Tab 1

Modelling the behaviour of Environmentally

friendly financial markets: Understanding the
moderation of climate risk factors

Submitted To
Dr. Aswini Kumar Mishra
FRAM : F414

Submitted By Group 4:
 1

Highlights:

● Climate risk factors significantly influence the behavior of environmentally friendly

financial markets, impacting investment decisions and asset pricing.

● The study employs multivariate quantile-on-quantile regression (m-QQR) to analyze

the moderating effects of climate risks on renewable energy returns.

● Investor sentiment, shaped by media coverage of climate change, leads to short-term

fluctuations in renewable energy asset prices.

● The paper highlights the need for systematic analysis of how environmental policy
fluctuations affect investor confidence and asset pricing in renewable energy markets.

Abstract:

This research paper examines the behavior of environmentally friendly financial markets by
modeling the moderation of climate risk factors. As climate change increasingly threatens
economic stability, understanding how these risks influence investment decisions and asset
pricing is crucial. The study employs multivariate quantile-on-quantile regression (m-QQR)
to analyze the interactions between climate risk factors and renewable energy returns,
revealing that physical and transition risks significantly impact market dynamics.
Additionally, the research highlights the role of investor sentiment, shaped by media coverage
and public interest in climate issues, in driving short-term fluctuations in renewable energy
asset prices. Innovative hedging strategies are proposed to manage environmental exposure
within investment portfolios, enhancing resilience against climate-related shocks. The paper
identifies critical research gaps, particularly the need for systematic analyses of how
fluctuations in environmental policies affect investor confidence and market behavior. By
integrating high-frequency environmental data into financial models, this study aims to
provide valuable insights for investors and policymakers seeking to navigate the complexities
of sustainable finance in an era marked by climate uncertainty. Ultimately, this research
contributes to a deeper understanding of the dynamics at play in environmentally friendly
financial markets and offers a framework for future studies in this evolving field.

Keywords:

Climate risk factors; Environmentally friendly financial markets; Renewable energy

investments; Multivariate quantile-on-quantile regression; Hedging strategies
 2

Table of Contents

1. Introduction 3
2. Literature Review 5
2.1 Theoretical Foundations of Environmental Financial Markets 5
2.2 Empirical Evidence on Climate Risk Factors 5
2.3 Moderation of Climate Risk in Financial Markets 6
2.4 Investor Sentiment and Market Behavior 6
2.5 Institutional Frameworks and Policy Implications 6
2.6 Emerging Trends and Future Research Directions 7
2.7 Research Gaps In Existing Literature 7
3. Model and data 8
3.1 Empirical Model 8
4. Results and Discussion 12
4.1 Cross Quantile Correlation estimate 12
4.2 Partial Cross Quantile Correlation estimates 15
4.3 Multivariate Quantile-On-Quantile Regression 19
4.4 Quantile on Quantile Regression (QQR) 25
5. Conclusion 29
6. Policy Implications 30
7. References 31
 3

1. Introduction

The intersection of finance and environmental sustainability has garnered increasing attention
in recent years, particularly as climate change poses significant risks to economic stability. The
need for environmentally friendly financial markets is underscored by the growing awareness
of climate-related risks and the urgent call for sustainable investment practices. This research
paper seeks to model the behavior of environmentally friendly financial markets, focusing on
the moderation of climate risk factors and how these factors influence market dynamics.

The concept of sustainable finance encompasses a range of investment strategies that prioritize
environmental, social, and governance (ESG) criteria. As investors become more cognizant of
the impacts their investments have on the environment, there has been a marked shift towards
green assets. These assets are not only viewed as a means to achieve financial returns but also
as a vehicle for promoting sustainability and mitigating climate change effects. This shift has
prompted researchers and practitioners alike to explore how climate risk factors moderate the
performance and behavior of these environmentally friendly financial markets.

Recent studies have highlighted the critical role that climate risk plays in shaping investment
decisions. For instance, Pastor et al. (2021) propose an equilibrium model that illustrates how
sustainable investing influences asset prices and corporate behavior. Their findings suggest that
while green assets may exhibit lower expected returns due to investors' environmental
preferences, they tend to perform better during climate-related events. This adaptation of
modern portfolio theory to include environmental factors has significant implications for risk
assessment and market behavior.

Moreover, Bolton et al. (2020) provide compelling evidence that investors demand
compensation for exposure to carbon emission risks, identifying a notable carbon premium in
stock returns. This is supported by Chava (2014), who finds that firms with strong
environmental concerns face higher costs of equity and debt capital, indicating that markets are
increasingly adept at pricing environmental risks, albeit with some inefficiencies.

The moderation of climate risk within environmentally friendly financial markets occurs
through various channels. Physical risks associated with climate change—such as extreme
weather events—can disrupt production capabilities and infrastructure within renewable
energy sectors (Perera et al., 2020). Transition risks, which arise from regulatory changes or
shifts in market demand for fossil fuels versus renewable energy sources, further complicate
investment landscapes (Wang et al., 2022). Understanding these moderating factors is essential
for developing robust models that accurately reflect market behaviors in response to climate
risks.

Investor sentiment also plays a pivotal role in shaping market dynamics. Recent literature
indicates that media coverage and public interest in climate issues can lead to fluctuations in
renewable energy asset prices (Esquivias et al., 2022). However, these effects are often
transient, highlighting the importance of sustained investor engagement and education
 4

regarding the long-term benefits of green investments. The psychological factors driving
investor sentiment toward sustainable assets warrant further exploration to fully understand
their impact on market behavior.

Moreover, institutional frameworks and policy implications significantly influence how

climate risks are moderated within financial markets. Engle et al. (2020) propose innovative
hedging strategies for managing environmental exposure within investment portfolios, which
are crucial for enhancing resilience against climate-related shocks. Baker et al. (2018) examine
green bonds as a financing mechanism for sustainable projects, revealing that these instruments
often trade at a premium due to their perceived environmental benefits. This suggests an
evolving landscape where investors increasingly prioritize sustainability over traditional
financial metrics.

Despite the progress made in understanding environmentally friendly financial markets,

several research gaps remain that warrant further investigation. The integration of high-
frequency environmental data into financial models is largely unexplored, limiting insights into
short-term market dynamics and investor behavior in response to climate risks. Additionally,
comprehensive analyses of cross-market spillover effects within environmental indices are
needed to capture the interconnectedness of various segments in sustainable investing.

This paper aims to contribute to the existing literature by modeling the behavior of
environmentally friendly financial markets through an empirical framework that considers both
drivers and moderators of climate risk. By examining how these factors interact and influence
market outcomes, we hope to provide valuable insights into developing effective strategies for
promoting sustainable investments.

2. Literature Review
 5

The intersection of finance and environmental sustainability has garnered increasing attention
in recent years, particularly as climate change poses significant risks to economic stability. This
literature review synthesizes key studies on the behavior of environmentally friendly financial
markets, focusing on how climate risk factors moderate these markets. The review highlights
theoretical foundations, empirical evidence, and emerging trends that shape our understanding
of this evolving field.

2.1 Theoretical Foundations of Environmental Financial Markets

The theoretical underpinnings of environmentally friendly financial markets have evolved from
traditional financial theories to incorporate sustainability considerations. In their paper (Pastor
et al., 2021) propose an equilibrium model that illustrates how sustainable investing influences
asset prices and corporate behavior. Their findings suggest that while green assets may exhibit
lower expected returns due to investors' environmental preferences, they tend to perform better
during climate-related events. This adaptation of modern portfolio theory to include
environmental factors has significant implications for risk assessment and market behavior.

Moreover, the pricing of climate risk has emerged as a critical area of study. Bolton et al.,
(2020) provide compelling evidence that investors demand compensation for exposure to
carbon emission risks, identifying a notable carbon premium in stock returns. This is supported
by Chava (2014), who finds that firms with strong environmental concerns face higher costs of
equity and debt capital, indicating that markets are increasingly adept at pricing environmental
risks, albeit with some inefficiencies.

2.2 Empirical Evidence on Climate Risk Factors

Recent empirical studies have highlighted the long-term impacts of climate change on asset
valuation and market behavior. Giglio et al. (2021) analyze real estate markets to reveal how
environmental risks are integrated into long-term asset prices, providing insights applicable to
other environmentally focused financial markets. Their work emphasizes the importance of
understanding discount rates under various climate scenarios.

