The effect of collateral-based monetary policy on green finance: Evidence from China

Abstract

Research background: Green finance is crucial to accelerating China’s green transition, and its growth depends largely on the corresponding monetary policy. To increase financial institutions’ support for the green economy, China’s central bank has recognized green bonds as eligible collateral for monetary policy tools since June 1, 2018. Purpose of the article: In this context, we investigate the effect of collateral-based monetary policy (CBMP) on green finance (GF) in China by utilizing a quasi-natural experiment approach. Methods: Using the Propensity Score Matching-Difference in Difference (PSM-DID) method and daily bond trading data, we investigated the impact of CBMP on the cost and availability of green finance. In further analysis, we employed bond issuance data and listed company data to examine the spillover effects of CBMP and its influence on the real economy. Findings & value added: Our results suggest that CBMP, in the secondary market, actively stimulated the growth of GF by reducing green bond spreads and expanding their financing scale. Furthermore, this beneficial outcome is particularly obvious for low-grade green bonds, bonds issued by state-owned enterprises (SOEs), and in regions with stringent environmental regulations and high government green attention. Particularly, we have also found that there exists a spillover effect across markets, i.e., endowing collateral eligibility to green bonds in the secondary market, can reduce bonds’ spreads and increase bonds’ financing scale in the primary market. Finally, we have found that CBMP effectively incentivizes corporate green behavior rather than “greenwashing”. Our findings suggest that China should further optimize CBMP, focus more on non-SOEs green finance difficulties, and strengthen local governments’ green attention and implementation capacity.

Full text

OeconomiA
copernicana
Volume 15 Issue 4 December 2024
p-ISSN 2083-1277, e-ISSN 2353-1827
www.oeconomia.pl
Copyright © Instytut Badań Gospodarczych / Institute of Economic Research (Poland)
This is an Open Access article distributed under the terms of the Creative Commons Attribu-
tion License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
ORIGINAL ARTICLE
Citation: Wang, P., & Lu, Z. (2024). The effect of collateral-based monetary policy on green
finance：Evidence from China. Oeconomia Copernicana, 15(4), 1223–1262.
https://doi.org/10.24136/oc.3001
Contact to corresponding author: Zheng Lu, zlu@scu.edu.cn
Article history: Received: 5.03.2024; Accepted: 10.12.2024; Published online: 30.12.2024
Penghao Wang
Zhongnan University of Economics and Law, China
orcid.org/0009-0006-2313-1769
Zheng Lu
Sichuan University, China
orcid.org/0000-0003-1363-2747
The effect of collateral-based monetary policy on green finance:
Evidence from China
JEL Classification: G12; G18; G21
Keywords: green finance; collateral-based monetary policy; financing costs; financing availability
Abstract
Research background: Green finance is crucial to accelerating China’s green transition, and its
growth depends largely on the corresponding monetary policy. To increase financial institu-
tions’ support for the green economy, China’s central bank has recognized green bonds as
eligible collateral for monetary policy tools since June 1, 2018.
Purpose of the article: In this context, we investigate the effect of collateral-based monetary
policy (CBMP) on green finance (GF) in China by utilizing a quasi-natural experiment ap-
proach.
Methods: Using the Propensity Score Matching-Difference in Difference (PSM-DID) method
and daily bond trading data, we investigated the impact of CBMP on the cost and availability

Oeconomia Copernicana, 15(4), 1223–1262
1224
of green finance. In further analysis, we employed bond issuance data and listed company
data to examine the spillover effects of CBMP and its influence on the real economy.
Findings & value added: Our results suggest that CBMP, in the secondary market, actively
stimulated the growth of GF by reducing green bond spreads and expanding their financing
scale. Furthermore, this beneficial outcome is particularly obvious for low-grade green bonds,
bonds issued by state-owned enterprises (SOEs), and in regions with stringent environmental
regulations and high government green attention. Particularly, we have also found that there
exists a spillover effect across markets, i.e., endowing collateral eligibility to green bonds in
the secondary market, can reduce bonds’ spreads and increase bonds’ financing scale in the
primary market. Finally, we have found that CBMP effectively incentivizes corporate green
behavior rather than “greenwashing”. Our findings suggest that China should further opti-
mize CBMP, focus more on non-SOEs green finance difficulties, and strengthen local govern-
ments’ green attention and implementation capacity.
Introduction
Since October 2017, China’s economic strategy has shifted from pursuing
high-speed growth to achieving high-quality development. This strategic
transformation particularly emphasizes the sustainability and aims to
achieve coordinated development of economic growth, social well-being
enhancement and environmental protection. To accomplish its purpose, the
Chinese government has adopted numerous initiatives, one of which is to
facilitate the development of green finance (GF). GF performs its function
by allocating financial resources towards environmentally friendly and
low-carbon projects, which requires the assistance from monetary policy
(Nawaz et al., 2021; Batrancea et al., 2020). However, the inability of tradi-
tional monetary policy to proactively allocate credit resources has resulted
in a large green financing gap (Batrancea et al., 2020; Debrah et al., 2022).
Particularly in China, traditional monetary policy directed large funds to
construction, infrastructure and manufacturing, resulting in other indus-
tries, such as green and environmental protection industries, suffering from
financing constraints and high costs (Fang et al., 2023). According to a re-
search report, the total demand for green funds in China was 2.1 trillion
yuan in 2018, but the total supply was only 1.3 trillion yuan, indicating that
there remains a large supply-demand gap in the green finance market
(CBNEditor, 2019). Different from traditional monetary policy, collateral-
based monetary policy (CBMP) requires commercial banks to provide eli-
gible collateral. Changes in the types of acceptable collateral affect both
financial institutions’ access to funding from the People’s Bank of China
(PBOC) and their preferred types of eligible collateral. As a result, the

