Levine, R., Lin, C., Xie, W.S., 2021. Corporate immunity to the COVID-19 pandemic.
pe
Journal of Financial Economics 141, 802-830
Doidge, C., Dyck, A., Mahmudi, H., Virani, A., 2019. Collective Action and Governance Activism.
Review of Finance 23, 893-933
Du, X.Q., 2015. Is Corporate Philanthropy Used as Environmental Misconduct Dressing? Evidence from
Chinese Family-Owned Firms. Journal of Business Ethics 129, 341-361
ot
Dyck, A., Lins, K.V., Roth, L., Wagner, H.E., 2019. Do institutional investors drive corporate social
responsibility? International evidence. Journal of Financial Economics 131, 693-714
Edmans, A., 2009. Blockholder Trading, Market Efficiency, and Managerial Myopia. Journal Of Finance
tn
64, 2481-2513
Edmans, A., Manso, G., 2011. Governance Through Trading and Intervention: A Theory of Multiple
Blockholders. Review Of Financial Studies 24, 2395-2428
El Ghoul, S., Guedhami, O., Kwok, C.C.Y., Mishra, D.R., 2011. Does corporate social responsibility
rin
affect the cost of capital? Journal of Banking & Finance 35, 2388-2406
Flugum, R., Souther, M.E., 2022. Stakeholder Value: A Convenient Excuse for Underperforming
Managers? Available at SSRN: https://ssrn.com/abstract=3725828
Friedman, M., 1970. The social responsibility of business is to increase its profits. New York Times
ep
Magazine
Ge, W.X., Ouyang, C.Y., Shi, Z.Y., Chen, Z.L., 2022. Can a not-for-profit minority institutional
shareholder make a big difference in corporate governance? A quasi-natural experiment. Journal
of Corporate Finance 72, 102125
Pr
Gillan, S.L., Koch, A., Starks, L.T., 2021. Firms and social responsibility: A review of ESG and CSR
research in corporate finance. Journal of Corporate Finance 66, 101889
Gonzalez, T.A., Calluzzo, P., 2019. Clustered shareholder activism. Corporate Governance-an
34
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
International Review 27, 210-225
Gu, Z.Y., Li, Z.Q., Yang, Y.G., Li, G.Q., 2019. Friends in Need Are Friends Indeed: An Analysis of
ed
Social Ties between Financial Analysts and Mutual Fund Managers. Accounting Review 94,
153-181
Gul, F.A., Kim, J.B., Qiu, A.A., 2010. Ownership concentration, foreign shareholding, audit quality, and
stock price synchronicity: Evidence from China. Journal Of Financial Economics 95, 425-442
iew
Guo, Y., Li, J., Lin, B., 2023. Corporate site visit and tax avoidance: The effects of monitoring and tax
knowledge dissemination. Journal of Corporate Finance, Forthcoming
Harjoto, M., Jo, H., Kim, Y., 2017. Is Institutional Ownership Related to Corporate Social Responsibility?
The Nonlinear Relation and Its Implication for Stock Return Volatility. Journal of Business
Ethics 146, 77-109
v
Harjoto, M., Laksmana, I., 2018. The Impact of Corporate Social Responsibility on Risk Taking and
Firm Value. Journal of Business Ethics 151, 353-373
re
Huang, B.H., Punzi, M.T., Wu, Y., 2021. Do banks price environmental transition risks? Evidence from
a quasi-natural experiment in China. Journal of Corporate Finance 69, 101983
Jiang, F.X., Kim, K.A., 2020. Corporate Governance in China: A Survey. Review of Finance 24, 733-
772
Jiang, X.Y., Yuan, Q.B., 2018. Institutional investors' corporate site visits and corporate innovation.
er
Journal of Corporate Finance 48, 148-168
Li, F.Z., Jiang, Y.X., 2022. Institutional investor networks and crash risk: Evidence from China. Finance
Research Letters 47, 102627
pe
Li, Z.F., Patel, S., Ramani, S., 2021. The Role of Mutual Funds in Corporate Social Responsibility.
