In Holdings We Trust: Uncovering the ESG Fund Lemons

Abstract

Using a novel survey of retail global equity funds offered in Australasia, we provide new insights into the evolving landscape of responsible investing (RI) and potential greenwashing. Our analysis has three components. First, we surveyed asset managers to elicit an understanding of how and why they integrate ESG information into investment decisions. We found that RI was primarily driven by performance and fund flow focused value, rather than ethical values. Further, we found that climate change themes were prioritised within the investment process, relative to other ESG sub-themes. Second, we compared survey responses to portfolio holdings data to evaluate whether fund managers were as environmentally responsible as they claimed to be. Surprisingly, we found that portfolio carbon intensity was significantly higher for respondents that were members of a climate initiative, and not significantly different for those that prioritised climate change themes or engaged in a decarbonisation strategy. The divergence between words and actions appears to be consistent with greenwashing funds ('lemons') seeking responsible investment flows without 'walking the talk'. Third, we evaluated the determinants of carbon and ESG performance across our entire sample of survey respondents and non-respondents. We found that the ESG named funds had similar emissions intensities and inconsistent ESG performance, across three major external rating providers, relative to non ESG named funds. JEL Classification: G11; G24; M14

Full text

In Holdings We Trust: Uncovering the ESG Fund
Lemons
Lachie McLean 1
Ivan Diaz-Rainey 1
Sebastian A. Gehricke 1,2
Renzhu Zhang 1
Using a novel survey of retail global equity funds offered in Australasia, we provide new insights into
the evolving landscape of responsible investing (RI) and potential greenwashing. Our analysis has three
components. First, we surveyed asset managers to elicit an understanding of how and why they integrate
ESG information into investment decisions. We found that RI was primarily driven by performance and
fund flow focused value, rather than ethical values. Further, we found that climate change themes were
prioritised within the investment process, relative to other ESG sub-themes. Second, we compared
survey responses to portfolio holdings data to evaluate whether fund managers were as environmentally
responsible as they claimed to be. Surprisingly, we found that portfolio carbon intensity was
significantly higher for respondents that were members of a climate initiative, and not significantly
different for those that prioritised climate change themes or engaged in a decarbonisation strategy. The
divergence between words and actions appears to be consistent with greenwashing funds (‘lemons’)
seeking responsible investment flows without ‘walking the talk’. Third, we evaluated the determinants
of carbon and ESG performance across our entire sample of survey respondents and non-respondents.
We found that the ESG named funds had similar emissions intensities and inconsistent ESG
performance, across three major external rating providers, relative to non ESG named funds.
JEL Classification: G11; G24; M14
KEYWORDS: ESG, Sustainable Investments, Greenwashing, Responsible Investments




1 University of Otago – Otago Business School
2 Corresponding author. Email address: sebastian.gehricke@otago.ac.nz (S. Gehricke); Phone: 0064 021 616012.

1

1. Introduction
The non-financial dimensions of corporate performance have been subject to greater public
scrutiny in recent periods. Concerned with climate change, product safety issues, poor working
conditions, and corporate scandals, among other issues, stakeholders are urging companies to
address their impact within environmental, social, and governance (ESG) contexts (Baldini et
al., 2018). Consequently, responsible investment strategies have been adopted rapidly within
the fund management industry and the global asset management industry over recent periods
(Alda, 2020; Amel-Zadeh & Serafeim, 2018; van Duuren et al., 2016). In addition to financial
performance, RI integrates ESG considerations into investment decisions to holistically
evaluate all dimensions of corporate performance (Sandberg et al., 2009). Recently, demand
for assets with superior ESG performance has been unprecedented. Between 2019 and 2021,
sustainable-labelled investments in Australasia have more than doubled (Kennaway, 2021).
Globally, stated responsible investments in 2020 reached US$35.3 trillion in assets,
representing over 36% of all professionally managed assets (GSIA, 2020). Correspondingly,
an increasing number of institutions are publicly pledging their commitment to incorporate
ESG criteria into investment decisions, such as through signing up to the United Nations
Principles of Responsible Investment (UNPRI) (Brandon et al., 2021).
Within the asset management industry, the rapid growth of responsible investments has been
driven by the creation of new ESG focused funds in addition to the ‘green rebranding’ of
existing conventional funds (Morningstar, 2021). In response to the surge in funds that have
been labelled as ‘ethical’ or ‘sustainable’, global regulators are increasing their scrutiny of such
classifications. In the United States, a review conducted by the Securities and Exchange
Commission found that some funds were misleading consumers by not adhering to their ESG
policies (Chin, 2021). Recently, the European Commission introduced standards for the eco-
labelling of investment vehicles through a taxonomy for responsible investment
(Schwartzkopff, 2021). In New Zealand, the Financial Markets Authority has developed a
disclosure framework for integrated financial products (FMA, 2020), while upcoming
legislation will mandate climate-related financial disclosures for investment schemes with
more than $1 billion in assets under management (Ministry for the Environment, 2021).
Following suit, the Australian Securities and Investments Commission has recently engaged a
review of the ‘sustainable’ products offered by its fund management industry (Kennaway,
2021). Clearly, there are concerns surrounding the substance and transparency of RI claims,
and consequently, a stream of literature focussed on institutional ‘greenwashing' has emerged.

2

To capitalise on the growing ESG investment market, funds that greenwash present themselves
in such a manner that they appear more responsible than they truly are (Brandon et al., 2021;
Liang et al., 2020). Recent studies have highlighted that institutions are signalling their
commitment to RI without exhibiting better ESG performance than their uncommitted peers,
particularly in the U.S. (Brandon et al., 2021; Liang et al., 2020; Kim & Yoon, 2020). If
investors cannot differentiate between greenwashing funds (‘lemons’) and genuine responsible
funds, this information failure is an example of adverse selection, which may negatively impact
investor welfare (Akerlof, 1970).
This paper investigates the ESG strategies and performance of global equity funds available to
Australian and New Zealand retail investors. First, we surveyed 105 asset managers to elicit an
understanding of how and why they integrate ESG information into investment decisions,
obtaining 44 usable responses. Second, we compared survey responses with portfolio holdings
data to evaluate the stated and actual carbon performance of fund managers. Finally, we
investigated the determinants of carbon and ESG performance across our entire sample of
survey respondents and non-respondents. To measure carbon and ESG performance, we
calculated value-weighted portfolio carbon intensities and ESG scores using recent portfolio
holdings data. To address the emerging issue of divergent ESG ratings (Berg et al., 2019), we
utilised three leading ESG rating providers in our analysis: Refinitiv, MSCI, and Sustainalytics.
Through our survey, we highlighted the wide adoption of RI strategies practised by global
equity funds available in Australasia, which has been primarily motivated by performance-
focused value, rather than ethical values. The divergence between value and values was
emphasised for funds with U.S-based headquarters. We documented that the most common RI
approaches were integrating ESG considerations into fundamental analysis, negative
screening/exclusions, and active ownership of portfolio companies. Within the investment
process, funds with U.S. headquarters tended to prioritise environmental themes, which have
objective data and strong links to performance through rising stakeholder emphasis (Credit
Suisse, 2021; Climate Action 100+, 2020; Trinks et al., 2020; TCFD, 2017). In contrast, funds
with Australasian headquarters tended to equally prioritise environmental, social, and
governance themes. Across various ESG subthemes, fund managers tended to place the highest
importance on climate change, followed by corporate behaviour. Despite this, the number of
respondents that responded with their portfolio-level carbon intensities was very low.

3

Through our comparison of portfolio holdings data to survey responses, we highlighted a
divergence between the stated and actual carbon performance of respondents. We found that
portfolio carbon intensity was significantly higher for respondents that were members of a
climate-related initiative (such as the Climate Action 100+, Net Zero Asset Managers, and the
Investor Group on Climate Change), and not significantly different for respondents that placed
a high level of importance on climate change in their response to the survey, or those that had
a portfolio decarbonisation strategy. This finding appears to be consistent with greenwashing,
rather than being a consequence of active engagement. To attract responsible investment flows,
funds are overstating their environmental commitments without ‘walking the talk’. This result
has significant implications for investor welfare and highlights that regulatory intervention may
be required to address the adverse selection problem caused by ESG ‘lemons’ (greenwashing
funds).
In the final part of our analysis, we found that the carbon and ESG performance of our survey
respondents was not significantly different to that of the non-respondents. This provides
evidence that our survey is robust to self-selection bias. Alongside this, we found that the ESG
performance of ESG named funds was only significantly higher, relative to non-ESG named
funds, when using external ratings from MSCI and Sustainalytics, but not Refinitiv. This
prompts further discussion into the divergence of different ESG ratings, which is driven by the
different scope, measurement, and weighting methodologies of the providers (Berg et al., 2019).
This paper contributes to the RI literature, building from key institutional investor surveys that
examine the ESG motivations and strategies of asset managers (Amel-Zadeh and Serafeim,
2018; van Duuren et al., 2016). Our analysis provides updated insights into the evolving
landscape of RI, and we present novel findings relating to the importance placed by global
equity funds on various ESG issues. This paper also contributes to the emerging asset
management literature on greenwashing. While complementing many of the recent working
papers in this field of research (Brandon et al., 2021; Raghunandan & Rajgopal, 2021; Liang
et al., 2020; Kim and Yoon, 2020), this paper exhibits several key differences. Specifically, we
addressed fund-level heterogeneity through our own survey of global equity funds, comparing
survey responses to measures of actual carbon performance. In contrast to other studies, we
focused on the Australasian context, given the recent explosion of ESG assets and investment
flows in that region (RIAA, 2021a; RIAA, 2021b), which has not received attention within the
academic literature. Finally, we focused on funds that are available to retail investors, given

4

that this investor type is more likely to fall for the obfuscation that is associated with
greenwashing (Carlin & Manso, 2011; deHaan et al., 2021).
The remainder of this paper is structured as follows. Section 2 contains the Literature Review.
Section 3 outlines the Data and Methodology, while Section 4 details the Survey Descriptive
Results. We address the Stated vs. Actual Carbon Performance in Section 5, and the
Determinants of Carbon and ESG Performance in Section 6. Finally, Section 7 presents the
Conclusion.
2. Literature Review
2.1 Responsible Investing
Among fund managers, approaches to ESG vary significantly and can include exclusion-based
screening, best-in-class screening, decarbonisation, thematic investing, quantitative ESG factor
investing, direct engagement with companies, and the incorporation of ESG risks into financial
valuations (RIAA, 2021a). 3 RI approaches are not mutually exclusive, and a full ESG
integration strategy seeks to identify risks and exploit opportunities while evaluating ESG
factors throughout the entire investment process (Eccles et al., 2017). A recent body of
literature has highlighted that ESG considerations are not only being incorporated by
institutions that explicitly label themselves as socially responsible, but also by ‘conventional’
fund managers (van Duuren et al., 2016; Amel-Zadeh & Serafeim, 2018; Alda, 2020). Potential
explanations for this include the evolving role of fiduciary duty to also consider material non-
financial information (Waygood et al., 2009), the improved ability to mitigate risks (Boubaker
et al., 2020), the opportunity to generate higher financial returns (Albuquerque et al., 2019;
Revelli, 2017), and finally, stakeholders’ demand for responsible investments (Kim & Yoon,
2021; Bauer et al., 2021; Alda, 2020). Through their survey of mainstream investment
organisations, which represented 43% of the total global assets under management in 2015,
Amel-Zadeh and Serafeim (2018) investigated the motivations behind ESG investing. In their
results, most respondents chose to incorporate ESG considerations into investment decisions
because they were financially material to investment performance (value). In comparison, a
much smaller proportion of respondents chose to consider such information because they
considered it their ethical responsibility to do so (values). While ESG investing approaches
differed greatly across institutions, the study found that the most common techniques were
incorporating ESG information into financial valuations, engaging with firms to make positive

3 A full list of definitions for these approaches can be found in Appendix 1.

5

change, and defining the investment universe through negative screening. Within the
Australasian context, these approaches have also been documented as being dominant among
Australian and New Zealand fund managers (RIAA, 2021a; RIAA 2021b). Alongside this, a
survey conducted by van Duuren et al. (2016) investigated the ESG strategies of international
fund managers. In their analysis, they highlighted similarities between ESG investing and
fundamental investing, given that ESG investors tended to focus on company-level analysis
over industry-level analysis. Out of the three ESG dimensions, fund managers were most
focused on governance factors, emphasising the importance of quality management over other
social and environmental issues. Both Amel-Zadeh and Serafeim (2018) and van Duuren et al.
(2016) found significant differences in the geographic perceptions of ESG investing between
the United States and Europe. For example, they found that U.S.-based fund managers tended
to be more sceptical about the social benefits of RI, while European-based managers were more
likely to view ESG integration as their ethical responsibility. In the U.S., there is an ongoing
regulatory debate around whether RI falls within the scope of fiduciary duty for institutional
investors (Brandon et al., 2021). In this paper, we extended the prior analysis to also investigate
institutions that are available to Australia and New Zealand retail investors and compared the
approaches of funds that have Australasian-based headquarters to those based in the U.S. and
other regions.
Through his literature review, Cappucci (2018) evaluated the incorporation of RI strategies
among fund managers, proposing an ‘ESG investing paradox’. While many fund managers
touted their ESG capabilities, they tended to fall short of their commitment to full ESG
integration, often only implementing half-measures. For example, a survey conducted by
Eccles et al. (2016) found that only 21% of institutional investors incorporated a strategy of
full ESG integration. Similarly, van Duuren et al. (2016) found that fund managers preferred
to incorporate ESG information through modified inputs such as company analysis (81%) and
external ratings (45%) over raw ESG data (30%), implying that they did not truly integrate
ESG into their bottom-up fundamental analysis (Cappucci, 2018). Notably, voluntary surveys
often overstate true ESG integration strategies given that ESG underperformers are less likely
to respond. While the adoption of RI strategies has expanded rapidly within the fund
management industry, the movement has arguably been diminished by managers who focus
solely on exclusions. Statman (2020) argued that these institutions are ‘waving banners’,
promoting their social conscience while “doing nothing to enhance the utilitarian, expressive
and emotional benefits of others” (p.g. 23). Under this interpretation, funds that divest from

6

fossil fuel firms are merely selling them to other investors. In contrast, Heinkel et al. (2001)
argued that divestments place downward pressure on stock valuations, which increases the
costs of raising capital. Within the Australasian context, negative screening is widely
incorporated, with tobacco production and controversial weapons being the most frequently
excluded sectors as traditional non-sin screening (RIAA, 2021a; RIAA, 2021b). There are
various potential explanations for the ‘ESG investing paradox’. For example, institutions are
limited in their ability to integrate ESG factors given a lack of standards that govern ESG
reporting (Amel-Zadeh & Serafeim, 2018). Similarly, they are constrained by a lack of high-
quality company data and cross-company comparability (Eccles et al., 2016). According to
Cappucci (2018), many fund managers also face misaligned incentives given that the
investment performance of managers is typically measured in 1-, 3-, and 5- year time horizons,
while many of the benefits of ESG investments are likely to materialise beyond these. Finally,
while processed ESG ratings play a crucial role in guiding a fund’s portfolio allocations, several
studies have highlighted a divergence in external ESG ratings between different providers
(Chatterji et al., 2016; Berg et al., 2019). This is because different providers measure and weigh
ESG issues differently, as well as have varying scopes of what to measure and aggregate (Berg
et al., 2019). This is a consequence of the lack of generally accepted ESG standards. For
example, some providers may consider activities such as lobbying to be a relevant ESG issue,
while others may not. According to Berg et al. (2021), ESG data tends to be gathered from firm
Corporate Social Responsibility (CSR) reports, regulatory filings, modelled data,
questionnaires, and the media. For example, the measurement of a firm’s product safety could
be based on the information provided in a CSR report, media reports about the firm, or
complaints made to the regulator. Hence, a provider’s choice of measurement may lead to
divergent assessments of performance, which decreases the reliability of the ratings. More
transparent ESG rating provider methodologies could help elevate some of the issues as
investors/managers can choose a provider that aligns with their values/goals. Further, the
development of disclosure standards, taxonomies of sustainable activities, and spatial finance
data as well as more providers may help alleviate some of these issues.
Several research papers have investigated the relationship between firm ESG performance and
empirical stock returns. However, many studies rely on external ESG ratings as their proxy for
performance, leading to inconsistent conclusions given the divergence bias discussed above.
For example, some studies found that firms with better ESG performance had higher stock
returns (Albuquerque et al., 2019; Lins et al., 2017) while some found that they had lower stock

7

returns (Chava, 2014; El Ghoul et al., 2011). Despite this, the modern financial landscape now
recognises that ESG risks can cause substantial losses for both investors and lenders. For
example, climate risks may reduce the value of a firm’s assets, driven by physical damage from
weather events or the stranding of assets due to the global energy transition (Chenet, 2021).
Alongside this, reputational risks may lead to litigation costs and lower operating performance
for firms. Earnings may be permanently reduced if consumers choose to boycott firms that act
unethically (Tamayo-Torres et al., 2019). Given this, a series of recent literature has established
a positive relationship between CSR and creditworthiness (Jiraporn et al., 2014), access to
finance (Cheng et al., 2013), and a negative relationship between CSR and default risk
(Boubaker, 2020). These studies support the incorporation of ESG considerations as a risk-
mitigation strategy.
2.2 Greenwashing
The rise in financial products branded as ‘ethical’ or ‘responsible’ has prompted global
regulators, such as those in the U.S., Europe, Australia, and New Zealand, to review their asset
management industries. Within the U.S., the Securities Exchange Commission recently found
evidence that funds were not adhering to their own ESG claims (Chin, 2021), while the
European Commission adopted a taxonomy for responsible investment to address the
sustainable labelling of investment products (Schwartzkopff, 2021). Consequently, an
emerging stream of literature has investigated asset manager ‘greenwashing’, where
institutions overstate their commitment to RI (Liang et al., 2020; Kim & Yoon, 2020; Brandon
et al., 2021). By appearing more responsible than they truly are, greenwashing enables funds
to profit from the increased demand for ESG investing (Brandon et al., 2021). Investors may
not be able to differentiate between greenwashing funds and genuine responsible funds because
they suffer from asymmetric information, where fund managers possess more private
information about the quality of their ESG claims. This may be emphasised for retail investors
who are more likely to fall for the obfuscation that is associated with greenwashing (Carlin &
Manso, 2011; deHaan et al., 2021). Consequently, this may create an adverse selection problem
that drives genuine funds out of the market, leaving only ESG ‘lemons’ (greenwashing funds),
to the detriment of investor welfare (Akalof, 1970).
Corresponding to the significant rise in signatories to the Principles of Responsible Investment
(PRI), recent studies have evaluated whether member institutions are fulfilling their ESG
commitments or merely using their status to attract socially conscious investment flows.