Oeconomia Copernicana, 15(4), 1223–1262
1225
PBOC can influence the flow of funds and resource allocation in the finan-
cial market through appropriate collateral policies. On June 1, 2018, the
PBOC expanded the scope of collateral for the Medium-Term Lending Fa-
cilities (MLF) in the interbank market and innovatively used green bonds
as collateral for MLF. Therefore, it is quite valuable to examine empirically
whether the CBMP can address the supply-demand gap and support the
development of GF.
However, existing evidence on this topic is surprisingly sparse (Fang et
al., 2023; Dikau &Volz, 2021; Fang et al., 2020; Liu et al., 2023). The reasons
may be twofold. On the one hand, GF is a new concept, and historically,
central banks in most countries have prioritized supporting government
economic goals while ignoring the sustainability and greening of the econ-
omy (Dikau & Volz, 2021). On the other hand, lacking policy counterfactual
and causal identification strategies makes us unable to separate the effects
of CBMP from other macro-control policies introduced simultaneously
(Fang et al., 2020).
By examining the Chinese bond market’s response to the inclusion of
green bonds as collateral in MLF, our works may address above important
gap in knowledge. To support the green economic development, the PBOC
incorporated green bonds into the collateral pool for the MLF on June 1st,
2018. After the expansion, commercial banks can acquire liquidity by
pledging green bonds to the PBOC. The expansion of MLF eligible collat-
eral provides us with a rare and ideal natural experiment in which we can
investigate the effect of CBMP utilizing the Propensity Score Matching-
Difference in Difference (PSM-DID) approach. To be concrete, we focus on
corporate bonds and investigate the CBMP’s differential influence on green
bonds (treated bonds) and non-green bonds (control bonds). The analysis
proceeds in four steps: (1) using bond transaction data, we determine the
causal impacts of CBMP on GF by comparing changes in spreads and fi-
nancing scale between treated bonds and control bonds before and after the
expansion; (2) we explored the heterogeneous effects of CBMP under dif-
ferent conditions; (3) using bond issuance data, we determine the primary
market’s behavior after the CBMP goes into force to investigate the spillo-
ver effects of the CBMP; and (4) we investigate the impact of CBMP on
green finance practices by evaluating the green behaviors of companies
that receive green funds.
Compared with the existing literature, this study has three marginal
contributions. Firstly, this study provides new evidence on the determi-