Journal of Business Ethics 174, 715-737
Li, Z.S., Lin, B.X., Zhang, T., Chen, C., 2018. Does short selling improve stock price efficiency and
liquidity? Evidence from a natural experiment in China. European Journal of Finance 24, 1350-
1368
ot
Lin, Y.R., Fu, X.M., 2017. Does institutional ownership influence firm performance? Evidence from
China. International Review of Economics & Finance 49, 17-57
Lins, K.V., Servaes, H., Tamayo, A., 2017. Social Capital, Trust, and Firm Performance: The Value of
tn
Corporate Social Responsibility during the Financial Crisis. Journal Of Finance 72, 1785-1823
Manchiraju, H., Rajgopal, S., 2017. Does Corporate Social Responsibility (CSR) Create Shareholder
Value? Evidence from the Indian Companies Act 2013. Journal of Accounting Research 55,
1257-1300
rin
Pareek, A., 2012. Information Networks: Implications for Mutual Fund Trading Behavior and Stock
Returns. Available at SSRN: https://ssrn.com/abstract=1361779
Pool, V.K., Stoffman, N., Yonker, S.E., 2015. The People in Your Neighborhood: Social Interactions
and Mutual Fund Portfolios. Journal of Finance 70, 2679-2732
ep
Rossi, A.G., Blake, D., Timmermann, A., Tonks, I., Wermers, R., 2018. Network centrality and delegated
investment performance. Journal of Financial Economics 128, 183-206
Shiu, Y.M., Yang, S.L., 2017. Does Engagement in Corporate Social Responsibility Provide Strategic
Insurance-Like Effects? Strategic Management Journal 38, 455-470
Pr
Staiger, D., Stock, J.H., 1997. Instrumental variables regression with weak instruments. Econometrica
65, 557-586
Tian, H.W., Tian, G.L., 2022. Corporate sustainability and trade credit financing: Evidence from
35
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
environmental, social, and governance ratings. Corporate Social Responsibility and
Environmental Management 29, 1896-1908
ed
Tian, Z.L., Wang, R., Yang, W., 2011. Consumer Responses to Corporate Social Responsibility (CSR)
in China. Journal of Business Ethics 101, 197-212
Xiang, C., Chen, F., Jones, P., Xia, S., 2021. The effect of institutional investors' distraction on firms'
corporate social responsibility engagement: evidence from China. Review of Managerial
iew
Science 15, 1645-1681
Xu, Q.P., Kim, T., 2022. Financial Constraints and Corporate Environmental Policies. Review of
Financial Studies 35, 576-635
Yi, Y.Y., Zhang, Z.Y., Xiang, C., 2022. The value of CSR during the COVID-19 crisis: Evidence from
Chinese firms. Pacific-Basin Finance Journal 74, 101795
v
Zhang, J., Zi, S., Shao, P., Xiao, Y.C., 2020. The value of corporate social responsibility during the crisis:
Chinese evidence. Pacific-Basin Finance Journal 64, 101432
re
Zhang, M., Ma, L.J., Su, J., Zhang, W., 2014. Do Suppliers Applaud Corporate Social Performance?
Journal of Business Ethics 121, 543-557
er
pe
ot
tn
rin
ep
Pr
36
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Appendix Variable definition
Variable Definition
ed
ESGi,t Firm i’s annual ESG score issued by the Sino-Securities Index Information
Service (Shanghai) Co. Ltd.
CliqueOwni,t The total fraction of firm i owned by all institutional investor cliques
identified with the Louvain algorithm.
iew
CliqueTop1i,t The fraction of firm i owned by its largest institutional investor clique
identified with the Louvain algorithm.
CliqueHHIi,t The sum of the squares of the year-end fraction of firm i owned by each
institutional investor clique identified with the Louvain algorithm.
CliqueOwn_BKi,t The total fraction of firm i owned by all institutional investor cliques
v
identified with the Bron-Kerbosch algorithm.
BMi,t Firm i’s year-end book-to-market ratio.
re
LnAsseti,t The natural logarithm of firm i’s year-end total assets (in Chinese Yuan).
TobinQi,t The ratio of the firm’s year-end market value to the replacement cost of its
assets.
Tani,t The ratio of firm i’s year-end tangible assets to total assets.
Profiti,t The ratio of firm i’s earnings before interests and taxes to total assets.
Levi,t
er
The ratio of firm i’s year-end total debt to total assets.
Top1i,t The fraction of firm i owned by its largest shareholder.
IOi,t The fraction of firm i owned by institutional investors.
pe
ManOwni,t The fraction of firm i owned by top executives.
SOEi,t The binary variable that equals 1 for state-owned enterprises and 0
otherwise.
BIndepi,t The fraction of independent directors on firm i’s board.
BSizei,t The number of firm i’s board directors.
ot
Treatmenti,t The instrumental variable of clique ownership measures. See Equation (3)
for details.
