Bus Strat Env - 2022 - Atif - The Effect of Corporate Environmental Social and Governance Disclosure On Cash Holdings

Received: 21 April 2021 Revised: 17 January 2022 Accepted: 2 February 2022
DOI: 10.1002/bse.3016
RESEARCH ARTICLE
The effect of corporate environmental, social and governance

disclosure on cash holdings: Life-cycle perspective
Muhammad Atif1 | Benjamin Liu2 | Sivathaasan Nadarajah2
1
Department of Accounting and Corporate
Governance, Macquarie Business School, Abstract
Macquarie University, Sydney, New South This paper examines the impact of corporate environmental, social and governance
Wales, Australia
2 (ESG) disclosure on cash holdings, specifically during various stages of the firm life
Department of Accounting, Finance and
Economics, Griffith University, Brisbane, cycle of S&P 1500 indexed firms. Using a sample of 9811 firm-year observations
Queensland, Australia
from 2006 to 2015, we document a significantly negative relation between ESG dis-
Correspondence closure and cash holdings in the introduction, growth and shake-out/decline stages,
Muhammad Atif, Department of Accounting
and those lower cash holdings are associated with higher firm performance and a
and Corporate Governance, Macquarie
Business School, Macquarie University, North positive value of cash. Our findings are robust to alternative econometric specifica-
Ryde, Sydney, NSW 2109, Australia.
tions, alternative measures, additional control variables, propensity score matching
Email: [email protected]
and the use of an instrumental variable approach. Overall, our study offers useful
insights into the global debate on ESG.
KEYWORDS
cash holdings, corporate governance, ESG, life cycle
1 | I N T RO DU CT I O N Jensen, 1986). Self-interested managers may hold more cash on hand

to safeguard themselves against market scrutiny that would occur in
All over the world, environmental, social and governance (ESG) fac- the case of a firm's submission for external funding. Due to a return
tors, commonly known as non-financial performance, have received on cash holdings that is lower than the required cost of capital and
considerable attention from practitioners and regulators. The prior double taxation (Opler et al., 1999), such levels of cash holdings can
research studying the influence of ESG on firm-level outcomes such cause an agency problem (Jensen, 1986). The potential solution to the
as financial performance and market value (e.g., Arouri & agency problem is to have an efficient monitoring mechanism (internal
Pijourlet, 2017; Cheng et al., 2014; Jiao, 2010; Lu et al., 2017), cost of and external) in place (Fama & Jensen, 1983), and an ESG disclosure
equity (El Ghoul et al., 2011), default risk (Atif & Ali, 2021) and finan- serves as both an internal and external monitoring mechanism (Lu
cial risk (Lee & Faff, 2009; Oikonomou et al., 2012) becomes more et al., 2017).1 However, the effect of ESG disclosure as a governance
apparent. Corporate cash holdings, which account for a significant instrument on cash holding decisions has received inadequate atten-
portion of a firm's total assets (Bates et al., 2009), on the other hand, tion in the literature.
allow for managers' discretionary spending (Harford et al., 2008; The corporate life-cycle theory proposes that firms advance from
one stage of development to the next stage in a predictable pattern
(Porter, 2008), with each stage being distinctive and irreversible
Abbreviations: 2SLS, two-stage least squares; CSR, corporate social responsibility; ESG,
environmental, social and governance; FE, fixed effect; FLC, firm life cycle; GFC, global
(Miller & Friesen, 1984; Porter, 2008). Depending on where a firm is
financial crisis; GICS, Global Industry Classification Standards; GRI, Global Responsible
1
Initiative; IV, instrumental variable; OLS, ordinary least squares; PRI, Principles for For instance, a higher G score indicates the strength of internal governance mechanism, and
Responsible Investment; PSM, propensity score matching; S&P, Standard & Poor's; SASB, the E and S scores provide firm-specific incremental information for external monitoring and
Sustainability Accounting Standards Board; VIF, variance inflation factor. decision making (Lu et al., 2017).
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2022 The Authors. Business Strategy and The Environment published by ERP Environment and John Wiley & Sons Ltd.
Bus Strat Env. 2022;31:2193–2212. wileyonlinelibrary.com/journal/bse 2193