Oeconomia Copernicana, 15(4), 1223–1262
1226
nants of GF. Existing empirical studies showed that macroeconomic condi-
tions (Campiglio, 2016; Chang et al., 2021), central bank functions (Campi-
glio et al., 2018; Aloui et al., 2023), policy uncertainty (Hafner et al., 2020),
and traditional monetary policy (Chen et al., 2019; Campiglio, 2016; De-
salegn et al., 2022) had important impacts on GF. In this paper, we instead
investigate the CBMP, an important and previously unexplored factor, and
find its association with GF. In addition, because of the endogeneity of
asset pledging, determining the causal impact of CBMP on GF is difficult.
However, we effectively overcome this endogeneity issue by constructing
a quasi-natural experiment using a Chinese monetary policy shock.
Secondly, this article enriches the research on collateral’s influence on
asset prices and financial stability (Chen et al., 2023; Fang et al., 2023; Ny-
borg, 2017). There is controversy surrounding existing studies about collat-
eral value. On the one hand, collateral can increase asset prices by alleviat-
ing frictions arising from incomplete contracts and insufficient information
(Gârleanu & Pedersen, 2011), or by increasing investors’ willingness to pay
(Broer& Kero, 2021). On the other hand, banks tend to pledge the lowest
quality collateral to the central bank and allocate highly liquid assets in the
market (Fecht et al., 2016), which undermines market discipline and leads
to distortions in the financial markets and overall economy (Choi et al.,
2021). Our findings add to this conventional wisdom by demonstrating that
using green bonds as collateral for MLF significantly promotes the ad-
vancement of GF.
Thirdly, we present unique insights into the macroeconomic effects of
monetary policy. Our study is one of the few analyses that considers the
impact of unconventional monetary policies on the green economy. In ex-
isting literature, monetary policy has primarily been considered as an in-
fluence on economic stability (Sui et al., 2022; Grilli et al., 2020), employ-
ment (Bahaj et al., 2022; Liu et al., 2023), asset pricing (Schmidt, 2020), and
financing costs (Geng et al., 2024). In spite of this, monetary policy is largely
understudied when it comes to its impact on GF. It is surprising given the
importance of GF in achieving green economic transformation and high-
quality economic growth. Aiming to fill this gap, this paper further investi-
gates the impact of CBMP on GF and corporate environmental governance
behavior.
The remainder of this article is presented as follows. Section 2 discusses
the implementation backdrop and primary features of CBMP in China.
Section 3 reviews the literature. Section 4 introduces the research design

Oeconomia Copernicana, 15(4), 1223–1262
1227
and empirical strategy. Section 5 presents the baseline findings, robustness
tests and heterogeneity analysis. Section 6 serves as a further study, dis-
cussing the spillover effects of CBMP and its impact on the real economy.
Section 7 is the discussion of the empirical results. The final section of the
study presents our conclusions.
Policy background
Prior to the outbreak of the global financial crisis, central banks around the
world typically regulated the scale and pace of economic activity by using
the two primary tools of traditional monetary policy, namely interest rates
and money supply. During the crisis, banks’ asset quality deteriorated and
the transmission from short-term to long-term interest rates was impeded,
resulting in the failure of the traditional monetary policy. To address the
deficiencies of traditional monetary policies, developed countries have
adopted unconventional structural monetary policies and coped with nega-
tive shocks by expanding the scope of eligible collateral. For example, in
2008, the European Central Bank lowered the minimum rating requirement
for eligible collateral from A- to BBB-. Similarly, the Federal Reserve ex-
panded the scope of collateral to include auto loans, commercial real estate
loans, and credit card loans. Developed countries only use collateral-based
monetary policies as temporary relief tools. However, after drawing on the
experience of developed countries, the PBOC used CBMP as a conventional
tool, and successively launched the standing lending facilities (SLF),
pledged supplementary lending (PSL), and medium-term lending facilities
(MLF). A common feature of these policy tools is that the relevant banks
are required to pledge qualified assets. A common feature of these policy
tools is that the relevant banks are required to pledge qualified assets. Once
the central bank includes specific assets as eligible collateral, commercial
banks can use these assets as collateral to obtain funding from the central
bank while enjoying relatively lower financing costs. As a result, central
banks can influence commercial banks’ asset allocation and provide sup-
port to specific sectors by incentivizing investment and lending in these
types of assets. Thus, these monetary instruments not only ensure the asset
security of the PBOC, but also represent the government’s goal of reshap-
ing the economy.

Oeconomia Copernicana, 15(4), 1223–1262
1228
Since the late 1970s, scale-driven economic growth has resulted in ex-
cessive energy consumption and pollution emissions. In October 2017, the
Chinese government proposed that economic development strategies
should shift from pursuing speed to pursuing quality in order to resolve
the contradiction between economy and environment. The Chinese gov-
ernment also stated that promoting green development and strengthening
ecological civilization are the best ways to achieve high-quality growth.
However, the green and environmental protection industries in China are
confronted with challenges of funding difficulties and high financing costs
in recent years (Liu et al., 2022). To increase support for the green economy
and other related fields, the PBOC changed the classification of collateral
for MLF loans made by financial institutions on June 1, 2018. The expan-
sion enables green bonds and green credits with ratings above AA as col-
lateral for MLF and gives green bonds priority over other corporate bonds.
This expansion represents three characteristics: First, this is the first time
that the PBOC has used high-quality assets as collateral. Prior to June 1,
2018, the central bank only accepted collateral for MLF loans that was rela-
tively high credit-rated and steady, such as Treasury bonds (Fang et al.,
2020); Second, the purpose of expansion is not to increase liquidity, but to
optimize capital investment; Third, it makes no distinction between old and
new bonds, allowing all qualified bonds to be utilized as collateral. As the
eligibility of assets for collateral is determined by the central bank, this
reform provides an excellent ʺquasi-naturalʺ experiment for our analysis,
enabling us to effectively overcome the influence of other factors that inter-
fere with causal relationships and omitted variables by using the differ-
ence-in-differences method.
Literature review and hypotheses development
The duties of central banks and GF
Green transformation goals require coordinated efforts from the green fi-
nancial system. In response to the substantial green finance gap, some re-
searchers have delved into the role of central banks in improving the green
finance market（Campiglio et al., 2018). Some hold that the economic green
transformation falls under the purview of the market and the government.
The central bank’s primary responsibility is to maintain price stability, and