AdjROEi,t The firm’s return on equity adjusted for the industry median.
tn
AdjROAi,t The firm’s return on assets adjusted for the industry median.
DVisiti,t The binary variable that equals 1 if firm i is visited by institutional
investors at least once over the year and 0 otherwise.
LnNVisiti,t The natural logarithm of one plus the number of times for which firm i is
rin
dCliqueSharei,h The change in firm i’s shares owned by all institutional investors connected
to institution j via any clique.
dNonCliqueSharei,h The change in firm i’s shares owned by all institutional investors not
connected to institution j via any clique.
Pr
37
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 1 Summary statistics
This table reports the number of observations, mean values, 25 percentiles, medians, 75 percentiles,
ed
standard deviations, minimum values, and maximum values of the main variables. Detailed
definitions of the variables are presented in the Appendix. All continuous variables are winsorized
at the 1% and 99% levels.
Variable N mean p25 p50 p75 sd min max
iew
ESGi,t+1 27,594 73.360 70.040 73.690 77.230 5.647 56.400 85.310
CliqueOwni,t 27,594 0.058 0.002 0.016 0.069 0.094 0 0.475
BMi,t 27,594 0.624 0.437 0.624 0.809 0.245 0.120 1.155
LnAsseti,t 27,594 22.211 21.275 22.029 22.946 1.291 19.959 26.215
TobinQi,t 27,594 2.011 1.236 1.602 2.290 1.263 0.866 8.353
v
Tani,t 27,594 0.927 0.915 0.957 0.979 0.0880 0.528 1
Profiti,t 27,594 0.059 0.032 0.056 0.087 0.060 -0.196 0.236
re
Levi,t 27,594 0.422 0.255 0.415 0.580 0.207 0.0500 0.884
Top1i,t 27,594 0.353 0.235 0.334 0.456 0.150 0.0880 0.750
IOi,t 27,594 0.160 0.018 0.071 0.214 0.205 0 0.829
ManOwni,t 27,594 0.137 0 0.004 0.253 0.202 0 0.691
SOEi,t
BIndepi,t
27,594
27,594
0.392
0.374
er0
0.333
0
0.333
1
0.429
0.488
0.053
0
0.333
1
0.571
BSizei,t 27,594 8.680 7.000 9.000 9.000 1.709 5.000 15.000
pe
ot
tn
rin
ep
Pr
38
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 2 Clique ownership and ESG performance
This table reports the estimation results of Equation (1) in the text. ESGi,t+1 is the firm’s one-year
ed
lead ESG score issued by SNSI, and RKS_ESGi,t+1 is that issued by RKS. CliqueOwni,t is the total
fraction of firm i owned by all institutional investor cliques identified with the Louvain algorithm.
Detailed definitions of all variables are reported in Appendix. Observations during 2020-2021 are
excluded in Column (2). Blockholders with at least 5% ownership are excluded while calculating
CliqueOwni,t in Column (3). Ind FE and Year FE denote the industry and year fixed effects,
iew
respectively. Standard errors clustered at the firm level are disclosed in parentheses. *, **, and ***
denote the significance level of 10%, 5%, and 1%, respectively.
(1) (2) (3) (4)
ESGi,t+1 ESGi,t+1 ESGi,t+1 RKS_ESGi,t+1
CliqueOwni,t 2.121*** 2.210*** 4.267*** 6.214***
(0.53) (0.54) (0.97) (1.75)
v
BMi,t -1.303*** -1.069** -1.263*** -10.459***
(0.44) (0.48) (0.45) (1.10)
re
LnAsseti,t 1.343*** 1.286*** 1.345*** 4.450***
(0.07) (0.08) (0.08) (0.18)
TobinQi,t -0.221*** -0.289*** -0.231*** -0.445***
(0.06) er (0.07) (0.06) (0.16)
Tani,t 5.639*** 5.345*** 5.836*** 6.958***
(0.72) (0.76) (0.76) (1.58)
Profiti,t 18.296*** 16.803*** 18.305*** 56.704***
(0.93) (1.04) (0.93) (2.24)
pe
Levi,t -5.193*** -5.136*** -5.396*** -8.290***
(0.37) (0.38) (0.38) (0.89)
Top1i,t 0.557 0.167 0.606 2.356**
(0.45) (0.47) (0.46) (1.08)
IOi,t -0.036 -0.395 -0.001 2.389***
(0.28) (0.29) (0.00) (0.74)
ot
39
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 3 Alternative measures of clique ownership
This table reports the regression results of Equation (1) in the text using alternative clique ownership
ed
measures. Panel A (Panel B) uses SNSI ESG scores (RKS CSR scores) to proxy for the firm’s ESG
performance. CliqueTop1i,t is the fraction of firm i owned by its largest institutional investor clique
identified with the Louvain algorithm. CliqueHHIi,t is the sum of the squares of the fraction of firm
i owned by each clique identified with the Louvain algorithm. Finally, CliqueOwn_BK i,t is the total
iew
fraction of firm i owned by all institutional investor cliques identified with the Bron-Kerbosch
algorithm. The control variables are identical to those in Table 2. Ind FE and Year FE denote the
industry and year fixed effects, respectively. Standard errors clustered at the firm level are disclosed
in parentheses. *, **, and *** denote the significance level of 10%, 5%, and 1%, respectively.