Additionally, Hong et al. (2020) present a comprehensive framework for understanding climate
finance that underscores the integration of climate risk into investment decisions. Their
research indicates that institutional investors are increasingly prioritizing climate risks in their
portfolios, although methodologies for assessing these risks vary widely (Krueger et al., 2020).

2.3 Moderation of Climate Risk in Financial Markets

The moderation of climate risk within environmentally friendly financial markets occurs
through multiple channels. Physical risks associated with climate change—such as extreme
weather events—can directly impact renewable energy infrastructure and production
capabilities (Perera et al., 2020). Transition risks stemming from regulatory changes or shifts
in market demand also pose challenges for investors, particularly in jurisdictions with
fluctuating climate policies (Wang et al., 2022).
 6

The role of green finance has been identified as a crucial element in mitigating climate risks
and promoting sustainable investments. Demiralay et al. (2023) highlight the resilience of
green finance products during economic downturns, noting that these instruments have
outperformed fossil fuel investments during crises like the COVID-19 pandemic. The
interaction between oil prices and renewable energy investments further complicates market
dynamics; for instance, rising oil prices often lead oil-importing countries to increase
investments in renewable energy sources (Shah et al., 2018; Bowden & Payne, 2015).

2.4 Investor Sentiment and Market Behavior

Investor sentiment plays a pivotal role in shaping the behavior of environmentally friendly
financial markets. Ahmed et al. (2024) demonstrate that media coverage and public interest in
climate change can lead to temporary surges in renewable energy asset prices. However, these
effects are often short-lived, highlighting the need for sustained investor engagement and
education regarding the long-term benefits of green investments.

Methodological advancements have also contributed to our understanding of these dynamics.

Quantile-on-Quantile Regression (QQR) techniques have been employed to capture complex
relationships between climate risk factors and financial market returns (Sim & Zhou, 2015).
Recent applications by Esquivias et al. (2022) illustrate how these methodologies can reveal
varying dependencies across different stages of economic development.

2.5 Institutional Frameworks and Policy Implications

The structure of financial markets and institutional frameworks significantly influence the
moderation of climate risks. Engle et al. (2020) propose innovative hedging strategies for
managing environmental exposure within investment portfolios. These strategies are crucial
for enhancing resilience against climate-related shocks.

Moreover, Baker et al. (2018) examine green bonds as a mechanism for financing sustainable
projects, finding that these instruments often trade at a premium due to their perceived
environmental benefits. This suggests that investors are increasingly willing to prioritize
sustainability over purely financial returns.

Policy uncertainty remains a critical factor affecting environmental markets. Hsu et al. (2022)
investigate how fluctuations in environmental policy impact corporate behavior and asset
prices, revealing significant implications for firm valuations and investment decisions. Their
findings underscore the necessity for consistent and resilient policy frameworks to support
renewable energy markets and mitigate climate risks.

2.6 Emerging Trends and Future Research Directions

Recent literature highlights several gaps in our understanding of environmentally friendly

financial markets that warrant further exploration. For instance, integrating high-frequency
 7

environmental data into market models remains underexplored, as does the analysis of cross-
market spillover effects within environmental indices.

Additionally, the impact of policy uncertainty on market behavior represents an important

frontier for research given the rapidly evolving regulatory landscape surrounding climate
finance (Gernaat et al., 2021; Zeppini & Van Den Bergh, 2020; Chew et al., 2021). Future
studies should also consider the role of technological innovations in shaping investment
strategies related to sustainability.

2.7 Research Gaps In Existing Literature

The literature on environmentally friendly financial markets reveals several critical research
gaps that warrant further investigation. First, the integration of high-frequency environmental
data into financial models remains largely unexplored, limiting insights into short-term market
dynamics and investor behavior in response to climate risks. Additionally, there is a need for
comprehensive analyses of cross-market spillover effects within environmental indices, as
most studies have focused on isolated markets, neglecting the interconnectedness of various
segments in sustainable investing. The impact of policy uncertainty on market behavior also
represents a significant frontier; while some research has begun to address this issue, a
systematic analysis of how fluctuations in environmental policies affect investor confidence
and asset pricing is lacking. Longitudinal studies examining the long-term valuation impacts
of climate risks on asset prices are scarce, as existing research often emphasizes short-term
effects or specific events. Furthermore, while investor sentiment has been acknowledged as
influential in renewable energy markets, deeper exploration into the psychological factors
driving this sentiment is needed. The role of technological innovations in shaping investment
strategies related to sustainability is another underexplored area, particularly as new
technologies emerge that could alter risk perceptions. Comparative studies across different
regions are also needed to understand how varying regulatory environments and cultural
contexts influence environmentally friendly financial markets globally. Finally, advancements
in quantitative methodologies that effectively model nonlinear relationships in environmental
finance could enhance the robustness of findings related to climate risk moderation. Addressing
these gaps through targeted research could significantly contribute to a more comprehensive
understanding of the dynamics at play in environmentally friendly financial markets

3. Model and data

3.1 Empirical Model
In the COP27 discussion, there is a strong argument for utilising innovation and green financing
channels to expand the renewable energy industry. The use of carbon credits as a hedging
strategy and possible source of funding for renewable energy projects is gaining popularity
around the world (Demiralay et al., 2022). However, there are two main reasons why the carbon
credits are criticised: first, they are misused in greenwashing; second, they are not universally
acknowledged. Green bonds satisfy the excess demand for climate finance for renewable
energy projects (Alharbi et al., 2023). The renewable energy sector has expanded at a notable
rate thanks to these two green financing sources. According to Cheng et al. (2021), the
 8

continuation of a renewable energy project is now dependent on research and development

related to energy innovation. Thus, the market for renewable energy development is thus
dependent on the anticipated profits from these advances.

Unexpected weather shocks and uncertainties, however, can have an impact on how well the
renewable energy sector performs and who drives it.
Climate risk poses a significant threat to the development of renewable energy projects, hence
diminishing their investment attraction. Global evidence for this element has been shown by
Lee et al. (2021). Wang et al. (2022b) speculate that the volatility seen in the renewable energy
markets may be partially attributed to the unusual climatic risks. Firm-level innovations aiming
at developing renewable energy technologies are similarly impacted by the uncertainty
surrounding climate legislation (Bai et al., 2023). Consequently, taking into account the
possibility of lowering climate risks brought on by natural catastrophes and the ambiguity
surrounding climate policy might produce a more accurate image of how the renewable energy
market might develop in the future. Additionally, there is a correlation between these suggested
moderators and the physical and transition risk categories. To better understand the behaviour
of the renewable energy market and green financing patterns, these components must be
included in the mathematical specification when creating a climate resilient policy framework.
The two climate scenarios might lead to different operations of the green funding channels.
Developing a more thorough comprehension of these differences can help when designing an
investment strategy for renewable energy projects.

The empirical model by assuming the drivers (DR) and moderators (M) is as follows:
RENt = f(DRt, Mt) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1)
DR = {RGB,RCC, INN}
M = {ND,CPU}
REN = Returns on Renewable Energy Prices.
RGB = Returns on Green Bonds.
RCC = Returns on Carbon Credits.
INN = Returns on Energy Innovation.
ND = Natural Disasters.
CPU = Climate Policy Uncertainty.
t = Study period

Moderator M will use DR in an additive and multiplicative manner. It will guarantee that M is
present in REN's elasticity with respect to any DR member. This will assist in identifying the
moderating responsibilities of M's members. The first order differentiation of Eq. (1) may be
used to express the marginal influence of returns on the ith renewable energy generation driver
on returns on renewable energy pricing under the moderation of the jth member of M.

𝜕𝑅𝐸𝑁! /𝜕𝐷𝑅",! = 𝑓 ′ (𝐷𝑅",! , 𝑀$,! ), ∀𝑖, 𝑗 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (2)

The criterion in Eq. (2) suggests that the expected influence of drivers is determined by the
magnitude of M at time t. On the other hand, this magnitude could alter over time and show
aggressive conduct. An X(qxq) matrix, in which the values are distributed along q quantiles
over time t, will be used to depict these effects.