Oeconomia Copernicana, 15(4), 1223–1262
1229
excessive intervention in the green economic transformation may disrupt
the market’s normal operations. This situation has led to a widespread lack
of green responsibilities in central banks and significant controversies sur-
rounding monetary policies related to green development. Dikau and Volz
(2021) used data from a global survey to analyze the responsibilities of
central banks and found that only 12% of all central banks surveyed had
clear sustainability responsibilities. However, the opposing view contends
that a neutral monetary policy exhibits a carbon bias, that is, it allocates
more resources to high-pollution and high-energy-consuming industries,
impeding the economic green transformation (Desalegn et al., 2022). There-
fore, scholars holding this view advocate the “Green Central Bank” model,
implying the integration of green development support into the central
bank’s duties and the promotion of a green structural monetary policy.
From the perspective of actual implementation results, the European cen-
tral banks’ green policies successfully encourage investment in sustainable
projects and firms, encouraging the growth of green finance markets (Aloui
et al., 2023). According to Chen et al. (2019), low borrowing rate can incen-
tivize borrowers to choose green projects, hence driving the growth of GF
and green innovation. Similarly, Campiglio (2016) believes that financing
policies that consider interest rates will enhance finance availability for
environmentally friendly industries. Therefore, we propose the following
hypothesis:
H1: Including green bonds in the scope of eligible collateral for MLF will help
promote the development of GF.
CBMP and the Financing Scale of GF
Traditional monetary policy works through credit mechanisms, whereas
non-traditional monetary policy operates through collateral mechanisms.
In a market environment with complete information, traditional monetary
policy can achieve both aggregate and structural equilibrium. However,
under the constraints of information asymmetry and bank credit controls,
especially after the global economic recession in 2008, traditional monetary
policy cannot automatically regulate the allocation of credit resources
(Vespignani, 2015; Borio & Zhu, 2012). Therefore, in the green finance field,
where information asymmetry is prominent, it is difficult to fill the huge
green financing gap by relying on traditional monetary policy. In addition,

Oeconomia Copernicana, 15(4), 1223–1262
1230
traditional monetary policy mainly focuses on the aggregate regulation of
the macroeconomy, which makes it difficult to solve economic structural
problems (Vespignani, 2015). In this context, major economies commonly
opt to utilize CBMP to allocate financial resources to green and low-carbon
industries by adjusting the scope of collateral (McConnell et al., 2022).
In terms of mechanism, CBMP affects the financing scale of green fi-
nance via two pathways. First, the scarcity mechanism. Banks can obtain
liquidity by pledging collateral to the central bank, making eligible collat-
eral scarce and thereby increasing demand for such assets from financial
institutions (DʹAmico et al., 2018). Second, the credit enhancement mecha-
nism. On the one hand, by pledging qualified collateral, commercial banks
can obtain liquidity from the central bank, implying that the central bank’s
issuance of currency, as a national credit action, is linked to the eligible
collateral. Therefore, CBMP can enhance the credibility of eligible collat-
eral. On the other hand, the pledging mechanism serves as a form of risk
guarantee for the collateral to a certain extent, thereby it can achieve the
credit enhancement effect. The increased credibility of green bonds relative
to other assets has attracted more capital providers to finance green pro-
jects, thereby increasing the financing scale of GF. Therefore, we propose
the following hypothesis:
H2: Including green bonds in the scope of eligible collateral for MLF facilitates
increasing the financing scale of GF.
CBMP and the Financing cost of GF
CBMP can reduce the financing cost of GF through signal transmission
mechanism, social recognition mechanism and asset pledge value mecha-
nism. First, the information asymmetry between bondholders and issuers
results in creditors demanding higher risk premiums when pricing bonds
(Zou et al., 2023; Palea & Drogo, 2020). Incorporating assets into the collat-
eral pool can transmit positive signals to investors, which is beneficial for
improving their attitudes and suppressing their demands for risk premi-
ums, thereby lowering the cost of collateral (Tang et al., 2023; Zhang et al.,
2021). Second, social recognition is an important factor affecting investors’
investment decisions (Green & Jame, 2013). Including green assets in the
collateral pool can convey a positive signal to investors regarding green
bonds’ future development potential. This aids in increasing the market