Panel A Dep var= ESGi,t+1 Panel B Dep var= RKS_ESGi,t+1
v
(1) (2) (3) (4) (5) (6)
CliqueTop1i,t 3.619*** 5.291*
re
(0.90) (2.72)
CliqueHHIi,t 6.998** 9.001
(2.86) (9.74)
CliqueOwn_BK i,t 1.944*** 6.451***
Controls Yes
er
Yes
(0.54)
Yes Yes Yes
(1.87)
Yes
Ind FE Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes
pe
F-statistics 132.430 131.704 129.519 163.118 163.273 161.824
Adj. R2 0.219 0.218 0.218 0.303 0.303 0.304
Obs 27,594 27,594 27,258 23,998 23,998 23,710
ot
tn
rin
ep
Pr
40
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 4 Alternative ownership cutoffs to identify institutional investor cliques
In Columns (1) and (2), CliqueOwni,t is the fraction of firm i owned by institutional investor cliques
ed
that are identified with the cutoff of a 3% common ownership between pairs of institutions.
Additionally, the cutoff to identify connected pairs of institutions is set to a 2.5% common
ownership in Columns (3) and (4) and a 1% common ownership in Columns (5) and (6). Firm FE
and Year FE denote the firm and year fixed effects, respectively. The control variables are identical
iew
to those in Table 2. Standard errors clustered at the firm level are disclosed in parentheses. *, **,
and *** denote the significance level of 10%, 5%, and 1%, respectively.
3% common 2.5% common 1% common
ownership ownership ownership
Dep var= ESGi,t+1 (1) (2) (3) (4) (5) (6)
v
CliqueOwni,t 7.662*** 2.135*** 7.634*** 2.038*** 7.710*** 2.052***
(0.50) (0.53) (0.49) (0.52) (0.48) (0.51)
re
Controls No Yes No Yes No Yes
Firm FE Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes
F-statistics 234.733 134.301
er 240.374 136.485 253.446 138.376
Adj. R2 0.080 0.219 0.080 0.220 0.081 0.220
Obs 28,154 28,154 28,402 28,402 28,576 28,576
pe
ot
tn
rin
ep
Pr
41
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 5 Controlling for firm fixed effects
This table reports the regression results of Equation (1) in the text while controlling for firm fixed
ed
effects. The control variables are identical to those in Table 2. All variables are defined in detail in
the Appendix. Firm FE and Year FE denote the firm and year fixed effects, respectively. Standard
errors clustered at the firm level are disclosed in parentheses. *, **, and *** denote the significance
level of 10%, 5%, and 1%, respectively.
iew
(1) (2) (3) (4)
Dep var= ESGi,t+1 ESGi,t+1 ESGi,t+1 ESGi,t+1
CliqueOwni,t 2.121***
(0.53)
CliqueTop1i,t 3.619***
v
(0.90)
CliqueHHIi,t 6.998**
re
(2.86)
CliqueOwn_BKi,t 1.944***
(0.54)
Controls Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes
Year FE Yes
er Yes Yes Yes
F-statistics 132.364 132.430 131.704 129.519
Adj. R2 0.219 0.219 0.218 0.218
pe
Obs 27,594 27,594 27,594 27,258
ot
tn
rin
ep
Pr
42
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 6 Clique ownership and ESG performance: Exogenous network shocks
e d
w
This table reports the two-stage least square regression results using Treatmenti,t as the instrumental variable of clique ownership measures. We define Treatmenti,t in
e
Equation (3) in the text. The control variables are identical to those in Table 2. All variables are defined in detail in the Appendix. Firm FE and Year FE denote the
i
firm and year fixed effects, respectively. Standard errors clustered at the firm level are disclosed in parentheses. *, **, and *** denote the significance level of 10%,
v
5%, and 1%, respectively.