10990836, 2022, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/bse.3016 by Cochrane Philippines, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2194 ATIF ET AL.
in its life cycle, its structures, capabilities and strengths, strategies and decisions throughout the firm's life cycle. Therefore, this paper sheds
cash flow unpredictability will vary. In line with life-cycle theory, prior new light on ESG disclosure and cash holdings at various stages of a
studies investigate and confirm the significance of the firm life cycle firm's life cycle.
in determining policies and decisions (Atif & Ali, 2021; DeAngelo Using a panel of S&P 1500 indexed firms with 9811 firm-year
et al., 2006; Faff et al., 2016; Grullon et al., 2002; Hasan & observations for the period 2006–2015, we find a significantly nega-
Habib, 2017). Given that firms at various phases of their life cycle tive relation between overall ESG disclosure and cash holdings, as well
adopt different financial policies and may have varying levels of gov- as specific ESG components (E, S and G) and cash holdings, consistent
ernance mechanisms, it is plausible to expect that the effect of ESG with prior literature. We further find higher (lower) cash holdings in
disclosure on corporate cash holdings is likely to differ across the life- the introduction, growth and shake-out/decline (mature) stages. How-
cycle stages. This study attempts to fill this void in the literature by ever, when the ESG is considered, the formerly positive relation
investigating the effect of ESG disclosure on cash holdings throughout between cash holdings and the three life-cycle stages
the four stages of the life cycle. (i.e., introduction, growth and shake-out/decline) turns negative,
ESG disclosure outlines a firm's overall initiatives towards differ- whereas the relation between cash holdings and the maturity stage
ent stakeholders, including employees, investor communities, regula- becomes positive but insignificant. These findings indicate that firms
tors and society (Campbell, 2007; Malik, 2015). These activities with better ESG disclosure have lower cash holdings because they
differentiate firms apart from the competition, build investor loyalty, have easier access to the capital market as a result of their moral capi-
increase employee retention and enhance operating efficiency tal (Godfrey, 2005) (in the introduction and growth stages) and lower
(Greening & Turban, 2000). Consequently, investors use ESG disclo- agency cost (in the shake-out/decline stage). Our further analysis
sure as a powerful tool when evaluating the firm value and making shows better performance as well as a positive value of cash holdings
investment decisions.2 Recent efforts by high-profile investment in the introduction, growth and shake-out/decline stages for firms
banks further demonstrate the importance of ESG in the investment with better ESG disclosure. Furthermore, our results remain robust
screening process. Goldman Sachs (2007), for example, has launched and consistent when using alternative measures, specifications
the GS SUSTAIN platform, which integrates the features of ESG into (e.g., Tobit model) and additional controls.
its financial analysis. The firm's E and S factor disclosure represents Our study may be prone to endogeneity concerns because man-
the firm's outlook on the environmental (E) and social (S) risk profile, agers who respond to shareholders' demands for lower cash holdings
and it also serves as a tool of external monitoring for investors and may also respond to external concerns about ESG disclosure, thus
other stakeholders. Similarly, better governance (G) indicates the biasing our results. To rule out the concerns of endogeneity, we adopt
strength of the internal governance mechanism. Therefore, the focus several strategies. First, we employ a propensity score matching
that firms have on greater ESG disclosure helps them to be more sus- (PSM) approach to mitigate selection bias caused by unobserved firm-
tainable, boosts investor trust and ultimately influences their financial specific factors. To construct a treatment versus control group, we
decisions (Cheng et al., 2014; Gramlich & Finster, 2013; Wanderley use firm-years of ESG disclosure to rank sample into terciles and
et al., 2008). Given the growing importance and awareness of ESG assign 1 (0) if a firm belongs to the top (bottom) tercile. Our results
disclosure, this study investigates how it affects one of the key corpo- still demonstrate a significant difference in cash holdings between the
rate policies—cash holdings—at different stages of the firm's life cycle. treated and controlled groups, suggesting that greater ESG disclosure
Recent research reveals that cash holdings are related to firm life- has a pronounced negative impact on cash holdings than less ESG dis-
cycle stages in a non-monotonic manner (Faff et al., 2016). According closure. Second, we use an instrumental variable (IV) approach that
to Faff et al. (2016), cash holdings are higher (lower) for firms in the uses two-stage least squares (2SLS) regression analysis and continue
introduction and growth (mature and shake-out/decline) stages due to find a significantly negative effect of ESG disclosure on cash hold-
to more future investment opportunities and limited access to the ings. Overall, our results hold even after adjusting for possible
external capital market (fewer investment opportunities and reduced endogeneity bias.
cash flows). Building on Faff et al.'s (2016) study, we investigate the Our core contributions to the literature are twofold. First, we
differential effect of ESG disclosure on cash holdings across the four contribute to the diverse literature on ESG disclosure, firm life cycle
stages of firm life cycle. Extending the current literature, our study and corporate cash policy. There is a growing body of literature linking
subsequently explores the value of cash holdings and firm perfor- corporate policies with various stages of a firm's life cycle (e.g., Atif &
mance in the different stages of the firm life cycle. Extending Ali, 2021; Dickinson, 2011; Faff et al., 2016; Owen & Yawson, 2010).
Cheung's (2016) findings, we deploy ESG disclosure as the holistic However, prior research into the influence of ESG disclosure on cash
governance mechanism and examine its impact on cash holding holdings as a firm progresses through the different phases of its life
cycle has received scant attention in the literature. Hence, there is a
2
The United Nations introduced the Principles for Responsible Investment (PRI) in 2006 dearth of understanding on how much cash firms with better ESG dis-
(UN PRI, 2006). The PRI has achieved over 1000 signatories from around the world with a
net worth of US$30 trillion under management (UN PRI, 2011). The PRI are considered as
closure hold across the life-cycle stages. Extending these strands of
standards for taking ESG issues into consideration when determining mainstream investment diverse literature, our study shows a significantly negative association
decisions. For instance, climate change policies, carbon emissions, employee rights and
between ESG disclosure and cash holdings in the introduction, growth
welfare, and management strengths have been given more weight in the investment decision
process. and shake-out/decline stages. Overall, our findings shed new light on
ATIF ET AL. 2195
the relation between cash holdings and the individual stages of the the investment screening process (Capelle-Blancard & Monjon, 2014)
firm life cycle in the context of ESG disclosure. compared to conventional methods that simply employ financial infor-
Second, we contribute to the emergent literature that non- mation.4 The comprehensive information provided by ESG disclosure
financial disclosures provide incremental information to investors and provides both internal and external monitoring purposes. For instance,
other stakeholders. A recent study by Lu et al. (2017) finds that better disclosure on E and S factors signals to investors and stake-
extra-financial information in the CSR report (i.e., non-financial holders that firms have firm-specific policies in place to manage envi-
reports) curbs the agency issues connected with managers' misuse of ronmental and social risks. This extra-financial information is useful
corporate resources and helps investors to evaluate firms' investment for decision making and investor monitoring of corporate resources
decisions. Other studies suggest that the issuance of a CSR report (Healy & Palepu, 2001; Lu et al., 2017), providing an external monitor-
reduces the cost of equity capital (Dhaliwal et al., 2011) and analysts' ing mechanism for firms. Similarly, better G factor disclosure reflects
forecast dispersion (Dhaliwal et al., 2012). Building on this line of the strength of the internal governance mechanism, thus providing
research, our study demonstrates that firms with better ESG disclo- information about the internal monitoring mechanism. Better internal
sure result in higher firm performance and positive value of cash and external monitoring mechanisms help to reduce the agency prob-
holdings, especially during the introduction, growth and shake-out/ lem. In addition, ESG helps in stabilising the financial performance
decline stages. Our study shows ESG disclosure as a holistic gover- through corporate image and reputation and in maintaining a better
nance mechanism that provides both internal and external monitoring relationship with financial markets (Carter, 2005; Cornell &
purposes. Overall, our research adds to the ongoing debate about Shapiro, 1987; McGuire et al., 1988). In the case of a negative event,
the importance of ESG disclosure for firms, investors and other stakeholders are likely to punish the firm; however, a firm with ESG
stakeholders. initiatives is less likely to be affected due to its moral capital among
The rest of the paper is organised as follows: Section 2 reviews stakeholders (Atif & Ali, 2021; Godfrey, 2005). ESG fosters investor
the literature and develops the hypothesis; Section 3 describes the loyalty and trust, resulting in funds being available even during a crisis
sample and summary statistics and elaborates on the research (Godfrey et al., 2009). Therefore, ESG acts as a form of ‘insurance’ for
method; Section 4 presents the empirical results, as well as the sensi- firms.
tivity and endogeneity tests; and Section 5 concludes the study. Following the growing interest in sustainability, the need for
information about firms' ESG activities and policies has steadily
increased (e.g., Amel-Zadeh & Serafeim, 2018; Cohen &
2 | LITERATURE REVIEW AND Simnett, 2015). In response to this demand, several countries, includ-
HYPOTHESIS DEVELOPMENT ing Australia (2003), China (2008), South Africa (2010) and the UK
(2013), have enacted mandatory ESG disclosure regulations, requiring
2.1 | Background: ESG factors firms to properly disclose information on ESG issues in traditional
financial disclosures or in specialised standalone reports (Krueger
ESG initiatives broadly cover a firm's environmental, social and gov- et al., 2021). However, in the USA, 83% of firms registered with the
ernance factors.3 The environmental factor refers to the firm's use Securities and Exchange Commission (SEC) have disclosed some sus-
of energy, use of water, carbon emissions, pollution and natural tainability information in their regulatory filings, even though much of
resources conservation. It involves an assessment of environmental the sustainability information is voluntary in the context of the USA
risks to the firm's income and how the firm is managing such poten- (Christensen et al., 2021; SASB, 2017). Investors and other stake-
tial risks. For instance, a firm might face risks related to an oil spill, holders frequently rely on ESG information for investment perfor-
waste water disposal, toxic emissions or potential fines due to non- mance, the product development process (e.g., green bonds), ethical
compliance. The social factor, referring to the firm's relationship issues and product strategy (i.e., differentiation vs. low-price strategy).
with society, includes employee health, safety and rights at the Research that explicitly examines ESG disclosure in the USA is rela-
workplace, as well as product safety and social philanthropy. It eval- tively limited to one factor of ESG (e.g., governance).5 The sole focus
uates the possible risks such as labour inefficiency and compliance on the governance factor could undermine the ethical/moral and
costs as a result of inadequate workplace health and safety. The social considerations that are essential for societies and future gener-
governance factor includes board independence, diversity, share- ations' common good (Richardson, 2009). Therefore, ESG disclosure
holder rights and disclosure of information. It assesses the informa- presents an interesting backdrop to investigate its effect on firm
tion required by investors and stakeholders on accounting policies.
transparency, conflict of interest and strength of governance
structure.
ESG has recently emerged as an important and powerful mea-
surement for making investment decisions. Investors use ESG disclo- 4
According to a survey conducted by Yeldar (2013), analysts and individual investors describe
sure as an ‘extra-financial’ evaluation and ‘material’ information in non-financial performance as extra-financial information that allows them to get in-depth
knowledge of the advantages and challenges associated with firm future.
5
Extant literature examines the effect of issuance of a CSR report on cash holdings
3
More details about ESG coverage can be found at https://www.msci.com/what-is-esg. (e.g., Arouri & Pijourlet, 2017; Lu et al., 2017).
2196 ATIF ET AL.
2.2 | ESG and cash holdings profitability ratio as proxies (e.g., DeAngelo et al., 2010; Khan &
Watts, 2009; Porter, 2008). Realising the limitations of the univariate
Cash holdings, the most liquid among corporate assets, provide measure and the reality that the life cycle of a firm is influenced by a
managers' discretion over their expenditure, with the two major variety of factors, subsequent research has turned to a composite of
motivations falling into operational requirements and the agency variables (e.g., age, assets/sales growth, dividend payout and capital
problem. Cash holdings, in terms of operational necessities, help to expenditures) to derive a life-cycle proxy (e.g., Anthony &
reduce the transaction costs and, more importantly, protect against Ramesh, 1992; Faff et al., 2016). Dickinson (2011) provides a parsimo-
future funding and underinvestment risk, commonly known as pre- nious firm-level life-cycle measure based on the predicted behaviour
cautionary motives (Bates et al., 2009; Han & Qiu, 2007). The of cash flows at various stages of the life cycle. According to
agency problem, on the other hand, which comes from separation Dickinson (2011) and Livnat and Zarowin (1990), cash flows from
of ownership and control, leads to increased cash holdings, with operating, investing and financing activities capture a firm's profitabil-
opportunistic managers retaining cash reserves to extract private ity, growth and risk, and the combination of the three types of cash
benefits (Jensen, 1986; Malmendier & Tate, 2008). In the absence flows provides a reliable proxy for classifying firms into life-cycle
of market scrutiny, managers retain cash levels higher than those stages. As a result, recent empirical research on accounting and
required to finance positive net present value (NPV) projects, at the finance deploys this measure as a suitable proxy for firm life-cycle
expense of shareholders. Higher cash reserves are also considered stages (e.g., Faff et al., 2016; Hasan et al., 2015; Koh et al., 2015).
to be detrimental to the firm value and shareholder wealth because Faff et al. (2016) and Hasan and Habib (2017) assert that a firm's
interest income on cash deposits earns a return less than the firm's life-cycle stage is an important determinant of corporate policies.
cost of capital (Tong, 2010) and is subject to double taxation (Opler Recent research has documented that dividend policy (DeAngelo
et al., 1999).6 et al., 2006; Grullon et al., 2002), equity offerings (DeAngelo
An efficient monitoring mechanism could be a potential solution et al., 2010), takeover activity (Owen & Yawson, 2010), financing,
to the agency problem (Fama & Jensen, 1983). Given the importance investment and cash policies (Faff et al., 2016), CSR activity involve-
of ESG disclosure associated with internal as well as external monitor- ment (Hasan & Habib, 2017) and default risk (Atif & Ali, 2021) are pre-
ing mechanisms, ESG disclosure provides firm-specific information to dictable and related to a firm's life-cycle stage.
capital providers to monitor the use of cash holdings. External moni- Corporate cash holdings, an important policy that bears firm strat-
toring mechanisms (E and S) influence managers' opportunistic behav- egy and managerial motivation, are related to life-cycle stages in a
iour, resulting in lower cash holdings. Similarly, the G factor appraises non-monotonic manner (Faff et al., 2016). Firms at different life-cycle
the structure of the internal governance mechanism that restrains the stages have distinct investment opportunities and cash flow patterns
managerial rent-seeking in cash holdings. Fama (1980) argues that (Dickinson, 2011; Miller & Friesen, 1984), and they adopt cash policies
strong governance is critical in alleviating the agency problem. There- accordingly (Faff et al., 2016). Firms in the early stages of develop-
fore, improved ESG disclosure leads to better monitoring mechanisms ment lack a customer base and loyalty, as well as lack of knowledge
that mitigate the agency problem of holding excess cash. about potential revenue, and operational parameters and industry
dynamics (Jovanovic, 1982). These firms, which have negative cash
flows from operating and investing activities (Dickinson, 2011;
2.3 | ESG, cash holdings and firm life cycle Jovanovic, 1982), rely on external financing (via debt and equity) for
cash supply (Barclay & Smith, 2005; Jensen, 1986; Myers, 1984).
According to firm life-cycle theory, changes in internal (e.g., strategy These firms, on the other hand, make an early large investment to
choice, resources and managerial capability) and external factors capture opportunities and to deter rival entry (Spence, 1977, 1979,
(e.g., market competition and macroeconomic conditions) cause dis- 1981). Due to lack of capabilities at this stage, introduction firms have
tinct phases in a firm's development, and firms inevitably evolve and limited access to external resources and are therefore compelled to
transit from one phase to the next (Dickinson, 2011; Porter, 2008). hold high cash reserves to meet operational needs (Faff et al., 2016).
The theory also claims that firms would follow a predictable pattern Firms in the growth stage enjoy higher market reputation and sales
characterised by development phases that are difficult to reverse turnover, and improved profit margins (Hay & Ginter, 1979;
(Miller & Friesen, 1984; Porter, 2008). Wernerfelt, 1985), resulting in positive operating cash flows
The extant literature (e.g., Dickinson, 2011; Gort & (Dickinson, 2011; Spence, 1977, 1979, 1981). Growth firms continue
Klepper, 1982) identifies and classifies life-cycle stages as ‘introduc- to make large investments to overcome entry barriers and experience
tion’, ‘growth’, ‘maturity’ and ‘shake-out and decline’. Much research negative investing cash flows (Dickinson, 2011; Spence, 1977, 1979,
effort has been devoted to measuring firm life-cycle stages and cap- 1981). As these firms continue to rely on external resources, resulting
turing their effects on firm policies and performance. Some studies in positive financing cash flows (Dickinson, 2011), growth firms are
use univariate firm life-cycle indicators such as firm age, size and the further motivated to have high cash holdings to capture investment
opportunities (Faff et al., 2016).
6 Firms in the mature stage enjoy the maximum operating efficien-
Lower cash holdings motivate managers to run the firm more efficiently, resulting in higher
shareholder value (La Porta et al., 2000). cies and thus have positive and stable operating cash flows
ATIF ET AL. 2197
(Dickinson, 2011; Hasan & Habib, 2017). These firms have fewer H1. ESG disclosure negatively affects cash holdings in
investment opportunities and still have negative investing cash flows, the introduction stage of the firm life cycle.
although they are less than the earlier stages' investing cash outflows
(Dickinson, 2011; Jovanovic, 1982; Wernerfelt, 1985). In terms of H2. ESG disclosure negatively affects cash holdings in
financing, mature firms shift their focus away from acquiring financing the growth stage of the firm life cycle.
to servicing debt and distributing excess funds to shareholders, lead-
ing to negative financing cash flows (Barclay & Smith, 2005; H3. ESG disclosure does not affect cash holdings in the
Dickinson, 2011; Jensen, 1986; Myers, 1977). Cash holdings for mature stage of the firm life cycle.
mature firms are predicted to be lower (Faff et al., 2016) than for
other life-cycle stages, due to the fact that their investments are more H4. ESG disclosure negatively affects cash holdings in
likely to be geared towards the maintenance of asset value (Hasan & the shake-out/decline stage of the firm life cycle.
Habib, 2017) and their corporate governance mechanisms are in place
and functioning, reducing the agency problem. Firms in the shake-
out/decline stage face a variety of issues, including reduced market 3 | RESEARCH DESIGN
shares and declining growth rates, and thus declining prices and prof-
itability, downgraded resources and managerial capabilities, and thus 3.1 | Sample
fragile financial performance, resulting in unstable operating cash
flows (Dickinson, 2011; Hasan & Habib, 2017; Wernerfelt, 1985). Our initial sample consists of Standard & Poor's (S&P) 500, S&P
These firms see no future investment opportunities; instead, they liq- MidCap 400 and S&P SmallCap 600 (thus S&P 1500) indexed firms,
uidate assets to service existing debt and to support operations, which obtained from Bloomberg, covering the period 2006–2015.8 The
results in positive investing cash flows (Dickinson, 2011). In terms of Bloomberg scoring framework incorporates uniform international
cash policy, Faff et al. (2016) find that these firms will reduce cash guidelines such as the Global Responsible Initiative (GRI) and the
holdings due to a reduction in the internal cash flow and external United Nations. Disregarding industry variances, we initially choose
financing. In contrast, we argue that shake-out/decline firms, despite 14,494 firm-years containing information on ESG disclosure and
reduced cash flow, will hold high cash reserves to reduce firm risk due accounting measures. Consistent with prior works (e.g., Atif
to the agency problem.7 Risk-eschewing managers are motivated to et al., 2019; Bates et al., 2009), we exclude firms operating in the
hold excess cash, a risk-free asset, to reduce firm (thus their own) risk financial services industry due to the stringent liquidity conditions.
(Jensen & Meckling, 1976; Tong, 2010) to avoid market scrutiny. We also require the sample firm-years to have ESG and accounting
The impact of life-cycle stages on cash holdings is moderated by data in the final sample.9 Therefore, our final sample has 9811 firm-
corporate ESG disclosure because the ESG factors serve both internal years of data on 1290 non-financial firms.
and external monitoring purposes. Firms in the introduction and
growth stages with better ESG disclosure have better internal gover-
nance, which mitigates the opportunistic behaviour of managers for 3.2 | Variables and descriptive statistics
holding excess cash; they enjoy the trust and loyalty of external inves-
tors and stakeholders, which enables them to garner support and 3.2.1 | Corporate cash holdings
access to the capital market (Cheng et al., 2014; Clarkson et al., 2011;
Ferrell et al., 2016; Hasan & Habib, 2017; Udayasankar, 2008). Hence, To measure the level of cash holdings (ratio_of_cash), we follow prior
young firms (introduction and growth) with better ESG initiatives are studies (e.g., Atif et al., 2019; Bates et al., 2009; Nikolov &
encouraged to keep their cash holdings low. Mature firms are more Whited, 2014) and use the ratio of cash and marketable securities to
likely to have good governance mechanisms in place and a well- net assets, where net assets are defined as the book value of the total
established market reputation; therefore, ESG disclosure is likely to assets minus cash and marketable securities. Notably, such a measure
have a minimal effect on cash holdings. Shake-out/decline firms are of cash represents the cash reserves available at the disposal of man-
more likely to observe the agency problem due to deteriorated finan- agers in proportion to the assets. During our sample period, the aver-
cial performance and managerial opportunism (Faff et al., 2016); those age cash holdings ratio is 15.3%, as shown in Table 2 (Panel A), based
firms with better ESG disclosure are likely to hold fewer cash reserves on the full sample. Further, we divide the sample into high and low
due to the mitigating effect of ESG disclosure. Therefore, we hypo- ESG disclosure, based on the sample median. The subsample of high
thesise that ESG disclosure has a distinct effect on cash holdings at ESG disclosure shows 13.3% of average cash holdings compared to
different stages of a firm's life cycle: 18.1% for that of low ESG. The mean difference between high and
low ESG disclosure is significant at 1% level. Our study controls
7 8
Cash flow and cash holdings are two related but different issues. While cash flow indicates We start our sample from 2006 as ESG disclosure data from Bloomberg are mainly available
the source of cash, cash holdings are the result of a firm's cash policy decision, which is from this year.
9
determined by the firm's predicted future needs for cash and the agency motive for We collect firm age data from the OSIRIS database before merging it with the Bloomberg
holding cash. data.
2198 ATIF ET AL.
several firm-specific variables that may influence the cash level, which analysis, we also employ the individual scores of environmental (E),
we introduce in Section 3.2.3 (see Table 1 for definitions). social (S) and governance (G) disclosures, which represent the firm dis-
closure on each factor. For instance, environmental scores are based
on firm disclosure on climate change policies, hazardous wastes,
3.2.2 | Measuring ESG disclosure and firm life cycle nuclear energy and sustainability indicators, among others; social
scores are based on consumer protections, diversity, human rights,
Following prior studies (e.g., Atif & Ali, 2021; Cheng et al., 2014), we animal rights, welfare, child labour, and employee health and safety
use Bloomberg's ESG score as an umbrella measure to assess the indicators, among others; and governance scores are based on man-
firm's ESG disclosure. ESG score ranges from 0.1 to 100, with high agement structure, diversity, executive compensation and conflict of
score indicating more transparency and disclosure. In our additional interest indicators, among others. Moreover, we identify firms in the
TABLE 1 Definitions of variables
Notation Variable name Measure