Oeconomia Copernicana, 15(4), 1223–1262
1231
acceptance of green bonds, raising their relative price, and consequently
reducing the financing cost (Chen et al., 2025). Thirdly, the pledged value of
bond assets is an important way to reduce the financing cost of green
bonds. Existing literature on asset pricing suggests that increasing the col-
lateral ratio of financial assets raises asset prices (Chen et al., 2018). For
financial asset issuers, an increase in asset prices implies a decline in fi-
nancing costs (Ashcraft et al., 2011). The central bank’s endowment of col-
lateral eligibility to green bonds has effectively enhanced the pledge rate of
green bonds. Studies demonstrate that the central bankʹs expansion of the
eligible collateral scope reduces the credit spreads of newly incorporated
collateral bonds, subsequently lowering their financing costs (Fang et al.,
2020; Geng et al., 2024). Therefore, we propose the following hypothesis:
H3: Including green bonds in the scope of eligible collateral for MLF facilitates
reducing the financing cost of GF.
Although a large number of scholars have explored the economic effects
of CBMP, most of them are purely theoretical analyses (Dikau & Volz, 2021;
Choi et al., 2021; Ashcraft et al., 2011; Akomea-Frimpong et al., 2022). This is
mainly due to the absence of appropriate policy counterfactuals and causal
relationship identification strategies. In addition, China’s monetary policy
environment is complex, with multiple monetary policy tools operating in
parallel, and their policy effects influencing each other. This poses a chal-
lenge to separating the CBMP policy effects from other macroeconomic
control policies during the same period. However, on June 1, 2018, the cen-
tral bank’s inclusion of green bonds in the eligible collateral scope of the
MLF generated a favorable quasi-natural experiment for this research. This
allows us to employ the difference-in-differences model (DID) to resolve
the causal identification conundrums that have long troubled the existing
literature. In this study, we take the PBOC’s inclusion of green bonds in the
MLF collateral range as a policy shock. Using the DID method, we investi-
gate the impact of CBMP on the scale and cost of green financial financing.
Based on this, the model design of this paper can effectively overcome the
deficiencies of existing literature in endogenous treatment.

Oeconomia Copernicana, 15(4), 1223–1262
1232
Methods
Model setting
Benchmark regression model
Referring to Fang et al. (2023), we construct the following two models to
examine the impact of the CBMP on GF.


 

   



  

 



 

 

(1)


 

   



  

  

 !"

 

 

(2)
where 

represents the bond credit spread of bond # on day , which
measures the financing costs of GF.

indicates the financing scale
of bond # on day , measuring the availability of GF. 

is an identifica-
tion symbol for bonds that takes the value 1 if the bond # is green; other-
wise, it takes the value 0. The CBMP policy shock is measured by a time
dummy variable, 

, which is defined as 1 following CBMP implemen-
tation (on and after June 1, 2018) and 0 before CBMP implementation.
 refers to a collection of control variables associated with issuer
characteristics and the macroeconomic environment, including enterprise
size (), asset-liability ratio ($), current ratio (%), the quarterly GDP
growth rate (&') and the monthly M2 growth rate (()). () absorbs the
day-by-day time-fixing effect, which is necessary in the *+* model. 

are
the individual fixed effects, controlling the impact of bond characteristics,
including bond rating () and remaining bond maturity (). 

represents the random error term.
The models’ major parameters of interest are 

and 

, which directly
estimate the influence of CBMP on green finance. If 

is significantly nega-
tive, it means CBMP can reduce green financing cost. If 

is greater than 0,
it means that CBMP can effectively increase GF availability.
Parallel trend model
A valid estimation of the DID model requires that the treatment and
control groups meet the parallel trend assumption. We employ the event
study method to examine whether the differences in spreads and financing

Oeconomia Copernicana, 15(4), 1223–1262
1233
scales between green and ordinary bonds exhibit comparable trends before
the CBMP takes effect. Referring to Guo and Zhang(2023), the models are
constructed as follows:


,
-
-./
-.012-30
'

-


 

 4 !"

 

 

(3)
5

,
-
-./
-.012-30
'

-


 

 4 !"