(1) (2) (3) (4) (5) (6) (7) (8)
Treatmenti,t
CliqueOwni,t
5.583***
ESGi,t+1 CliqueTop1i,t
3.678***
ESGi,t+1 CliqueHHIi,t
0.733***
r e ESGi,t+1 CliqueOwn_BKi,t
5.293***
ESGi,t+1
CliqueOwni,t
(0.53)
12.729***
(4.01)
(0.32)
er (0.09) (0.55)
CliqueTop1i,t
CliqueHHIi,t
p e
19.320***
(6.19)
96.941***
CliqueOwn_BKi,t
o t (30.87)
13.450***
(4.22)
Controls
Firm FE
Year FE
Yes
Yes
Yes
t nYes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
ir n
F-statistics/Wald chi2 54.812 2153.999 55.356 2177.132 23.450 2096.772 53.486 2113.608
Adj. R2 0.406 0.180 0.349 0.187 0.285 0.160 0.295 0.178
p
Obs 21,457 21,457 21,457 21,457 21,457 21,457 21,196 21,196
r e
P 43
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 7 Coordination in trades within institutional investor cliques
dSharei,j,h is the change in firm i’s shares owned by institution j over half-year h. dCliqSharei,h is
ed
the change in firm i’s shares owned by all institutional investors connected to institution j via any
clique, and dNonCliqSharei,h is the change in firm i’s shares owned by all institutional investors not
connected to institution j. We control for firm, institution, and half-year fixed effects in all columns.
Standard errors clustered at the institution level are disclosed in parentheses. *, **, and *** denote
iew
the significance level of 10%, 5%, and 1%, respectively.
(1) (2) (3) (4)
dSharei,j,h dSharei,j,h dSharei,j,h dSharei,j,h
dCliqSharei,h 0.568*** 0.559*** 0.549*** 0.542***
(0.04) (0.04) (0.04) (0.04)
v
dNonCliqSharei,h 0.042*** 0.038***
(0.01) (0.01)
re
Constant 0.017*** 0.011*** 0.020*** 0.015**
(0.00) (0.00) (0.01) (0.01)
Firm FE No No Yes Yes
Institution FE No No Yes Yes
Half-year FE
F-statistics
Yes
217.345
er Yes
309.965
Yes
118.881
Yes
213.261
Adj. R2 0.310 0.309 0.350 0.349
Obs 695,704 690,098 695,667 690,071
pe
ot
tn
rin
ep
Pr
44
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 8 ESG performance and firm value
This table reports the results of regressing firm value proxies on the firm’s lagged SNSI ESG score
ed
(ESGi,t-1) and control variables. TobinQi,t is the firm’s year-end market value to the replacement cost
of its assets. AdjROEi,t is the firm’s return on equity adjusted for the industry median, and AdjROAi,t
is its return on assets adjusted for the industry median. Firm FE and Year FE denote the firm and
year fixed effects, respectively. Standard errors clustered at the firm level are disclosed in
iew
parentheses. *, **, and *** denote the significance level of 10%, 5%, and 1%, respectively.
(1) (2) (3)
TobinQi,t AdjROEi,t AdjROAi,t
ESGi,t-1 0.005*** 0.021* 0.012**
(0.00) (0.01) (0.01)
v
BMi,t -4.247*** -3.871*** -0.019***
(0.07) (0.33) (0.00)
re
LnAsseti,t -0.049** 0.587*** 0.002***
(0.02) (0.13) (0.00)
Levi,t 0.076 -3.220*** -0.019***
(0.08) (0.52) (0.00)
Top1i,t 0.222* -0.697 0.001
(0.12)
er (0.78) (0.00)
IOi,t 0.000 0.013*** 0.000***
(0.00) (0.00) (0.00)
pe
SOEi,t -0.150*** -0.116 -0.001
(0.04) (0.32) (0.00)
CapExi,t -0.233 2.892** 0.010*
(0.15) (1.13) (0.01)
Constant 5.370*** -2.904 0.003
ot
45
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 9 Cliques and governance via voice
e d
w
This table reports the results of examing the impact of clique ownership on institutional investors’ site visits to the firm. DVisiti,t is a binary variable that equals 1 if
e
firm i is visited by institutional investors at least once over year t and 0 otherwise. LnNVisiti,t is the natural logarithm of one plus the number of times for which firm i
i
is visited by institutional investors over year t. The control variables are identical to those in Table 8. Firm FE and Year FE denote the firm and year fixed effects,
v
respectively. Standard errors clustered at the firm level are disclosed in parentheses. *, **, and *** denote the significance level of 10%, 5%, and 1%, respectively.