8

Through their analysis of global hedge funds, Liang et al. (2020) found that one-fifth of PRI
signatories had portfolio ESG scores that fell below the median score for all hedge funds in the
sample. Despite possessing relatively low ESG scores, most of these signatories promoted
ESG-related terminology on their websites and were 92.9% likely to exhibit low ESG scores
in the following year, consistent with greenwashing. Furthermore, an analysis of U.S. active
fund managers by Kim and Yoon (2020) highlighted that PRI signatories did not exhibit
material improvements in their fund-level ESG performance post endorsement. However, they
found that flows into signatory funds increased significantly regardless of prior ESG
performance, suggesting that U.S. institutions were using their PRI status to capitalise on the
growing responsible investment market. While both studies found evidence of greenwashing,
they treated PRI signatories as a homogenous group of investors, failing to address the
heterogeneity in the type and intensity of ESG investing strategies implemented.
To overcome the issue of institutional heterogeneity, Brandon et al. (2021) compared equity
portfolio holdings data to a survey of the RI strategies of global institutions. This enabled the
researchers to examine whether stated RI strategies lead to better portfolio-level ESG outcomes,
applying a similar intuition to our study. In their methodology, institutions were placed into
groups based on their reported level of commitment: (1) full ESG incorporation into 100% of
their equity assets under management (AUM); (2) partial ESG incorporation; and (3) no
reported ESG incorporation. To address the widely cited divergence in ESG ratings, Brandon
et al. (2021) used an average of three external ESG providers when analysing an institution’s
actual ESG performance. In their results, they highlighted significant disparities between U.S.
and non-U.S. signatories, reinforcing the findings by Amel-Zadeh and Serafeim (2018) and
van Duuren et al. (2016). For example, U.S.-domiciled PRI signatories who reported partial or
full ESG integration did not obtain better portfolio ESG scores than non-PRI institutions, while
non-U.S. signatories did obtain better scores than non-signatories. Consistent with
greenwashing, U.S.-domiciled PRI signatories who reported no ESG incorporation obtained
significantly worse portfolio ESG scores than non-PRI institutions (Brandon et al., 2021).
While evaluating greenwashing, Brandon et al. (2021), Liang et al. (2020) and Kim and Yoon
(2020) conducted analysis at the institution level. Like our study, Raghunandan and Rajgopal
(2021) analysed greenwashing at the fund level. They evaluated whether self-labelled ESG
mutual funds in the U.S. make stakeholder-friendly investments. Rather than using portfolio-
level ESG scores, they measured a fund’s performance through a series of weighted-average
environmental, social and governance metrics. Their results highlighted that ESG funds had

9

significantly more labour and environmental law violations, alongside a higher CO2 emissions
intensity, relative to the non-ESG funds issued by the same asset manager in the same year.
Raghunandan and Rajgopal (2021) argued that this result was consistent with ESG funds being
more “concerned about the existence of firm disclosures rather than the content of the
information being disclosed” (p.g. 3). Like Raghunandan and Rajgopal (2021), this paper also
uses our own calculated portfolio carbon intensity metrics, rather than relying solely on
external ESG ratings for our analysis.
As discussed in Delmas and Burbano (2011), an organisation’s tendency to greenwash
corresponds to their consumer-, investor-, and competitor-induced incentives, firm-level
characteristics, and most notably, the regulatory environment that they operate in. External
market drivers such as investor demand for sustainable investments have placed pressure on
‘brown’ institutions to overstate their commitment to RI, particularly when there are no
regulatory consequences (Delmas and Burbano, 2011). Among institutions, the competitive
landscape is converging - more organisations are modelling themselves as sustainable leaders
to appear legitimate and reputable among various stakeholders, even if they truly are not
(Delmas and Burbano, 2011). Finally, firm-level characteristics such as incentive structures
and ethical climate are relevant determinants. Until recently, institutional greenwashing has
been characterised by limited regulation and uncertain enforcement, allowing funds to profit
from the expanding ESG market while simultaneously avoiding punitive consequences.
However, recent investor welfare concerns have prompted regulators to intensify their scrutiny
of such practices (Schwartzkopff, 2021; Chin, 2021; Kennaway, 2021).
To measure greenwashing, academic studies control for the common fund-level characteristics
that drive ESG performance. In a matched sample of conventional and SRI funds, Alda (2021)
found that fund size, turnover, and expense ratios were positively related to ESG scores, while
fund age was negatively related. Under a Legitimacy Theory perspective, larger funds are more
visible to stakeholders, and face greater pressures to act with transparency (Singh et al., 1986).
Other fund-level controls within the literature include risk and return, manager experience, and
finally, the region of an institution’s headquarters (Alda, 2021; Raghunandan & Rajgopal, 2021;
Brandon et al., 2021). Specifically, evidence suggests that institutional investors tend to
incorporate more sustainable practices when regions have high environmental and social norms,
particularly in European countries (Dyck et al., 2019). Under the ‘Institutional Theory’
perspective, divergent social structures such as the cultural system, labour system and political

10

system can heavily influence an institution’s ethical behaviour given varying levels of
corporate monitoring (Campbell, 2007).

3. Data and Methodology
3.1 Primary Data: Survey Design
Following our review of the RI literature, we developed a survey to understand the ESG
motivations and strategies of global equity funds available to Australian and New Zealand retail
investors. Several key questions were inspired or adapted from the existing literature (Amel-
Zadeh and Serafeim, 2018; van Duuren et al., 2016; Krueger et al., 2020), alongside our own
novel contributions. The draft version of the survey received feedback from leading academic
researchers and our external partners at Morningstar, MyFiduciary and Saturn Advice. We also
solicited feedback from an asset management institution to ensure that the questions were
appropriate for practitioners. Based on this feedback, we redrafted, added, and removed survey
questions. The final version of the survey contained a total of 36 questions (see Appendix 2).
However, aware of the potential for survey fatigue, we designed a survey using ‘branching
logic’, where some questions were only displayed conditional on their previous responses (see
Figure 1).
The final version of the survey received approval from the University of Otago Human Ethics
Committee. It was added to the internet-based survey instrument Qualtrics and was sent out to
105 institutions via email on the 24th of September 2021. Our initial sample was constructed
by our external team partners from Morningstar and MyFiduciary, who identified the names
and contact details of institutions that offered global equity funds to Australian and New
Zealand retail investors. Any non-respondents were sent reminders to complete the survey on
the 26th of October and the 6th of November and the survey formally closed on the 12th of
November 2021. We received 44 eligible responses, reflecting an overall response rate of
41.9%. However, given that one institution outsources the management of its global equity
fund, there are only 43 eligible responses for the fund-level analysis. Compared to other
academic surveys in finance and accounting that were distributed via email (Amel-Zadeh and
Serafeim, 2018; Dichev et al., 2013; Graham et al., 2005), our response rate was significantly
higher.

11

Given the nature of this study, all respondents were required to disclose the name of their
institution and relevant global equity fund. To test whether institutions ‘walk the talk’, all
responses were required to be identifiable so that they could be matched to the portfolio
holdings data. However, to ensure anonymity, the research team emphasised that it would not
identify any respondent or respondent firm in any report or published papers arising from this
project. The survey contained a range of question types including text entry, multiple-choice,
and constant sum (point allocation) questions. Multiple-choice questions were accompanied
with free text response options to provide information and clarification on responses, which
would be referenced for additional context in the analysis. Constant sum questions directed
respondents to allocate one hundred points between various alternatives to establish relative
weightings (Mühlbacher and Botschen, 1988).
Referring to Figure 1, the survey began by asking institution-level demographic questions. For
example, we asked for the respondent’s job title, the institution’s size and headquarters location,
level of ESG incorporation across the institution, ESG memberships, attainment of ESG
capabilities, and finally the ESG motivations of the firm. The second part of the survey asked
respondents to provide their institution’s flagship retail ESG global equity fund. This could be
a conventional global equity fund if an institution does not offer an ESG specific fund. Enabling
respondents to choose their own flagship ESG fund is likely to bias the results against obtaining
a ‘greenwashing’ conclusion, making any such results even more credible. The remainder of
the survey asked questions relating to the fund provided. For example, we asked demographic
questions relating to the fund size, average holding period, investment style, benchmark, and
manager demographics. We then asked about the RI approaches incorporated, types of ESG
data used, engagement strategies, and voting approaches. Next, we instructed respondents to
weigh various ESG issues based on the importance that their fund places on them within the
investment process. These ESG issues were identified based on a review of the common
subthemes represented in various external ESG provider databases and white papers, such as
Bloomberg, Refinitiv, and MSCI (Refinitiv, 2021; MSCI, 2020). Finally, we asked respondents
to provide calculations that measure the carbon emissions intensity of their fund. The full
survey can be found in Appendix 2.
In any survey, there is a risk that responses are strategic or untruthful (Krueger et al., 2020).
This may be amplified when respondents are identifiable to researchers. However, given that
our respondents were aware that their responses would be evaluated relative to objective
portfolio holdings data, any untruthful responses would help us to establish evidence in favour

12

of greenwashing. It is possible that our survey suffered from self-selection bias, where our
respondents were more likely to have higher ESG credentials relative to non-respondents.
However, we evaluated this later in the paper and found that responding funds did not have
significantly different portfolio carbon intensities or portfolio ESG scores relative to non-
responding funds (see Sections 4.2 and 4.3)
3.2 Secondary Data
We obtained portfolio holdings data for our sample of global equity funds from our external
partner at Morningstar. For survey respondents, we gathered holdings data on the specific fund
provided by the individual institution in the survey (i.e., their self-identified flagship ESG fund).
For non-respondents, we gathered holdings data based on the most ESG incorporated fund
offered by the individual institution, as identified by our external partners at MyFiduciary and
Morningstar. However, when an institution does not offer an ESG-specific global equity fund,
the funds chosen by respondents, and those selected for non-respondents, could be general
global equity funds. Of the 105 institutions in our sample, 78 funds had holdings data available
in the Morningstar database. This includes 33 funds who had responded to our survey and 45
funds who did not respond. The holdings data was dated 30 June 2021, or as close to this as
possible when unavailable.
To measure carbon performance, we obtained firm-level emissions data from Refinitiv Eikon.
This data source has been employed for analysis in related studies (Liang et al., 2020; Dyck et
al., 2019). For each security, we downloaded the reported and estimated Scope 1 and 2 CO2-
equivalent emissions data measured in tonnes of CO2 equivalent (tCO2e) and total revenue
data measured in millions of U.S. dollars.4 The reported data is based on company filings, while
estimated data is based on Refinitiv’s carbon estimate models. We examined Scope 1 and 2
CO2 equivalent emissions intensity, which is a metric commonly used within the industry to
measure a firm’s emission efficiency (Garvey et al., 2018; Raghunandan & Rajgopal, 2021).
In this calculation, emissions are scaled by revenues, allowing for a comparison between firms
of different sizes. The carbon intensity for each stock 𝑖 is defined in Equation (1) below:
𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦𝑆𝑐𝑜𝑝𝑒 1 & 2 𝐶𝑂2 𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 𝐸𝑚𝑖𝑠𝑠𝑖𝑜𝑛𝑠 𝑡𝐶𝑂2𝑒
𝑇𝑜𝑡𝑎𝑙 𝑅𝑒𝑣𝑒𝑛𝑢𝑒 $𝑈𝑆𝑚, 1

4 Scope 1 refers to the greenhouse gas (GHG) emissions that occur from company-owned and controlled resources,
such as manufacturing plants. Scope 2 refers to the indirect GHG emissions that are generated from the
consumption of purchased energy, such as from a utility provider.

13

The carbon intensity calculation initially covered 74% of the securities on our aggregated list.
For the remaining securities, we estimated the Scope 1 and 2 CO2 equivalent emissions
intensity using industry averages based on GICS sector codes. Given a delay in reporting and
data availability, carbon intensity figures were calculated for each security based on the fiscal
year ending 2020. We acknowledge that this data slightly predates our portfolio holdings date
of 30 June 2021. However, we assume that emission intensity figures are ‘sticky’ and are not
materially different to those on the date of our holdings data. We also note that fund managers
are likely to rely on predated data when making informed portfolio decisions, given the delayed
reporting of emissions.
To measure the carbon efficiency of each fund in our sample, we calculated the value-weighted
portfolio carbon intensity as in Equation (2).
𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦𝑤,∙𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦,,


2
where, the 𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 quantifies the weighted average carbon intensity for
fund 𝑗 at the holdings date. The variable 𝑤, denotes the long-only value-weighting of stock 𝑖
in fund 𝑗‘s portfolio. 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦, is the Scope 1 and 2 carbon intensity of stock 𝑖 in
fund 𝑗‘s portfolio. 𝑁 denotes the total number of stocks invested in fund 𝑗.5 This metric was
chosen as a proxy for the carbon performance of the global equity funds in our sample, given
it measures the carbon efficiency of a fund.
Next, we analysed ESG performance, obtaining firm-level ESG scores based on the ISIN codes
of the securities held by the 78 global equity funds. We attempted to overcome the issue of
divergent ESG ratings by utilising three leading ESG providers in our analysis: Refinitiv, MSCI,
and Sustainalytics. Refinitiv measures a company’s relative ESG performance, commitment,
and effectiveness (Refinitiv, 2021), scoring firms on a percentile scale from 0 (‘poor’) to 100
(‘excellent’). MSCI measures a firm’s resilience to long-term, financially relevant ESG risks
(MSCI, 2020), scoring companies on a scale from 0 (‘laggard) to 10 (‘leader’). Finally,
Sustainalytics measures the unmanaged risks of a company concerning ESG issues that are
considered material (Garz et al., 2019), scoring firms on a scale from 0 (‘negligible risk’) to
40+ (‘severe risk’). We inverted the Sustainalytics scores for our regressions analyses to ensure

5 Despite using long-only portfolio positions in our calculation, our results are almost identical when calculating
net-positions, given the lack of short positions in the sample of portfolio holdings.

14

that a higher score implies better ESG performance and coefficients are comparable. Based on
the available data, Refinitiv ESG scores cover 76% of the securities on our aggregated list,
while MSCI and Sustainalytics ESG scores have data coverage of 78% each. Due to data
availability, we also obtained ESG rating data based on the fiscal year ending 2020, assuming
that ESG scores are not materially different to the date of our fund holdings data. As above, we
noted that fund managers are likely to rely on predated ESG data when making portfolio
decisions.
For each fund, we calculated the value-weighted portfolio ESG Score as in Equation (3).
𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐸𝑆𝐺 𝑆𝑐𝑜𝑟𝑒𝑤,∙𝐸𝑆𝐺 𝑆𝑐𝑜𝑟𝑒,,


3
where 𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐸𝑆𝐺 𝑆𝑐𝑜𝑟𝑒 quantifies the weighted average Refinitiv, MSCI or
Sustainalytics ESG score for fund 𝑗 at the holdings date. The variable 𝑤, represents the
normalised long-only value-weighting of stock 𝑖 in fund 𝑗‘s portfolio. For each fund 𝑗,
individual weightings have been normalised (rescaled) based on the holdings that have
available ESG ratings. This method is commonly used by fund ESG rating providers to address
missing data that is difficult to estimate (MSCI, 2021; Barr et al., 2021). 𝐸𝑆𝐺 𝑆𝑐𝑜𝑟𝑒 , is the
Refinitiv, MSCI or Sustainalytics ESG Score of stock 𝑖 in fund 𝑗‘s portfolio. 𝑁 denotes the
total number of stocks invested in fund 𝑗. A higher weighted-average ESG Score indicates that
a fund has relatively better ESG performance, as measured by our three different rating
providers (including our inverted Sustainalytics scores).
Finally, we utilised the Morningstar Direct database to gather portfolio-level data for our
sample of global equity funds to use in regression analysis. For example, we obtained data
relating to the institution’s location of headquarters, fund style, size, management fees, fund
age, 12-month financial performance and volatility. This data, which is consistent across survey
respondents and non-respondents, were used in our subsequent analysis.
3.3 Determinants of Respondent Carbon Performance
In this section, we analysed the drivers of carbon performance of our survey. To do this, we
integrated survey responses into OLS regression analysis to address the heterogeneity between
different funds. However, as we only had holdings data for 33 responding funds, this limits the

15

number of explanatory variables that we could use in our analysis. Consequently, caution
should be maintained given that our models lack a complete set of fund-level controls. To
overcome this concern, we individually rotated each variable of interest with various sets of
control variables. We analysed the determinants of portfolio carbon intensity for respondents
in Equations (4) and (5).6
𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦
𝛽
𝛽
𝑋𝛽
𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠𝛽
𝐸𝑆𝐺 𝑁𝑎𝑚𝑒𝜀
, 4
𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦
𝛽
𝛽
𝑋𝛽
𝑆𝑡𝑦𝑙𝑒𝛽
𝐿𝑛𝑆𝑖𝑧𝑒𝜀
, 5
Here, we tested various explanatory variables individually, represented by 𝑋, after controlling
for 𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠 and 𝐸𝑆𝐺 𝑁𝑎𝑚𝑒 in Equation (4), and 𝑆𝑡𝑦𝑙𝑒 and 𝐿𝑛𝑆𝑖𝑧𝑒 in Equation
(5). 𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠 is a categorical variable that reflects the region of the institution’s
headquarters for fund 𝑗. 𝐸𝑆𝐺 𝑁𝑎𝑚𝑒 is a binary variable equal to one if fund 𝑗 has an ESG-
related name, and zero otherwise. 𝑆𝑡𝑦𝑙𝑒 is a categorical variable that reflects the investment
style of fund 𝑗. This compares Value and Growth investment styles relative to Blend styles.
Ln𝐹𝑢𝑛𝑑 𝑆𝑖𝑧𝑒 represents the natural logarithm of net assets (USD $m). All control variables
have been sourced from the Morningstar Direct database. 𝑋 represents the following binary
variables of interest: 𝐶𝑙𝑖𝑚𝑎𝑡𝑒 𝐼𝑛𝑖𝑡𝑖𝑎𝑡𝑖𝑣𝑒 takes the value of one if the institution related to
fund 𝑗 is a member/signatory of a climate-related initiative (such as the Climate Action 100+,
Net Zero Asset Managers, and the Investor Group on Climate Change) and zero otherwise.
𝑀𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝑡𝑜 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 takes the value of one if fund 𝑗 indicates that ESG
information is material to investment performance and zero otherwise.
𝐸𝑡ℎ𝑖𝑐𝑎𝑙 𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑖𝑏𝑖𝑙𝑖𝑡𝑦 takes the value of one if fund 𝑗 view ESG considerations as an
ethical responsibility and zero otherwise. 𝐷𝑒𝑐𝑎𝑟𝑏𝑜𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 takes the value of one if fund 𝑗
targets portfolio decarbonisation as a RI approach and zero otherwise. 𝐹𝑢𝑡𝑢𝑟𝑒 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒
takes the value of one if fund 𝑗 believes that ESG risks, although not yet priced, will soon
impact investment performance and zero otherwise. 𝐸𝑛𝑔𝑎𝑔𝑒𝑚𝑒𝑛𝑡 takes the value of one if
fund 𝑗 actively engages with its portfolio companies on ESG issues and zero otherwise. 𝑋 also

6 We also analysed the actual vs. stated ESG performance. However, these results lack statistical power. Relative
to carbon intensity, portfolios had less ESG data coverage given that this data is hard to estimate. In addition to
Equations (4) and (5), we also tested the results using different combinations of control variables. Results are
available upon request.