 

 

(4)
where we divided the entire sample period into 12 sub-intervals, each cor-
responding to a one-month window. The dummy variable '

-
returns 1 if
date  is within subperiod 6, otherwise it returns 0. Fig. 1 plots the point
estimate of 
-
, 6 7 {−5, ... -2, 0, ..., 6} and the corresponding 95% confidence
intervals. The definitions of other variables remain consistent with those of
the benchmark regression model.
Further analysis model
As the secondary market is not directly involved in financing activities,
in the further analysis section, we utilize data from the primary bond mar-
ket to analyze the impact of CBMP on green bond financing activities. Fol-
lowing Fang et al. (2023), models employed in this section are constructed
as follows:
8

 

   



  

 

9

 

 

(5)
:(;<=

 

   



  

  

9

 

 

 (6)
8

represents the cost of green finance, described by the difference be-
tween the issuance yields of corporate bonds and government bonds.
:(;<=

indicates the logarithm of the face value of bond# issued on date
, indicating the availability of green finance in the primary market. The
variables 

and 

have the same definitions as in the benchmark
regression. > represents many control variables related to the macroeco-
nomic environment, including the monthly M2 growth rate (?)) and quar-
terly GDP growth rate (&'). () absorbs the day-by-day time-fixing effect,
which is necessary in the *+* model. 

controls for the potential impact of
bond rating () and remaining bond maturity () on the regres-
sion results. 

represents the random error term.

Oeconomia Copernicana, 15(4), 1223–1262
1234
In the further analysis section, we also examine whether CBMP im-
proves corporate green performance or promotes corporate “greenwash-
ing” behavior. The models are constructed as follows:
%#

 

   



  

 

9

 

 @

 

(7)
A#$

 

   



  

 

9

 

 @

 

(8)
#$

 

   



  

 

9

 

 @

 

(9)
The subscripts # and  represent the enterprise and quarter respectively.
We measure the corporate green performance from three aspects: environ-
mental information disclosure level (%#

), green innovation level
(A#$

), and pollution control investment level (#$

). The val-
ue of %#

is 3 when environmental information is revealed in all
three documents (annual report, CSR report, and environmental report). It
is 2 if disclosed in any two of these documents, and 1 if disclosed in any
one of them. A#$

is equal to the logarithm of the number of green
patents granted plus 1. #$

is measured by the logarithm of the
company’s investment in pollution control. 

and 

have the same
definitions as in the benchmark regression. We selected five control varia-
bles, represented by >, including: (1) the corporate leverage ($) is defined
as the ratio of total liabilities to total assets; (2) enterprise size () is
the logarithm of total assets; (3) current ratio (%) is calculated as current
assets divided by current liabilities; (4) profitability (), measured by net
profit divided by total assets; (5) asset structure (B#C#), the proportion
of fixed assets to total assets. 

and @

represent individual and quarter
fixed effects. 

is the random error term.
In addition, to verify whether enterprises conceal their low green in-
vestment behavior by acquiring green enterprises with higher scores in the
environmental dimension, this paper constructs the following model:
DA

 

   

(:

 

9

 

 @

 

(10)
DA

denotes the environmental dimension (E) score within the envi-
ronmental, social, and governance (ESG) framework of enterprise # in quar-
ter. (:

is a dummy variable indicating whether the enterprise has en-
gaged in merger and acquisition activities (M&A). If enterprise # has en-

Oeconomia Copernicana, 15(4), 1223–1262
1235
gaged in M&A at time, then (:

is 1 starting from quarter . Otherwise,
it is 0. The definitions of other variables remain consistent with model (7).
Variables
Explained variable
The explained variable is the level of GF, which is measured from the
perspectives of cost and accessibility in this study. Specifically, the cost of
GF() is assessed using the bond credit spread, indicated by the yield dif-
ference between corporate and government bonds. It is worth noting that
the reference government bonds have the same issuance and maturity
dates as corporate bonds. The availability of GF () is measured by
the bond financing scale, calculated as the logarithm of the daily total trad-
ing volume of bonds.
Explanatory variables
The interaction term of treat and time dummy variables (

E 

)
is our core explanatory variable. 

is used to identify whether corpo-
rate bonds are influenced by the CBMP. Specifically, if bond # belongs to a
green bond, we classify it as the treatment group and define 

 F;
otherwise, we classify it as the control group and define 

 G. 

equals 1 for the observations after June 1, 2018; otherwise, it equals 0.
Control variables
Referring to the existing literature (Fang et al., 2023; Cheng et al., 2024;
Chen et al., 2025), we introduce seven control variables in our regression
analysis to account for potential confounding factors that may influence the
relationship between the CBMP and GF. Two of the variables are related to
bond characteristics (BC): (1) bond rating () is 1 if the bond is rated
AA, 2 if AA+, and 3 if AAA; (2) remaining bond maturity (). Three
variables are related to issuer characteristics (IC): (3) enterprise size
() is the logarithm of total assets at the end of 2017; (4) the debt-
servicing capacity ($) is defined as the asset-liability ratio of the bond
issuer at the end of 2017; (5) current ratio (%) is the ratio of current assets to
current liabilities at the end of 2017. Two of the variables are related to