Panel A. Dependent variables
ratio_of_cash Cash holdings The ratio of cash and marketable securities
to net assets
cashta Cash holdings The ratio of cash and marketable securities
to total assets
cash (industry adj) Cash holdings The ratio of cash and marketable securities
scaled by industry adjusted cash holdings
Ln_cash Cash holdings Log of cash and marketable securities
Panel B. Independent variables
ESG_score Environmental, social and governance score ESG disclosure score ranging from 0.1 to
100
E_score Environmental score Environment disclosure score ranging from
0.1 to 100
S_score Social score Social disclosure score ranging from 0.1 to
100
G_score Governance score Governance disclosure score ranging from
0.1 to 100
ESG_index ESG index A variable ranging from 0 to 1 measuring
the average extent of ESG disclosure
FLC Firm life cycle A vector of dummy variables that captures
different stages of firm life cycle
(Dickinson, 2011)
Panel C. Control variables
RETA Retained earnings to total assets Retained earnings to total asset
LEV Leverage The sum of short- and long-term debt
divided by total assets
CFTA Cash flow Cash flow to total assets
CAPEX Capital expenditure The ratio of capital expenditure to total
assets
MTB Growth opportunities The market value of equity scaled by the
book value of equity
Div_Dum Dividend A dummy variable equal to 1 if the dividend
was paid and 0 otherwise
ROA Return on assets Firm net income divided by total assets
R&D Research and development Total research and development
expenditure scaled by total assets
Size Size of firm Log of total assets
Firm_age Age of firm The number of years from the year of
incorporate as reported in OSIRIS
ATIF ET AL. 2199
TABLE 2 Descriptive statistics
Full sample (N = 9811) High ESG Low ESG
Std. 1st 3rd Std. Std.

Variable Mean dev. quartile Median quartile Mean dev. Mean dev. Mean diff. t-stat
Panel A. Dependent variable
ratio_of_cash 0.153 0.272 0.024 0.079 0.186 0.133 0.234 0.181 0.314 0.048*** 9.703
Panel B. Independent variables
ESG_score 19.389 11.860 11.842 14.050 21.488 28.152 13.041 12.017 1.323 16.135*** 87.759
E_score 20.743 17.075 6.207 15.625 33.103 21.628 16.966 1.995 1.410 19.632*** 15.649
S_score 18.580 15.108 8.772 14.035 24.561 23.806 15.123 6.145 2.747 17.660*** 51.671
G_score 52.275 5.978 48.214 51.786 55.357 55.925 5.914 48.882 3.544 7.043*** 72.111
ESG_index 0.699 0.283 0.333 0.667 1.000 0.940 0.128 0.475 0.187 0.465*** 14.001
Intro 0.025 0.155 0.000 0.000 0.000 0.020 0.141 0.031 0.173 0.010*** 3.609
Growth 0.307 0.461 0.000 0.000 1.000 0.291 0.454 0.329 0.470 0.038*** 4.392
Maturity 0.575 0.494 0.000 1.000 1.000 0.595 0.491 0.546 0.498 0.048*** 5.209
Shake-out/ 0.094 0.292 0.000 0.000 0.000 0.094 0.292 0.094 0.292 0.000 0.038
decline
Panel C. Control variables
RETA 0.146 1.309 0.000 0.228 0.436 0.180 1.257 0.099 1.378 0.080*** 3.357
LEV 0.237 0.212 0.064 0.218 0.354 0.249 0.216 0.221 0.206 0.280*** 7.202
CFTA 0.051 0.113 0.010 0.055 0.098 0.052 0.121 0.050 0.101 0.002 1.176
CAPEX 0.054 0.061 0.067 0.035 0.019 0.055 0.056 0.052 0.067 0.002** 2.374
MTB 4.454 6.187 1.531 2.319 3.757 4.652 6.842 4.190 5.283 0.461 0.940
Div_Dum 0.605 0.489 0.000 1.000 1.000 0.675 0.468 0.508 0.500 0.166*** 18.772
ROA 5.653 11.153 2.553 5.572 9.530 5.901 11.461 5.312 10.706 0.588*** 2.856
R&D 0.031 0.073 0.000 0.003 0.040 0.030 0.079 0.033 0.063 0.003*** 2.307
Size 3.375 0.720 2.860 3.334 3.840 3.580 0.770 3.090 0.523 0.489*** 39.331
Firm_age 29.735 27.491 12.000 20.000 38.000 32.779 30.101 25.470 22.674 7.308*** 13.729
Note: This table presents the descriptive statistics on the full sample and subsamples of high and low ESG disclosure. ***, **, and *, represent significance at
the 1%, 5%, and 10% levels, respectively. All variables are defined in Table 1.
Bloomberg ESG data set that fail to provide information (E, S and G • Growth: if ONCF > 0, INCF < 0 and FNCF > 0;
scores) on the three factors. We then employ the ESG index ranging • Maturity: if ONCF > 0, INCF < 0 and FNCF < 0; and
from 0 to 1, based on the average extent to which ESG scores are • Shake-out/decline: if ONCF < 0, INCF > 0 and FNCF ≤ 0 or ≥0.
disclosed.
To quantify the firm life cycle, we use the proxies of We also use the life-cycle proxy of retained earnings to total
Dickinson (2011) and DeAngelo et al. (2006),10 which are both exten- assets ratio (RETA) developed by DeAngelo et al. (2006), which
sively used in the literature (e.g., Faff et al., 2016; Owen & gauges a firm's reliance on external funds or internal liquidity.
Yawson, 2010). Our cash flow patterned-based classification of firms DeAngelo et al. (2006) argue that RETA is a good proxy for a firm's
into life-cycle stages, based mainly on Dickinson (2011), is consistent life cycle because of its composition of equity and external finance. A
with the theory. Specifically, based on the following net cash flow pat- higher (lower) than average RETA denotes that the firm is mature
terns from operating (ONCF), investing (INCF) and financing (FNCF) (young and growing), as a result of the accumulation of profits over
activities, we classify firms into four stages: introduction, growth, time and investment activities (DeAngelo et al., 2006). We use a
maturity and shake-out/decline: dummy variable equal to 1 if a firm has a high RETA and 0 otherwise.
To demonstrate the impact of ESG on cash holdings, we use the
• Introduction: if ONCF < 0, INCF < 0 and FNCF > 0; interaction between ESG (umbrella measure) and the business life-
cycle stages.
10
For two reasons, we do not adopt Anthony and Ramesh's (1992) classification procedure Panel B in Table 2 presents the summary statistics of ESG disclo-
for firm life cycle. First, that procedure leads to erroneous classification of firms in the life
sure and the firm life cycle. Based on the full sample, we find an aver-
cycle (Dickinson, 2011). Second, the classification procedure is ‘ad hoc’, relying on portfolio
sorts to categorise firms into various life-cycle stages. age of 19.3 ESG_score, 20.7 E_score, 18.5 S_score and 52.2 G_score,
2200
TABLE 3 Correlation matrix
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1 ratio_of_cash 1.000
2 ESG_score .009 1.000
3 E_score .003 .971 1.000
4 S_score .024 .856 .723 1.000

5 G_score .052 .750 .663 .640 1.000
6 Intro .028 .037 .036 .040 .007 1.000
7 Growth .099 .128 .134 .078 .061 .055 1.000
8 Maturity .027 .114 .124 .063 .055 .134 .019 1.000
9 Shake-out/decline .106 .019 .012 .029 .004 .027 .163 .397 1.000
10 RETA .002 .051 .049 .037 .041 .044 .052 .056 .002 1.000
11 LEV .346 .031 .013 .065 .068 .073 .194 .198 .010 .056 1.000
12 CFTA .401 .120 .143 .045 .020 .231 .339 .381 .042 .110 .328 1.000
13 CAPEX .230 .055 .072 .032 .012 .014 .230 .139 .127 .001 .050 .497 1.000
14 MTB .035 .045 .046 .024 .038 .002 .036 .038 .009 .003 .108 .089 .010 1.000
15 Div_Dum .196 .168 .144 .174 .195 .055 .098 .105 .007 .117 .083 .021 .007 .016 1.000
16 ROA .267 .095 .099 .057 .049 .140 .177 .203 .021 .155 .313 .573 .099 .083 .027 1.000
17 R&D .382 .136 .171 .047 .046 .096 .110 .036 .083 .093 .234 .358 .215 .016 .249 .176 1.000
18 Size .234 .471 .450 .400 .430 .056 .015 .003 .001 .020 .144 .057 .032 .002 .218 .033 .079 1.000
19 Firm_age .129 .116 .129 .061 .096 .014 .054 .046 .012 .092 .068 .022 .082 .022 .240 .027 .039 .115 1.000
Note: The table presents the correlation matrix among all the variables used in this study. Bold coefficients show high correlations. All variables are defined in Table 1.
ATIF ET AL.
ATIF ET AL. 2201
while ESG_index shows a 0.69 average disclosure. In the subsamples 3.3 | Correlation matrix
of high versus low ESG disclosure, all the ESG-related variables have
higher averages, and the mean differences are significant at 1% level. Table 3 reports the correlations among the variables in our regression
Across the firm life-cycle stages, our sample shows that the mature model to check the possibility of multicollinearity problem. The highest
stage has the highest per cent of firms (57.5%), followed by the correlations exist between ESG_score and each individual score E_score,
growth stage (30.7%) and the shake-out/decline stage (9.4%), with S_score and G_score (.9, .8 and .7, respectively). Individual E, S and G
only 2.5% of firms in the introduction stage. scores also have strong correlation. In general, a correlation greater
The subsamples show the per cent of firms in high versus low than 0.7 may suggest a problem with multicollinearity (e.g., Liu
ESG disclosure (e.g., intro: 3.1% vs. 2.0%; growth: 32.9% vs. 29.1%). et al., 2014). The correlations among these variables are not an issue
The difference in mean comparisons in the firm life-cycle stages is also for our study because we use highly correlated variables in different
significant at 1% level, except for the shake-out/decline stage. regressions instead of simultaneously in the model. There are no corre-
Table S1 shows the summary statistics in the subsamples of the four lation coefficient values higher than .7 for the remaining variables. Fur-
life-cycle stages. Shake-out/decline is led by cash holdings (25.5%), ther, we calculate the variance inflation factor (VIF) and find VIF less
followed by the introduction (21.9%). Mature firms have an average than 3.5, while the overall mean value is 3.5, implying that
ESG_score of 21.0, followed by shake-out/decline firms (19.2). All the multicollinearity is not an issue in the model.11
variables are defined in Table 1.
3.4 | Model and estimation method