The Xqxq matrix will be estimated by assuming tail dependence between the series. The
conduct of the series in very cold settings lends weight to this notion.
 9

Multivariate Quantile-on-Quantile Regression (m-QQR), as outlined by Sinha et al. (2023),

will be the estimate technique. This approach is a multivariate implementation of the Quantile-
on-Quantile Regression method developed by Sim and Zhou (2015). As suggested by Farzin
and Bond (2006), the m-QQR technique may consider the moderating effects through the use
of marginal impact models. The returns on pricing for renewable energy are assumed to have
a quantile distribution. The ρth quantile of returns on pricing for renewable energy, as a
function of returns on drivers and moderators of renewable energy generation, is described as
follows:

𝑅𝐸𝑁! = 𝛽 ! (𝐷𝑅! ) + ∑% 𝛼 & (𝑀! ) + ∑% 𝛾 & (𝐷𝑅! ∗ 𝑀! ) + 𝜖 & . . . . . . . . . . . . . . . . . .

(3)

In this case, stochastic error is taken to be i.i.d. [≈(0,1)] of null ρ-

quantile. REN displays returns on renewable energy prices, DR
displays the matrix of returns on renewable energy generating
drivers, and M displays the matrix of m moderators. The inclusion of
moderation in Eq. (3) looks like this:

𝑅𝐸𝑁! = (𝛽 & + 8 𝛾 & 𝑀! )𝐷𝑅! + 8 𝛼 & (𝑀! ) + 𝜖 &

% %
& & &
𝑅𝐸𝑁! = 𝛿 (𝐷𝑅! ) + ∑% 𝛼 (𝑀! ) + 𝜖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (4)

Because it is unknown how moderation may affect the DR, δρ() is left undetermined in Eq. (4).
Eq. (4) must be examined for DRφ in order to evaluate the impact of the φth quantile of returns
on renewable energy generating drivers (DRφ) on the ρth quantile of REN. Thus, δρ()'s first
order Taylor expansion may be expressed as follows:

𝛿& (𝐷𝑅! ) ≈ 𝛿& (𝐷𝑅' ) + 𝛿& (𝐷𝑅( ) (𝐷𝑅! − 𝐷𝑅( ) . . . . . . . . . . . . . . . . . . . . . . . . . . (5)

While DRφ is solely indexed in φ, δρ (DRφ) and δρ′ (DRφ) are indexed in both ρ and φ.
Similarly, it is also shown by the first order Taylor expansion of 𝛾 & 𝑀! that γρ(Mφ) and γρ′
(Mφ) are indexed in ρ and φ. If βρ and γρ(Mt) are nested functions, then δρ(DRφ) and δρ′
(DRφ) may be expressed as δ0(ρ, φ) and δ1(ρ, φ). Changing them out in Equation (4) yields:

𝑅𝐸𝑁! = (𝛽 & + ∑% 𝛾 & 𝑀! )𝐷𝑅! + ∑% 𝛼 & (𝑀! ) + 𝜖 & . . . . . . . . . . . . . . . . . . . . . . . . .

(6)

δ0 and δ1 are indexed in ρ and φ, and the (*) component of Eq. (6) indicates the ρth quantile
of REN, based on the φth quantile of DR. As a result, (*) illustrates the entire dependency
between REN and DR. Additionally, Eq. (4)'s (*) component accounts for the consequences of
unexplained moderations. In this way, Eq. (6) may be written as:
 10

𝑅𝐸𝑁! = 𝛽0 (𝜌, 𝜙) + 𝛽1 (𝜌, 𝜙)(𝐷𝑅! − 𝐷𝑅" ) + / {𝛾0 (𝜌, 𝜙) + 𝛾1 (𝜌, 𝜙)(𝑀! 𝐷𝑅! − 𝑀" 𝐷𝑅" )}
#

+ / 𝛼(𝜌)(𝑀! ) + 𝜖 $
#
. . . . . . . . . . . . . . . . . . . . (7)

Here, the (*) part of Eq. (7) captures the mitigated influence of DR on REN. To assess Equation
(7), the optimisation problem may now be stated as follows:

𝑚𝑖𝑛 % /
𝜉" [𝑅𝐸𝑁# − 𝛽$ − 𝛽1 (𝐷𝑅 / & {𝛾0 + 𝛾1 (𝑀/ /
& &
# − 𝐷𝑅 ) − % # 𝐷𝑅# − 𝑀 𝐷𝑅 )}
! '

;# ) − 𝜌
𝐹! (𝐷𝑅
+% 𝛼(𝜌)(𝑀# )]𝐾( )
𝑥
'
. . . . . . . . . . . . . . . .(8)

̂DRt, ̂DRφ, M̂tDRt, and M̂φDRφ stand for the estimated values of DRt, DRφ, MtDRt, and
MφDRφ, respectively. ξρ needs to be solved in order to obtain the regression coefficients.
Gaussian Kernel K() is created to provide the weightage of ˂DRφ throughout its bandwidth
neighbourhood given the local influence of the φth quantile of DR. The following is a depiction
of the difference between ̂DRt and ̂DRφ:

= 1
∑)*+1 = @
𝐹) (𝐷𝑅! ) = )
𝐼(𝐷𝑅* < 𝐷𝑅! ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (9)

3.2 Description of Data

The S&P Global Clean Energy Index returns serve as the basis for measuring the returns on
renewable energy pricing, green bonds, carbon credits, and energy innovation. The Green Bond
Index, the WilderHill New Energy Global Innovation Index, and the IHS MARKIT Global
Carbon Credit, in that order. The whole damage caused by a natural catastrophe is expressed
in US dollars (Centre for Research on the Epidemiology of Disasters, 2023). The statistics have
been accessible throughout the whole period of the USA's unfavourable weather occurrences.
To generate the daily data for this variable, the overall damage that happened during the event
has been evenly dispersed across its duration. For instance, the daily damage is determined to
be D/X (in dollar terms) if a natural catastrophe has an impact that lasts for X days and the
overall damage incurred is D (in dollar terms). This assessment is predicated on the idea that a
disaster's aftermath might get worse or better. Moreover, it depends on the geographical and
structural features unique to that province or region. As a result, the impact's distributive nature
will never be the same for two disasters. If the effects have a uniform distribution, it could be
feasible to average them across time. The reason for not normalising the distribution to assign
tail values is that there is not enough data available to determine the impact's extremes for
different locations over a specific period of time. Lastly, the degree of ambiguity surrounding
 11

climate policy is measured using the Climate Policy ambiguity Index (Gavriilidis, 2021). The
daily statistics for the United States of America (USA) as of August 1, 2014 to July 6, 2023.

The lack of homoskedasticity or normalcy in the model parameters is shown by the rejection
of the null hypothesis in the Jarque-Bera and ARCH-LM tests. Results for kurtosis and
skewness point to the possibility of asymmetrical characteristics in the variables. The idea that
the variables have leptokurtotic distributions is confirmed by comparing them to the kernel
density distribution displayed in Figure 1. The time-varying correlation and distributional
characteristics of the variables are displayed in Figs. 2-3. The correlation estimates shift
directions at regular intervals, which may indicate a tail dependency among the variables, as
shown by the dynamic conditional correlation values in Fig. 3 and the 30-day rolling window
correlation estimates in Fig. 2. It facilitates the use of quantile-based techniques.

CleanEnerg World_Sus Carbon_Eff Green_Bon MAC cpu_index Total Daily

y_Index tainability_ icient_Inde d_Index (SONX) Damage due
(CLNX) Index x (LCEX) (GBDX) to ND
dependent (SUSX) Carbon Returns
Credit Green
Bond
Mean 0.0003 0.0003 0.0003 0 0.0003 0.0036 221392.6352
Median 0.0003 0.0005 0.0005 0 0.0001 0 10416.6667
SD 0.0152 0.0092 0.0096 0.0038 0.0209 0.094 1204079.725
Variance 14498079832
0.0002 0.0001 0.0001 0 0.0004 0.0088 17
Min -0.1175 -0.1006 -0.0992 -0.0238 -0.139 -0.6109 0
Max 0.1166 0.08 0.0892 0.023 0.1198 2.4304 20000000

Table 1: Basic Descriptive Statistics of all the variables used.