16

represents the numeric variables 𝐸𝑛𝑣𝑖𝑟𝑜𝑛𝑚𝑒𝑛𝑡 𝑊𝑒𝑖𝑔ℎ𝑡𝑖𝑛𝑔 and
𝐶𝑙𝑖𝑚𝑎𝑡𝑒 𝐶ℎ𝑎𝑛𝑔𝑒 𝑊𝑒𝑖𝑔ℎ𝑡𝑖𝑛𝑔, which reflects the relative importance placed by fund 𝑗 on
environment and climate change (weighted) themes, respectively (see Section 4.1.4). The
explanatory variables represented by 𝑋, have been chosen to reflect the stated RI measures of
each fund.
3.4 Determinants of Carbon and ESG Performance for all funds
Next, we estimated an OLS regression model to analyse the determinants of portfolio carbon
intensity for our entire sample of 78 responding and non-responding funds in Equation (6).
𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐶𝑎𝑟𝑏𝑜𝑛 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦
𝛽
𝛽
𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠𝛽
𝑆𝑡𝑦𝑙𝑒𝛽
𝐸𝑆𝐺 𝑁𝑎𝑚𝑒
𝛽
Ln𝐹𝑢𝑛𝑑 𝑆𝑖𝑧𝑒𝛽
𝑆𝑢𝑟𝑣𝑒𝑦 𝑅𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡𝛽
𝐴𝑔𝑒
 𝛽𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒𝛽
𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝛽
𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 𝐶𝑜𝑣𝑒𝑟𝑎𝑔𝑒
𝜀
 6
where the explanatory variables 𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠,𝐸𝑆𝐺 𝑁𝑎𝑚𝑒, 𝑆𝑡𝑦𝑙𝑒, and Ln𝐹𝑢𝑛𝑑 𝑆𝑖𝑧𝑒
have been defined above (see Section 3.3). 𝑆𝑢𝑟𝑣𝑒𝑦 𝑅𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡 is a binary variable equalling
to one if fund 𝑗 has responded to our survey, and zero otherwise. We employ other fund-level
controls such as 𝐴𝑔𝑒, 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒, and 𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦, which represent the fund
age (months), 12-month financial performance, and the standard deviation of the 12-month
financial performance of fund 𝑗, respectively, at the respective date of holdings disclosure. All
control variables have been sourced from Morningstar Direct. For consistency, variables that
were self-reported by the respondents in our survey were replaced with those from the
Morningstar Direct database for this analysis. Finally, 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 𝐶𝑜𝑣𝑒𝑟𝑎𝑔𝑒 is defined as the
percentage of fund 𝑗’s gross weight that comes from securities with carbon intensity data
available.7
Finally, to analyse the determinants of a fund’s portfolio ESG score, we utilised an OLS
regression model for the entire sample of respondents and non-respondents in Equation (7).

7 We excluded funds that have an intensity coverage value of less than 70% from our regression analysis.

17

𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝐸𝑆𝐺 𝑆𝑐𝑜𝑟𝑒
𝛽
𝛽
𝐻𝑒𝑎𝑑𝑞𝑢𝑎𝑟𝑡𝑒𝑟𝑠𝛽
𝑆𝑡𝑦𝑙𝑒𝛽
𝐸𝑆𝐺 𝑁𝑎𝑚𝑒
𝛽
Ln𝐹𝑢𝑛𝑑 𝑆𝑖𝑧𝑒𝛽
𝑆𝑢𝑟𝑣𝑒𝑦 𝑅𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡𝛽
𝐴𝑔𝑒
 𝛽𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒𝛽
𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝛽
𝐸𝑆𝐺 𝐶𝑜𝑣𝑒𝑟𝑎𝑔𝑒
𝜀
 7
Here, 𝐸𝑆𝐺 𝐶𝑜𝑣𝑒𝑟𝑎𝑔𝑒 is a control variable that reflects the percentage of fund 𝑗’s gross weight
of securities with Refinitiv, MSCI, or Sustainalytics ESG data available.8 All other control
variables are consistent with those defined earlier in this paper.
The control variables used in Equations (6) and (7) have been adopted from the literature that
analyses greenwashing at the fund level (Raghunandan & Rajgopal, 2021; Alda, 2020). Other
variables of interest, such as the region of headquarters, stemmed from our review of the RI
literature. Specifically, prior studies have found that U.S.-based institutions were more
sceptical about the social benefits of ESG considerations (Amel-Zadeh & Serafeim, 2018; van
Duuren et al. 2016), and were more commonly engaged in greenwashing (Brandon et al., 2021;
Kim & Yoon, 2020).
4. Results
4.1 Survey Descriptive Results
4.1.1 Survey Summary Statistics
The left-hand panel of Table 1 reports the demographic characteristics of respondents and their
associated institutions. Despite our Australasian investor focus, the responding institutions
were diverse with headquarter locations spread among Australia (36%), New Zealand (27%),
the U.S. (23%), the U.K. (5%) and other countries (9%). Among the responding institutions,
80% reported that they incorporated ESG considerations across all of their funds, while 20%
reported that they did across some funds. This shows that ESG considerations have become
more important since previous surveys, such as in Amel-Zadeh and Serafeim (2018), who
found that 35% of respondents did not allocate any portion of their AUM to ESG-related
investments. Our finding could be driven by the loose definition of ESG considerations. For
example, some funds may believe classical Governance practices and traditional exclusions,
such as pornography and tobacco, are ESG considerations. It may also reflect the growing

8 We excluded funds that have an ESG coverage value of less than 70% from our regression analysis, as it is a
common practice in the fund ratings industry (MSCI, 2021).

18

importance of ESG issues over time given that our survey was undertaken in 2021 while their
survey was distributed in 2015.
[INSERT TABLE 1 HERE]
When asked how their institution builds ESG capabilities, most respondents indicated that they
conducted in-house training (95%), while only a third (34%) used external training
providers. We asked the respondents to report their institution’s commitments to ESG
initiatives. Most institutions were signatories of the PRI (91%), and just over half (59%) were
members of at least one climate-related initiative such as the Carbon Disclosure Project (39%),
Climate Action 100+ (30%), and the Investor Group on Climate Change (27%).
The right-hand side of Table 1 illustrates the characteristics of the flagship ESG global equity
funds provided by respondents in the survey. Of the 43 funds, the median size was US$340
million, 91% were actively managed, and only one indicated that their lead portfolio manager
was female (2%), highlighting a significant gender disparity. Over half of the funds (56%) had
a holding period of two to five years, while the most popular investment style was quality
(47%), followed by growth (37%) and fundamental strategies (35%).
For our subsequent analyses, we divided the sample into different groups to compare different
demographic characteristics. We conditioned the survey responses based on (1) the region of
the institution’s headquarters (Australasia, US, and other); (2) the size of the global equity fund
(greater than 50% of the median size and less than or equal to 50% of the median size), and;
(3) whether the fund has an ESG-related name (yes and no).
4.1.2 Responsible Investing Motivations
Corresponding to the rise of RI strategies within the asset management industry, we assessed
the key motivations of the institutions in our sample. Table 2 presents the survey responses to
the question “Why does your institution consider ESG information when making investment
decisions”, where respondents could choose one or more alternatives that represented their
institution. This question was adapted from Amel-Zadeh and Serafeim (2018).9
[INSERT TABLE 2 HERE]

9 Respondents were allowed to provide text entry answers for reasons that were not included in this list.

19

Table 2 highlights that RI appears to be primarily driven by value rather than values
Specifically, most respondents indicated that ESG risks are material to financial performance
(81%), followed by growing client/stakeholder demand (74%). This result is consistent with
the findings of Amel-Zadeh and Serafeim (2018). Fewer respondents cited having an ethical
responsibility (51%) or encouraging positive change in individual companies (65%) as
motivations for RI.
The divergence between performance-based value and ethical values as a motivation for RI is
emphasised for the U.S. region in Column 3 of Table 2. While 89% of U.S. institutions used
ESG information because it is material for financial performance, only 11% did so because it
is their ethical responsibility. Across the regions, there is a large disparity in the proportion of
respondents who selected ‘We see it as an ethical responsibility’. We conducted a Chi-Square
Test of Independence to assess the relationship between ethical responsibility and region of
headquarters (comparing Australasia and the U.S. only due to sample size constraints). At the
1% level, we concluded that there is a significant relationship between the two variables (𝜒
(1, N=44) = 8.69, p = .003). Australasian institutions were more likely to view RI as their
ethical responsibility, relative to U.S. institutions. These findings relate to both Amel-Zadeh
and Serafeim (2018) and van Duuren et al. (2016) who found that US-based funds were more
sceptical about ESG investing and more likely to adopt RI practices for performance-based
reasons, in comparison to non -U.S. funds. This appears to be consistent with our results.
4.1.3 Responsible Investing Approaches
In Table 3, we assessed the responsible investment approaches adopted by the global equity
funds provided by respondents (see Question 23 in Appendix 2, which was adapted from Amel-
Zadeh and Serafeim (2018)). Definitions for RI approaches can be found in Appendix 1.
[INSERT TABLE 3 HERE]
The results in Table 3 reveal that the most common RI approach among respondents was
incorporating ESG considerations into fundamental analysis (84%). Following this, 81% of
funds negatively screened (excluded) stocks and 79% employed active engagement strategies
with corporations on their ESG issues. Within the survey, only 37% of funds utilised best-in-
class screening, which reflects a large divergence between positive and negative screening.
Surprisingly, only 28% targeted the decarbonization of their portfolios. Compared to Amel-

20

Zadeh and Serafeim (2018), our respondents more widely incorporated RI approaches. This
may indicate an increasing uptake in ESG adoption among fund managers since their 2015
survey was conducted, alongside our focus on the most ESG ambitious global equity funds.
Column 4 of Table 3 highlights that a significantly lower proportion of large funds incorporated
ESG considerations into fundamental analysis relative to small funds, at the 5% level. This
result is surprising, given that larger funds likely have more resources to integrate ESG
information into financial forecasts and analysis. Column 7 highlights that screening
approaches (both positive and negative) were significantly more common among ESG named
funds, relative to non-ESG named funds, at the 5% level. While not significant, ESG named
funds tended to utilise a greater variety of RI approaches such as thematic investment, portfolio
overlay, quantitative ESG factor and decarbonisation, but also incorporated ESG factors into
fundamental analysis less commonly.
4.1.4 ESG Theme Priority
Within RI, ESG considerations are broad and encompass many different themes, which may
be treated with varying levels of importance within the investment process. Table 4 presents
the relative weightings of ESG themes and subthemes by the funds in our sample. Using
constant sum (point allocation) questions, respondents were asked to allocate 100 points
between environmental, social and governance themes based on the importance placed by their
fund in the investment process. This question was repeated for four environmental subthemes
(climate change, pollution and waste, natural capital, and environmental opportunities), four
governance subthemes (board composition, remuneration, corporate behaviour, and
shareholder rights), and four social subthemes (supply chain and community, health and safety,
product liability, and human capital management). By allocating 100 points between various
alternatives, we could establish relative weightings. Unless stated otherwise, Table 4 reports
the average (mean) weighting of each theme/subtheme.
[INSERT TABLE 4 HERE]
Panel A of Table 4 highlights that on average, the global equity funds in our survey placed the
highest importance on environmental themes within the investment process, followed by
governance and then social themes. This result differs from van Duuren et al. (2016) who found
that fund managers were relatively more focused on governance factors relative to social and

21

environmental factors. Our result is likely attributable to the rising stakeholder and policymaker
emphasis on Environmental issues, particularly climate change. However, the environmental
weighting also had the largest variability among the distribution of the responses, reflected by
the standard deviation of 0.15. It is likely that funds placed a lower importance on social themes
given that the related metrics are more difficult to measure and less objective than
environmental and governance issues.
In Panel A, we found that the distribution of ESG named and non-ESG named funds, with
respect to the relative importance placed on environmental and governance themes, was
significantly different at the 5% level. Specifically, ESG named funds placed a higher
weighting on environmental themes and a lower weighting on governance themes, relative to
non-ESG named funds. This fits with the traditional view that conventional asset managers
value strong corporate governance given that it is essential for reducing agency problems
(Picou & Rubach, 2006). In Panel A, our regional analysis highlighted that on average, funds
with Australasian headquarters placed equal importance on environmental and governance
themes when selecting investments, while funds with U.S. headquarters tended to provide a
higher weighting for environment themes. Given that respondents with U.S. headquarters were
primarily motivated by financial performance, it is likely that they prioritised environmental
themes given that there is better data and stronger links to performance, relative to social
themes, which is driven by rising consumer demand through climate change awareness (Credit
Suisse, 2021; Climate Action 100+, 2020; Trinks et al., 2020; TCFD, 2017).
Panel B of Table 4 establishes the relative weightings of environmental, social and governance
subthemes. On average, the most important environmental subtheme was climate change (0.35),
followed by environmental opportunities (0.24). The most prioritised governance subtheme
was corporate behaviour (0.34), followed by shareholder rights (0.23). Finally, the most
important social subtheme was supply chain & community (0.28), followed equally by health
& safety and product liability (0.24).
Panel C of Table 4 presents the weighted ESG subthemes, which have been ordered based on
their overall level of importance. For each fund in our sample, we multiplied the weighting of
each subtheme (as represented in Panel B) by the weighting of the associated theme (as
represented in Panel A). Panel C highlights that on average, global equity funds in our sample
placed the highest importance on climate change (0.15), followed by corporate behaviour
(0.11), supply chain & community (0.09) and environmental opportunities (0.09). We noted

22

that these subthemes had considerably higher variability in their distribution of responses
relative to other subthemes. On average, fund managers placed the least importance on natural
capital (including biodiversity) within the investment decision making process, which is cause
for concern as environmental hazards go far beyond climate change and are interconnected
(Chandellier & Malacain, 2021). While fund managers tended to prioritise climate change, it
is somewhat contradictory that few respondents engaged in portfolio decarbonisation as a RI
approach (as shown in Table 3).
Panel C emphasises that ESG named funds shared a strong climate focus. In the investment
process, they placed relatively greater importance on climate change themes, relative to non-
ESG named funds, who addressed climate change and corporate behaviour with equal
importance. Our regional analysis indicates that funds with U.S.- based headquarters ranked
climate change as the most important subtheme, followed by environmental opportunities.
Again, this is likely due to the strong thematic links to financial performance, which is
prioritised in the U.S. In comparison, funds with Australasian headquarters ranked corporate
behaviour first, followed by climate change.
4.1.5 Reported Portfolio Carbon Intensity
Rising stakeholder demand for addressing climate change coupled with improved stock-level
data coverage has led many fund managers to measure their portfolio’s exposure to emissions
In Table 5, we present the reported weighted average portfolio carbon intensity of respondents’
funds, with reference to Scope 1 and 2 emissions in Panel A, and Scope 1, 2 and 3 emissions
in Panel B (see Question 34 and 35 in Appendix 2). Respondents were able to select that they
used a different intensity measure or did not calculate this metric with the opportunity to
explain their answer in free text.
[INSERT TABLE 5 HERE]
In Panel A of Table 5, only 49% of funds in the sample provided their Scope 1 and 2 portfolio
carbon intensity, while a further 9% selected that they use a different intensity calculation.
While ESG investing appears to be widely adopted across our sample of global equity funds,
this result is surprising given that carbon exposure is a material ESG risk that can cause
substantial losses for investors. Namely, assets may become stranded due to the global energy
transition (Chenet, 2021). The proportion of ESG funds that provided their weighted average

23

carbon intensity was significantly higher than non-ESG funds at the 10% level, reflecting the
larger importance placed on climate change themes in Section 4.1.4. Interestingly, some ESG
named funds indicated that they did not calculate an emissions intensity metric, citing that “It
is not a specific objective of the fund” or “We do not currently have any clients who have
requested this information. However, we could calculate this if a client requested.” Across the
regions, respondents with Australasian headquarters lagged behind other funds in our sample
with respect to reporting this information.
In Panel B, most respondents did not calculate Scope 1, 2 and 3 portfolio carbon intensity, with
more than half citing data reliability and coverage issues with Scope 3 emissions. The nature
of Scope 3 emissions, which result from the activities that are not controlled by the reporting
corporate, can lead to double-counting, as two or more organisations may account for the same
emissions. A significantly higher proportion of ESG named funds (68%) explained why they
did not calculate this intensity measure at the 1% level, relative to non-ESG named funds (54%).
4.1.6 Further Survey Results
In Appendix 3, we also investigated the types of ESG data used by respondents, highlighting
that fund managers more commonly incorporate analysis at the individual firm level rather than
at the aggregated sector or country level. Appendix 4 details the types of active engagement
approaches implemented, revealing that fund managers tended to prefer private interactions
with firms first and only took public actions once private interventions fail. Finally, Appendix
5 highlights the types of ESG voting approaches used, finding that most respondents used proxy
voting.