Oeconomia Copernicana, 15(4), 1223–1262
1236
macroeconomic environment (ME): (6) GDP growth rate (&'). It is equal
to the previous quarter’s GDP of the province where the bond issuer is
located divided by the GDP of the same period last year, minus 1; (7) broad
money supply growth rate (()). It is calculated as the previous monthʹs
broad money supply divided by the amount of the same period last year
minus 1.
Data source
We utilize daily trading data of corporate bonds in Chinese secondary
market to estimate the models. Bond transaction data comes from the
CSMAR database, including bond yield to maturity, bond issuance loca-
tion, credit ratings, etc.; GDP growth rate data comes from the National
Bureau of Statistics, and M2 data comes from the official website of PBOC.
Our sample period spans from January 1st, 2018 to December 31st, cov-
ering the entire year. Since bonds rated under AA may be associated with
higher default risk and greater volatility, we focus solely on bonds rated at
or above AA to ensure that our findings are not confounded by these high-
er-risk bonds (Fang et al., 2020). Subsequently, we excluded samples with
maturities shorter than one year because, within such a short duration,
even minor price fluctuations have the potential to induce significant devi-
ations in implied yields (Fang et al., 2020). In addition, a bond will be re-
moved from our sample if it does not trade on a given day (Fang et al.,
2020).
In the end, we obtain a total of 50,800 sample observations, with 285
green bond observations accounting for 0.5% of the total. Nevertheless, the
given data has some limitations, particularly the relatively small sample
size of the treatment group when compared to the control group. Using the
PSM method to match samples helps improve the sample distribution of
distinct groups, thereby overcoming data bias and confounding variables.
First, we estimate the probability of a bond being a green bond based on
observable characteristics:, $, %, , , &', and ().
The control group for each green bond contains ten ordinary bonds in the
same trading month whose predicted score is closest to the predicted score
of the green bond. Second, we keep observations with non-empty weights.
Finally, we obtain 1345 observations, including 1093 ordinary bonds and

Oeconomia Copernicana, 15(4), 1223–1262
1237
252 green bonds, accounting for 23% of the sample1. Table 1 reports the
descriptive statistics of the variables.
In the further analysis section, data on bond issuance information is ob-
tained from the China Bond Information Network. According to the sample
selection criteria outlined in the baseline regression, we only retain newly
issued bonds with a rating of at least AA in 2018.
The research subject for further analysis is Chinese A-share listed com-
panies from 2016 to 2019. Choosing such an interval is for the following
considerations: first, taking 2016 as the starting year is because, in August
of the year, China issued the ʺGuidance on Building a Green Financial Sys-
temʺ, making it the first country to implement a systematic green financial
policy framework; second, taking 2019 as the ending year is to exclude the
potential interference of the COVID-19 pandemic. The dataset utilized in
this section comes from the CSMAR database.
Empirical method
This study evaluates the impact of including green bonds in the MLF
collateral scope on the scale and cost of green finance financing. Given that
MLF and green finance may influence each other, this may cause the results
to be biased due to endogeneity. The DID model is one of the most com-
monly used econometric methods for estimating the impact of public poli-
cies (Guo & Zhang, 2023; Fang et al., 2023; Liu et al., 2023). It effectively
overcomes the endogeneity problem caused by mutual causality via com-
paring the differences in the impact of policy shocks on the treatment and
control group (Ma et al., 2024). Therefore, this study uses the DID model to
analyze the impact of CBMP on GF based on a quasi-natural experiment in
which the PBOC includes green bonds in the scope of eligible collateral for
MLF. Furthermore, to alleviate the interference of the sample size disparity
between the treatment and control groups on the research outcomes, this
study employs the propensity score matching (PSM) method proposed by
Rosenbaum and Rubin (2023) to identify control group enterprises with the
closest corporate characteristics to those in the treatment group. To this
end, we use the PSM-DID model to analyze the impact of CBMP on GF.
A valid estimation of the DID model requires that the treatment and
control groups meet the parallel trend assumption. Event analysis method
is a common tool for testing the effectiveness of the DID model. Therefore,
1
Some green bonds have less than 10 sample matches.