Critical insights regarding the dynamics of environmentally sustainable financial markets as

well as the effect of climate-related variables arise in descriptive statistics. Variations, which
are very scarce between 0.0092 and 0.0152 standard deviations apart, for key indices such as
Clean Energy Index (CLNX), World Sustainability Index (SUSX), and Carbon-Efficient Index
(LCEX), ensure stable and constant values over time. These indices play a vital role while
modeling market behaviors related to moderate climate risk scenarios. Although it is a
fluctuating index with the standard deviation of 0.094 and ranges significantly from -0.6109 to
2.4304. Such a wide range shows the volatile nature along with higher risk-return
characteristics of green bonds. On the contrary, metrics such as Total Daily Damage due to
Natural Disasters are highly volatile with variance up to over 1.4 trillion and peaking at 20
million, meaning that climatic phenomena have very destructive and non-predictable results.

This would, thereby, highlight the importance that climate risk plays in its significant factor
concerning performance affecting environmentally sustainable financial markets. Moreover,
the cpu_index has the highest difference between its mean and median, indicating that it also
contains outliers and requires very sophisticated models to account for such variability in the
data. The findings point toward contrasting dynamics of stability entering into environmental
 12

indices and volatility created by the assumptions of risk and natural disasters. They stress
integrating risk-adjusted models and disaster-resilient strategies in understanding the behavior
of environmentally friendly financial markets under varied climate conditions.

CleanEnerg World_Sus Carbon_Eff Green_Bon MAC cpu_index Total Daily

y_Index tainability_ icient_Inde d_Index (SONX) Damage
(CLNX) Index x (LCEX) (GBDX) due to ND
dependent (SUSX) Carbon Returns
Credit Green
Bond
Skewness -0.1241 -0.9112 -0.8192 -0.0632 -0.0752 11.9363 12.9163
Kurtosis 7.4535 14.8673 16.4278 4.3151 3.9474 258.5169 188.2086
p25 -0.0072 -0.0038 -0.0036 -0.0019 -0.0107 0 0
p75 80555.555
0.0075 0.0049 0.0047 0.002 0.0116 0 6
Jarque-Bera 22279.934 27066.768 6695236.5 3585042.7
5520.8505 5 9 1849.0161 1548.1157 49 68

Table 2 : Distributional properties of all the variables used.

The statistical distributions offer significant insight into marked asymmetries, heavy-tailed
features, and sample realizations of outliers in environment-friendly financial markets. CLNX,
SUSX, and LCEX indices indicate a slightly negatively skewed and normally slightly more
kurtotic behaviour with very few extreme negative values. Indices like SONX, or returns on
green bonds, and Total Daily Damage due to Natural Disasters, show substantial positive
skewness and unusually high kurtosis-concepts that reflect the existence of infrequent
significant outlier values that include extreme returns and devastating economic consequences
from natural disasters. The interquartile ranges (IQRs) for most of the indices are relatively
thin, implying stability in their central distributions, while metrics like Total Daily Damage
show wide dispersions, which reflect variability. High Jarque-Bera statistics show that all
variables involved deviate significantly from normality, so extreme methodology must be
utilised on these variables using GARCH models when estimating the volatility or quantile
regressions for the identification of tail risks and also to disclose non-linear interactions. Such
results unequivocally reinforce the need for climate risk factors and extreme events when
building models for sustainable financial markets.
 13

Figure 1 : Returns of CLNX over time

Figure 2 : Returns of GBDX over time

Figure 3 : Returns of LCEX over time

 14

Figure 4 : Returns of SONX over time

Figure 5 : Returns of SUSX over time

 15

Figure 6 : Total Damage due to ND

4. Results and Discussion

Variable VIF 1/VIF Tolerance R-squared

CleanEnergy_I
ndex (CLNX)
dependent 6.893716819 0.1450596284 0.1450596284 0.8549403716
World_Sustain
ability_Index
(SUSX) 14.91252805 0.06705771127 0.06705771127 0.9329422887
Carbon_Efficie
nt_Index
(LCEX)
Carbon Credit 15.36425034 0.06508615637 0.06508615637 0.9349138436
Green_Bond_I
ndex (GBDX)
Returns Green
Bond 1.145251934 0.8731703223 0.8731703223 0.1268296777

MAC (SONX) 5.657365933 0.1767607066 0.1767607066 0.8232392934

cpu_index 1.016149056 0.9841075916 0.9841075916 0.01589240843
Total Daily
Damage due
to ND 1.018372905 0.9819585686 0.9819585686 0.0180414314
Table 3 : Multicollinearity test results

The multicollinearity analysis indicates that the variables utilized in the model interrelate with
one another at different levels of dependence. The VIF values for SUSX and LCEX are
unusually high at 14.91 and 15.36, thereby reflecting strong multicollinearity, whereby both
variables share significant variance with others in the model (R-squared > 0.93). Similarly, the
Clean Energy Index (CLNX) had a relatively high VIF of 6.89, indicating the presence of some
multicollinearity but to a smaller extent. Variables like Green Bond Index (GBDX), Total Daily
Damage (ND) and cpu_index had relatively low VIF values, indicating their independence in
terms of multicollinearity while providing different information to the model. Tolerance values
were also significantly high for GBDX at 0.87 and for ND at 0.98 further supporting their
independence. The findings lead to the conclusion that while combining SUSX and LCEX,
some kind of careful variable selection or regularization is compulsory for alleviating
multicollinearity and to improve the robustness of the model.
 16

Wu- Wu- Durbin- Durbin-

Variable Hausman_stat Hausman_pval Wu_stat Wu_pval

cpu_index 3.855290283 0.04958934503 2046.411173 0

Total Daily
Damage due
to ND 68.76090593 1.11E-16 2046.064799 0

Both 68.27517795 1.44E-15 2046.625751 0

Table 4 : Multicollinearity test results

The results of the exogeneity test by the Wu-Hausman and Durbin-Wu statistics provide
evidence for significant endogeneity issues concerning the relevant variables. The cpu_index
has a Wu-Hausman statistic of 3.86 together with its p-value of 0.0496, thus providing slight
evidence against the null hypothesis of exogeneity. Now, Total Daily Damage due to Natural
Disasters shows a much stronger denial of exogeneity as the Wu-Hausman statistic is 68.76
and the p-value is 1.11E-16. When taken both variables together, test results are extremely
significant since the Wu-Hausman statistic is 68.28 and the p-value is 1.44E-15, which supports
the theory of endogeneity further. The Durbin-Wu statistics support these conclusions as
having exceedingly high values along with p-values of 0, which suggests that the variables are
probably affected by unobserved elements that are associated with the error term.

Variable_Pair Quantile Beta sup_stat CV1 CV5 CV10

(CLNX)
dependent_v 0.089835822 0.045829081 0.029588630 0.023683433
s_(SUSX) 0.1 1.056824513 93 39 37 92
(CLNX)
dependent_v 0.078211174 0.039812789 0.023744267 0.017258617
s_(SUSX) 0.5 1.019630124 02 32 86 04
(CLNX)
dependent_v 0.066625964 0.045313146 0.030190774 0.024114203
s_(SUSX) 0.9 1.128158122 7 83 88 21
(CLNX)
dependent_v 0.089664975 0.044869039 0.029126282 0.023107506
s_(LCEX) 0.1 1.065007547 12 96 8 71
(CLNX)
dependent_v 0.078148908 0.038164841 0.022493208 0.017104911
s_(LCEX) 0.5 1.046192742 52 56 55 27
(CLNX)
dependent_v 0.066887574 0.045170888 0.029504600 0.023668492
s_(LCEX) 0.9 1.146659753 79 11 38 5
 17

(CLNX)
dependent_v 0.870139015 0.131240204 0.059753390 0.038139506 0.031135535
s_(GBDX) 0.1 1 2 68 52 57
(CLNX)
dependent_v 0.839760653 0.115917289 0.051934298 0.030153668 0.021915158
s_(GBDX) 0.5 5 9 53 99 39
(CLNX)
dependent_v 0.998997293 0.124138474 0.057293262 0.038068610 0.030980646
s_(GBDX) 0.9 1 3 35 35 98
(CLNX)
dependent_v
s_MAC 0.653230952 0.045924881 0.024640779 0.015335956
(SONX) 0.1 6 57 29 0.018086669 47
(CLNX)
dependent_v
s_MAC 0.632977867 0.046668980 0.020116429 0.013165512 0.010605365
(SONX) 0.5 1 41 01 9 92
(CLNX)
dependent_v
s_MAC 0.651206123 0.053830765 0.025737193 0.018276811 0.015393468
(SONX) 0.9 3 67 5 23 69