(1) (2) (3) (4) (5) (6) (7) (8)
Dep var=
CliqueOwni,t
DVisiti,t
3.953***
DVisiti,t DVisiti,t DVisiti,t LnNVisiti,t
3.770***
CliqueTop1i,t
(0.55)
7.110***
(0.88)
er (0.27)
6.423***
(0.44)
CliqueHHIi,t
CliqueOwn_BKi,t
16.416***
(3.69)
p e 3.923***
14.148***
(1.62)
3.709***
Controls
Firm FE
Yes
Yes
Yes
Yes
o t Yes
Yes
(0.59)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
(0.30)
Yes
Yes
Year FE
F- statistics /Wald Chi2
Adj. R2/Pseudo R2
Yes
1100.499
0.135
t n Yes
1122.308
0.137
Yes
1091.193
0.131
Yes
1089.893
0.134
Yes
133.149
0.300
Yes
137.266
0.301
Yes
123.205
0.288
Yes
127.467
0.295
ir n
Obs 13515 13515 13515 13295 13540 13540 13,540 13,320
e p
P r 46
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 10 Cliques and governance via the threat of exit
This table reports the results of regressing the firm’s Tobin’s Q on lagged clique ownership, the
ed
indicator for short-selling (MTi,t), their interaction term, and control variables. MTi,t equals 1 if firm
i is allowed for short-selling over year t and 0 otherwise. Detailed definitions of all variables are
reported in the Appendix. The control variables are identical to those in Table 8. Firm FE and Year
FE denote the firm and year fixed effects, respectively. Standard errors clustered at the firm level
iew
are disclosed in parentheses. *, **, and *** denote the significance level of 10%, 5%, and 1%,
respectively.
(1) (2) (3) (4)
TobinQi,t TobinQi,t TobinQi,t TobinQi,t
CliqueOwni,t-1 0.178*
(0.09)
v
CliqueOwni,t-1×MTi,t 0.575***
(0.16)
re
CliqueTop1i,t-1 0.191
(0.15)
CliqueTop1i,t-1×MTi,t 0.985***
er (0.29)
CliqueHHIi,t-1 0.394
(0.54)
CliqueHHIi,t-1×MTi,t 3.452***
(1.15)
pe
CliqueOwn_BKi,t-1 0.217**
(0.10)
CliqueOwn_BKi,t-1×MTi,t 0.607***
(0.17)
Controls Yes Yes Yes Yes
ot
47
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850
Table 11 Cross-sectional analyses
This table reports the subsample regression results of Equation (1) in the text. In Panels A and B,
ed
we split the sample into the Low (High) subsample if the firm’s internal control quality or analyst
coverage is below (above) the sample median. We measure internal control quality with the internal
control index developed by the DIB internal control and risk management database, and analyst
coverage is the number of analysts following the firm over the year. In Panel C, we divide the sample
iew
into subsamples depending on whether the firm belongs to a heavily polluting industry. Finally, we
define firms as industry followers in Panel D if their gross profit margins are below the industry
median. Ind FE and Year FE denote the industry and year fixed effects, respectively. The control
variables are identical to those in Table 2. Standard errors clustered at the firm level are disclosed
in parentheses. *, **, and *** denote the significance level of 10%, 5%, and 1%, respectively.
v
Panel A Internal control Panel B Analyst Coverage
Dep var=ESGi,t+1 (1) Low (2) High (3) Low (4) High
re
CliqueOwni,t 3.920*** 2.168*** 4.472*** 1.601***
(0.98) (0.670) (1.30) (0.611)
Controls Yes Yes Yes Yes
Ind FE Yes Yes Yes Yes
Year FE
F-statistics
Yes
er
24.053
Yes
27.798
Yes
27.029
Yes
22.783
Adj. R2 0.188 0.219 0.202 0.197
Obs 12,799 12,882 14,084 13,510
pe
Coefficients (Low-High) 1.752* 2.872**
Wald Chi2 2.82 4.28
Panel C Heavily polluting firms Panel D Industry followers
Dep var=ESGi,t+1 (1) Yes (2) No (3) Yes (4) No
CliqueOwni,t 3.976*** 1.547** 3.598*** 1.777***
ot
48
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4384850