4.2 Determinants of Respondent Carbon Performance
Table 6 presents the fund summary statistics relating to the 33 survey respondents with
available portfolio holdings data. The mean portfolio carbon intensity was 103.52 tonnes of
CO2 equivalent per million USD of revenue, and there was significant variability across
responding funds. Interestingly, the fund with the largest portfolio carbon intensity value had
a climate-related name. For the responding funds, the mean weighted-average ESG score using
Refinitiv data was 66.36, reflecting the ‘Third Quantile’ or an above-average ESG performance,
6.19 using MSCI data, reflecting the ‘Average’ category, and 21.51 using Sustainalytics data,

24

in the ‘Medium’ category of ESG risk. The mean age of the responding funds was
approximately 8 years old.
[INSERT TABLE 6 HERE]
We compare the stated and actual Scope 1 and 2 portfolio carbon intensity of respondents in
Figure 2. On average, the (few) funds who provided this metric in the survey are underreporting
their portfolio’s exposure to carbon-intensive firms. This is attributable to the severe
underreporting by a select group of respondents. The divergence between stated and actual
carbon performance is further emphasised in Table 7. After establishing the relative weightings
of various ESG subthemes in Section 4.4, we rank each fund from one to thirty-three based on
the level of importance they place on climate change during the investment process (from
highest to lowest). Alongside this, we rank each fund from one to thirty-three based on the size
of their actual portfolio carbon intensity (from lowest to highest). In Table 7, we highlight that
some of the funds that place the highest importance on climate change themes have a relatively
poor ranking concerning their carbon performance. Notably, the fund that places the greatest
importance on climate change only has the 22nd lowest portfolio carbon intensity (out of the 33
responding funds), indicating that their portfolio is relatively more exposed to carbon-intensive
firms than most other respondents. Across the 33 responding funds, stated and actual rankings
have a correlation of –0.056, which highlights a negligible relationship between words and
actions. These findings are likely driven by one of two possible reasons: On one hand, some
respondents may be overstating their commitment to climate change to attract sustainable
investment fund flows, consistent with greenwashing. On the other hand, funds that prioritise
climate change themes may be actively engaging with their portfolio companies, which have
higher emission intensities, to improve their carbon efficiency. Whether these funds engage
actively is left for further research.
[INSERT TABLE 7 HERE]
Table 8 presents the results for the multilinear regression models in Equations (4) and (5), as
defined in Section 3.3. The results in Table 8 highlight several interesting relationships. Firstly,
respondents that actively engage with portfolio companies as a RI approach have a higher
portfolio carbon intensity, at the 5% level of significance, relative to those that do not. In the
investment process, fund managers face a choice of engaging with or divesting away from

25

environmentally damaging companies (Atta-Darkua et al., 2020). Intuitively, those that choose
to engage wish to instil a positive environmental change into companies that are currently
unsustainable.
Surprisingly, Table 8 highlights that portfolio emission intensity does not significantly differ
for funds that place a high level of importance on climate change themes (Climate Change
Weighting), and for funds that undertake a portfolio decarbonisation strategy
(Decarbonisation). These results appear to be counterintuitive, as one would expect that
managers that truly value climate change, and those seeking to reduce their carbon footprint,
will have relatively lower exposure to carbon-intensive companies. Another key finding is that
the portfolio carbon intensity is higher when respondents are a member of a climate-related
initiative (Climate Initiative), and this is significantly worse, at the 10% level, across the
various regressions. This may indicate that institutions are joining climate initiatives to appear
more sustainable without truly embracing sustainable investing approaches. Finally, portfolio
carbon intensity is higher for those respondents that believe that ESG risks, although not yet
priced, will soon impact investment performance at the 5% level of significance (Future
Performance). Again, this is surprising given that carbon intensity is an objective dimension
of ESG risk (De Spiegeleer et al., 2021). The above results are robust when employing financial
year 2019 revenue data, to avoid the COVID impact on revenues.
[INSERT TABLE 8 HERE]
A recent survey of institutional investors by Krueger et al. (2020) found that larger, longer-
term, and ESG-focused investors consider engagement and risk management to be a better
method for addressing climate-related risks, relative to divestment. Hence, active engagement
strategies may potentially explain some of the unintuitive findings above. To test whether the
results above are driven by engagement, we conduct further analysis. Firstly, we rerun our
regressions controlling for active engagement as a RI strategy, where 𝐸𝑛𝑔𝑎𝑔𝑒𝑚𝑒𝑛𝑡1. To
extend this analysis, we also create a new control variable 𝐴𝑐𝑡𝑖𝑣𝑖𝑠𝑡 𝐹𝑢𝑛𝑑, which looks
beyond voting or holding private discussions with management, which most funds do (see
Appendix 4). Specifically, 𝐴𝑐𝑡𝑖𝑣𝑖𝑠𝑡 𝐹𝑢𝑛𝑑1 if fund 𝑗 indicated that they take legal action
against management on ESG issues, submit shareholder proposals on ESG issues, or publicly
criticise management on ESG issues, and 0 otherwise. We define activist funds as those who
are demanding and proactively engage with management to create change. As with engagement,
we also control for activist funds. Table 9 presents the results for these analyses.

26

[INSERT TABLE 9 HERE]
In Table 9, it is evident that our earlier results hold, even after controlling for actively engaging,
and activist funds. Specifically, the portfolio carbon intensity does not significantly differ for
funds that place a high level of importance on climate change themes, and for those that have
a portfolio decarbonisation strategy. Hence, these results offer some evidence that funds are
overstating their RI strategies, again hinting at greenwashing. Table 9 also highlights that
portfolio carbon intensity is higher for the funds who are members/signatories of a climate
initiative at the 5% level of significance. Notably, funds that practice engagement have higher
emissions intensity, but activist funds do not. Given that we have controlled for engagement
and activist funds, this result is consistent with greenwashing. Specifically, funds are signalling
their commitment to global emission objectives to capitalise on the growing sustainable
investment market, without ‘walking the talk’. Given that retail investors are more likely to fall
for the obfuscation associated with greenwashing (Carlin & Manso, 2011; deHaan et al., 2021),
this information failure is an example of adverse selection, which may negatively impact
investor welfare (Akerlof, 1970). The results in Table 9 are also robust when we incorporate
our initial control variables (headquarters, style, ESG name, and size), alongside portfolio
carbon intensity coverage.
4.3 Determinants of Carbon and ESG Performance
Next, we turned our attention to the carbon and ESG performance of global equity funds in our
full sample of respondents and non-respondents. Table 10 presents the fund summary statistics
relating to the 33 respondents and 45 non-respondents. Across the entire sample, the mean
portfolio carbon intensity was 109.30 tonnes of CO2 equivalent per million USD of revenue,
with a significant spread observed between funds. The mean weighted-average ESG score
using Refinitiv data was 66.84, 6.22 using MSCI data, and 21.54 using Sustainalytics data.
These scores were very similar to the respondent-only sample (refer to Section 5.1). Of the 78
funds, there were 23 ESG named funds, or 29% of the sample. The mean fund age was
approximately 18 years old, which more than doubled that of the respondents-only sample. For
the subsequent regression analyses, we required that at least 70% of a fund’s gross weight came
from securities with carbon intensity or ESG rating data available.
[INSERT TABLE 10 HERE]

27

Table 11 investigates the common fund-level characteristics that drive the carbon performance
of global equity funds in our full sample, presenting the results to the multi-linear model in
Equation (6) (see Section 3.4). Accounting for our sample size, we utilised two variations of
the OLS regression model to test the effect of all explanatory variables of interest on portfolio
carbon intensity.
[INSERT TABLE 11 HERE]
The results in Table 11 highlight that survey respondents did not have a significantly different
portfolio carbon intensity relative to non-respondents. This is surprising, given our prior
expectation that the respondents would be biased towards institutions that have a high
awareness of ESG issues and risks, such as exposure to carbon-intensive companies. Alongside
this, we noted that our survey respondents had a greater preference for addressing
environmental themes within the investment process, on average, relative to previous surveys
(van Duuren et al., 2016). Another interesting result from Table 11 is that portfolio carbon
intensity was not significantly different for ESG named funds, relative to non-ESG named
funds. This is surprising as the ESG named funds in our survey were more climate-focused, on
average, relative to non-ESG labelled funds who placed a relatively high level of importance
on governance and corporate behaviour themes.
In Table 11, it is evident that value style funds had a significantly higher portfolio carbon
intensity relative to blend style funds at the 5% level. Value investing is a style that targets
companies that trade at a significant discount to their intrinsic value and such investors tend to
focus on the fundamental aspects of a company, including price-to-earnings multiples and free
cash flow that can be used to pay out dividends. Hence, a potential explanation for the result
above is that value-style funds were accumulating carbon-inefficient firms that are temporarily
undervalued due to divestment. This interpretation is consistent with Heinkel et al. (2001), who
argued that divestments place downward pressure on stock valuations, as higher discount rates
are applied to future cash flows. Alongside this, firms within the oil, gas, and coal sectors have
historically high yields relative to other equities, and their profitability is secure due to their
ownership of extraction rights (Bullard, 2014). Consequently, such companies are likely to
appeal to value investors when their prices are cheap, especially in the short investment horizon
these managers operate within.
In this paper, our survey highlighted that RI was largely driven by performance-based
expectations and stakeholder demand rather than by ethical motivations, particularly for

28

institutions based in the United States. Consequently, we assessed the portfolio carbon intensity
by the region of headquarters. While the results in Table 11 indicate that asset managers with
headquarters in the U.S. and other regions had higher portfolio carbon intensities, relative to
those with Australasian headquarters, the results are not consistently significant across both
variations of our model. Hence, there is insufficient evidence to conclude that portfolio carbon
intensity is significantly different across the regions.
As a robustness check, we recalculated carbon intensity for each security using FY19 revenues
(maintaining FY20 emissions) to account for the potential sales impact of the COVID-19
Pandemic. These results are presented in Appendix 7 and are consistent with the results above.
The final part of this analysis explores the determinants of ESG performance for our entire
sample of global equity funds. Table 12 presents the results of the multi-linear regression model
in Equation (7) (see Section 3.4). We investigated two variations of this model due to our
sample size while testing all explanatory variables of interest on the portfolio-level ESG scores.
In our analysis, we compared these models using portfolio ESG scores calculated from three
different rating providers: Refinitiv, MSCI, and Sustainalytics. This attempts to overcome the
noise attributable to each individual provider, given that they measure and aggregate ESG
issues differently (Chatterji et al., 2016; Berg et al., 2019).
[INSERT TABLE 12 HERE]
Table 12 highlights the divergence in results when using different ESG providers. Utilising
Refinitiv ratings, ESG named funds within our sample did not have significantly different
portfolio ESG scores to non-ESG named funds. However, with MSCI ratings, ESG named
funds had relatively higher portfolio ESG performance, and this is significant at the 5% or
lower levels across the model variations. The result is similar for Sustainalytics data, which is
significant at the 10% or lower levels. Globally, it is estimated that 60% of all retail investment
into ESG funds has flowed into those that are built on MSCI’s ratings (Simpson et al., 2021).
This explains the significant and positive relationship between ESG named funds and MSCI
portfolio scores. The MSCI methodology has been criticised given that “the ratings don’t
measure a company’s impact on the Earth and society. In fact, they gauge the opposite: the
potential impact of the world on the company and its shareholders” (Simpson et al., 2021, p.g.
1). Consequently, if regulations aimed at alleviating climate change do not threaten a
company’s bottom line, emissions are deemed irrelevant by MSCI within their framework

29

(Simpson et al., 2021), which may explain the surprising carbon intensity determinant results
earlier.
Further inconsistencies between the different ESG providers can be observed in Table 12. For
example, growth style funds had significantly lower portfolio ESG performance relative to
blend style funds at the 1% level when using Refinitiv ratings. Growth investors are those who
focus on capital appreciation, often looking for smaller companies with long term prospects.
Such companies often lack coverage by ESG providers or are disadvantaged by the ESG rating
size bias (Drempetic et al., 2019). Contrastingly, growth style funds had better portfolio ESG
performance at the 5% level of significance, when using Sustainalytics data. A potential
explanation is that growth funds are pursuing rapidly advancing industries that offer sustainable
alternatives, such as green technologies. However, the differences observed between the
Sustainalytics and Refinitiv scores are likely being driven by the noise associated with the
individual rating methodologies. In Table 12, there is some evidence that asset managers with
headquarters based in the U.S. had significantly higher portfolio Refinitiv and MSCI ESG
scores, relative to those based in Australasia. Furthermore, those with headquarters in Europe
and other regions had significantly higher Refinitiv ESG scores, relative to those based in
Australasia. This may suggest that asset managers based in Australasia were trailing behind
their international peers with respect to their externally measured ESG performance. However,
this is not robust across all three rating providers and the Australasian funds had lower carbon
intensities.
Finally, the results in Table 12 indicate that responding funds did not have a significantly
different portfolio ESG performance relative to non-respondents, across all three ESG
providers. As with portfolio carbon intensity, this result is surprising as we anticipated that
respondents would be biased towards ESG ‘adopters’ with awareness of ESG issues and risks,
relative to ESG ‘sceptics’. Across the three ESG providers, the coverage of ESG scores within
a portfolio is significantly and positively related to the portfolio-level score, despite our
normalisation of asset weightings based on the available data. Missing data presents a key
challenge when evaluating portfolio ESG performance, and this is an issue that could receive
further attention within the literature.
In our analysis, we also investigated the non-normalised value-weighted portfolio ESG scores
as a robustness check in Appendix 8. While producing similar results, we found that ESG

30

named funds did not have significantly different portfolio ESG performance, relative to non-
ESG named funds, when using Sustainalytics ratings.
5. Conclusion
Through our survey of retail global equity funds offered in Australasia, we elicited an
understanding of how and why asset managers integrate ESG information into the investment
process. We provided insights into the underlying motivations behind RI, the types of
information and investment approaches used, and finally, the relative importance placed by
fund managers on various ESG issues. Next, we evaluated the stated and actual carbon
performance of respondents, before assessing the determinants of carbon and ESG performance.
Our survey highlighted that RI was largely driven by expected value (performance-based
expectations and client demand) rather than values (having an ethical responsibility to make a
positive difference), particularly for institutions that are based in the United States. We
documented that the top three RI approaches were incorporating ESG considerations into
fundamental analysis, negative/exclusionary screening, and active engagement, consistent with
previous literature (Amel-Zadeh & Serafeim, 2018) and industry reports (RIAA, 2021a; RIAA;
2021b). We found that funds with U.S. headquarters tended to place a relatively greater level
of importance on environmental themes within the investment process, which have more
objective data, and stronger links to financial performance (Credit Suisse, 2021; Climate
Action 100+, 2020; Trinks et al., 2020; TCFD, 2017). In contrast, funds with Australasian
headquarters prioritised environmental, social, and governance themes more equally. Across
all ESG subthemes, fund managers tended to place the highest importance on climate change,
followed by corporate behaviour. This was emphasised for ESG named funds, which share a
strong climate focus. Among institutions, RI strategies are now being widely adopted into
investment decisions. However, the magnitude of these changes is not yet clear, particularly
when many managers are not even measuring portfolio carbon intensity.
We documented a divergence between the stated and actual carbon performance of respondents,
highlighting that fund managers were overstating their commitment to global emission
objectives, without ‘walking the talk’. We found that portfolio carbon intensity was
significantly higher for respondents that were members of a climate initiative, and not
significantly different for those that prioritised climate change themes or engaged in a
decarbonisation strategy. This finding does not appear to be driven by actively engaging or
activist funds, rather, it is consistent with greenwashing funds seeking to attract responsible

31

investment flows. This is concerning for retail investors, who may not be able to distinguish
between greenwashing funds (‘lemons’) and genuine responsible funds.
Finally, we found no significant differences in the carbon and ESG performance of respondents
and non-respondents, which indicates that our survey is robust to self-selection bias. We also
found that ESG named funds only obtain significantly better portfolio ESG scores, relative to
non-ESG named funds, using ratings from MSCI and Sustainalytics, but not Refinitiv. We
contributed to the discussion surrounding the divergence of external ESG ratings, highlighting
that some of the inconsistencies found in our results are likely attributable to the different scope,
measurement, and weighting methodologies of the providers, as shown by Berg et al. (2019).
Given that we evaluated each fund’s portfolio at one point in time, it would be interesting to
explore the evolution of portfolio carbon intensities and ESG scores over time in further
research. Alongside this, additional analysis could compare the divergence of each fund’s ESG
performance from its stated benchmark, alongside a general global equity benchmark given
that our sample features both ESG focussed funds and conventional funds.