Table 5 : Cointegration test results

The results for the cointegration test of the Clean Energy Index (CLNX) and the other key
variables by means of quantile regression analysis describe their long-term relationships. For
CLNX and SUSX, the beta coefficients are consistently near or above 1 across quantiles,
indicating a strong long-run positive relationship. The sup_stat values generally decline with
increasing quantiles; therefore, at every significance level, they remain above CV, indicating
cointegration. Correspondingly, for CLNX and LCEX, beta coefficients keep constant in a
range close to 1, and sup_stat has remarkably higher values of cointegration, notably in lower
and middle quantiles. In contrast, the interaction between CLNX and GBDX shows smaller
beta coefficients (around about 0.87 up to 0.99), meaning a weaker but still significant long-
term relationship and sup_stat values are certainly larger than the critical values on all levels.
Finally, CLNX and MAC (SONX) demonstrate the lowest beta coefficients (approximately
0.63 to 0.65), suggesting a comparatively weaker long-term association; however, the sup_stat
values continue to signify cointegration across the majority of quantiles.
 18

Carbon_Effici Green_Bond
ent_Index CleanEnergy _Index
(LCEX) _Index (GBDX) Total Daily World_Sustai
Carbon (CLNX) Returns Damage due nability_Inde
Quantile Credit dependent Green Bond MAC (SONX) to ND x (SUSX)
0.002325428
0.05 1.83150466 2.58222462 1.223284197 4.497794787 324 2.756517033
0.1 3.82568018 4.145604685 1.860654278 4.114345667 7.037350199 4.587782194

0.15 2.938657922 6.643359252 2.43074528 4.383132265 9.779195601 5.921202752

0.2 4.21659709 7.007898639 3.502580321 4.99886944 63.12653457 6.153334094
0.25 5.590887349 8.041270635 4.123131917 7.303389855 - 6.671015393

0.3 6.336956809 7.581834338 3.425994611 7.196652778 - 7.772788138

0.35 7.247254556 8.206554281 3.125831436 7.27172319 - 7.642926049

0.4 7.063204481 9.42163421 2.930653654 7.015577535 - 6.730069989

0.45 5.801573465 9.296067377 3.02935008 5.701002458 - 6.746213955

0.5 5.680196988 8.375742052 3.844572379 5.101477189 - 6.142113334

0.55 6.049902591 7.291748655 4.137380158 4.792625993 - 4.841600376

0.6 3.99568008 6.235259878 4.171644805 4.18583127 - 3.753596369

0.65 2.259923709 5.091195995 4.010604642 3.693401803 - 2.402199877

0.794921698
0.7 2.061034786 4.295260103 4.117843804 3.839129999 - 4
0.633500060 0.473497330
0.75 2 3.62928343 4.210607817 4.015093056 - 5
0.680905637 0.212446911
0.8 6 3.783717532 3.970256705 4.178510436 - 2
0.675944713 0.342247481
0.85 1 3.205307237 3.329634757 3.198113899 - 4
0.591782051 0.269131836
0.9 1 1.883869793 3.685333493 1.941426858 - 3
-
- 0.438273782
0.95 1.186467092 1.407000367 1.790082284 2.238269122 - 2
Table 6 : Quantile Autoregression
 19

4.1 Cross Quantile Correlation estimate

Fig 7: Clean Energy Index (CLNX) vs. Carbon Efficient Index (LCEX)

The correlations between the Clean Energy Index (CLNX) and the Carbon Efficient Index (LCEX) (Fig
1) carbon credit exhibit a generally positive pattern, with values typically ranging between 0.05 and 0.5
across various quantile combinations. This indicates a relatively consistent positive relationship
between clean energy investments and carbon-efficient credits. The highest correlations are observed
in the upper quantiles, particularly between the 0.75 and 0.95 range of both indices, with values reaching
up to 0.46. This suggests that under favorable market conditions, there is a stronger positive correlation,
meaning both assets tend to perform well simultaneously. Conversely, the lower quantiles (0.05 to 0.25)
show weaker correlations, especially in the bottom-left region of the heatmap, indicating that carbon-
efficient credits may not always offer significant downside protection for clean energy investments in
less favorable market conditions.
 20

Fig 8: Clean Energy Index (CLNX) vs. Green Bond Index(GBDX)

The correlations between the Clean Energy Index and the Green Bond Index (Fig 2) are relatively low
across most quantile combinations, typically ranging between 0.00 and 0.18. Negative correlations
emerge in some lower quantiles, suggesting that under certain market conditions, returns on green bonds
might move inversely with clean energy investments. The highest correlations are observed in the
middle-upper quantiles, around 0.75 to 0.85, indicating a slight positive relationship in relatively
favorable market conditions. This implies that green bonds provide only a weak hedge against clean
energy investments, with occasional negative correlations in lower quantiles. While they may offer
some stability in favorable conditions, green bonds may not serve effectively as counter-cyclical
investments for clean energy during market downturns.
 21

Fig 9: Clean Energy Index (CLNX) vs. Solar Index (SONX/MAC)

The heatmap (Fig 3) reveals strong positive correlations across most quantiles, with values peaking
around 0.71, particularly in the lower to middle quantiles. In the lower quantiles (approximately 0.05
to 0.3 for both indices), the notably high correlation indicates that during market downturns, returns for
clean energy and solar markets tend to move in the same direction. While the correlation decreases
slightly in the upper quantiles, it remains relatively strong, showing that even in favorable market
conditions, clean energy and solar energy investments are closely aligned. This suggests that solar
energy plays a significant role within clean energy portfolios, particularly during uncertain market
conditions, highlighting its importance in aligning with broader clean energy trends.
 22

Fig 10: Clean Energy Index (CLNX) vs. World Sustainability Index (SUSX)

The heatmap shows a moderate correlation between the Clean Energy Index (CLNX) and the World
Sustainability Index (SUSX) across most quantile pairs. The highest correlations occur around the
middle quantiles, approximately 0.25 to 0.75 for both indices, with values reaching up to 0.45,
indicating a strong positive association between clean energy and sustainability performance during
moderate market conditions. In contrast, correlations drop to as low as 0.05 in the lower and upper
quantiles, suggesting that extreme market conditions weaken the relationship between the two indices.
This may be due to market volatility or sector-specific risks in clean energy that do not align directly
with broader sustainability trends. For investors, this indicates that while sustainable investments are
moderately linked to clean energy, they may be less effective as hedges during market extremes.
 23

4.2 Partial Cross Quantile Correlation estimates

Partial Cross Quantile Correlational analysis is performed in order to filter out the effects
caused by other external variables, in this case Climate Policy Uncertainty and Damage
by Natural Disasters. This allows us to examine the correlation between the two indices in
a much clearer lens.

Fig 11 : Clean Energy Index (CLNX) vs. World Sustainability Index (SUSX) Carbon Credit

The partial cross quantile correlation coefficients between the Clean Energy Index (CLNX)
and the World Sustainability Index (SUSX) show a varied picture. We see that similar quantile
levelled quantiles of both indices have move together in the positive direction, i.e the high,mid
and low quantiles of CLENX and SUSX, all show significantly positive correlation coefficients,
ranging from 0.25 to 0.31, while high quantiles of CLENX and low quantiles of SUSX as well as low
quantiles of CLENX and high quantiles of SUSX both show negative correlation coefficients,
ranging from -.17 to -.14.
 24

Fig 12 : Clean Energy Index (CLNX) vs. Carbon Efficient Index (LCEX) Carbon Credit

The correlations between the Clean Energy Index (CLNX) and the Carbon Efficient Index (LCEX)
carbon credit, while controlling for other variables, generally show low to moderate positive
correlations in favorable market conditions, with increasing values observed in the upper
quantiles. The highest correlations occur in the 0.85–0.95 quantile range, where values reach up
to 0.36. This indicates a positive relationship in favorable market scenarios, suggesting that clean
energy and carbon-efficient investments tend to move in the same direction during periods of
high returns.