3.2.3 | Controls
As in prior studies, we begin by testing the general relation between
Our firm-specific controls, which are expected to influence ESG and cash holdings and ESG using the baseline regression model below:
cash holdings, are based on previous research (e.g., Atif et al., 2019;
X
Faff et al., 2016; Hasan & Habib, 2017). To begin, we control for RETA ratio_of_cashit ¼ α þ β1 ðESGÞit þ δ2 ðZÞit þ δ3 ðindustry effectsÞi
X
to determine how reliant a firm is on external capital (14.6% mean þ δ4 ðyear effectsÞt þ εit ð1Þ
value). In the comparison of subsamples (18.0% vs. 9.9%), firms with
high ESG have a higher RETA ratio. Second, we account for leverage, Then, we use a modified baseline regression model to test our
measured as the book value of total debt to total assets (LEV), with an hypotheses:
average value of 23.7%. The relation between cash holdings and
leverage is ambiguous (Ozkan & Ozkan, 2004) because higher debt ratio_of_cashit ¼ α þ βX
1 ðESGÞit þ β2 ðFLCÞit þ δ3 ðX
ESG FLCÞit þ δ4 ðZÞit
exposes firms to bankruptcy, while highly leveraged firms use higher þ δ5 ðindustry effectsÞi þ δ6 ðyear effectsÞt þ εit
cash holdings to hedge the risk of financial distress. Third, we include ð2Þ
cash flow (CFTA), defined as the ratio of cash flows to total assets,
because of its sensitivity to cash holdings; it has an average value of where we measure ratio_of_cash (in both models) as cash and market-
5.1%. Fourth, we consider capital expenditure (mean value 5.4%), as able securities to net assets, and net assets are defined as the book
the ratio of capital expenditure to total assets (CAPEX), by taking value of total assets minus cash and marketable securities. ESG (in both
future expenditures into account. Fifth, to add growth opportunities models) is the score ranging from 0.1 to 100 (provided by Bloomberg).
(MTB), we use the market value of equity divided by the book value of Following prior studies (e.g., Faff et al., 2016; Hasan & Habib, 2017),
equity with an average of 4.4. Sixth, as a control measure, we include we use FLC (in model 2) that Principles for Responsible Investment rep-
dividend payment (Div_Dum), a dummy equal to 1 if paid and 0 other- resents the firm life-cycle stages (introduction, growth, maturity and
wise (0.6 mean value) because firms decide how much cash to keep shake-out/decline). The vector Z (in both models) represents controls,
on hand to pay dividends to shareholders. Seventh, we control for such as LEV, CFTA, CAPEX and MTB. The industry effects are based on
return on assets (ROA) as a measure of internal return, with an aver- the two-digit code of the Global Industry Classification Standards
age of 5.6%. Eighth, we consider research and development (R&D) as (GICS), and the year effects are the sample year (2006–2015).
the ratio of R&D to total assets, which has consequences for ESG We estimate the baseline using ordinary least squares (OLS) with
(McWilliams & Siegel, 2001), with a mean value of 3.1. Ninth, we industry and year fixed effects. Further, we replace the contempora-
include size of the firm (Size), defined as the natural logarithm of total neous variables with 1-year lagged (t 1) controls. The rationale for
assets (3.3 on average). Additionally, we use firm age (Firm_age) as the this is that ESG disclosure and firm-level characteristics may require
number of years since incorporation, with a mean value of 29.7. Older time to influence the cash holding decisions of a firm. To control for
and mature firms are better able to engage in ESG activities to seize heteroscedasticity, we correct the standard errors for clustering of
the opportunities (Attig et al., 2014), and our sample shows that older residuals at the firm level (Petersen, 2009).
firms have higher ESG scores (32.7 vs. 25.4). The difference of mean
comparison in all the firm-specific variables is significant at 1% level, 11
Lardaro (1993) suggests that multicollinearity can cause an issue if VIF exceeds 10. For the
except for CFTA and MTB. sake of brevity, we do not report VIF results.
2202 ATIF ET AL.
TABLE 4 The effect of ESG disclosure on cash holdings
Panel A Panel B Panel C
ratio_of_cash
Variable OLS (1) OLS (2) Lagged (3) OLS (4) OLS (5) Lagged (6) OLS (7) OLS (8) OLS (9)
ESG_score 0.002*** 0.001*** 0.001** — — — — — —
(15.36) (2.74) (2.10)
ESG_index — — — 0.135*** 0.017** 0.027* — — —
(14.05) (2.05) (1.83)
E_score — — — — — — 0.001* — —
(1.91)
S_score — — — — — — — 0.002*** —
(2.41)
G_score — — — — — — — — 0.002***
(2.84)
RETA — 0.004* 0.000 — 0.004 0.000 0.000 0.006*** 0.004*
(1.88) (0.04) (0.97) (0.01) (0.30) (4.16) (1.89)
LEV — 0.151*** 0.038 — 0.155*** 0.035 0.114*** 0.123*** 0.152***
(8.12) (1.60) (9.24) (1.49) (8.29) (8.08) (8.20)
CFTA — 0.074 (1.61) 0.050 — 0.081 (0.51) 0.046 0.235*** 0.226*** 0.075 (1.62)
(0.85) (0.79) (5.96) (5.58)
CAPEX — 0.232*** 0.051 — 0.221** 0.050 0.163*** 0.295*** 0.229***
(3.73) (0.71) (2.28) (0.70) (3.29) (5.18) (3.68)
MTB — 0.000** 0.000 — 0.000*** 0.000 0.000* 0.000 (1.25) 0.000**
(2.16) (0.19) (3.47) (0.17) (1.71) (2.17)
Div_Dum — 0.000 0.017** — 0.000 (0.04) 0.017** 0.007 0.002 0.000
(0.08) (2.12) (2.13) (1.46) (0.28) (0.06)
ROA — 0.001*** 0.000 — 0.001 0.000 0.001*** 0.001*** 0.001***
(3.48) (0.10) (0.83) (0.17) (3.70) (4.17) (3.51)
R&D — 1.450*** 0.228*** — 1.471*** 0.216*** 0.450*** 0.708*** 1.456***
(23.63) (3.14) (3.46) (2.98) (7.78) (14.13) (23.82)
Size — 0.065*** 0.029*** — 0.060*** 0.027*** 0.032*** 0.054*** 0.064***
(11.60) (4.52) (10.32) (4.23) (8.62) (12.90) (12.14)
Firm_age — 0.000*** 0.000 — 0.000*** 0.000 0.000*** 0.000*** 0.000***
(4.08) (0.04) (6.32) (0.06) (4.86) (3.87) (4.07)
Constant 0.193*** 0.368*** 0.187*** 0.241*** 0.353*** 0.191*** 0.204*** 0.308*** 0.322***
(38.32) (11.07) (3.97) (27.97) (12.40) (4.10) (7.12) (11.52) (7.76)
Industry Yes Yes Yes Yes Yes Yes Yes Yes Yes
effects
Year Yes Yes Yes Yes Yes Yes Yes Yes Yes
effects
N 9811 9811 8643 9811 9811 8643 9811 9811 8643
2
Adj. R .131 .228 .192 .121 .227 .193 .345 .286 .227
Note: This table presents the regression results where ESG disclosure is measured by a score ranging from 0 to 100. Panel A presents the results when
cash holdings are measured by the ratio of cash and cash equivalents to net assets using OLS and 1-year lagged control variables. Panel B presents the
results when ESG is measured as the extent of disclosure ranging from 0 to 1 using OLS and 1-year lagged variables. Panel C presents the results with
individual ESG scores (E, S and G) using OLS. OLS method employs the regression while controlling for industry and year effects. The 1-year lagged control
variables employ the industry- and year-effect estimations where contemporaneous variables are replaced by lagged variables in the regression model. The
robust t-statistic of each coefficient is shown in parentheses. All variables are defined in Table 1.
Coefficients are reported at 1%, 5% and 10% levels of significance with ***, ** and *, respectively.
ATIF ET AL. 2203
TABLE 5 The effect of ESG disclosure on cash holdings: Life-cycle perspective
ratio_of_cash
Panel A OLS (1) Lagged (2) OLS (3) Lagged (4)

Intro 0.001* (1.82) 0.001* (1.89) — —
Growth 0.023*** (3.50) 0.025*** (2.46) — —
Maturity 0.044*** (4.78) 0.048*** (3.14) — —
Shake-out/decline 0.057*** (2.92) 0.056*** (2.73) — —
High_RETA — — 0.014*** (2.33) 0.019*** (2.39)
RETA 0.006 (1.50) 0.007 (1.53) — —
LEV 0.159*** (9.62) 0.167*** (9.74) 0.163*** (8.89) 0.156*** (9.55)
CFTA 0.109 (1.02) 0.121 (1.00) 0.103*** (2.51) 0.142 (1.11)
CAPEX 0.204*** (2.74) 0.219*** (2.72) 0.208*** (3.46) 0.154* (1.66)
MTB 0.003*** (3.31) 0.005*** (3.56) 0.004*** (2.04) 0.003*** (3.40)
Div_Dum 0.001 (0.16) 0.002 (0.26) 0.000 (0.05) 0.003 (0.29)
ROA 0.011 (0.23) 0.009 (0.29) 0.010 (0.61) 0.012 (0.28)
R&D 1.350*** (3.80) 1.361*** (3.83) 1.371*** (23.97) 1.319*** (3.59)
Size 0.058*** (13.78) 0.057*** (13.59) 0.057*** (13.57) 0.057*** (13.29)
Firm_age 0.020*** (7.50) 0.022*** (7.22) 0.027*** (3.95) 0.032*** (7.86)
Constant — — 0.393*** (12.60) 0.341*** (12.95)
Industry effects Yes Yes Yes Yes
Year effects Yes Yes Yes Yes
N 9811 7643 9811 7643
2
Adj. R .224 .226 .219 .224
ratio_of_cash
Panel B. The effect of ESG disclosure and life
cycle on cash holdings OLS (1) Lagged (2) OLS (3) Lagged (4)
ESG_score 0.001*** (2.26) 0.001*** (3.42) 0.000 (0.97) 0.001*** (2.73)
Intro 0.053*** (2.62) 0.049*** (2.43) — —
Intro ESG_score 0.006*** (2.76) 0.009*** (2.39) — —
Growth 0.019 (1.48) 0.020 (1.17) — —
Growth ESG_score 0.001*** (2.18) 0.002*** (2.09) — —
Maturity 0.049*** (3.57) 0.051*** (2.17) — —
Maturity ESG_score 0.001 (1.41) 0.003 (1.67) — —
Shake-out/decline 0.046*** (2.58) 0.049*** (2.36) — —
Shake-out/decline ESG_score 0.002*** (2.17) 0.003*** (2.12) — —
High_RETA — — 0.017 (1.19) 0.019 (1.33)
High_RETA ESG_score — — 0.001* (1.89) 0.003* (1.96)
RETA 0.004* (1.88) 0.005 (0.99) — —
LEV 0.159*** (8.37) 0.166*** (9.18) 0.154*** (8.49) 0.150*** (7.94)
CFTA 0.133*** (2.66) 0.091 (0.56) 0.075 (0.47) 0.101*** (2.14)
CAPEX 0.202*** (3.11) 0.242*** (2.52) 0.231*** (2.36) 0.197*** (3.07)
MTB 0.007*** (2.16) 0.009*** (3.39) 0.010*** (3.18) 0.011*** (2.13)
Div_Dum 0.002 (0.26) 0.003 (0.27) 0.000 (0.00) 0.001 (0.20)
ROA 0.002*** (4.00) 0.001 (0.93) 0.003 (0.79) 0.002*** (3.82)
R&D 1.441*** (23.29) 1.445*** (3.41) 1.444*** (3.36) 1.419*** (22.87)
Size 0.066*** (11.55) 0.067*** (10.27) 0.065*** (10.13) 0.066*** (11.57)
Firm_age 0.031*** (3.97) 0.032*** (6.47) 0.030*** (5.31) 0.027*** (4.17)
Constant — — 0.379*** (12.57) 0.365*** (9.45)
(Continues)
2204 ATIF ET AL.
TABLE 5 (Continued)
ratio_of_cash
Panel B. The effect of ESG disclosure and life
cycle on cash holdings OLS (1) Lagged (2) OLS (3) Lagged (4)
N 9811 7643 9811 7643
2
Adj. R .235 .229 .228 .231
Note: This table presents the regression results of model (2),
X X
ratio_of_cashit ¼ α þ β1 ðESGÞit þ β2 ðFLCÞit þ β3 ðESG FLCÞit þ δ4 ðZÞit þ δ5 ðindustry effectsÞi þ δ6 ðyear effectsÞt þ εit
where ESG disclosure is measured by a score ranging from 0 to 100 and FLC is firm life cycle based on Dickinson (2011) approach. Panel A presents the
cash holdings in different stages of firm life cycle (based on Dickinson, 2011, and RETA) using OLS. Panel B presents the results of interactions between
Intro ESG_score, Growth ESG_score, Maturity ESG_score and Shake-out/decline ESG_score on cash holdings in different stages of firm life cycle. All
the regressions employ industry and year effects. The robust t-statistic of each coefficient is shown in parentheses. All variables are defined in Table 1.
Coefficients are reported at 1%, 5% and 10% levels of significance with ***, *** and *, respectively.
4 | RESULTS mitigates the agency problem by improving internal and external