32

References
Akerlof, G. A. (1970). Quality uncertainty and the market mechanism. Quarterly Journal of
Economics, 84(3), 488-500.
Albuquerque, R., Yrjö K. & Chendi Z. (2019). Corporate social
responsibility and firm risk: Theory and empirical evidence. Management Science,
65(10), 4451–4469.
Alda, M. (2020). ESG fund scores in UK SRI and conventional pension funds: Are the ESG
concerns of the SRI niche affecting the conventional mainstream? Finance Research
Letters, 36, 101313.
Amel-Zadeh, A., & Serafeim, G. (2018). Why and how investors use ESG information: Evidence
from a global survey. Financial Analysts Journal, 74(3), 87-103.
Atta-Darkua, V., Chambers, D., Dimson, E., Ran, Z., & Yu, T. (2020). Strategies for responsible
investing: Emerging academic evidence. Journal of Portfolio Management, 46(3), 26-35.
Baldini, M., Maso, L. D., Liberatore, G., Mazzi, F., & Terzani, S. (2018). Role of country- and
firm-level determinants in environmental, social, and governance disclosure. Journal of
Business Ethics, 150(1), 79–98.
Barr et al. (2021). Morningstar Sustainability Ratings Methodology. Morningstar.
morningstar.com/content/dam/marketing/shared/research/methodology/744156_Morningst
ar_Sustainability_Rating_for_Funds_Methodology.pdf.
Bauer, R., Ruof, T., & Smeets, P. (2021). Get real! Individuals prefer more sustainable
investments. Review of Financial Studies, 34(8), 3976-4043
Berg, F., Koelbel, J. F., & Rigobon, R. (2019). Aggregate confusion: The divergence of ESG
ratings. MIT Sloan School of Management Working Paper 5822-19.
Berg, F., Koelbel, J. F., Pavlova, A., & Rigobon, R. (2021). ESG confusion and stock returns:
Tackling the problem of noise. SSRN Working Paper.
Boubaker, S., Cellier, A., Manita, R., & Saeed, A. (2020). Does corporate social responsibility
reduce financial distress risk? Economic Modelling, 91, 835-851.

33

Brandon, R. G., Glossner, S., Krueger, P., Matos, P., & Steffen, T. (2021). Do responsible
investors invest responsibly? Swiss Finance Institute Research Paper No. 20-13, European
Corporate Governance Institute – Finance Working Paper 712/2020.
Bullard, N. (2014). Fossil fuel divestment: a $5 trillion challenge. Bloomberg New Energy
Finance. https://data.bloomberglp.com/bnef/sites/4/2014/08/BNEF_DOC_2014-08-25-
Fossil-Fuel-Divestment.pdf.
Campbell, J. L. (2007). Why would corporations behave in socially responsible ways? An
institutional theory of corporate social responsibility. Academy of Management
Review, 32(3), 946-967.
Cappucci, M. (2018). The ESG integration paradox. Journal of Applied Corporate Finance, 30(2),
22-28.
Chandellier, J., & Malacain, M. (2021). Biodiversity and Re/insurance: An Ecosystem at Risk.
HAL Working Paper.
Chatterji, A. K., Durand, R., Levine, D. I., & Touboul, S. (2016). Do ratings of firms converge?
Implications for managers, investors, and strategy researchers. Strategic Management
Journal, 37(8), 1597-1614.
Chava, S. (2014). Environmental externalities and cost of capital. Management Science, 60(9),
2223-2247.
Chenet, H. (2021). Climate change and financial risk. In C. Zopounidis, R. Benkraiem, & I.
Kalaitzoglou (Ed.), Financial Risk Management and Modeling (pp. 393-419). Springer.
Cheng, B., Ioannou, I., & Serafeim, G. (2014). Corporate social responsibility and access to
finance. Strategic Management Journal, 35(1), 1-23.
Chin, K. (2021). SEC Review Highlights Potentially Misleading ESG Practices Among Funds.
Wall Street Journal. Retrieved December 1, 2021, from https://www.wsj.com/articles/sec-
review-highlights-potentially-misleading-esg-practices-among-funds-11618019507.
Climate Action 100+. (2020). 2020 Progress Report. Retrieved January 25, 2022, from
https://www.climateaction100.org/wp-content/uploads/2020/12/CA100-Progress-
Report.pdf

34

Credit Suisse. (2021). The decarbonizing portfolio: A sustainable investment strategy for a low
carbon future. https://www.credit suisse.com/media/assets/corporate/docs/publications
/white-papers/whitepaper-the-decarbonizing-portfolio-en.pdf.
Delmas, M. A., & Burbano, V. C. (2011). The drivers of greenwashing. California Management
Review, 54(1), 64-87.
De Spiegeleer, J., Höcht, S., Jakubowski, D., Reyners, S., & Schoutens, W. (2021). ESG: a new
dimension in portfolio allocation. Journal of Sustainable Finance & Investment, 1-41.
Dichev, I. D., Graham, J. R., Harvey, C. R., & Rajgopal, S. (2013). Earnings quality: Evidence
from the field. Journal of Accounting and Economics, 56(2-3), 1-33
Drempetic, S., Klein, C., & Zwergel, B. (2020). The influence of firm size on the ESG score:
Corporate sustainability ratings under review. Journal of Business Ethics, 167(2), 333-360.
Dyck, A., Lins, K. V., Roth, L., & Wagner, H. F. (2019). Do institutional investors drive corporate
social responsibility? International evidence. Journal of Financial Economics, 131(3), 693-
714.
Eccles, R. G., Kastrapeli, M. D., & Potter, S. J. (2017). How to integrate ESG into investment
decision‐making: Results of a global Survey of institutional investors. Journal of Applied
Corporate Finance, 29(4), 125-133.
El Ghoul, S., Guedhami, O., Kwok, C. C., & Mishra, D. R. (2011). Does corporate social
responsibility affect the cost of capital? Journal of Banking & Finance, 35(9), 2388-2406.
FMA. (2020). Disclosure framework for integrated financial products.
https://www.fma.govt.nz/assets/Guidance/Disclosure-framework-for-integrated-financial-
products.pdf
Garvey, G. T., Iyer, M., & Nash, J. (2018). Carbon footprint and productivity: does the “E” in
ESG capture efficiency as well as environment. Journal of Investment Management, 16(1),
59-69.
Garz, H., Volk, C., & Morrow, D. (2018). The ESG risk ratings: Moving up the innovation curve.
Sustainalytics White Paper.
GSIA. (2020). Global Sustainable Investment Review 2020. http://www.gsi-alliance.org/wp-
content/uploads/2021/08/GSIR-20201.pdf.

35

Graham, J. R., Harvey, C. R., & Rajgopal, S. (2005). The economic implications of corporate
financial reporting. Journal of Accounting and Economics, 40(1-3), 3-73.
Heinkel, R., Kraus, A., & Zechner, J. (2001). The effect of green investment on corporate
behaviour. Journal of Financial and Quantitative Analysis, 36(4), 431-449.
Jiraporn, P., Jiraporn, N., Boeprasert, A., & Chang, K. (2014). Does corporate social responsibility
(CSR) improve credit ratings? Evidence from geographic identification. Financial
Management, 43(3), 505-531.
Kennaway, G. (2021). Australian sustainable investing fund landscape - Q3 2021.Morningstar.
Retrieved December 5, 2021, from https://www.morningstar.com.au/funds/article/
australian-sustainable-investing-fund/216617.
Kim, S., & Yoon, A. (2020). Analyzing active managers' commitment to ESG: Evidence from
United Nations Principles for Responsible Investment. Management Science (Forthcoming).
Krueger, P., Sautner, Z., & Starks, L. T. (2020). The importance of climate risks for institutional
investors. Review of Financial Studies, 33(3), 1067–1111.
Liang, H., Sun, L., & Teo, M. (2020). Greenwashing: Evidence from hedge funds. Research
Collection Lee Kong Chian School of Business Working Paper.
Lins, K. V., Servaes, H., & Tamayo, A. (2017). Social capital, trust, and firm performance: The
value of corporate social responsibility during the financial crisis. Journal of Finance, 72(4),
1785-1824.
McCahery, J. A., Sautner, Z., & Starks, L. T. (2016). Behind the scenes: The corporate governance
preferences of institutional investors. Journal of Finance, 71(6), 2905-2932.
Ministry for the Environment. (2021). Mandatory climate-related disclosures. Ministry for the
Environment. Retrieved January 14, 2022, from https://environment.govt.nz/what-
government-is-doing/areas-of-work/climate-change/mandatory-climate-related-financial-
disclosures/
Morningstar. (2021). Global Sustainable Fund Flows: Q3 2021 in Review.
https://www.morningstar.com/content/dam/marketing/shared/pdfs/Research/Global-ESG-
Q3-2021Flows.pdf?utm_source=eloqua&utm_medium=email&utm_campaign=none
&utm_c ontent=27223.

36

MSCI. (2020). MSCI ESG Ratings Methodology. MSCI ESG Research.
https://www.msci.com/documents/1296102/21901542/MSCI+ESG+Ratings+Methodolog
y+-+Exec+Summary+Nov+2020.pdf.
MSCI. (2021). MSCI ESG Fund Ratings Methodology. MSCI ESG Research.
https://www.msci.com/documents/1296102/15388113/MSCI+ESG+Fund+Ratings+Exec+
Summary+Methodology.pdf.
Mühlbacher, H., & Botschen, G. (1988). The use of trade-off analysis for the design of holiday
travel packages. Journal of Business Research, 17(2), 117-131.
Picou, A., & Rubach, M. J. (2006). Does good governance matter to institutional investors?
Evidence from the enactment of corporate governance guidelines. Journal of Business
Ethics, 65(1), 55-67.
Raghunandan, A., & Rajgopal, S. (2021). Do ESG funds make stakeholder-friendly investments?
Columbia Business School Working Paper.
Refinitiv. (2021). Environment, Social, and Governance Scores from Refinitiv.
https://www.refinitiv.com/content/dam/marketing/en_us/documents/methodology/refinitiv
-esg-scores-methodology.pdf.
Revelli, C. (2017). Socially responsible investing (SRI): From mainstream to margin? Research
in International Business and Finance, 39, 711-717.
RIAA. (2021a). Responsible Investment Benchmark Report Australia 2021.
https://responsibleinvestment.org/wp-content/uploads/2021/09/Responsible-Investment-
Benchmark-Report-Australia-2021.pdf.
RIAA. (2021b). Responsible Investment Benchmark Report Aotearoa New Zealand 2021.
https://responsibleinvestment.org/wp-content/uploads/2021/09/Responsible-Investment-
Benchmark-Report-Aotearoa-New-Zealand-2021.pdf.
Sandberg, J., Juravle, C., Hedesström, T. M., & Hamilton, I. (2009). The heterogeneity of socially
responsible investment. Journal of Business Ethics, 87(4), 519-533.
Schwartzkopff, F. (2021). Fund Managers Face Delays as EU Greenwash Rules Hit Hurdles.
Retrieved December 1, 2021, from https://www.bloomberg.com/news/articles/2021-11-
22/fund-managers-face-more-delays-as-eu-greenwash-rules-hit-hurdles.

37

Simpson, C., Rathi, A., & Kishan, S. (2021). The ESG Mirage. Retrieved January 13, 2021, from
https://www.bloomberg.com/graphics/2021-what-is-esg-investing-msci-ratings-focus-on-
corporate-bottom-line/
Singh, J. V., Tucker, D. J., & House, R. J. (1986). Organizational legitimacy and the liability of
newness. Administrative Science Quarterly, 171-193.
Statman, M. (2020). ESG as waving banners and as pulling plows. The Journal of Portfolio
Management, 46(3), 16-25.
Tamayo-Torres, I., Gutierrez-Gutierrez, L., & Ruiz-Moreno, A. (2019). Boosting sustainability
and financial performance: the role of supply chain controversies. International Journal of
Production Research, 57(11), 3719-3734.
TCFD. (2017). Recommendations of the Task Force on Climate-Related Financial Disclosures.
Retrieved January 25, 2022, from https://www.fsb-tcfd.org.
Trinks, A., Mulder, M., & Scholtens, B. (2020). An efficiency perspective on carbon emissions
and financial performance. Ecological Economics, 175, 106632
van Duuren, E., Plantinga, A., & Scholtens, B. (2016). ESG integration and the investment
management process: Fundamental investing reinvented. Journal of Business Ethics, 138(3),
525-533.
Waygood, S., Hilton, P., McQuillen, M. J., Joly, C., & Knight, E. (2009). Fiduciary responsibility:
Legal and practical aspects of integrating environmental, social and governance issues into
institutional investment. United Nations Environment Programme Finance Initiative.

38

Figures
Figure 1: Survey Flow Diagram
Figure 1 presents the order flow of our survey questions. Arrows and boxes represented by dotted lines indicate
the questions that were displayed to respondents conditional on their previous response.
Engagement Selected
Institution Demographic Questions
Why do you consider ESG information when
making investment decisions?
Adapted from Amel-Zadeh & Serafeim (2018)
Fund Demographic Questions
Which type of ESG data do you use?
Adapted from van Duuren et al. (2016)
What is Your Flagship Retail ESG Global
Equity Fund?
What measures of direct engagement over ESG
issues have you taken in the past five years?
Adapted from Krueger et al. (2018)
Which ESG investment approaches do you
incorporate within this fund?
Adapted from Amel-Zadeh & Serafeim (2018)
How do you undertake voting on ESG issues?
Voting Selected
Weighting of ESG Themes & Subthemes
What is Your Fund’s Carbon Intensity?
Institution-Level
Fund Level

39

Figure2:Statedvs.ActualPortfolioCarbonIntensity
Figure 2 illustrates the percentage difference between stated and calculated Scope 1 and 2 portfolio carbon intensity of respondents. Of the thirty-three responding funds with
available portfolio holdings data, only twelve provided their Scope 1 and 2 portfolio carbon intensity in the survey. The percentage difference is calculated as (Stated –
Actual)/Actual. A positive percentage difference indicates overreporting, while a negative percentage difference indicates underreporting.
‐80%
‐60%
‐40%
‐20%
0%
20%
40%
Fund 1 Fund 2 Fund 3 Fund 4 Fund 5 Fund 6 Fund 7 Fund 8 Fund 9 Fund 10 Fund 11 Fund 12
Difference Between Stated and Actual Portfolio
Carbon Intensity (%)
UNDERREPORTED
OVERREPORTED

40

Tables
Table 1: Summary Statistics
Institution-Level Fund-Level
N
%
N
%
Respondent Job Title/Position Fund Size (in US $ million)
ESG/Responsible Investment Specialist 9 20% < 100 11 26%
Investment Analyst/Strategist 7 16% 100 - 500 12 28%
Fund/Portfolio Manage
r
6 14%
500 - 1000 8 19%
Executive/Managing Directo
r
5 11%
1000 - 5000 9 21%
Chief Executive Office
r
2 5%
> 5000 3 7%
Chief Investment Office
r
3 7%
Othe
r
12 27%
Fund Type
Actively Manage
d
39 91%
Location of Head
q
uarters Passivel
y
Mana
g
e
d
4 9%
Australia 16 36%
N
ew Zealand 12 27% Lead Fund Manager Gender
United States 10 23% Male 41 95%
United Kingdom 2 5% Female 1 2%
Othe
r
4 9%
Other 1 2%
ESG Considerations in Investment Decisions ESG Named Fund
Across All Funds 35 80%
N
o 24 56%
Across Some Funds 9 20% Yes 19 44%
N
one 0 0%
Average Fund Holding Period
Attainment of ESG Capabilities Short (less than 6 months) 0 0%
Internal Training 42 95% Medium (6 months to 2 years) 9 21%
Hiring ESG Experts 21 48% Long (2 years to 5 years) 24 56%
Hiring Investment Experts with Some
ESG Training 17 39% Very Long (more than 5 years) 9 21%
External Training 15 34%
Industry and Academic Groups 14 32% Fund Investment Styles
Othe
r
9 20%
Quality 20 47%
Growth 16 37%
Signatories/Memberships Value/Fundamental 15 35%
Principle for Responsible Investment 40 91% Specific Theme 12 28%
Climate-Related Initiatives 26 59% Concentrated (less than 50 holdings) 10 23%
Responsible Investment Association of
Australasia 19 43% Broad Market 8 19%
Othe
r
19 43%
Factor/Quantitative 7 16%
Momentu
m
2 5%
Othe
r
8 19%
Notes: Table 1 presents the institution-level summary statistics on the left-hand side, and the fund-level summary statistics on the right-hand
side, as provided by respondents in the survey.

41

Table 2: Responsible Investing Motivations (n=44)
Re
g
ion of Head
q
uarters
(1) (2) (3) (4) (5)
All Australasia U.S. Other Range
ESG information is material to investment performance 81% 79% 89% 83% 10%
Growing client/stakeholder demand 74% 71% 78% 83% 12%
ESG risk and opportunities, although not yet priced, will soon
affect investment performance 67% 64% 67% 83% 19%
We believe this will encourage positive change in individual
firm ESG practices 65% 71% 33% 83% 50%
We see it as an ethical responsibility 51% 64% 11% 50% 53%
It is
p
art of our mandated investment strate
gy
/SIPO 40% 43% 22% 50% 28%
Othe
r
9% 4% 22% 17% 19%
Notes: Table 2 presents the percentage of survey responses to the question “Why does your institution consider ESG information when making
investment decisions?”, where respondents could choose one or more alternatives that represented their institution. Column (1) reports the
percentage of respondents that selected the response for a given row. Columns (2), (3), and (4) report the percentage of respondents with
headquarters in Australasia, the United States, and Other regions, who selected the response for a given row. Across the region of headquarters,
the range (high minus low) of percentages are reported in Column (5). The responses in Table 2 have been ordered from highest to lowest
based on the proportion of respondents that chose each reason in Column (1).