In the central quantiles (0.30 to 0.65), the correlations vary widely, ranging from negative to
positive coefficients between -0.14 and 0.27. The peak correlation is observed around the central
quantile, gradually diminishing as values move away from this midpoint. This mixed range
highlights varying market dynamics in moderate return scenarios.
In contrast, the lower quantiles (0.05 to 0.25) exhibit negative correlations, with values ranging
from -0.10 to -0.18. This pattern suggests that during downturns, clean energy investments and
carbon-efficient credits often move in opposite directions, reflecting a potential divergence in
investment performance under adverse market conditions.
 25

Fig 13: Clean Energy Index (CLNX) vs. Green Bond Index (GBDX) Returns (Controlling for Other Variables)

The partial correlations between the Clean Energy Index (CLNX) and the Green Bond Index
(GBDX) are generally very low across the quantiles, highlighting a weak relationship and
suggesting that green bonds are relatively independent of clean energy investments. In the upper
quantiles (0.75 to 0.95), correlations show a slight increase, peaking at approximately 0.11,
indicating a mild positive association under favorable market conditions, though still weak. The
central quantiles (0.3 to 0.65) display a range of values from -0.03 to 0.11, with a pattern of peaks
and valleys at the edges and a gradual plateau in the middle, reflecting only a slight positive
correlation that occasionally turns negative during uncertain times. In the lower quantiles (0.05
to 0.25), correlations are close to zero or slightly negative, particularly between 0.05 and 0.15,
where values range from -0.05 to -0.07, suggesting that green bonds and clean energy
investments are largely uncorrelated or may even diverge in poor market conditions.
 26

Fig 14 : Clean Energy Index (CLNX) vs. Solar Energy Index (SONX) Returns

The partial correlations between the Clean Energy Index (CLNX) and the Solar Energy Index
(SONX) generally exhibit moderate to high positive correlations under favorable market
conditions, with increasing values in the upper quantiles. In the upper quantiles (0.75 to 0.95),
the correlations reach their highest in the 0.85–0.95 range, peaking at 0.68, indicating a strong
positive relationship where clean energy and solar energy investments tend to move in the
same direction during periods of high returns. In the central quantiles (0.3 to 0.65), the
correlations range mostly positively from 0.0 to 0.61, with the highest values at the center
gradually decreasing toward the edges. However, in the lower quantiles (0.05 to 0.25),
correlations are slight, ranging from -0.04 to -0.18, suggesting minimal or negligible
dependence between the indices during market downturns, with occasional divergence.

4.3 Quantile on Quantile Regression (QQR)

 27

Figure 15 : QQR Contour Plot of GBDX (Green Bonds Index)

The QQR output for GBDX (Figure 15) demonstrates the impact of green bond returns on
renewable energy prices across various quantile combinations. The results reveal that the
impact is mostly positive across the spectrum, suggesting that investments in green bonds
generally align with rising renewable energy prices. Interestingly, the influence of green
bond returns declines along their own quantiles, which may indicate that higher green bond
returns correlate less with renewable energy price increases. Conversely, the effect
strengthens along the quantiles of renewable energy prices, particularly between the 80th
to 95th percentiles. This implies a growing preference among investors for green bonds
as a financing instrument for renewable energy projects as market interest in renewable
energy prices grows. The rising trajectory of green bond returns in higher quantiles of
renewable energy prices indicates investor confidence in green bonds as a stable,
impactful choice for environmentally sustainable projects. However, the diminishing effect
at the upper quantiles of green bonds points to an untapped potential, possibly due to
regulatory constraints and increasing scrutiny over greenwashing practices. Studies,
including those by Smetana (2023) and Pronina and Freke (2019), underscore the
necessity for transparency in the green bond market to maximize its potential and build
investor trust.
 28

Figure 16 : QQR Contour Plot of LCEX (Carbon Credit Index)

The QQR output for LCEX (Figure 16) shows the influence of carbon credit returns on
renewable energy prices. The findings suggest a diminishing positive impact from lower to
higher quantile combinations, highlighting that the early stages of renewable energy
investments are better supported by carbon credits. Specifically, the effect remains
positive within the 5th to 55th quantiles for both carbon credits and renewable energy
prices, but it turns negative at higher quantiles, indicating a shift in influence. The negative
impact in higher quantiles suggests that as renewable energy projects progress, the
relevance of carbon credits diminishes, potentially due to market saturation or concerns
over the legitimacy of carbon credit schemes. Instances of greenwashing and double
carbon offsetting have been condemned by the United Nations, as seen at the COP27
summit, and the European Parliament has also pushed for restrictions on carbon offset
schemes. This scrutiny likely impacts investor confidence in carbon credits for supporting
renewable energy projects. Nevertheless, carbon credits play a foundational role in the
early stages of renewable energy adoption, but their efficacy wanes as projects mature
and more sophisticated financial instruments, like green bonds, take precedence.
 29

Figure 17 : QQR Contour Plot of SONX (Solar Index)

The QQR analysis for SONX (Figure 17) reveals the relationship between solar index
returns and renewable energy prices. The effect of solar returns on renewable energy
prices appears mostly positive in the lower quantiles, particularly between the 5th to 50th
quantiles. However, as both variables increase to higher quantiles, the impact becomes
negative, indicating a possible saturation or diminishing marginal returns from solar
investments in driving renewable energy prices at higher levels. This trend may be
attributed to the high initial cost and investment associated with solar infrastructure, which
can yield substantial returns initially but may not sustain the same level of impact as the
market matures. Solar index investments continue to hold value in lower quantiles,
promoting renewable energy adoption. However, beyond a certain threshold, factors such
as operational inefficiencies or market competition may reduce solar energy's influence on
overall renewable energy prices. This trend underscores the need for continuous
innovation and policy support to sustain the solar sector's contribution to renewable energy
goals.
 30

Figure 18 : QQR Contour Plot of SUSX (Sustainability Index)

The QQR output for SUSX(Figure 18) shows the effect of sustainability index returns on
renewable energy prices. The results indicate a strong, positive influence from the 5th to
the 50th quantiles, suggesting that sustainability-related investments are highly supportive
of renewable energy prices at the initial stages. However, as both sustainability index
returns and renewable energy prices reach higher quantiles, the influence diminishes, with
some negative impacts observable at the highest quantiles. This trend suggests that while
sustainability investments play a crucial role in the early development of renewable energy
projects, their relative impact on driving renewable energy prices becomes less significant
as the market matures. This might be due to factors like market saturation or limited
scalability of some sustainability-focused initiatives. The trend also highlights the evolving
financial landscape, where investors may shift focus from generalized sustainability
indices to more specialized instruments, such as green bonds, as the renewable energy
sector progresses. The findings underscore the need for innovative financial solutions to
maintain growth in the renewable sector and caution against over-reliance on traditional
sustainability metrics alone.

4.4 Multivariate Quantile-On-Quantile Regression

The transition to renewable energy is increasingly essential in addressing climate change,
reducing carbon emissions, and promoting sustainable development. Financing mechanisms
such as carbon credits, green bonds, solar energy investments, and sustainability-focused
initiatives play crucial roles in supporting the growth of the renewable energy sector.
Understanding the effectiveness and resilience of these mechanisms under varying economic
 31

and environmental conditions is key for investors and policymakers alike. Using a multivariate
quantile-on-quantile regression (m-QQR) approach, this study examines the interactions
between renewable energy returns and these diverse financing tools. By analyzing how
returns on renewable energy are influenced across different quantile combinations,
particularly in the face of climate policy uncertainty and natural disasters, we gain insights into
the viability and limitations of these financial instruments in promoting renewable energy. The
results provide valuable guidance on the strategic use of these mechanisms to foster long-
term growth and resilience in renewable energy investments.