monitoring.
4.1 | ESG disclosure, cash holdings and firm life After establishing the negative relation between ESG disclosure
cycle and cash holdings, we move to investigate the relation at various
stages of a firm's life cycle. To examine the relation between cash
First, we examine the impact of the ESG disclosure (ESG_score) on firm holdings and life-cycle stages (a vector of dummy variables based on
cash holdings and present the OLS regression results in Table 4. Our Dickinson, 2011), we use contemporaneous as well as 1-year lagged
analysis begins by regressing cash holdings (ratio_of_cash) on ESG dis- variables in Columns 1 and 2, respectively. In Table 5, Panel A shows
closure (ESG_score). In Column 1 (without control variables), we show the level of cash holdings across the stages of the life cycle. In both
a significantly negative relation along with industry and year-fixed the introduction and growth stages, cash holdings are positive and sig-
effects. Columns 2 and 3 include contemporaneous as well as 1-year nificant at 10% or higher in both columns. The reasons for high cash
lagged (t 1) control variables, respectively. Across all the specifica- holdings in young (introduction and growth) firms could include fre-
tions (Panel A), ESG disclosure (ESG_score) has a significantly (at 5% or quent future investments, volatile cash flows, less stable customer
better level) negative impact on cash holdings. For instance, a 1-point base and limited access to the capital market (Barclay & Smith, 2005;
increase in the ESG disclosure results in a 0.10% decrease in cash Jovanovic, 1982; Myers, 1984). These findings are consistent with the
holdings as shown in Column 2. The economic significance is also literature (i.e., Faff et al., 2016). The maturity stage has negative (at 1%
strong and important. For instance, an increase in ESG_score by one level) cash holdings, which is also consistent with Faff et al. (2016),
(sample) standard deviation would decrease cash holdings by approxi- indicating stable cash flows and lower funding requirements with
mately 7.7% (0.001 11.860/0.153). There is consistent and statisti- established corporate governance. However, despite few investment
cally strong evidence that ESG disclosure has a negative influence on opportunities and reduced cash flow, the shake-out/decline stage
cash holdings, demonstrating that ESG provides a strong governance shows positive and significant (at 1% level) cash holdings (Columns
mechanism and helps reduce the agency problem. 1 and 2), contrary to findings by Faff et al. (2016). Overall, these find-
As a robustness check, we then use ESG_index (a variable ranging ings suggest higher cash holdings at all life-cycle stages, except matu-
from 0 to 1, based on the average extent of ESG disclosure) as an rity. For a robustness check, we also determine cash holdings across
independent variable in Panel B. As in Panel A, we regress cash hold- the life-cycle stages based on DeAngelo et al. (2006), using a dummy
ings on ESG_index, where we use no controls in Column 4, contempo- variable RETA, and find consistent results in Columns 3 and 4.
raneous controls in Column 5 and 1-year lagged (t 1) controls in Panel B of Table 5 reports the results based on the interactions
Column 6. All the models (Panel B) suggest that the extent of ESG dis- between firm life-cycle stages and ESG disclosure on cash holdings. In
closure (ESG_index) has a significantly (at 10% or better level) negative both the introduction and growth stages, Columns 1 (contemporane-
impact on cash holdings. Moreover, we regress cash holdings on indi- ous measures) and 2 (1-year lagged measures) show that the interac-
vidual ESG factors (E_score, S_score and G_score) in Panel C (Columns tion with ESG disclosure (Intro ESG_score and Growth ESG_score)
7–9). These individual scores have a negative impact on cash holdings has a negative impact on cash holdings (at 1% and 5% levels, respec-
as well (significant at 10% or better level). Our findings, which are in tively), indicating that firms with better ESG disclosure have lower
line with previous research, suggest that greater ESG disclosure cash holdings, which is consistent with H1 and H2. However, firms in
ATIF ET AL. 2205
the mature stage remain unaffected, supporting H3. The interaction stage (significant at 10% level), based on DeAngelo et al.'s (2006)
between the shake-out/decline stage and ESG disclosure (Shake-out/ classification.
decline ESG_score) is significantly negative at 5% level, showing According to Faff et al. (2016), firms in the introduction and
lower cash holdings for firms with better ESG disclosure supporting growth stages, compared to firms in other stages, have higher cash
H4. Overall, these findings suggest that ESG disclosure reduces cash holdings due to ongoing investment opportunities. Our findings indi-
holdings for firms in three life-cycle stages, which would otherwise cate that firms in the introduction and growth stages with better ESG
hold high cash reserves. As a robustness check, Columns 3 and 4 show disclosure hold less cash, and the reason for this could be due to
the positive impact of ESG disclosure on cash holdings in the mature investors' trust and moral capital, which allow these firms to access
TABLE 6 Sensitivity analysis
Part A: Variable ratio_of_cash (1) cashta (2) cash (industry adj) (3) Ln_cash (4)
Panel A
OLS regression (N = 9811)
ESG_score 0.001*** (2.74) 0.002** (2.06) 0.001** (2.09) 0.002*** (3.82)
Controls Yes Yes Yes Yes
Panel B
OLS regression (N = 9811)
ESG_index 0.017** (2.05) 0.004** (2.15) 0.100** (2.09) 0.045* (1.82)
Panel C
Tobit regression (N = 9811)
ESG_score 0.007*** (2.75) 0.001* (1.85) 0.001** (2.14) 0.002*** (3.16)
Panel D
Controlling for board variables, i.e., board size, board independence, CEO duality and GFC (N = 9811)
ESG_score 0.004** (2.22) 0.003** (2.31) 0.007*** (3.18) 0.026** (2.29)
Panel E
Tobit regression (N = 9811)
FLC 0.105*** (3.97) 0.002** (2.25) 0.058*** (2.99) 0.013*** (2.82)
Intro ESG_score 0.005*** (2.77) 0.007* (1.97) 0.007 (1.67) 0.002** (2.96)
Growth ESG_score 0.002** (2.17) 0.004** (2.11) 0.011* (1.97) 0.004** (2.27)
Maturity ESG_score 0.007 (1.47) 0.009 (1.67) 0.012* (1.89) 0.009 (1.53)
Shake-out/decline ESG_score 0.007** (2.19) 0.005** (2.09) 0.003* (1.87) 0.006** (2.12)
Note: This table presents the results of sensitivity analyses using alternative variables, alternative methods and additional control variables in five panels
(A–E). Industry and year effects are included in all the regressions. The robust t-statistic of each coefficient is shown in parentheses. All variables are
defined in Table 1.
2206 ATIF ET AL.
funds from the external market (Cheng et al., 2014; Ferrell TABLE 7 Propensity score matching
et al., 2016). Moreover, ESG disclosure moderates increased cash
Panel A
holdings of firms in the shake-out/decline stage, due to a higher
agency problem (managerial opportunism). Through strong internal as Pre-match Post-match
well as external monitoring, better ESG disclosure mitigates the Variable ESG_tercile
agency-related issues, resulting in lower cash holdings. RETA 0.004 (0.16) 0.056 (1.67)
LEV 0.688** (2.02) 0.526 (1.70)
CFTA 7.540*** (7.91) 4.433 (1.78)
4.2 | Sensitivity checks
CAPEX 4.930*** (4.56) 1.262 (1.20)
MTB 0.001 (0.53) 0.002 (0.52)
This section includes a battery of sensitivity checks to test the robust-
Div_Dum 0.546*** (4.92) 0.251 (1.41)
ness of our results. First, we use alternative definitions to define cash
ROA 0.014* (1.89) 0.021 (1.45)
holdings (our dependent variable of interest). As an alternative to
R&D 12.143*** (8.87) 9.774* (1.86)
ratio_of_cash, we use the ratio of cash and cash equivalents to total
assets (cashta), the cash holdings adjusted by industry average and the Size 4.510*** (33.49) 0.594* (1.89)
12 Firm_age 0.005*** (2.76) 0.010* (1.98)
natural logarithm of the cash holdings. Second, we explore the
impact of ESG_score and ESG_index on different measures of cash Constant 15.876*** (26.58) 1.673*** (3.76)
holdings using a different estimating method (e.g., Tobit regression).13 Industry effects Yes Yes
Third, following prior studies (e.g., Atif et al., 2019), we control for Year effects Yes Yes
additional corporate governance variables that may influence corpo- N 5313 3344
rate policies and decisions of cash holdings. Specifically, we include Pseudo-R2 .590 .153
board size (measured as the total number of directors), board indepen-
Panel B. Difference in firm characteristics
dence (measured as the percentage of independent directors) and
CEO duality (a dummy variable is equal to 1 if CEO is the chairman of Variable Treatment Control Difference t-stat
board and 0 otherwise). Fourth, since global financial crisis (GFC) RETA 0.289 0.243 0.050 1.724
occurred during our sample period, we also include GFC as a dummy LEV 0.248 0.227 0.021 1.200
variable with a value of 1 for years 2007–2009 and 0 otherwise CFTA 0.060 0.058 0.002 0.610
because one could argue that during the GFC, the relation between CAPEX 0.052 0.051 0.001 0.480
cash holdings and ESG could be different due to costly external MTB 5.285 4.488 0.797 0.990
financing (Duchin, 2010). Finally, to study the impact of ESG on cash Div_Dum 0.748 0.685 0.063 1.830
holdings, we use the Tobit regression across the life-cycle stages
ROA 6.243 6.495 0.252 1.410
(Panel E).
R&D 0.022 0.021 0.001 0.390
Table 6 reports the results (Panels A–E), with Column 1 in Panel
Size 3.780 3.798 0.018 1.390
A repeating the OLS regression as in Table 4 but using each different
Firm_age 37.168 33.19 3.978 1.156
measure for our dependent variable of interest. Nevertheless, in
Panels A and B, the regression in Column 1 is identical to the regres- Panel C. Propensity score estimator
sions in Columns 2 and 5 of Table 4 (reported for the sake of compari- Variable Treatment Control Difference t-stat
son). When we use various measures of cash holdings (Panel A), we ratio_of_cash 0.090 0.120 0.030*** 11.450
can validate the negative effect of ESG disclosure (ESG_score) on cash
Note: The table presents the results of the propensity score matching in
holdings (significant at 5% or better level). Panel B confirms a negative
three panels. Panel A shows the pre- and post-sample results, Panel B
impact of ESG_index (alternative measure of ESG disclosure) on cash presents the differences in firm characteristics for the matched sample
holdings (significant at 10% or better level). Panel C also shows a neg- and Panel C shows the propensity score matching estimator. Robust t-
ative impact in all the columns at 10% or better level of significance statistics are shown in parentheses. All variables are defined in Table 1.
Coefficients are reported at 1%, 5% and 10% levels of significance with
when using the Tobit regressions. Panel D suggests consistent results
***, ** and *, respectively.
even after controlling for board characteristics and the GFC. Across all
panel specifications, we account for industry and year fixed effects, in
addition to controls, as in Table 4 (Column 2). Panel E, which reports introduction, growth and shake-out/decline stages, as well as ESG dis-
results based on Equation 2, shows that the interactions between the closure, have a negative impact on cash holdings using Tobit
regression.
12
We also use an excess cash holdings measure following Dittmar and Mahrt-Smith (2007)
To summarise, we find that ESG disclosure has a negative influ-
and Frésard and Salva (2010): Our results (untabulated) remain consistent. ence on cash holdings by using alternative proxies for our dependent
13
Tobit regression measures variables with censored data and avoids the biases that result
and independent variables, an alternative estimation technique and
from other approaches. We also employ firm fixed-effect (FE) regression and find consistent
results. FE avoids the misspecification of the model, due to the omitted variable bias. controlling for additional variables. The negative impact of ESG
ATIF ET AL. 2207
disclosure is mainly noticeable throughout the introduction, growth TABLE 8 Two-stage least squares
and shake-out/decline stages of a firm's life cycle. Overall, these find- First-stage Second-stage
ings are consistent with the earlier results (Tables 4 and 5). ESG_score ratio_of_cash
Variable (1) (2)
Industry_ESG 0.019** (2.10) —
4.3 | Endogeneity tests ESG_score-fitted — 0.001*** (3.24)
RETA 0.238 (0.93) 0.004** (2.09)
Our baseline results may be prone to endogeneity bias because of LEV 1.420** (2.29) 0.166*** (9.22)
causality issue. It is possible that firm-level governance influences
CFTA 2.897 0.119*** (2.62)
firm-level event of ESG disclosure or vice versa. For example, man- (5.06)
agers who respond to shareholders' demands for lower cash holdings CAPEX 7.194*** 0.216*** (3.81)
may also respond to external concerns about ESG disclosure, thus (10.12)
biasing our results. To rule out the concerns of endogeneity, we use MTB 0.001 (0.88) 0.004** (2.32)
two econometric techniques, that is, PSM and 2SLS. Div_Dum 1.248*** 0.009 (1.42)
First, we consider PSM to explore that two similar firms are likely (16.26)
to follow the same trend without any treatment. In case treatment ROA 0.004 0.003*** (3.70)
occurs, the impact should be reflected in the difference between the (0.97)
changes in the two firms (e.g., treatment vs. control groups) R&D 3.696 1.471*** (26.69)
(Roberts & Whited, 2012). (5.30)
We execute a two-step PSM to compare changes in cash holdings Size 1.152*** 0.067***
(17.22) (12.08)
between two groups of firms with similar characteristics but differ-
Firm_age 0.006*** (5.78) 0.005*** (4.91)
ences in terms of ESG disclosure. Following prior studies (Atif
et al., 2021; Brogaard et al., 2017; Fang et al., 2014), we rank firm- Constant 2.113*** 0.384*** (21.69)
(2.322)
years of ESG disclosure to construct a treatment versus control group
Industry effects No No
and retain the firms in the top and bottom terciles. Then, we create a
dummy variable (ESG_tercile) that is 1 if a firm is in the top tercile Year effects Yes Yes
(i.e., treated group) and 0 if a firm is in the bottom tercile N 7502 7502
(i.e., controlled group). The treatment (control) group consists of firms Model fits
with the highest (lowest) ESG disclosure. F-statistics 71.53***
In the first step, we begin by running the logit regression for [0.00]
ESG_tercile with all our explanatory variables indicated in Equation 1. Cragg–Donald Wald F- 72.04
statistics
The predicted estimates from the logit regression are used as the pro-
pensity scores for each firm-year observation. Table 7 shows the Stock–Yogo weak ID test 16.38
critical values at 10% IV
results of the logit regression in Panel A (pre-match column). The
size
results suggest that firms with higher ESG disclosure are larger and
Note: This table presents the results of the 2SLS. Column 1 reports the
older, have lower leverage, pay more dividends and spend more on
2
first-stage regression, and Column 2 shows the second-stage regression
R&D expenditures. The pseudo-R for this is high (59.0%). results. The robust t-statistic of each coefficient is shown in parentheses.
In the second step, we use the propensity scores to form one-to- All variables are defined in Table 1.
one matched pairs, ensuring that the firms with higher ESG disclosure Coefficients are reported at 1%, 5% and 10% levels of significance with
(the treatment group) are sufficiently similar to those with lower ESG ***, ** and *, respectively.
disclosure (the control group). Specifically, each firm-year with higher