42

Table 3: Responsible Investing Approaches (n=43)
Notes: Table 3 presents the percentage of survey responses to the question “Which ESG investment approaches do you incorporate within this fund?”, where multiple responses were allowed. Column (1) reports the
percentage of respondents that selected the response for a given row. Columns (2) and (3) report the percentages for funds greater than 50% of the median fund size (‘>50%’) and less than or equal to 50% of the median
fund size (50%) respectively. Column (4) reports the difference between Column (2) and (3), and the results of a test of the null hypothesis that the two percentages are equal. Columns (5) and (6) report the percentages
for ESG named funds (‘Yes’) and non-ESG named funds (‘No’) respectively. Column (7) reports the difference between Column (5) and (6), alongside the results of a test of the null hypothesis that the two percentages
are equal. Columns (8), (9), and (10) report the percentage of respondents with headquarters in Australasia, the United States, and other regions, who selected the response for a given row. Across the region of headquarters,
the range (high minus low) of percentages are reported in Column (11). Finally, ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively. The responses in Table 3 have been ordered from
highest to lowest based on the proportion of respondents that chose each reason in Column (1).
All Fund Size ESG Named Fund Region of Headquarters
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
>50% 50% Diff
(2) - (3) Yes No Diff
(5) - (6) Australasia U.S. Other Range
Fundamental analysis incorporating ESG considerations 84% 71% 96% -25%** 79% 88% -9% 85% 70% 100% 30%
N
egative screening 81% 81% 82% -1% 95% 71% 24%** 85% 80% 67% 19%
Engagement/active ownership with companies on ESG 79% 81% 77% 4% 79% 79% 0% 78% 80% 83% 5%
Positive (best in class) screening 37% 33% 41% -8% 58% 21% 37%** 37% 40% 33% 7%
Decarbonization of
p
ortfolio 28% 38% 18% 20%
37% 21% 16% 22% 30% 50% 28%
Thematic investment 21% 29% 14% 15% 32% 13% 19% 11% 40% 33% 29%
Overlay/Portfolio tilt 12% 14% 9% 5% 26% 0% 26% 4% 20% 33% 30%
Quantitative ESG factor investin
g
12% 14% 9% 5%
21% 4% 17% 7% 20% 17% 13%
Impact investing 5% 10% 0% 10% 11% 0% 11% 7% 0% 0% 7%
Othe
r
5% 5% 5% 0% 5% 4% 1% 0% 10% 17% 17%

43

Table 4: ESG Theme Priority (n=43)
All Fund Size ESG Named Fund Region of Headquarters
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Mean SD
>50% 50% Diff
Yes No Diff Australasia U.S. Other Range
Panel A: Themes
Environmen
t
[E] 38% 0.15 40% 37% 3%
44% 34% 10%** 34% 46% 44% 12%
Governance [G] 32% 0.10 32% 32% 0%
29% 35% -5%** 32% 26% 33% 7%
Social [S] 29% 0.08 28% 31% -3% 27% 31% -5% 32% 26% 24% 8%
Panel B: Sub-Themes
[E] Climate Chan
g
e 35% 0.22
37% 34% 2%
36% 35% 2% 33% 38% 41% 8%
[E] Environmental Opportunities 24% 0.10 23% 25% -2%
23% 24% -1% 24% 27% 17% 10%
[E] Pollu
t
ion & Waste 23% 0.12 21% 25% -4%
23% 23% -1% 26% 17% 20% 9%
[E]
N
atural Capital (Includin
g
Biodiversit
y
) 18% 0.11
20% 16% 3%
18% 18% 0% 16% 19% 23% 6%
[G] Corporate Behaviou
r
34% 0.18
30% 37% -7%** 31% 36% -6%** 39% 25% 28% 14%
[G] Shareholder Ri
g
hts 23% 0.08
23% 23% 1%
23% 23% 1% 22% 25% 23% 3%
[G] Remuneration 22% 0.09 24% 20% 4%
24% 20% 4% 20% 27% 23% 7%
[G] Board Composition 21% 0.07 23% 20% 3%* 22% 21% 1% 19% 24% 28% 8%
[S] Suppl
y
Chain & Communit
y
28% 0.15 24% 32% -8%
27% 29% -3% 30% 25% 27% 5%
[S] Heal
t
h & Safet
y
24% 0.07
24% 24% 0%
23% 24% -1% 24% 24% 23% 1%
[S] Product Liabilit
y
24% 0.14
27% 22% 6%
27% 22% 5% 24% 25% 23% 2%
[S] Human Capital Mana
g
emen
t
23% 0.10 25% 22% 2% 23% 24% -1% 22% 27% 27% 5%
Panel C: Wei
g
hted Sub-Themes
[E] Climate Chan
g
e 15% 0.18
16% 15% 1%
20% 12% 8% 11% 20% 24% 13%
[G] Corporate Behaviou
r
11% 0.07
10% 12% -3%* 9% 12% -3%** 13% 7% 9% 6%
[S] Suppl
y
Chain & Communit
y
9% 0.06 7% 10% -4%* 7% 10% -2% 10% 7% 6% 3%
[E] Environmental Opportunities 9% 0.05 9% 8% 1%
9% 8% 1% 8% 12% 5% 6%
[E] Pollution & Waste 8% 0.04 8% 8% -1%
8% 8% 0% 9% 6% 7% 3%
[G] Shareholder Ri
g
hts 7% 0.03
7% 7% 0%
7% 8% -1%
8% 7% 7% 1%
[S] Product Liabilit
y
7% 0.05
8% 7% 1%
7% 7% 0% 8% 7% 5% 2%
[G] Remuneration 7% 0.03 7% 7% 1%
7% 7% 0% 7% 7% 7% 0%
[S] Health & Safet
y
7% 0.03
7% 7% -1%
6% 8% -1%
8% 6% 5% 2%
[G] Board Composition 7% 0.03 7% 7% 1% 6% 7% -1%
7% 7% 9% 3%
[S] Human Capital Mana
g
emen
t
7% 0.03
7% 6% 0%
6% 7% -1%
7% 7% 6% 0%
[E]
N
atural Capital (Includin
g
Biodiversit
y
) 6% 0.04 7% 6% 2%* 7% 6% 1% 6% 8% 7%% 2%
Notes: Survey respondents were asked to allocate 100 points between various ESG themes and subthemes, based on the relative importance that the fund places on them within the investment process (See survey questions 30-33 in Appendix 2). Table 4 presents the relative
weightings of ESG themes and subthemes by the funds in our sample. Column (1) reports the mean percentage of respondents that selected the response for a given row, while Column (2) reports the standard deviation. Columns (3) and (4) report the mean percentages for funds
greater than 50% of the median fund size (‘>50%’) and less than or equal to 50% of the median fund size (50%) respectively. Column (5) reports the difference between Column (3) and (4), and the results of a non-parametric Mann-Whitney U Test of the null hypothesis that
the two distributions are equal (represented in brackets). Columns (6) and (7) report the mean percentages for ESG named funds (‘Yes’) and non-ESG named funds (‘No’) respectively. Column (8) reports the difference between Column (6) and (7), alongside the results of a non-
parametric Mann-Whitney U Test of the null hypothesis that the two distributions are equal (represented in brackets). Columns (9), (10), and (11) report the mean percentage of respondents with headquarters in Australasia, the United States, and other regions, who selected the
response for a given row. Across the region of headquarters, the range (high minus low) of percentages are reported in Column (12). Finally, ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively. Across the rows, Panel A reports the results for
ESG themes, Panel B for individual E, S, and G subthemes, and finally, Panel C reports the results for the weighted E, S, and G subthemes.

44

Table 5: Reported Portfolio Carbon Intensity (n=43)
All Fund Size ESG Named Fund Region of Headquarters
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
>50% 50% Diff Yes
N
o Diff
Australasia U.S. Othe
r
Range
Panel A: Sco
p
e 1 and 2 Emissions Intensit
y
Metric provided 49% 52% 46% 7%
63% 38% 26%* 37% 60% 83% 46%
A different intensity measure is used 9% 0% 18% -18%
5% 13% -7% 15% 0% 0% 15%
Metric not
p
rovided
E
xplained 23%
19% 27% -8%
16% 29% -13% 30% 20% 0% 30%
Unexplained 19% 29% 9% 20%
16% 21% -5% 19% 20% 17% 3%
Total 42% 48% 36% 11% 32% 50% -18% 48% 40% 17% 31%
Panel B: Scope 1, 2 and 3 Emissions Intensity
Metric
p
rovided 9%
5% 14% -9%
5% 13% -7% 7% 10% 17% 9%
A different intensity measure is used 9% 10% 9% 0%
16% 4% 12% 11% 0% 17% 17%
Metric not provided
E
x
p
lained 61%
62% 59% 3%
68% 54% 14%*** 63% 60% 50% 13%
Unexplained 21% 24% 18% 6%
11% 29% -19% 19% 30% 17% 13%
Total 81% 86% 77% 8% 79% 83% -19% 82% 90% 67% 26%
Notes: Table 5 presents the percentage of survey response to the question "What is the weighted-average emissions (tonnes of CO2e) intensity per sales ($USD) of the fund (as at Q2 2021)?”. Column (1)
reports the percentage of respondents that selected the response for a given row. Columns (2) and (3) report the percentages for funds greater than 50% of the median fund size (‘>50%’) and less than or
equal to 50% of the median fund size (50%) respectively. Column (4) reports the difference between Column (2) and (3), and the results of a test of the null hypothesis that the two percentages are equal.
Columns (5) and (6) report the percentages for ESG named funds (‘Yes’) and non-ESG named funds (‘No’) respectively. Column (7) reports the difference between Column (5) and (6), alongside the results
of a test of the null hypothesis that the two percentages are equal. Columns (8), (9), and (10) report the percentage of respondents with headquarters in Australasia, the United States, and other regions,
who selected the response for a given row. Across the region of headquarters, the range (high minus low) of percentages are reported in Column (11). Finally, ***, **, and * represent the significance at the
1%, 5%, and 10% levels respectively. Across the rows, Panel A reports the results relating to Weighted Average Portfolio Scope 1 and 2 Carbon Intensity, while Panel B reports the results relating to
Weighted Average Portfolio Scope 1, 2, and 3 Carbon Intensity.

45

Table 6: Respondent Fund Summary Statistics
Variables Obs Mean Std Dev Min Max Skew Kurt
Portfolio Weighted Average Scope 1 & 2
Carbon Intensity (tCO2e/US$M) 33 103.52 64.62 12.07 320.61 1.22 5.28
Carbon Intensity Coverage (%) 33 98.26 5.00 72.26 100.00 -4.54 23.72
Refinitiv ESG Score 33 66.36 6.26 52.13 76.80 -0.35 2.49
Refinitiv ESG Score Coverage (%) 33 91.26 5.26 69.76 100.00 -2.05 9.67
MSCI ESG Score 33 6.19 0.60 4.87 7.53 0.05 3.10
MSCI ESG Score Coverage (%) 33 93.42 7.32 60.60 100.00 -2.91 13.40
Sustainalytics ESG Score 33 21.61 2.41 14.67 26.11 0.44 3.75
Sustainalytics ESG Score Coverage (%) 33 93.42 7.32 60.60 100.00 -2.91 13.40
U.S. Headquarters Dummy 33 0.15 0.36 0.00 1.00 1.94 4.78
Other Region Headquarters Dummy 33 0.18 0.39 0.00 1.00 1.65 3.72
Value Style Dummy 33 0.09 0.29 0.00 1.00 2.85 9.10
Growth Style Dummy 33 0.42 0.50 0.00 1.00 0.31 1.09
ESG Name Dummy 33 0.39 0.50 0.00 1.00 0.43 1.19
Size (Ln($)) 33 18.56 2.33 12.03 22.22 -0.80 3.35
Age (Months) 33 101.36 85.74 6.00 328.00 0.97 3.17
Financial Performance (%) 31 38.71 13.84 7.06 75.02 -0.09 3.86
Volatility (%) 31 4.45 1.10 2.48 7.18 0.52 3.04
Management Fees (%) 33 0.90 0.31 0.20 1.50 -0.27 2.67
Notes: Table 6 presents the fund summary statistics relating to the 33 survey respondents with available portfolio holdings data. The fund
characteristics for survey respondents are presented at the holdings date of 30 June 2021 (or as close to this date as possible). These include
calculated measures of carbon performance and ESG performance, alongside various fund-level control variables that have been obtained
from Morningstar Direct.

46

Table 7: Actual vs. Stated Carbon Performance of Respondents (n=33)
Panel A: Stated vs. Actual Rankings
Stated Importance Placed
on Climate Change
Portfolio Scope 1 & 2
Carbon Intensity
(
tCO2e/US$M
)
(1) (2)
Fund A 1 22
Fund B 2 8
Fund C 3 24
Fund D 4 17
Fund E 5= 3
Fund G 5= 18
Panel B: Correlation Between Stated and Actual Rankin
g
s
(
Full Sam
p
le
)
Portfolio Scope 1 & 2 Carbon Intensity (tCO2e/US$M)
Rankin
Stated Importance Placed on Climate Change
Ranking –0.056
Notes: In Panel A of Table 7, Column (1) presents the top five respondents that place the highest relative importance (weighted) on Climate
Change themes among the thirty-three responding funds (ranking 1 reflects the fund that places the highest importance on climate change).
Column (2) reports the ranking of these respondents with respect to their Weighted Average Scope 1 and 2 Carbon Intensity, relative to all
thirty-three responding funds (ranking 1 reflects the fund that has the lowest portfolio carbon intensity). Panel B presents the correlation
between the stated and actual rankings of all thirty-three responding funds.

47

Table 8: Determinants of Respondent Portfolio Carbon Intensity (n=33)
Portfolio Carbon Intensity
(1) (2) (3) (4) (5) (6) (7) (8)
Panel A:
M
odel 1
Material to Financial Performance 47.35*
Ethical Responsibility 19.94
Future Performance
55.20***
Decarbonisation
-14.00
Engagement
54.90**
Environment Weighting
23.10
Climate Change Weighting
-0.223
Climate Initiative
53.20**
Headquarters
U.S. -14.07 0.07 0.46 -6.65 -16.45 -12.66 -10.93 -14.46
Othe
r
17.53 26.93 9.5 25.89 14.53 18.06 20.19 17.20
ESG Name 9.34 -4.37 8.52 1.72 7.96 -0.25 1.60 -16.30
Constant 61.49** 89.46*** 63.24*** 102.10*** 58.65** 93.46*** 100.90*** 83.21***
R
-Squared 0.11 0.04 0.18 0.03 0.15 0.02 0.02 0.18
Panel B:
M
odel 2
Material to Financial Performance 45.71
Ethical Responsibility 16.71
Future Performance
52.37**
Decarbonisation
-12.04
Engagement
54.64**
Environment Weighting
-22.87
Climate Change Weighting
-31.48
Climate Initiative
43.41*
Style
Value -6.76 3.03 2.17 -2.82 -12.17 -0.82 -2.23 13.63
Growth -22.13 -30.68 -21.33 -29.59 -25.02 -29.23 -30.62 -15.90
Ln(Size) -4.78 -2.62 -3.77 -3.59 -4.33 -4.16 -4.45 -3.61
Constant 166.30* 155.80 148.90 185.50* 154.20 202.00* 204.10* 155.00*
R
-Squared 0.15 0.09 0.23 0.08 0.20 0.07 0.08 0.18
Notes: Table 8 presents the results to the regression models that explore the determinants of respondent carbon performance using ordinary least squares. In Panel A, Model 1 correspondents to Equation (4). Across
Columns (1) to (8), we rotate the variable of interest, maintaining Location of Headquarters and ESG Name as control variables. In Panel B, Model 2 correspondents to Equation (5). Across Columns (1) to (8), we rotate
our variables of interest while maintaining Style and Ln(Size) as control variables. Here, standard errors are robust to heteroskedasticity, while ***, **, and * represent the significance at the 1%, 5%, and 10% levels
respectively.

48

Table 9: Determinants of Portfolio Carbon Intensity for Respondents – Additional Controls (n=33)
Portfolio Carbon Intensity
(
1
)
(
2
)
(
3
)
(
4
)
(
5
)
(
6
)
(
7
)
(
8
)
Future Performance 44.52**
49.26**
(2.347)
(2.627)
Decarbonisation -25.89
-21.61
(-0.949)
(-0.866)
Climate Change Weighting
-26.44 -2.762
(-0.445) (-0.046)
Climate Initiative
41.91** 42.94**
(2.070) (2.045)
Engagement 38.94* 61.55** 57.33** 48.22***
(1.907) (2.367) (2.200) (2.807)
Activist Fund
21.61 39.84 34.51 27.94
(0.962) (1.500) (1.513) (1.231)
Constant 45.69*** 62.39*** 64.16*** 46.67*** 63.01*** 91.21*** 89.31*** 70.85***
(3.064) (2.883) (2.884) (3.222) (5.082) (6.909) (5.304) (5.484)
N
33 33 33 33 33 33 33 33
R
-Squared 0.24 0.16 0.14 0.24 0.20 0.09 0.07 0.18
Notes: For additional controls, Table 9 uses ordinary least squares to individually regress Future Performance, Decarbonisation, Climate Change Weighting, and Climate Initiative against Engagement in Columns (1),
(2), (3), and (4) respectively. It also individually regresses Future Performance, Decarbonisation, Climate Change Weighting, and Climate Initiative against Activist Fund in Columns (5), (6), (7), and (8) respectively.
Robust t-statistics are presented in parenthesis, while ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively.