Figure 19 : QoQ contour plot of carbon credits (cc)

In Figure 19, which presents the multivariate quantile-on-quantile (m-QQR) contour plot
carbon credits (cc) against renewable energy returns under the control of natural
disasters, we observe varied impacts across different quantiles. The coefficients show a
range of both positive and negative values, spanning from approximately -0.12 to 0.05.
This suggests that, while carbon credits have some influence on renewable energy returns
under the impact of natural disasters, their effect is relatively inconsistent and moderate.
The fluctuating values around zero indicate that carbon credits may not provide a
significant hedge for renewable energy returns when natural disasters are a major risk
factor. This could be due to carbon credits being perceived more as a regulatory tool than
a stable funding source for renewable energy projects, especially under adverse
environmental conditions.
 32

Figure 20 : QoQ contour plot of green bonds (gb)

In contrast, Figure 20, which represents the m-QQR contour plot for green bonds (gb) with
natural disasters as the control, displays a more stable and consistently positive influence
across higher quantiles of renewable energy returns. The coefficients predominantly range
between 0.24 and 0.34, with a notable upward trend from the 65th to the 95th quantile. This
trend implies that green bonds provide a stronger positive impact on renewable energy returns
in scenarios where natural disasters are a significant factor. This positive relationship
highlights green bonds as a resilient investment tool for renewable energy, potentially due to
their structured repayment and increasing market penetration. The stable influence of green
bonds under adverse conditions reinforces their role as a viable investment vehicle, supporting
renewable energy projects even when natural disasters pose substantial risks.

Figure 21 : QoQ contour plot of solar index (SO)

Figure 21, which shows the multivariate quantile-on-quantile (m-QQR) contour plot of the
Solar Index (so) against renewable energy returns under the control of natural
disasters, we observe a positive impact across various quantiles of renewable energy returns.
The regression coefficients range from approximately -0.53 to 0.59, indicating that the Solar
Index has a fairly stable influence on renewable energy returns even under adverse conditions
 33

caused by natural disasters. However, this influence is not uniform across all quantiles. While
the impact generally strengthens at higher quantiles, there is some fluctuation in the mid-range
quantiles, possibly reflecting the volatility in solar energy production and investment in
response to unpredictable climatic events. The results suggest that solar energy investments,
represented by the Solar Index, offer a resilient yet slightly variable hedge for renewable
energy returns when natural disaster risks are present. The overall upward trend in the higher
quantiles indicates that solar energy could play a crucial role in promoting renewable energy
generation, though this potential may be somewhat tempered by environmental disruptions.

Figure 22 : QoQ contour plot of Sustainability Index (sus)

In Figure 22, which depicts the m-QQR contour plot of the Sustainability Index (sus)
against renewable energy returns, with natural disasters as the control, a somewhat
contrasting trend is visible. The coefficients range from around -0.34 to 0.48, with a gradual
increase observed across higher quantiles of renewable energy returns. This positive trend
suggests that the Sustainability Index exhibits a meaningful and strengthening impact on
renewable energy returns, especially at higher quantiles, even amidst the challenges posed
by natural disasters. However, the influence remains less intense at lower quantiles, which
might be due to the limited integration of sustainable practices in certain segments of the
renewable energy market. As the quantiles increase, the effect of the Sustainability Index
becomes more pronounced, indicating that sustainable investments are likely better suited to
mitigate risks and enhance returns in high-risk scenarios. This aligns with findings by Chen et
al. (2023) and JPMorgan Chase’s 2022 Climate Report, which highlighted the importance of
sustainability as a hedge against climate-related financial risks. Such evidence suggests that
sustainability-linked investments, while challenged by climate risks, hold strong potential as
reliable instruments for supporting renewable energy returns under extreme conditions.
 34

Carbon Credits and Their Impact on Renewable Energy Returns

Figure 23 : QoQ contour plot of Carbon Credits(cc)

The multivariate quantile-on-quantile regression (m-QQR) analysis between returns on

renewable energy prices and carbon credits (Figure 23) reveals that the influence of
carbon credits on renewable energy returns diminishes across different quantile levels,
particularly under climate policy uncertainty and natural disasters. The analysis shows
a negative or low positive association across most quantiles, especially at higher quantiles
where the impact becomes weakest. This suggests that carbon credits might not be a
sustainable financing mechanism for renewable energy projects under adverse climatic
conditions. At higher quantiles (around the 75th percentile and above), the regression
coefficients for carbon credits turn negative or near zero, indicating limited financial viability
and effectiveness. This trend underscores the challenges of relying solely on carbon credits
for renewable energy financing, as their sensitivity to climate risk factors reduces their
influence on renewable energy returns

Green Bonds as a Resilient Financing Mechanism for Renewable Energy

 35

Figure 24 : QoQ contour plot of green bonds(gb)

The m-QQR results between returns on renewable energy prices and green bonds
(Figure 24) indicate that green bonds have a more resilient and positive influence on
renewable energy returns compared to carbon credits, especially under conditions of
climate policy uncertainty and natural disasters. Green bonds maintain a steady or slightly
increasing positive association with renewable energy returns across most quantiles. Notably,
the regression coefficients for green bonds show an upward trend at higher quantiles (from
the 35th to 95th percentiles), indicating their viability as a sustainable financing mechanism.
Unlike carbon credits, green bonds do not experience a sharp decline in effectiveness at
elevated quantiles. Instead, the positive influence strengthens, suggesting that green bonds
can support renewable energy projects even in challenging market conditions. This resilience
is likely due to the structured payment schedules of green bonds, which provide stability and
hedge against climate risks

Solar Index as a Stable Contributor to Renewable Energy Performance

 36

Fig 15 : QoQ contour plot of Solar index(so)

Figure 25 : QoQ contour plot of Solar Index(So)

The m-QQR analysis of the Solar Index (So) in figure 25 shows a consistent positive
association with renewable energy returns across different quantiles, underscoring
its role as a stable contributor to renewable energy performance. The contour plot
highlights that even at lower quantiles (below the 5th percentile) of renewable energy
returns, the Solar Index maintains a positive, though slightly subdued, influence. As we
move to higher quantiles (especially above the 75th percentile), the regression coefficients
increase significantly, reaching around 0.57. This upward trend suggests that solar energy
is an essential component for renewable energy generation, especially when market
returns are high, reflecting favorable conditions or periods of strong solar energy
production. The growing market share and technological advancements in solar energy
likely enhance its production efficiency, further amplifying its impact on renewable energy
returns.

Sustainability Index and Its Quantile-Dependent Impact on Renewable Energy

 37

Fig 16 : QoQ contour plot of sustainability index(sus)

Figure 26 : QoQ contour plot of Sustainability Index (Sus)

The analysis of the Sustainability Index (Sus) in Firgure 26 reveals a quantile-dependent

impact on renewable energy returns, with a relatively low influence in low-return conditions
and a stronger effect as returns increase. At lower quantiles (approximately from the 5th
to the 35th), the regression coefficients for sustainability investments are minimal,
suggesting limited impact on renewable energy returns in less favorable market scenarios.
However, from the 35th to the 65th quantiles, the influence of sustainability investments
grows, indicating a moderate positive impact on renewable energy projects under stable
market conditions. At higher quantiles (from the 65th percentile onward), the influence of
the Sustainability Index peaks, with coefficients reaching values as high as 0.475. This
trend highlights that sustainability-focused investments are particularly valuable in high-
return markets, where they enhance the financial attractiveness of renewable energy
projects, likely due to increased investor demand for sustainable investments and the
rising importance of ESG factors.

5. Conclusion
This study offers a nuanced understanding of how different financial tools—green bonds,
carbon credits, solar energy investments, and sustainability-linked investments—affect
renewable energy returns under varying conditions, especially when considering the impact
of climate risks. The findings indicate that green bonds stand out as a reliable source of
support, consistently strengthening renewable energy returns, particularly at higher quantiles.
 38

Their stability across these quantiles, even under challenging conditions like natural disasters,
suggests that green bonds are a robust mechanism for financing renewable energy. This
stability is likely due to their structured repayment schedules and increasing acceptance in
financial markets, providing investors with confidence and dependability during volatile times
(Fang et al., 2022; Nakajima et al., 2023).

In contrast, carbon credits show a more limited and inconsistent impact on renewable energy
returns. With coefficients that are often close to zero, carbon credits serve as a relatively
uncertain hedge, likely due to their primary role as regulatory mechanisms rather than stable
sources of funding. This limited resilience may reduce their effectiveness in directly supporting
renewable energy, especially in high-risk environmental scenarios (Demiralay et al., 2022;
Blaufelder et al., 2021).