disclosure is matched to a firm-year with lower ESG disclosure using tests. In the first diagnostic test, we re-estimate the logit regression
the closest propensity score. We use matching with replacement and for the post-match sample and report the results in Panel A (post-
require that the difference between the treatment and matched firms' match column) of Table 7. The magnitude of the regression coeffi-
propensity scores is less than 0.1% in absolute value.14 Using these cients is smaller than those in the pre-match and is statistically insig-
criteria, we effectively match 3344 firm-year observations; thus, our nificant, indicating a decline in the degree of freedom in the restricted
treatment and control groups are nearly identical along with all sample. This also suggests that in terms of observable characteristics,
explanatory variables except ESG disclosure. both groups are similar. The pseudo-R2 drops substantially from
To verify that the treatment and control groups are indistinguish- 59.0% (pre-match) to 15.3% (post-match), demonstrating that PSM
able in terms of observable characteristics, we run two diagnostic removes all observable differences other than the difference in ESG
disclosure. In the second diagnostic test, we look at the differences
14 for each observable characteristic between the treatment and control
As a robustness test, we use nearest neighbour matching, changing the matching calliper
from 0.1% to 0.5%. The results qualitatively remain the same. firms and report the results in Panel B of Table 7. Again, we find no
2208 ATIF ET AL.
TABLE 9 The effect of ESG disclosure on firm performance and value of cash holdings
FPERF
Variable Tobin's q ROS Excess return