49

Table 10: Full Sample Fund Summary Statistics
Variables Obs Mean Std Dev Min Max Skew Kurt
Portfolio Weighted Average Scope 1 & 2
Carbon Intensity (tCO2e/US$M) 78 109.30 70.85 12.07 320.61 0.88 3.64
Carbon Intensity Coverage (%) 78 98.23 5.43 62.79 100.00 -5.15 30.97
Portfolio Refinitiv ESG Score 78 66.84 7.36 37.28 81.70 -1.38 6.10
Refinitiv ESG Score Coverage (%) 78 90.54 6.83 62.79 100.00 -2.12 7.94
Portfolio MSCI ESG Score 78 6.22 0.69 4.80 8.48 0.14 3.62
MSCI ESG Score Coverage (%) 78 93.06 9.39 51.39 100.00 -2.93 11.73
Portfolio Sustainalytics ESG Score 78 21.54 2.18 14.67 27.96 0.13 4.17
Sustainalytics ESG Score Coverage (%) 78 93.06 9.39 51.39 100.00 2.93 11.73
Respondent Dummy 78 0.42 0.50 0.00 1.00 0.31 1.10
U.S. Headquarters Dummy 78 0.18 0.39 0.00 1.00 1.67 3.79
Other Region Headquarters Dummy 78 0.27 0.45 0.00 1.00 1.04 2.08
Value Style Dummy 78 0.15 0.36 0.00 1.00 1.92 4.68
Growth Style Dummy 78 0.37 0.49 0.00 1.00 0.53 1.28
ESG Name Dummy 78 0.29 0.46 0.00 1.00 0.90 1.81
Size (Ln($)) 78 18.85 2.21 12.03 24.01 -0.60 3.59
Age (Months) 78 219.09 377.24 6.00 1846.00 3.27 12.68
Financial Performance (%) 74 41.61 15.51 7.06 108.10 1.32 7.71
Volatility (%) 74 4.57 1.06 2.48 8.16 1.01 4.51
Management Fees (%) 77 0.87 0.37 0.09 1.52 -0.24 2.11
Notes: Table 10 presents the fund characteristics for the entire sample of respondents and non-respondents at the holdings date of 30 June
2021. These include calculated measures of carbon performance and ESG performance, alongside various fund-level control variables that
have been obtained from Morningstar Direct.

50

Table 11: Determinants of Portfolio Carbon Intensity
Portfolio Carbon Intensity
Model 1 Model 2
(1) (2)
Respondent 8.02
14.38
(0.503) (0.858)
ESG Name -7.06 -4.61
(-0.378) (-0.243)
Headquarters
U.S. 28.11
29.33
(1.363) (1.389)
Othe
r
33.31
37.38*
(1.667) (1.984)
Style
Value 54.81**
66.64**
(2.235) (2.640)
Growth -27.95
-32.80*
(-1.621) (-1.905)
Age 0.00
(0.038)
Ln(Size) -4.38
(-1.144)
Intensity Coverage 1.49 1.30
(0.709) (0.607)
Financial Performance
1.02
(1.424)
Volatility
-18.13*
(-1.674)
Constant 29.37
3.49
(
0.135
)
(
0.016
)
N
77
73
R
-Squared 0.23 0.25
Notes: Table 11 presents the results to the regression models that explore the determinants of fund carbon performance using ordinary least
squares. Column (1) presents Model (1), which corresponds to the first variation of Equation (6), while Column (2) presents Model (2), which
corresponds to the second variation of Equation (6). T-statistics are presented in parenthesis, while ***, **, and * represent the significance
at the 1%, 5%, and 10% levels respectively.

51

Table 12: Determinants of Portfolio ESG Score
Portfolio ESG Score
Model 1 Model 2
Refinitiv
(1)
MSCI
(2)
Sustainalytics
(3)
Refinitiv
(4)
MSCI
(5)
Sustainalytics
(6)
Respondent -0.03 -0.15 -0.47
-0.40 -0.14 -0.74
(-0.028) (-1.082) (-1.045) (-0.314) (-1.030) (-1.628)
ESG Name 0.62 0.42** 1.05* 1.02 0.48*** 1.13**
(0.435) (2.530) (1.960) (0.697) (2.983) (2.122)
Headquarters
U.S. 4.00*** 0.43** 0.15
3.85** 0.39** 0.37
(2.832) (2.449) (0.257) (2.439) (2.290) (0.650)
Othe
r
4.46*** 0.32* -0.89
3.08** 0.22 -0.77
(2.952) (1.857) (-1.606) (2.160) (1.458) (-1.511)
Style
Value -2.58 -0.51** -0.88
-2.46 -0.43** -0.82
(-1.534) (-2.439) (-1.309) (-1.258) (-2.115) (-1.224)
Growth -6.71*** -0.02 1.01** -5.79*** -0.07 0.94**
(-4.639) (-0.147) (2.076) (-4.423) (-0.499) (2.011)
Age -0.00 -0.00 0.00
(-0.954) (-0.184) (0.760)
Ln(Size) 0.33 -0.02 -0.27**
(1.031) (-0.617) (-2.546)
ESG Coverage 0.83*** 0.03** 0.14*** 0.69*** 0.03** 0.11**
(5.874) (2.316) (2.687) (5.364) (2.252) (2.104)
Financial Performance
-0.16*** -0.00 -0.00
(-2.998) (-0.811) (-0.334)
Volatilit
y
1.78** -0.06 -0.07
(2.181) (-0.637) (-0.218)
Constant -14.01 3.11** -29.53*** 3.28 3.41** -31.16***
(
-0.956
)
(
2.027
)
(
-5.923
)
(
0.268
)
(
2.331
)
(
-6.394
)
N
75 74 74 71 70 70
R
-Squared 0.59 0.41 0.38 0.64 0.46 0.32
Notes: Table 12 presents the results to the regression models that explore the determinants of fund ESG performance using ordinary least
squares. Columns (1), (2) and (3) relate to Model (1), which corresponds to the first variation of Equation (7), while Columns (4), (5), and
(6) relate to Model (2), which corresponds to the second variation of Equation (7). For each variation of Equation (7), we utilise portfolio
ESG scores as based on Refinitiv, MSCI, and Sustainalytics data as dependent variables. Robust t-statistics are presented in parenthesis,
while ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively

52

Appendices
Appendix 1: Definitions of Responsible Investing Approaches
Responsible Investing Approach Definition
Fundamental analysis
incorporating ESG considerations
Incorporating ESG factors into the financial analysis of individual firms.
For example, using ESG factors as inputs into cost of capital estimates,
or financial forecasts.
Negative screening
Excluding firms within certain industries, that engage in certain
economic activities or score relatively badly on ESG factors relative to
their peers.
Engagement/active ownership
with companies on ESG
Influencing corporate activities or behaviour through shareholder power.
For example, holding discussions with management, submitting
shareholder proposals, and voting on ESG issues at annual meetings.
Positive (best in class) screening
Including firms within certain industries, that engage in certain
economic activities, or score relatively well on ESG factors relative to
their peers.
Decarbonization of portfolio
Actively reducing the exposure of a portfolio to carbon risk to align with
a low carbon future. For example, divesting stocks that are highly
exposed to carbon emissions, actively engaging with portfolio
companies to reduce their emissions, or purchasing carbon credits to
offset emissions.
Thematic investment
Investing in specific themes or assets that are related to ESG factors. For
example, investing only in firms focussed on green technologies and
clean energies.
Overlay/Portfolio tilt
Using specific investment strategies to tilt the overall ESG performance
of a fund to reach a targeted level. For example, a portfolio could be
tilted towards a targeted carbon footprint.
Quantitative ESG factor investing Selecting securities that score well on ESG factors, which have
historically to achieved above-market returns in return-factor analysis.
Impact investing
Investing with the intention of generating measurable social and
environmental return, alongside financial return.
- Company Impact is the measurable impact that a company has due to
its business activities. For example, a company building solar farms is
mitigating emissions in the Energy sector.
- Portfolio/Capital Impact is the measurable impact an investor has by
providing capital to a project or investment. Showing additionality is
key for this type of impact. For example, providing capital directly to a
solar farm developer (in a primary market transaction), which allows
them to build additional renewable generation.
Notes: In Appendix 1, the definitions of responsible investing approaches have been adapted from Amel-Zadeh & Serafeim (2018), Credit
Suisse (2021) and RIAA (2021a).

53

Appendix 2: Full Survey
1. What is your Institution’s name?
__________________________________
2. What is the geographical location of your Institution’s headquarters?
o Australia
o New Zealand
o United Kingdom
o United States
o Other (please state): __________________
3. What is your Job Title/Position?
o Fund/Portfolio Manager
o ESG/Responsible Investment Specialist
o Chief Executive Officer
o Investment Analyst/Strategist
o Executive/Managing Director
o Chief Investment Officer
o Other (please explain): _____________________
4. What was the total size of assets under management (in $USD) at your
institution as at Q2 2021?
USD$__________________
5. Which of the following organisations is your Institution a signatory/member of?
[Please select all that apply]
 United Nations Principles of Responsible Investing
 B-Corporation
 Carbon Disclosure Project
 Responsible Investment Association of Australasia
 Climate Action 100+
 Investor Agenda
 Net Zero Asset Managers
 Investor Group on Climate Change (IGCC)
 Asia Investor Group on Climate Change (AIGCC)
 Climate Investment Coalition
 Climate League 2030
 Science Based Targets Initiative for Financial Institutions
 Net-Zero Asset Owner Alliance
 Other (Please State): _______________

54

6. Do you consider any ESG information when making investment decisions?
o Yes – in all of our funds
o Yes – in more than half of our funds
o Yes – in less than half of our funds
o No
IF YES – IN ALL OF OUR FUNDS IS SELECTED, MOVE TO QUESTION 8:
ELSE, MOVE TO QUESTION 7:
7. Why do you not consider ESG information when making investment decisions
across all of your funds?
[Please select all that apply]
 Our SIPO/mandate does not allow it
 There is no stakeholder demand for such policy
 We lack access to reliable nonfinancial data
 ESG information is not material to investment performance
 We believe such policy to be ineffective in inducing change at firms
 Such information is not material to a diversified investment portfolio
 Including such information is detrimental to investment performance
 It would violate our fiduciary duty to our stakeholders
 It is not possible to reflect our client’s diverse ethical views
 Other (please state): ____________________
*IF NO WAS SELECTED IN QUESTION 6, THE SURVEY IS FINISHED.
FOR ALL OTHER ANSWERS SELECTED IN QUESTION 6, MOVE TO QUESTION
8*
8. What percentage (%) of your assets under management formally incorporate
Environmental, Social, AND Governance considerations?
_____________% of AUM
9. Does your institution have a written ESG policy?
 Yes
 No
IF YES, MOVE TO QUESTION 10:
IF NO, MOVE TO QUESTION 11:

55

10. What was the date of your first ESG policy?
__________________
11. How do you build ESG capabilities in your team?
[Please select all that apply]
 Internal training
 External training (please state provider): _________________
 Industry and academic groups e.g. CFA ESG mico-credentials. (please state
the qualifications): _____________
 Hiring ESG experts
 Hiring investment experts with some ESG training
 Other (please state): ______________
12. Which of the following applies to your institution?
[Please select all that apply]
 We have ESG specialist/s
 Everyone is trained in ESG
 None of the above
13. Why do you consider ESG information when making investment decisions?
[Please select all that apply]
 ESG information is material to investment performance
 Growing client/stakeholder demand
 We believe this will encourage positive change in individual firm ESG
practices
 It is part of our mandated investment strategy/SIPO
 We see it as an ethical responsibility
 ESG risk and opportunities, although not yet priced, will soon affect
investment performance
 Other (please state): _______________

56

14. What is your flagship retail ESG global equity fund? Please choose your global
equity fund that incorporates ESG principles the most (i.e. this could be one of
your general global equity funds)
____________________________________________________
The remainder of this survey will specifically focus on the fund that you have provided
above.
Please answer all questions below based on your Flagship retail ESG Fund identified
above.
15. How much money is invested in this fund (in $USD) as at Q2 2021?
___________ USD
16. What is the average holding period for equities in this fund?
o Short (less than 6 months)
o Medium (6 months to 2 years)
o Long (2 years to 5 years)
o Very long (more than 5 years)
17. What is the benchmark index for this fund?
_________________________
18. Which of the following applies to this fund?
o We aim to track our benchmark index
o We aim to outperform our benchmark index
19. What is the investment style of this fund?
[Please select all that apply]
 Value/Fundamental
 Momentum
 Growth
 Factor/Quantitative
 Broad Market
 Concentrated (less than 50 holdings)
 Quality
 Specific Theme (please state): _____________
 Other (please state): __________________

57

20. What is the gender of the lead Fund/Portfolio Manager for this fund?
o Male
o Female
o Non-binary
21. What was the start date of the lead Fund/Portfolio Manager in charge of this
fund?
____________________
22. Which of the following qualifications has the lead Fund/Portfolio Manager of the
fund attained?
[Please select all that apply]
 Bachelors in Finance/Accounting
 Masters in Finance/Accounting
 PhD in Finance/Accounting
 MBA
 CFA
 Other professional qualification (please state): ___________
 Bachelors in another subject (please state): ___________
 Masters in another subject (please state): ___________
 PhD in another subject (please state): ___________
23. Which ESG investment approaches do you incorporate within this fund?
[Please select all that apply]
 Negative screening: exposure based
 Negative screening: industry based
 Positive (best in class) screening
 Overlay/portfolio tilt
 Decarbonization of portfolio
 Quantitative ESG factor investing
 Fundamental analysis incorporating ESG considerations
 Impact Investing (1)
 Engagement/Active Ownership with companies on ESG (2)
 Thematic investment
 Other (Please state): _______________________
IF (1) IS SELECTED, MOVE TO QUESTION 24:
IF (2) IS SELECTED, MOVE TO QUESTION 25:
ELSE, MOVE TO QUESTION 27:

58

24. Which Impact Investing standard/methodology do you follow?
[Please select all that apply]
 Impact Investing and Reporting Standards (IRIS+)
 Global Impact Investing Rating System (GIIRS)
 Sustainability Accounting Standards Board (SASB)
 Global Reporting Initiative (GRI)
 Other (Please State): _____________________
IF (2) WAS SELECTED IN QUESTION 23, MOVE TO QUESTION 25:
ELSE, MOVE TO QUESTION 27:
25. What measures of direct engagement over ESG issues have you taken in the past
five years with any of your portfolio companies?
[Please select all that apply]
□ Questioning management on a conference call about ESG issues
□ Holding private discussions with management regarding the financial implications
of ESG issues
□ Publicly criticizing management on ESG issues
□ Privately proposing specific actions to management on ESG issues
□ Voting against management on proposals over ESG issues at the annual meeting (3)
□ Voting against re-election of any board directors due to ESG issues (3)
□ Submitting shareholder proposals on ESG issues
□ Legal action against management on ESG issues
□ Outsource to a third-party engagement provider (3)
□ Other (Please State): __________________________
IF (3) IS SELECTED, MOVE TO QUESTION 26
ELSE, MOVE TO QUESTION 27:
26. How do you undertake voting on ESG issues?
[Please select all that apply]
 Direct Voting
 Proxy Voting (please specify proxy provider): ___________________
 Proxy Voting generally, but Direct Voting on controversial issues (please
specify proxy provider): ___________________
 Other (Please State): ________________

59

27. Which type of ESG data do you use?
[Please select all that apply]
 ESG ratings (4)
 Raw data (e.g. emissions data)
 Analysis at firm level (e.g. incorporating stranded asset risk into valuation)
 Analysis at sector level (e.g. identifying sectors exposed to ESG risks)
 Analysis at country level (e.g. identifying countries exposed to ESG risks)
 Other (please state): ___________________
IF (4) IS SELECTED, MOVE TO QUESTION 28:
ELSE, MOVE TO QUESTION 29:
28. Which ESG rating provider do you use?
[Please select all that apply]
 Sustainalytics
 MSCI
 Refinitiv
 Bloomberg
 S&P Global
 FTSE Russel
 Other (please state): ___________
29. When emissions data is unavailable, how do you estimate/predict firm-level
emissions?
[Please select all that apply]
 Multiples/industry averages
 Regression analysis
 Machine learning
 From external provider e.g. Emmi
 Other (please state): ______________
 We do not estimate firm-level emissions

60

ESG Preferences
30. Allocate 100 points between Environment, Social and Governance themes based
on the importance this fund places on them in the investment process. [Place 100
points into the "Equally weighted" box if you place equal importance on each of
the three themes]
(Note: Please make sure that the sum of your allocated points equals 100)
31. Allocate 100 points between the specific Environmental themes below based on
the importance this fund places on them in the investment process.
(Note: Please make sure that the sum of your allocated points equals 100)
32. Allocate 100 points between the specific Social themes below based on the
importance this fund places on them in the investment process.
(Note: Please make sure that the sum of your allocated points equals 100)
Theme Points
Environment
Social
Governance
Equally weighted
Theme Points
Climate Chan
g
e
Pollution & Waste
Natural Capital (incl.
b
iodiversit
y)
Environmental
Opportunities
Theme Points
Health & Safet
y
Human Capital
Mana
g
ement
Product Liabilit
y
Supply Chain &
Communit
y

61

33. Allocate 100 points between the specific Governance themes below based on the
importance this fund places on them in the investment process.
(Note: Please make sure that the sum of your allocated points equals 100)
Stated Carbon Performance
34. What is the weighted-average Scope 1 and 2 emissions (tonnes of CO2e) intensity
per sales ($USD) of your fund (as at Q2 2021)?
o The fund's weighted-average Scope 1 and 2 emissions (tonnes of CO2e) intensity per
$USD of revenue is: ______________
o We do not calculate this because ________________
o We calculate a different intensity measure. Please explain: ________________
Theme Points
Board Composition
Remuneration
Cor
p
orate Behaviour
Shareholder Ri
g
hts

62

35. What is the weighted-average Scope 1, 2 and 3 emissions (tonnes of CO2e)
intensity per sales ($USD) of your fund (as at Q2 2021)?
o The fund's weighted-average Scope 1,2 and 3 emissions (tonnes of CO2e) intensity
per $USD of revenue is: _____________________
o We do not calculate this because ________________
o We calculate a different intensity measure. Please explain: ________________
36. What are the attributable fossil-fuel reserves of your flagship ESG global equity
fund (as at Q2 2021)? Attributable fossil-fuel reserves are the sum of all disclosed
reserves of each company multiplied by the percentage ownership of the fund in
that company.
o The fund's attributable fossil-fuel reserves measured in barrel of oil equivalent (BOE)
is: ____________________
o The fund's attributable fossil-fuel reserves measured in embedded carbon (tonnes of
CO2e) is: __________________
o We do not calculate this because: ____________________
o We calculate a different measure. Please explain: ________________

63

Appendix 3: Use of ESG Information
Various types of ESG information can be incorporated into the investment process, including
internal and external analysis. Table A3 details the types of ESG information used in the
specific global equity funds provided by respondents, presenting the survey responses to the
question “Which type of ESG data do you use”, adapted from van Duuren et al. (2016).
[INSERT TABLE A3 HERE]
Table A3 highlights that respondents more commonly incorporated analysis at the individual
firm level rather than at the aggregated sector or country level, consistent with van Duuren et
al. (2016). Specifically, the most common types of ESG information used among global equity
funds in our sample were analysis at the firm level (83%) and raw ESG data (79%). However,
the preference for raw ESG data over external ESG ratings differs from the findings of van
Duuren et al. (2016). Raw data, such as carbon emissions, can be sourced from external data
providers, company reports, and press statements, and often requires more internal resources
to process. Our finding likely corresponds to the growing concerns regarding the divergence of
different external ESG ratings, alongside the adoption of internal ESG scoring frameworks.
Our comparison of small and large funds in Table A3 highlights a significant difference in the
proportion of respondents who use external ESG ratings at the 1% level. While 86% of small
funds indicated that they use external ratings, only 50% of large funds selected this option.
Alongside this, large funds more commonly incorporated raw ESG data into their analysis
alongside financial data. These results are likely because larger funds have more internal
resources to process raw ESG data, reducing the need to rely on external ratings, relative to
smaller funds. This is consistent with many of the larger funds in our sample indicating that
they use their own internal ESG scoring framework for individual companies.