The Solar Index, representing solar energy investments, generally shows a positive influence,
particularly at higher quantiles. However, mid-range quantiles reveal some variability,
potentially due to the inherent volatility of solar energy production, which can be heavily
influenced by climatic factors. In favorable conditions, solar investments demonstrate
resilience, pointing to their potential as a viable hedging tool. However, this effectiveness may
be dampened by environmental unpredictability (Bloomfield et al., 2021; Auffhammer et al.,
2017).

Lastly, the Sustainability Index has relatively low influence in low-return conditions but gains
strength as returns increase, reaching peak impact at higher quantiles. This trend suggests
that sustainability-linked investments are especially valuable in high-return markets where
there is a greater demand for ESG-compliant assets. Their pronounced effect in such
conditions positions sustainability-linked investments as important tools for supporting
renewable energy projects when return potential is strong (Jankovic et al., 2022; Zhang, 2023).

These results underscore the need to align renewable energy financing strategies with specific
economic and environmental conditions. While green bonds demonstrate consistent reliability
across diverse scenarios, both carbon credits and solar investments may benefit from targeted
policy interventions to enhance their effectiveness in hedging against climate risks.

6. Policy Implications
A comprehensive and adaptable approach to renewable energy financing is essential to
address the diverse market conditions and challenges faced by the sector. Green bonds, with
their proven stability and increasing positive impact across various quantiles of renewable
energy returns, emerge as a key financing tool. Policymakers can enhance their effectiveness
by introducing tax incentives for green bond investors, especially those supporting renewable
energy projects. These incentives can lower the cost of capital for renewable energy firms,
encouraging sustained growth and investment in this sector. Public-private partnerships
further bolster green bond issuance by directing funds toward critical renewable energy
infrastructure, ensuring consistent development and expansion (Wilkes & Furness, 2022).
 39

Although carbon credits have historically demonstrated limited and inconsistent effects, they
hold potential when integrated with specific renewable energy projects or combined with
complementary financial instruments. For instance, allowing renewable energy firms to
redeem carbon credits as project funding would expand their role beyond regulatory
compliance. Bundling carbon credits with green bonds or other sustainable finance tools could
provide a more direct and reliable mechanism for channeling investments into renewable
energy, particularly in high-risk scenarios where traditional financing methods may fall short
(World Bank, 2022a; Lee et al., 2021).

In markets with high returns, promoting sustainability-focused investments such as ESG funds
can generate significant value. The Sustainability Index shows increased influence in these
conditions, highlighting the potential for growth when firms adopt and adhere to enhanced
ESG reporting standards. Financial incentives for companies meeting high ESG criteria can
attract a larger pool of investors, particularly in renewable energy markets, creating a virtuous
cycle of sustainable practices, improved financial returns, and sectoral expansion. This
approach not only benefits investors but also contributes to the broader goal of transitioning
toward a greener economy (Chen et al., 2023; International Energy Agency, 2021b).

Given the fluctuating impact of solar energy returns across different market conditions, tailored
risk mitigation strategies are crucial to stabilize investments in this segment. Policies such as
specialized insurance schemes for renewable energy projects and climate-risk funds that
provide financial support during natural disasters can help offset risks associated with volatile
environmental conditions. These measures create a safety net for investors, encouraging
them to support solar energy projects despite potential uncertainties, and solidifying solar’s
role as a cornerstone of the renewable energy portfolio (Baker & Adu-Bonnah, 2008; Zhao et
al., 2022).

The use of quantile-based analysis underscores the importance of aligning renewable energy
financing policies with specific market and environmental conditions. This analytical approach
allows policymakers to design targeted strategies, such as prioritizing green bonds in low-
return conditions where stability is critical and emphasizing sustainability-linked investments
in high-return markets where profitability can drive sector growth. By tailoring policies to
diverse scenarios, governments can ensure that renewable energy financing remains effective
and resilient, adapting dynamically to evolving market dynamics and climate risks, ultimately
fostering a sustainable transition to cleaner energy systems (Kanamura, 2020).

7. References
1. Pastor, L., Stambaugh, R. F., & Taylor, L. A. (2021). Sustainable investing in
equilibrium. Journal of Financial Economics, 142(2), 550–571.
[Link]
2. Bolton, P., Kacperczyk, M., & Kacperczyk, M. (2020). Do Investors Care About
Carbon Risk. Social Science Research Network.
[Link]
3. Chava, S. (2014). Environmental externalities and cost of capital. Management
Science, 60(9), 2223–2247. [Link]
 40

4. Giglio, S., Maggiori, M., Rao, K., Stroebel, J., & Weber, A. (2021). Climate Change
and Long-Run Discount Rates: Evidence from Real Estate. Review of Financial
Studies, 34(8), 3527–3571. [Link]
5. Hong, H., Karolyi, G. A., & Scheinkman, J. A. (2019). Climate finance. Review of
Financial Studies, 33(3), 1011–1023. [Link]
6. Krueger, P., Sautner, Z., & Starks, L. T. (2019). The importance of climate risks for
institutional investors. Review of Financial Studies, 33(3), 1067–1111.
[Link]
7. Perera, A. T. D., Wickremasinghe, W. L. D., & Perera, A. S. (2020). Physical climate
risks and renewable energy infrastructure resilience. Energy Policy, 137, 111083.
[Link]
8. Wang, C., Chen, Y., & Liu, J. (2022). Policy uncertainty and its effects on renewable
energy investor confidence. Energy Reports, 8, 505–518.
[Link]
9. Demiralay, S., Taner, O., & Yıldırım, A. B. (2023). Green finance and resilience
during the COVID-19 pandemic. Environmental Economics and Policy Studies, 25(4),
123–135. [Link]
10. Shah, N. A., Syed, S. H., & Butt, T. (2018). Oil prices, renewable investments, and
economic impacts. Renewable Energy, 131, 114–123.
[Link]
11. Organization for Economic Cooperation and Development (OECD). (2021). Building
a green and resilient recovery. [Link]
12. Bowden, N., & Payne, J. E. (2015). The interaction between oil prices and renewable
energy investments in emerging economies. Journal of Policy Modeling, 37(5), 804-
822. [Link]
13. Ahmed, F., Wang, Y., & Mukherjee, S. (2024). Climate risk and investor behavior:
The role of media and public attention in renewable energy markets. *Energy
Economics, 89*, 104872. [Link]
14. Sim, N., & Zhou, H. (2015). Oil prices, US stock return, and the dependence between
their quantiles. Journal of Banking & Finance, 55, 1–8.
[Link]
15. Esquivias, M. A., Sugiharti, L., Rohmawati, H., Rojas, O., & Sethi, N. (2022). Nexus
between Technological Innovation, Renewable Energy, and Human Capital on the
Environmental Sustainability in Emerging Asian Economies: A Panel Quantile
Regression Approach. Energies, 15(7), 2451. [Link]
16. Engle, R. F., Giglio, S., Lee, H., Kelly, B. T., & Stroebel, J. (2019). Hedging climate
change news. SSRN Electronic Journal. [Link]
17. Baker, M., Bergstresser, D., Serafeim, G., & Wurgler, J. (2018). Financing the
response to climate change: the pricing and ownership of U.S. green bonds.
[Link]
18. Hsu, P., Li, K., & Tsou, C. (2023). The pollution premium. The Journal of Finance,
78(3), 1343–1392. [Link]
19. Gernaat, D. E. H. J., Bogaart, P. W., & Doelman, J. C. (2021). Policy resilience and
green technologies in climate policies. *Global Environmental Change, 68*, 102262.
[Link]
20. Zeppini, P., & Van Den Bergh, J. C. J. M. (2020). Green finance instruments and
carbon pricing mechanisms. *Environmental Innovation and Societal Transitions, 37*,
123–135. [Link]
 41

21. Chew, T., Hui, C., & Shum, M. (2021). Climate risk in national energy policy: A
systematic review of emerging economies. *Renewable and Sustainable Energy
Reviews, 142*, 111753. [Link]
22. Naseer Z.A., Khan M.A.& Ali R.(2024). The impact of climate change risk on stock
market fluctuations: An empirical analysis. Journal of Financial Stability,54(1),
Article No.:101000.
23. Bolton, P., Despres, M., Pereira da Silva, L., Samama, F., & Svartzman, R.
(2020). The green swan: Central banking and financial stability in the age of
climate change. Bank for International Settlements.