ESG_score 0.001** (2.16) 0.005** (2.14) 0.003** (2.08)
ratio_of_cash 0.002* (1.89) 0.004* (1.99) 0.001 (1.45)
Δratio_of_cash — — 0.002 (1.15)
Intro 0.037* (1.91) 0.033* (1.89) 0.021** (2.11)
Intro ESG_score ratio_of_cash 0.014*** (2.33) 0.010*** (2.49) —
Intro ESG_score Δratio_of_cash — — 0.009** (2.14)
Growth 0.021** (2.13) 0.017** (2.14) 0.004** (2.10)
Growth ESG_score ratio_of_cash 0.008** (2.17) 0.010** (2.11) —
Growth ESG_score Δratio_of_cash — — 0.009** (2.04)
Maturity 0.030** (2.11) 0.023 (1.44) 1.001 (0.24)
Maturity ESG_score ratio_of_cash 0.040 (1.33) 0.013 (1.47) —
Maturity ESG_score Δratio_of_cash — — 0.026 (1.12)
Shake-out/decline 0.014* (1.88) 0.019** (2.16) 0.019 (1.10)
Shake-out/decline ESG_score ratio_of_cash 0.004* (1.91) 0.007* (1.88) —
Shake-out/decline ESG_score Δratio_of_cash — — 0.001** (2.20)
RETA 0.007* (1.98) 0.010* (1.99) 0.012** (2.20)
LEV 0.144** (2.17) 0.111** (2.14) 0.192* (1.94)
CFTA 0.135** (2.10) 0.112** (2.15) 0.099** (2.11)
CAPEX 0.191*** (2.99) 0.137** (2.14) 0.098* (1.84)
MTB 0.009** (2.14) 0.006** (2.13) 0.010** (2.19)
Div_Dum 0.004 (1.22) 0.005 (1.23) 0.001 (0.94)
R&D 1.130** (2.20) 1.103** (2.12) 0.212** (2.18)
Size 0.053*** (3.15) 0.023*** (2.43) 0.033*** (2.65)
Firm_age 0.022*** (2.47) 0.022*** (2.67) 0.025** (2.17)
LEV Δratio_of_casht 1 — — 0.005* (1.95)
ratio_of_cash Δratio_of_casht 1 — — 0.008** (2.10)
Constant 0.343*** (4.14) 0.339*** (5.13) 0.234*** (3.23)
Industry effects Yes Yes Yes
Year effects Yes Yes Yes
N 9811 9811 9811
Adj. R2 .232 .202 .211
Note: This table presents the results of the effect of ESG disclosure, cash holdings and firm life cycle (interaction term) on firm performance measured by
Tobin's q and ROS as dependent variables (Columns 1 and 2) and value of cash holdings (Column 3). Industry and year effects are included in all the
regressions. The robust t-statistic of each coefficient is shown in parentheses. All variables are defined in Table 1.
significant differences between the two groups in terms of observable The findings imply that the variations in cash holdings are due to a
characteristics.15 As a result of these findings, PSM appears to remove systematic difference in ESG disclosure. Hence, we infer that higher
other observable characteristics, increasing the probability that any ESG disclosure leads to lower cash holdings, thus mitigating
difference in cash holdings between the treatment and control groups agency cost.
can be attributed to changes in ESG disclosure, rather than differ- Second, we use IV approach, which uses 2SLS regression analysis.
ences in other variables. Specifically, we use IV as a source of exogenous variation in the
In Panel C, we report a PSM estimator and find a significant dif- ESG_score to account for potential causality issue. The construction of
ference in cash holdings between the treated and controlled groups. an IV that is correlated with the endogenous variable (i.e., ESG_score)
but that should not have a direct or indirect relation with the depen-
15 dent variable (exogenous variable) is a challenge when using the 2SLS
The mean differences between treatment and control groups are based on the average
treatment effect on the treated (Ahmed et al., 2021; Nadarajah et al., 2021). technique. A valid IV must meet two conditions: first, the relevance
ATIF ET AL. 2209
condition, which states that the IV correlates with ESG disclosure interaction between ESG, cash holdings and firm life-cycle vector
after controlling for the set of variables in our model, and, second, the (ESG_score FLC ratio_of_cash), which is the key variable of
exclusion restriction, which states that the IV influences cash holdings interest, the independent and control variables are the same as in
only through its correlations with ESG disclosure when the full set of Table 1. Table 9 reports the results using OLS for Tobin's q and
control variables is used. ROS, respectively. The interaction term shows a positive effect on
To construct IV for this study, we follow the approach employed firm performance, consistent with prior literature that concludes a
by Arouri and Pijourlet (2017) and Atif et al. (2020). Our IV is the ESG positive relation between firms' environmental and financial perfor-
industry mean, computed as the average score of ESG of all the firms mance (Alam et al., 2019; Endrikat et al., 2014). Overall, we con-
in a given year, excluding firm i's score of ESG in that year. The intui- clude that the level of cash holdings in the introduction, growth
tion behind this IV is that a firm's ESG disclosure might be highly and shake-out/decline stages, and ESG disclosure, lead to higher
related to its industry peers, due to their similar business mix and firm performance.
investment opportunities, but that such an industry average is unlikely Moreover, we examine the effect of ESG disclosure on the value
to directly affect a firm's cash holdings. Moreover, firm-level cash of cash holdings at various firm life-cycle stages. To investigate the
holdings may impact firm-level ESG disclosure but are unlikely to value of cash holdings, we use the method proposed by Faulkender
impact ESG disclosure at the industry level. Given these arguments, and Wang (2006).16 This approach is extensively employed in the lit-
the ESG industry mean should be a valid IV since it is related to firm- erature to observe the value of cash holdings (e.g., Lee &
level ESG disclosure but unrelated to firm-level cash holdings. Powell, 2011; Tong, 2010). As reported in Table 9 (Column 3), we find
Column 1 of Table 8 reports the results of a first-stage regression the coefficient on the interaction term between ESG disclosure and
with the ESG_score as the dependent variable. The IV, as well as the change in cash holdings to be positively associated with our depen-
same controls as in Equation 1, is the explanatory variables. As argued, dent variable (excess return) in the introduction, growth and shake-
the coefficient on industry_ESG is statistically significant at 5% level, out/decline stages of the life cycle. These findings are consistent and
suggesting that the IV is valid because of its relevance and its statisti- further support the argument that ESG disclosure reduces the agency
cal power to explain ESG_score. In addition, the F-statistics is high cost, resulting in better values of cash holdings.
(71.5), showing that IV is not weak. Further, the Cragg–Donald Wald
F weak-instrument test has a p value of .000, rejecting the null
hypothesis that the instrument is weak (Cragg & Donald, 1993). Col- 5 | CONC LU SION
umn 2 of Table 8 presents the second-stage regression results, where
we regress ESG_score-fitted on cash holdings along with other vari- ESG disclosure has emerged as a critical factor to consider when mak-
ables. Our results are in line with those of our main model: The coeffi- ing investment decisions. At the same time, allocation and use of a
cient on ESG_score-fitted is statistically significant and negative for firm's liquid assets (cash holdings) over the different phases of the life
cash holdings, indicating that greater ESG disclosure leads to lower cycle warrant further investigation. However, there is no evidence in
cash holdings. Thus, after minimising endogeneity concerns, we can the literature about how ESG disclosure influences cash holdings at
safely infer that ESG disclosure reduces cash holdings. different stages of a firm's life cycle. This study adds to the current
ESG literature by providing empirical evidence on the impact of ESG
disclosure (as well as those of individual E, S and G scores) on cash
4.4 | Further analysis holdings at various life-cycle stages. Using S&P 1500 firms, we initially
find higher (lower) cash holdings in the introduction, growth (mature)
In this section, we examine whether lower cash holdings as a result and shake-out/decline stages of the life cycle, which is largely consis-
of ESG disclosure affect firm performance at different periods of the tent with prior studies. We then find a significantly negative influence
firm life cycle. Given the impact of ESG disclosure on cash holdings of ESG disclosure on cash holdings in the introduction, growth and
at various stages of the firm life cycle, one may expect that firm per- shake-out/decline stages of firms (insignificant in the mature stage),
formance will suffer. To investigate the impact of ESG disclosure and resulting in higher firm performance and a positive value of cash hold-
cash holdings on firm performance at different stages of the life ings, in line with prior studies (e.g., Busch & Lewandowski, 2018;
cycle, we estimate the following regression model: Endrikat et al., 2014). Our results hold across various robustness
checks, including alternative measures, specifications, additional con-
FPERFit ¼ α þ β1 ðESGÞit þ β2 ðFLCÞit þ β3 ðratio_of_cashÞit trol variables and endogeneity techniques (i.e., PSM and 2SLS).
þ δ4 ðX
ESG FLC ratio_of_cash XÞit þ δ5 ðZÞit
Our study has implications for regulators, investors and firms.
þ δ6 ðindustry effectsÞi þ δ7 ðyear effectsÞt þ εit ð3Þ
Importantly, our research demonstrates how high ESG disclosure
leads to better cash holding policies as a result of the strong internal
To measure firm performance (FPERF), we use Tobin's and external monitoring mechanisms, as well as recognising its impact
q (market value of equity divided by the book value of equity) and at various stages of the firm life cycle. Even though 83% of SEC-
ROS (net income scaled by sales turnover) following prior studies
(e.g., Hossain et al., 2020; Liu et al., 2014). Except for the 16
See Appendix S1 for further details on the method.
2210 ATIF ET AL.
registered firms in the US report some sustainability information in Arouri, M., & Pijourlet, G. (2017). CSR performance and the value of cash
their regulatory filings, the voluntary nature of the ESG disclosure has holdings: International evidence. Journal of Business Ethics, 140, 263–
284. https://doi.org/10.1007/s10551-015-2658-5
hindered the ability of firms to disclose accurate information to some
Atif, M., Alam, M. S., & Hossain, M. (2020). Firm sustainable investment:
extent. Hence, as a crucial managerial and policy implication, regulat- Are female directors greener? Business Strategy and the Environment,
ing agencies, such as the Sustainability Accounting Standards Board 29(8), 3449–3469. https://doi.org/10.1002/bse.2588
(SASB), should consider revising disclosure policies related to ESG, in Atif, M., & Ali, S. (2021). Environmental, social and governance disclosure
and default risk. Business Strategy and the Environment, 30, 3937–
order to improve the accuracy and transparency of disclosure. This
3959. https://doi.org/10.1002/bse.2850
reform would ensure the availability and quality of firm-level ESG dis- Atif, M., Hossain, M., Alam, M. S., & Goergen, M. (2021). Does board gender
closure across all listed firms as well as support socially responsible diversity affect renewable energy consumption? Journal of Corporate
and sustainable business practices. This would further help reduce the Finance, 66, 101665. https://doi.org/10.1016/j.jcorpfin.2020.101665
Atif, M., Liu, B., & Huang, A. (2019). Does board gender diversity affect
gap between the demand for ESG information by investors and the
corporate cash holdings? Journal of Business Finance & Accounting, 46,
supply of information by firms (Krueger et al., 2021), as well as 1003–1029. https://doi.org/10.1111/jbfa.12397
address investor complaints about a lack of comparable and reliable Attig, N., Cleary, S. W., El Ghoul, S., & Guedhami, O. (2014). Corporate
sustainability information (Bernow et al., 2019). Future research, sub- legitimacy and investment–cash flow sensitivity. Journal of Business
Ethics, 121, 297–314.
ject to availability of data, may look into ESG disclosure and its impact
Barclay, M. J., & Smith, C. W. (2005). The capital structure puzzle: The evi-
on firm-level outcomes in developed and developing countries.
dence revisited. Journal of Applied Corporate Finance, 17, 1–17.
https://doi.org/10.1111/j.1745-6622.2005.012_2.x
ACKNOWLEDGEMEN TS Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009). Why do US firms hold so
We acknowledge feedback from participants and discussants at the 9th much more cash than they used to? The Journal of Finance, 64, 1985–
2021. https://doi.org/10.1111/j.1540-6261.2009.01492.x
Financial Markets and Corporate Governance Conference 2018, Mel-
Bernow, S., Godsall, J., Klempner, B., & Merten, C. (2019). More than
bourne; Accounting and Finance Association of Australia and values: The value-based sustainability reporting that investors want.
New Zealand (AFAANZ) Conference 2018, Auckland; Finance and McKinsey and Company.
Financial Planning Research Seminar 2018, Griffith University; Account- Brogaard, J., Li, D., & Xia, Y. (2017). Stock liquidity and default risk. Journal
of Financial Economics, 124, 486–502. https://doi.org/10.1016/j.
ing Conference 2019, University of Essex; and British Accounting and
jfineco.2017.03.003
Finance Association (BAFA) Corporate Finance and Asset Pricing Con- Busch, T., & Lewandowski, S. (2018). Corporate carbon and financial per-
ference 2019, Manchester. We especially thank Sumit Lodhia, formance: A meta-analysis. Journal of Industrial Ecology, 22, 745–759.
Jacqueline Birt, Robert Faff, Musa Mangena, Shahzad Uddin, Teerooven https://doi.org/10.1111/jiec.12591
Campbell, J. L. (2007). Why would corporations behave in socially respon-
Soobaroyen, Diogenis Baboukardos, Jia Liu, Rizwan Ahmad and Izidin El
sible ways? An institutional theory of corporate social responsibility.
Kalak for their useful comments on the earlier version of this paper. Academy of Management Review, 32, 946–967. https://doi.org/10.
Open access publishing facilitated by Macquarie University, as part of 5465/amr.2007.25275684
the Wiley - Macquarie University agreement via the Council of Austra- Capelle-Blancard, G., & Monjon, S. (2014). The performance of socially
responsible funds: Does the screening process matter? European
lian University Librarians. [Correction added on 18 May 2022, after first
Financial Management, 20, 494–520. https://doi.org/10.1111/j.1468-
online publication: CAUL funding statement has been added.] 036X.2012.00643.x
Carter, C. R. (2005). Purchasing social responsibility and firm performance:
CONF LICT OF IN TE RE ST S The key mediating roles of organizational learning and supplier perfor-
mance. International Journal of Physical Distribution and Logistics Manage-
The authors declare that they have no conflicts of interest.
ment, 35, 177–194. https://doi.org/10.1108/09600030510594567
Cheng, B., Ioannou, I., & Serafeim, G. (2014). Corporate social responsibil-
ORCID ity and access to finance. Strategic Management Journal, 35, 1–23.
Muhammad Atif https://orcid.org/0000-0001-8495-7886 https://doi.org/10.1002/smj.2131
Cheung, A. (2016). Corporate social responsibility and corporate cash
holdings. Journal of Corporate Finance, 37, 412–430. https://doi.org/
RE FE R ENC E S
10.1016/j.jcorpfin.2016.01.008
Ahmed, A., Atif, M., & Gyapong, E. (2021). Boardroom gender diversity Christensen, H. B., Hail, L., & Leuz, C. (2021). Mandatory CSR and sustain-
and CEO pay deviation: Australian evidence. Accounting and Finance, ability reporting: Economic analysis and literature review. Review of
61(2), 3135–3170. https://doi.org/10.1111/acfi.12696 Accounting Studies, 26(3), 1176–1248. https://doi.org/10.1007/
Alam, M. S., Atif, M., Chien-Chi, C., & Soytaş, U. (2019). Does corporate s11142-021-09609-5
R&D investment affect firm environmental performance? Evidence Clarkson, P. M., Walker, J., & Nicholls, S. (2011). Disclosure, shareholder
from G-6 countries. Energy Economics, 78, 401–411. https://doi.org/ oversight and the pay–performance link. Journal of Contemporary
10.1016/j.eneco.2018.11.031 Accounting and Economics, 7, 47–64. https://doi.org/10.1016/j.jcae.
Amel-Zadeh, A., & Serafeim, G. (2018). Why and how investors use ESG 2011.07.001
information: Evidence from a global survey. Financial Analysts Journal, Cohen, J. R., & Simnett, R. (2015). CSR and assurance services: A research
74(3), 87–103. https://doi.org/10.2469/faj.v74.n3.2 agenda. Auditing: A Journal of Practice & Theory, 34(1), 59–74. https://
Anthony, J. H., & Ramesh, K. (1992). Association between accounting per- doi.org/10.2308/ajpt-50876
formance measures and stock prices: A test of the life cycle hypothe- Cornell, B., & Shapiro, A. C. (1987). Corporate stakeholders and corporate
sis. Journal of Accounting and Economics, 15, 203–227. https://doi.org/ finance. Financial Management, 16, 5–14. https://doi.org/10.2307/
10.1016/0165-4101(92)90018-W 3665543
ATIF ET AL. 2211
Cragg, J. G., & Donald, S. G. (1993). Testing identifiability and specification empirical test of the risk management hypothesis. Strategic Manage-
in instrumental variable models. Econometric Theory, 9, 222–240. ment Journal, 30(4), 425–445. https://doi.org/10.1002/smj.750
https://doi.org/10.1017/S0266466600007519 Goldman Sachs. (2007). GS SUSTAIN targets sustainable corporate perfor-
DeAngelo, H., DeAngelo, L., & Stulz, R. M. (2006). Dividend policy and the mance. https://www.goldmansachs.com/our-firm/history/moments/
earned/contributed capital mix: A test of the life-cycle theory. Journal 2007-gs-sustain.html
of Financial Economics, 81, 227–254. https://doi.org/10.1016/j. Gort, M., & Klepper, S. (1982). Time paths in the diffusion of product inno-
jfineco.2005.07.005 vations. The Economic Journal, 92(367), 630–653. https://doi.org/10.
DeAngelo, H., DeAngelo, L., & Stulz, R. M. (2010). Seasoned equity offer- 2307/2232554
ings, market timing, and the corporate lifecycle. Journal of Financial Gramlich, D., & Finster, N. (2013). Corporate sustainability and risk. Journal
Economics, 95(3), 275–295. https://doi.org/10.1016/j.jfineco.2009. of Business Economics, 83(6), 631–664. https://doi.org/10.1007/
11.002 s11573-013-0666-4
Dhaliwal, D., Li, Z., Tsang, A., & Yang, Y. (2011). Voluntary nonfinancial dis- Greening, D. W., & Turban, D. B. (2000). Corporate social performance as
closure and the cost of equity capital: The initiation of corporate social a competitive advantage in attracting a quality workforce. Business &
responsibility reporting. The Accounting Review, 86, 59–100. https:// Society, 39(3), 254–280. https://doi.org/10.1177/
doi.org/10.2308/accr.00000005 000765030003900302
Dhaliwal, D., Radhakrishnan, S., Tsang, A., & Yang, Y. (2012). Nonfinancial Grullon, G., Michaely, R., & Swaminathan, B. (2002). Are dividend changes
disclosure and analyst forecast accuracy: International evidence on a sign of firm maturity? The Journal of Business, 75(3), 387–424.
corporate social responsibility disclosure. The Accounting Review, 87, https://doi.org/10.1086/339889
723–759. https://doi.org/10.2308/accr-10218 Han, S., & Qiu, J. (2007). Corporate precautionary cash holdings. Journal of
Dickinson, V. (2011). Cash flow patterns as a proxy for firm life cycle. The Corporate Finance, 13(1), 43–57. https://doi.org/10.1016/j.jcorpfin.
Accounting Review, 86(6), 1969–1994. https://doi.org/10.2308/accr- 2006.05.002
10130 Harford, J., Li, K., & Zhao, X. (2008). Corporate boards and the leverage
Dittmar, A., & Mahrt-Smith, J. (2007). Corporate governance and the value and debt maturity choices. International Journal of Corporate Gover-
of cash holdings. Journal of Financial Economics, 83(3), 599–634. nance, 1(1), 3–27. https://doi.org/10.1504/IJCG.2008.017648
https://doi.org/10.1016/j.jfineco.2005.12.006 Hasan, M. M., & Habib, A. (2017). Firm life cycle and idiosyncratic volatil-
Duchin, R. (2010). Cash holdings and corporate diversification. The Journal ity. International Review of Financial Analysis, 50, 164–175. https://doi.
of Finance, 65(3), 955–992. https://doi.org/10.1111/j.1540-6261. org/10.1016/j.irfa.2017.01.003
2010.01558.x Hasan, M. M., Hossain, M., & Habib, A. (2015). Corporate life cycle and
El Ghoul, S., Guedhami, O., Kwok, C. C., & Mishra, D. R. (2011). Does cor- cost of equity capital. Journal of Contemporary Accounting and Econom-
porate social responsibility affect the cost of capital? Journal of Bank- ics, 11(1), 46–60. https://doi.org/10.1016/j.jcae.2014.12.002
ing & Finance, 35(9), 2388–2406. https://doi.org/10.1016/j.jbankfin. Hay, R., & Ginter, P. (1979). Strategies for maintaining a share of the mar-
2011.02.007 ket. Paper presented at the Annual Meeting of the Southern Academy
Endrikat, J., Guenther, E., & Hoppe, H. (2014). Making sense of conflicting of Management, Atlanta.
empirical findings: A meta-analytic review of the relationship between Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate dis-
corporate environmental and financial performance. European Manage- closure, and the capital markets: A review of the empirical disclosure
ment Journal, 32(5), 735–751. https://doi.org/10.1016/j.emj.2013. literature. Journal of Accounting and Economics, 31(1–3), 405–440.
12.004 https://doi.org/10.1016/S0165-4101(01)00018-0
Faff, R., Kwok, W. C., Podolski, E. J., & Wong, G. (2016). Do corporate poli- Hossain, M., Atif, M., Ahmed, A., & Mia, L. (2020). Do LGBT workplace
cies follow a life-cycle? Journal of Banking & Finance, 69, 95–107. diversity policies create value for firms? Journal of Business Ethics,
https://doi.org/10.1016/j.jbankfin.2016.04.009 167(4), 775–791.
Fama, E. F. (1980). Agency problems and the theory of the firm. The Jour- Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance,
nal of Political Economy, 88, 288–307. https://doi.org/10.1086/ and takeovers. American Economic Review, 76(2), 323–329.
260866 Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial
Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. behavior, agency costs and ownership structure. Journal of Financial
The Journal of Law & Economics, 26(2), 301–325. https://doi.org/10. Economics, 3(4), 305–360. https://doi.org/10.1016/0304-405X(76)
1086/467037 90026-X
Fang, V. W., Tian, X., & Tice, S. (2014). Does stock liquidity enhance or Jiao, Y. (2010). Stakeholder welfare and firm value. Journal of Banking &
impede firm innovation? The Journal of Finance, 69(5), 2085–2125. Finance, 34(10), 2549–2561. https://doi.org/10.1016/j.jbankfin.2010.
https://doi.org/10.1111/jofi.12187 04.013
Faulkender, M., & Wang, R. (2006). Corporate financial policy and the Jovanovic, B. (1982). Selection and the evolution of industry. Eco-
value of cash. The Journal of Finance, 61, 1957–1990. https://doi.org/ nometrica: Journal of the Econometric Society, 50, 649–670. https://doi.
10.1111/j.1540-6261.2006.00894.x org/10.2307/1912606
Ferrell, A., Liang, H., & Renneboog, L. (2016). Socially responsible firms. Khan, M., & Watts, R. L. (2009). Estimation and empirical properties of a
Journal of Financial Economics, 122(3), 585–606. https://doi.org/10. firm-year measure of accounting conservatism. Journal of Accounting
1016/j.jfineco.2015.12.003 and Economics, 48(2–3), 132–150. https://doi.org/10.1016/j.jacceco.
Frésard, L., & Salva, C. (2010). The value of excess cash and corporate gov- 2009.08.002
ernance: Evidence from US cross-listings. Journal of Financial Koh, S., Durand, R. B., Dai, L., & Chang, M. (2015). Financial distress: Life-
Economics, 98(2), 359–384. https://doi.org/10.1016/j.jfineco.2010. cycle and corporate restructuring. Journal of Corporate Finance, 33,
04.004 19–33. https://doi.org/10.1016/j.jcorpfin.2015.04.004
Godfrey, P. C. (2005). The relationship between corporate philanthropy Krueger, P., Sautner, Z., Tang, D. Y., & Zhong, R. (2021). The effects of
and shareholder wealth: A risk management perspective. Academy of mandatory ESG disclosure around the world. Available at SSRN
Management Review, 30(4), 777–798. https://doi.org/10.5465/amr. 3832745. https://doi.org/10.2139/ssrn.3832745
2005.18378878 La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (2000).
Godfrey, P. C., Merrill, C. B., & Hansen, J. M. (2009). The relationship Agency problems and dividend policies around the world. The Journal
between corporate social responsibility and shareholder value: An of Finance, 55(1), 1–33. https://doi.org/10.1111/0022-1082.00199
2212 ATIF ET AL.
Lardaro, L. (1993). Applied econometrics. HarperCollins College Publishers. Porter, M. E. (2008). Competitive strategy: Techniques for analyzing indus-
Lee, D. D., & Faff, R. W. (2009). Corporate sustainability performance and tries and competitors. Simon and Schuster.
idiosyncratic risk: A global perspective. Financial Review, 44(2), 213– Richardson, B. J. (2009). Keeping ethical investment ethical: Regulatory
237. https://doi.org/10.1111/j.1540-6288.2009.00216.x issues for investing for sustainability. Journal of Business Ethics, 87(4),
Lee, E., & Powell, R. (2011). Excess cash holdings and shareholder value. 555–572. https://doi.org/10.1007/s10551-008-9958-y
Accounting and Finance, 51, 549–574. https://doi.org/10.1111/j. Roberts, M. R., & Whited, T. M. (2012). Endogeneity in empirical corporate
1467-629X.2010.00359.x finance. https://doi.org/10.2139/ssrn.1748604
Liu, Y., Wei, Z., & Xie, F. (2014). Do women directors improve firm perfor- SASB. (2017). The state of disclosure 2017: An analysis of the effective-
mance in China? Journal of Corporate Finance, 28, 169–184. https:// ness of sustainability disclosure in SEC filings. https://www.sasb.org/
doi.org/10.1016/j.jcorpfin.2013.11.016 wp-content/uploads/2017/12/2017State-of-DisclosureReport-
Livnat, J., & Zarowin, P. (1990). The incremental information content of web.pdf
cash-flow components. Journal of Accounting and Economics, 13(1), Spence, A. M. (1977). Entry, capacity, investment and oligopolistic pricing.
25–46. https://doi.org/10.1016/0165-4101(90)90066-D The Bell Journal of Economics, 8, 534–544. https://doi.org/10.2307/
Lu, L. Y., Shailer, G., & Yu, Y. (2017). Corporate social responsibility disclo- 3003302
sure and the value of cash holdings. The European Accounting Review, Spence, A. M. (1979). Investment strategy and growth in a new market.
26(4), 729–753. https://doi.org/10.1080/09638180.2016.1187074 The Journal of Reprints for Antitrust Law and Economics, 10, 330–345.
Malik, M. (2015). Value-enhancing capabilities of CSR: A brief review of https://doi.org/10.2307/3003316
contemporary literature. Journal of Business Ethics, 127(2), 419–438. Spence, A. M. (1981). The learning curve and competition. The Bell Journal
https://doi.org/10.1007/s10551-014-2051-9 of Economics, 12, 49–70. https://doi.org/10.2307/3003508
Malmendier, U., & Tate, G. (2008). Who makes acquisitions? CEO over- Tong, Z. (2010). CEO risk incentives and corporate cash holdings. Journal
confidence and the market's reaction. Journal of Financial Economics, of Business Finance & Accounting, 37(9–10), 1248–1280. https://doi.
89(1), 20–43. https://doi.org/10.1016/j.jfineco.2007.07.002 org/10.1111/j.1468-5957.2010.02208.x
McGuire, J. B., Sundgren, A., & Schneeweis, T. (1988). Corporate social Udayasankar, K. (2008). Corporate social responsibility and firm size. Jour-
responsibility and firm financial performance. Academy of Management nal of Business Ethics, 83(2), 167–175. https://doi.org/10.1007/
Journal, 31(4), 854–872. s10551-007-9609-8
McWilliams, A., & Siegel, D. (2001). Corporate social responsibility: A the- United Nations Principles for Responsible Investment. (2006). Secretary-
ory of the firm perspective. Academy of Management Review, 26(1), General launches ‘principles for responsible investment’ backed by
117–127. https://doi.org/10.5465/amr.2001.4011987 world's largest investors. https://www.un.org/press/en/2006/sg2111.
Miller, D., & Friesen, P. H. (1984). A longitudinal study of the corporate life doc.htm
cycle. Management Science, 30(10), 1161–1183. https://doi.org/10. United Nations Principles for Responsible Investment. (2011). Annual
1287/mnsc.30.10.1161 report 2011. https://www.unpri.org/about-the-pri/annual-report-
Myers, S. C. (1977). Determinants of corporate borrowing. Journal of 2011/706.article
Financial Economics, 5(2), 147–175. https://doi.org/10.1016/0304- Wanderley, L. S. O., Lucian, R., Farache, F., & de Sousa Filho, J. M. (2008).
405X(77)90015-0 CSR information disclosure on the web: A context-based approach
Myers, S. C. (1984). The capital structure puzzle. The Journal of Finance, analysing the influence of country of origin and industry sector. Journal
39(3), 574–592. https://doi.org/10.1111/j.1540-6261.1984. of Business Ethics, 82(2), 369–378. https://doi.org/10.1007/s10551-
tb03646.x 008-9892-z
Nadarajah, S., Atif, M., & Gull, A. A. (2021). State-level culture and work- Wernerfelt, B. (1985). Brand loyalty and user skills. Journal of Economic
place diversity policies: Evidence from US firms. Journal of Business Behavior & Organization, 6(4), 381–385. https://doi.org/10.1016/
Ethics, 1–20. https://doi.org/10.1007/s10551-021-04742-2 0167-2681(85)90005-8
Nikolov, B., & Whited, T. M. (2014). Agency conflicts and cash: Estimates Yeldar, R. (2013). The value of extra-financial disclosure: What investors
from a dynamic model. The Journal of Finance, 69(5), 1883–1921. and analysts said. https://www.globalreporting.org/resourcelibrary/
https://doi.org/10.1111/jofi.12183 The-value-of-extra-financial-disclosure.pdf
Oikonomou, I., Brooks, C., & Pavelin, S. (2012). The impact of corporate
social performance on financial risk and utility: A longitudinal analysis.
Financial Management, 41(2), 483–515. https://doi.org/10.1111/j. SUPPORTING INF ORMATION
1755-053X.2012.01190.x Additional supporting information may be found in the online version
Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants
of the article at the publisher's website.
and implications of corporate cash holdings. Journal of Financial Eco-
nomics, 52(1), 3–46. https://doi.org/10.1016/S0304-405X(99)
00003-3
Owen, S., & Yawson, A. (2010). Corporate life cycle and M&A activity. How to cite this article: Atif, M., Liu, B., & Nadarajah, S.
Journal of Banking & Finance, 34(2), 427–440. https://doi.org/10. (2022). The effect of corporate environmental, social and
1016/j.jbankfin.2009.08.003 governance disclosure on cash holdings: Life-cycle
Ozkan, A., & Ozkan, N. (2004). Corporate cash holdings: An empirical
perspective. Business Strategy and the Environment, 31(5),
investigation of UK companies. Journal of Banking & Finance, 28(9),
2103–2134. https://doi.org/10.1016/j.jbankfin.2003.08.003 2193–2212. https://doi.org/10.1002/bse.3016
Petersen, M. A. (2009). Estimating standard errors in finance panel data
sets: Comparing approaches. Review of Financial Studies, 22(1), 435–
480. https://doi.org/10.1093/rfs/hhn053