64

Table A3: Types of ESG Data Used (n=42)

All Fund Size ESG Named Fund Region of Headquarters
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
>50% 50% Diff Yes No Diff
Australasia U.S. Other Range
Analysis at firm level (e.g. incorporating stranded asset risk into
valuation) 83% 85% 82% 3% 78% 88% -10% 77% 90% 100% 23%
Raw data (e.g. emissions data) 79% 85% 73% 12% 83% 75% 8% 73% 90% 83% 17%
Analysis at sector level (e.g. identifying sectors exposed to ESG
risks) 69% 65% 73% -8% 78% 63% 15% 58% 90% 83% 32%
External ESG ratings 69% 50% 86% -36%** 67% 71% -4% 65% 70% 83% 18%
Analysis at country level (e.g. identifying countries exposed to
ESG risks) 43% 40% 46% -6% 44% 42% 3% 35% 40% 83% 49%
Othe
r
24% 35% 14% 21% 22% 25% -3% 23% 20% 33% 13%
Notes: Table A3 reports the percentage of survey responses to "Which type of ESG data do you use?", where Multiple responses were allowed. Column (1) reports the percentage of respondents that selected the response
for a given row. Columns (2) and (3) report the percentages for funds greater than 50% of the median fund size (‘>50%’) and less than or equal to 50% of the median fund size (50%) respectively. Column (4) reports
the difference between Column (2) and (3), and the results of a test of the null hypothesis that the two percentages are equal. Columns (5) and (6) report the percentages for ESG named funds (‘Yes’) and non-ESG named
funds (‘No’) respectively. Column (7) reports the difference between Column (5) and (6), alongside the results of a test of the null hypothesis that the two percentages are equal. Columns (8), (9), and (10) report the
percentage of respondents with headquarters in Australasia, the United States, and other regions, who selected the response for a given row. Across the region of headquarters, the range (high minus low) of percentages
are reported in Column (11). Finally, ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively. The responses in Table A3 have been ordered from highest to lowest based on the proportion of
respondents that chose each reason in Column (1).

65

Appendix 4: ESG Engagement Approaches
Active ESG ownership involves influencing corporate activities or behaviour through
shareholder power (Amel-Zadeh & Serafeim, 2018). Table A4 details the types of ESG
engagement strategies adopted by the specific global equity funds provided by respondents.
Adapted from Krueger et al. (2020), respondents were asked “What measures of direct
engagement over ESG issues have you taken in the past five years with any of your portfolio
companies?" Referring to Figure 1, this question was only available to respondents who
selected that they employ active engagement strategies with firms on their ESG issues, which
included 32 funds.
[INSERT TABLE A4 HERE]
The results in Table A4 indicate that the most frequent engagement approach was private
discussions with management regarding the financial implications of ESG issues, which was
selected by 97% of respondents within the subsample. This was followed by voting against
management on proposals over ESG issues at the annual meeting (91%) and questioning
management on a conference call about ESG issues (88%). While many responding funds
indicated that they privately propose specific actions to management (84%), the proportion of
funds that publicly submitted shareholder proposals was relatively low (47%). These results
support the interpretation that managers prefer private interactions with firms first, and only
take public actions once private interventions fail (McCahery et al., 2016). This divergence
was emphasised in both Australasian and U.S. regions. However, funds with headquarters in
other regions, including Europe, more actively submitted shareholder proposals and voted
against the re-election of the board of directors.

66

Table A4: Engagement Strategies (n=32)
Notes: Table A4 presents the survey responses to "What measures of direct engagement over ESG issues have you taken in the past five years with any of your portfolio companies?", where multiple responses were
allowed. Column (1) reports the percentage of respondents that selected the response for a given row. Columns (2) and (3) report the percentages for funds greater than 50% of the median fund size (‘>50%’) and less
than or equal to 50% of the median fund size (50%) respectively. Column (4) reports the difference between Column (2) and (3), and the results of a test of the null hypothesis that the two percentages are equal.
Columns (5) and (6) report the percentages for ESG named funds (‘Yes’) and non-ESG named funds (‘No’) respectively. Column (7) reports the difference between Column (5) and (6), alongside the results of a test of
the null hypothesis that the two percentages are equal. Columns (8), (9), and (10) report the percentage of respondents with headquarters in Australasia, the United States, and other regions, who selected the response for
a given row. Across the region of headquarters, the range (high minus low) of percentages are reported in Column (11). Finally, ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively. The
responses in Table A4 have been ordered from highest to lowest based on the proportion of respondents that chose each reason in Column (1).
All Fund Size ESG Named Fund Re
g
ion of Head
q
uarters
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
>50% 50% Diff Yes No Diff
Australasia U.S. Other Range
Holding private discussions with management regarding the
financial implications of ESG issues 97% 94% 100% -6% 93% 100% -7% 95% 100% 100% 5%
Voting against management on proposals over ESG issues at the
annual meeting 91% 88% 94% -6%
87% 94% -7% 90% 88% 100% 13%
Questioning management on a conference call about ESG issues 88% 88% 88% 0%
93% 82% 11% 84% 88% 100% 16%
Privately proposing specific actions to management on ESG
issues 84% 75% 94% -19%
80% 88% -8% 90% 63% 100% 38%
Voting against re-election of any board directors due to ESG
issues 75% 63% 88% -25%
80% 71% 9% 68% 75% 100% 32%
Submitting shareholder proposals on ESG issues 47% 50% 44% 6%
47% 47% 0% 42% 38% 80% 43%
Publicly criticizing management on ESG issues 22% 31% 13% 19%
13% 29% -16%
21% 25% 20% 5%
Outsourcing to a third-party engagement provider 13% 13% 13% 0%
20% 6% 14%
16% 13% 0% 16%
Legal action against management on ESG issues 3% 6% 0% 6%
0% 6% -6%
0% 0% 20% 20%
Other 9% 6% 13% -6% 7% 12% -5% 11% 13% 0% 13%

67

Appendix 5: ESG Voting Approaches
Voting can be used as a mechanism for institutions to influence an organisation’s corporate
behaviour. Table A5 details the survey responses to the question “How do you undertake voting
on ESG issues?”. Referring to Figure 1, this question was only available to respondents who
selected an engagement strategy related to voting or external outsourcing, which included 29
funds. Respondents were only able to select one option that applied to their fund and there was
an opportunity to provide text entry answers for choices that were not included in this list.
[INSERT TABLE A4 HERE]
The results in Table A5 highlight that the most common way to vote on ESG issues was through
proxy voting (59%), with many respondents citing ISS (Institutional Shareholder Services) or
Glass Lewis as their proxy providers. Interestingly, many respondents did not consider proxy
voting to be a form of outsourced engagement, which was only selected by 13% of respondents
in Table A5. There were large geographical differences in voting strategies, with 41% of funds
with Australasian headquarters using direct voting, compared to 0% with U.S. headquarters.
This was partially attributable to direct voting being more common among smaller funds,
relative to larger funds.

68

Table A5: Voting Approaches (n=42)
All Fund Size ESG Named Fund Region of Headquarters
(
1
)
(
2
)
(
3
)
(
4
)
(
5
)
(
6
)
(
7
)
(
8
)
(
9
)
(
10
)
(
11
)
>50% 50% Diff Yes No Diff
Australasia U.S. Other Range
Proxy Voting 59% 67% 50% 17% 57% 60% -3% 47% 71% 80% 33%
Direct Votin
g
24% 13% 36% -22%
29% 20% 9% 41% 0% 0% 41%
Proxy Voting generally, but Direct Voting on controversial issues 10% 7% 14% -8% 7% 13% -6%
12% 14% 0% 14%
Othe
r
7% 13% 0% 13% 7% 7% 0% 0% 14% 20% 20%
Notes: Table A5 presents the percentage of survey responses to “How do you undertake voting on ESG issues?”. Column (1) reports the percentage of respondents that selected the response for a given row. Columns (2)
and (3) report the percentages for funds greater than 50% of the median fund size (‘>50%’) and less than or equal to 50% of the median fund size (50%) respectively. Column (4) reports the difference between Column
(2) and (3), and the results of a test of the null hypothesis that the two percentages are equal. Columns (5) and (6) report the percentages for ESG named funds (‘Yes’) and non-ESG named funds (‘No’) respectively.
Column (7) reports the difference between Column (5) and (6), alongside the results of a test of the null hypothesis that the two percentages are equal. Columns (8), (9), and (10) report the percentage of respondents with
headquarters in Australasia, the United States, and other regions, who selected the response for a given row. Across the region of headquarters, the range (high minus low) of percentages are reported in Column (11).
Finally, ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively. The responses in Table A5 have been ordered from highest to lowest based on the proportion of respondents that chose each
reason in Column (1).

69

Appendix 6: Actual vs. Stated Carbon Performance – Robustness Check (n=33)
Portfolio Carbon Intensity (Using FY19 Revenues)
(1) (2) (3) (4) (5) (6) (7) (8)
Panel A: Model 1
Material to Financial Performance 54.55*
Ethical Responsibility 41.23
Future Performance
59.53**
Decarbonisation
-25.09
Engagement
63.27**
Environment Weighting
1.803
Climate Change Weighting
-18.48
Climate Initiative
63.21**
Headquarters
U.S. -18.75 7.627 -2.848 -7.453 -21.49 -15.29 -13.37 -19.32
Othe
r
42.97 60.00 34.51 56.28 39.52 45.84 48.68 42.49
ESG Name 31.14 9.881 29.68 22.44 29.55 22.06 23.48 0.950
Constant 47.91 69.68** 52.71* 95.50*** 44.62 92.73*** 94.89*** 72.30***
R
-Squared 0.150 0.125 0.194 0.087 0.188 0.073 0.075 0.215
Panel B: Model 2
Material to Financial Performance 54.65*
Ethical Responsibility 29.41
Future Performance
59.19**
Decarbonisation
-9.652
Engagement
64.60**
Environment Weighting
-13.74
Climate Change Weighting
-27.27
Climate Initiative
62.78*
Style
Value -8.19 4.63 2.40 -1.96 -14.49 0.02 -1.70 19.33
Growth -6.30 -18.35 -5.74 -14.37 -9.78 -13.72 -15.40 4.05
Ln(Size) -8.36 -5.15 -7.14 -6.98 -7.81 -7.35 -7.72 -6.92
Constant 223.60 195.00 205.10 245.90 209.50 255.60 262.20 202.70
R
-Squared 0.13 0.08 0.18 0.06 0.17 0.05 0.06 0.20
Notes: Appendix 6 presents a robustness check to the regression models that explore the determinants of respondent carbon performance using ordinary least squares. Portfolio Carbon Intensity has been recalculated
based on FY20 emissions and FY19 revenues, to account for the financial impact of the Covid-19 Pandemic. In Panel A, Model 1 correspondents to Equation (4). Across Columns (1) to (8), we rotate the variable of
interest, maintaining Location of Headquarters and ESG Name as control variables. In Panel B, Model 2 correspondents to Equation (5). Across Columns (1) to (8), we rotate our variables of interest while maintaining
Style and Ln(Size) as control variables. Here, standard errors are robust to heteroskedasticity, while ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively.

70

Appendix 7: Determinants of Carbon Performance – Robustness Check
Portfolio Carbon Intensity
Model 1 Model 2
(1) (2)
Respondent 8.420
13.32
(0.459) (0.704)
ESG Name 0.457 5.180
-0.023 (0.229)
Headquarters
U.S. 23.13
24.03
(1.145) (1.193)
Othe
r
40.19
47.34*
(1.575) (1.778)
Style
Value 63.10** 82.99**
(2.102) (2.485)
Growth -17.51
-25.61
(-0.862) (-1.432)
Age 0.00
(-0.058)
Ln(Size) -6.08
(-1.367)
Intensity Coverage 1.149 2.267
(0.800) (1.128)
Financial Performance
1.59
(1.380)
Volatility
-28.26**
(-2.035)
Constant 96.31
-68.20
(
0.622
)
(
-0.348
)
N
75
71
R
-Squared 0.209
0.243
Notes: Appendix 7 presents a robustness check to the regression models that explore the determinants of fund carbon performance using
ordinary least squares. Portfolio Carbon Intensity has been recalculated based on FY20 emissions and FY19 revenues, given the financial
impact of the Covid-19 Pandemic. Column (1) presents Model (1), which corresponds to the first variation of Equation (6), while Column (2)
presents Model (2), which corresponds to the second variation of Equation (6). Robust t-statistics are presented in parenthesis, while ***, **,
and * represent the significance at the 1%, 5%, and 10% levels respectively.

71

Appendix 8: Determinants of ESG Performance – Robustness Check
Portfolio ESG Score
Model 1 Model 2
Refinitiv
(1)
MSCI
(2)
Sustainalytics
(3) Refinitiv
(4)
MSCI
(5)
Sustainalytics
(6)
Respondent -0.69 -0.16 -0.33
-1.53 -0.20 -0.42
(-0.411) (-0.896) (-0.703) (-0.872) (-1.142) (-0.847)
ESG Name 1.51 0.46** 0.68
1.81 0.51** 0.79
(0.747) (2.236) (1.220) (0.889) (2.482) (1.396)
Headquarters
U.S. 4.65** 0.51** -0.23
4.30* 0.46** 0.05
(2.152) (2.305) (-0.393)
(1.968) (2.076) (0.079)
Othe
r
4.51** 0.36* -0.90
3.32* 0.28 -0.84
(
2.137
)
(
1.688
)
(
-1.555
)
(
1.684
)
(
1.394
)
(
-1.536
)
Style
Value -2.86 -0.53** -0.635
-3.03 -0.48* -0.50
(
-1.069
)
(
-2.021
)
(
-0.909
)
(
-1.121
)
(
-1.857
)
(
-0.698
)
Growth -6.38*** -0.07 1.13** -5.68*** -0.12 1.14**
(-3.476) (-0.380) (2.227) (-3.133) (-0.672) (2.275)
A
g
e -0.00 0.00 0.00
(-0.449) (0.077) (0.211)
Ln(Size) 0.49 -0.00 -0.30***
-1.202 (-0.006) (-2.726)
ESG Coverage 1.42*** 0.11*** -0.15*** 1.28*** 0.10*** -0.17***
(8.250) (5.460) (-2.822)
(7.194) (5.287) (-3.165)
Financial
Performance
-0.15* -0.01 -0.01
(-1.986) (-0.964) (-0.361)
Volatility
1.74 -0.03 -0.03
(1.544) (-0.228) (-0.078)
Constant -79.20*** -4.48** -0.19
-58.93*** -3.60* -3.30
(-4.333) (-2.325) (-0.038) (-3.481) (-1.915) (-0.632)
N
75 74 74 71 70 70
R
-Squared 0.63 0.53 0.33 0.65 0.56 0.27
Note: Appendix 8 presents a robustness check to the regression models that explore the determinants of fund ESG performance using ordinary
least squares. Portfolio ESG scores have been recalculated using non-normalised weightings (i.e., value-weighted averages based on all
available holdings). Columns (1), (2) and (3) relate to Model (1), which corresponds to the first variation of Equation (7), while Columns (4),
(5), and (6) relate to Model (2), which corresponds to the second variation of Equation (7). For each variation of Equation (7), we utilise
portfolio ESG scores as based on Refinitiv, MSCI, and Sustainalytics data as dependent variables. Robust t-statistics are presented in
parenthesis, while ***, **, and * represent the significance at the 1%, 5%, and 10% levels respectively.