Bus Strat Env - 2022 - Atif - The Effect of Corporate Environmental Social and Governance Disclosure On Cash Holdings

Uploaded by

Copyright:

Available Formats

Bus Strat Env - 2022 - Atif - The Effect of Corporate Environmental Social and Governance Disclosure On Cash Holdings

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

Bus Strat Env - 2022 - Atif - The Effect of Corporate Environmental Social and Governance Disclosure On Cash Holdings

Uploaded by

Copyright:

Available Formats

Received: 21 April 2021 Revised: 17 January 2022 Accepted: 2 February 2022

The effect of corporate environmental, social and governance

Muhammad Atif1 | Benjamin Liu2 | Sivathaasan Nadarajah2

1 | I N T RO DU CT I O N Jensen, 1986). Self-interested managers may hold more cash on hand

Bus Strat Env. 2022;31:2193–2212. wileyonlinelibrary.com/journal/bse 2193

TABLE 1 Definitions of variables

Notation Variable name Measure

TABLE 2 Descriptive statistics

Full sample (N = 9811) High ESG Low ESG

Std. 1st 3rd Std. Std.

TABLE 3 Correlation matrix

2 ESG_score .009 1.000

3 E_score .003 .971 1.000

4 S_score .024 .856 .723 1.000

6 Intro .028 .037 .036 .040 .007 1.000

7 Growth .099 .128 .134 .078 .061 .055 1.000

8 Maturity .027 .114 .124 .063 .055 .134 .019 1.000

3.4 | Model and estimation method

TABLE 4 The effect of ESG disclosure on cash holdings

Panel A Panel B Panel C

TABLE 5 The effect of ESG disclosure on cash holdings: Life-cycle perspective

Panel A OLS (1) Lagged (2) OLS (3) Lagged (4)

Note: This table presents the regression results of model (2),

4 | RESULTS mitigates the agency problem by improving internal and external

TABLE 6 Sensitivity analysis

disclosure (the control group). Specifically, each firm-year with higher

Variable Tobin's q ROS Excess return

You might also like