Oeconomia Copernicana, 15(4), 1223–1262
1238
referring to Guo and Zhang (2023), we utilize event analysis to test whether
there are significant differences between the experimental and control
groups before the shock and the dynamic effects after the policy.
In the robustness test section, this study employs the placebo model.
The placebo test is an idea based on the counterfactual hypothesis. It veri-
fies the authenticity and reliability of the obtained causal relationship by
simulating the situation where there is no real causal connection. With the
help of the placebo test model, we can verify whether the results obtained
in this article are due to random factors or other factors that have not been
considered. Prior to this, several scholars utilized the placebo model to
assess the robustness of benchmark regression results(Liu et al., 2023; Ma et
al., 2024).
Results
Benchmark regression results
The regression results of Eqs. (1) and (2) are presented in Table 2. We see
that when the bond spread () is the explained variable, the coefficient of
   is -0.079, indicating that CBMP effectively reduces bond
spreads. This conclusion is consistent with the results of Fang et al. (2023).
When bond financing scale () is the explained variable, the coeffi-
cient of    is 0.823, demonstrating that the application of CBMP
resulted in an of 0.823 in the financing scale of green bonds relative to other
bonds. The results without control variables in column (1) and (2) are con-
sistent with those in column (3) and (4) respectively. This conclusion is
consistent with the results of Ma et al. (2024), who discovered that after the
implementation of MLF, the loan volume of commercial banks with high
bond holdings increased significantly. The findings in Table 2 show that the
CBMP effectively reduces the costs of GF, increases its accessibility, and
thus promotes green finance development.
Parallel trend
To obtain unbiased estimates, the use of the PSM-DID model requires
parallel trend testing. We employ the event study method to examine
whether the differences in spreads and financing scales between green and

Oeconomia Copernicana, 15(4), 1223–1262
1239
ordinary bonds exhibit comparable trends before the CBMP takes effect.
Fig. 1 shows that prior to the implementation of the CBMP (June 1, 2018),
the average bond spreads and financing scales of green and ordinary bonds
maintained parallel trends. However, after CBMP takes effect, green bond
spreads began to fall much more than non-green bonds, and this trend
persists for five months. Meanwhile, green bonds enjoyed a much larger
financing scale than ordinary bonds, and this trend lasted for three months.
As shown in Fig. 1, before the PBOC adopted the CBMP, bonds with new
collateral requirements exhibited no significant differences in credit
spreads and financing scales compared with other bonds, satisfying the
DID model’s common trend assumption.
Robustness test
We performed the following robustness tests to support our conclu-
sions: changing the matching method, placebo tests and excluding other
policy interferences.
Changing the matching method
In this section, we change the sample matching method for robustness
testing. First, we match each green bond to five ordinary bonds from the
control group that were traded in the same month. In addition, to further
improve the comparability of the samples, we used the same method to
match the samples using a 1:2 ratio. Second, we conduct radius matching
within a caliper of 0.01. Third, we employ kernel matching, and last, we use
the unmatched full sample for regression. As shown in Table 3, under dif-
ferent matching methods, the significance levels and regression coefficients
of H" and 5IJ are basically consistent. Overall, results in Table 3 indicate
the reliability of the benchmark regression results.
Placebo tests
In this section, we conduct two placebo tests. One is to respond to the
robustness concerns resulted by non-policy shocks. Green bonds differ
from non-green bonds not only because green bonds are a sort of collateral
for MLF, but they also have additional advantages that attract investors.
For instance, green bonds may have better reputation and lower risk com-

Oeconomia Copernicana, 15(4), 1223–1262
1240
pared to ordinary bonds. Secured bonds, which provide certain collateral
or guarantees, can minimize investors’ risk and enhance investment securi-
ty, making them more popular with investors. For this reason, the findings
in benchmark regression may be resulted by the inherent characteristics of
green bonds rather than the policy shock. To address this concern, we carry
out placebo event studies in which secured bonds are defined as the treat-
ment group. Then, we implement Eqs. (3) and (4), where we define


 F if and only if the corporate bond is a secured bond. As illustrat-
ed in Fig. 2, prior to the implementation of CBMP, significant differences
were observed between secured bonds and other bonds in terms of both
financing cost and financing availability. Following the policy shock, the
placebo treatment group had a higher spread than the control group, but
their financing scale was smaller. This conclusion runs counter to the
benchmark regression results, indicating that bond characteristics such as
high reputation and low risk are unlikely to explain our findings.
The other is to respond to the robustness concerns resulted by unob-
servable factors. To further reduce the interference of grouping errors on
the empirical results and strengthen the causal relationship between acbmp
and GF, we randomly divided the samples into treatment groups and con-
trol groups. Afterwards, the baseline models (1) and (2) were estimated
1,000 times. The probability density distribution of coefficients is depicted
in Fig. 3. True    estimations are provided as vertical dotted
lines for comparison. The results reveal that all  coefficients are greater
than -0.079 (true estimation), while nearly all  regression coeffi-
cients are smaller than 0.823 (true estimation). Thus, the growth of GF is
not due to other unobservable factors, and the benchmark regression re-
sults remain robust.
Excluding other policy interferences
In 2018, seven provincial-level administrative regions in China, namely
Guangxi, Sichuan, Hainan, Jiangsu, Fujian, Gansu, and Guizhou, intro-
duced green credit interest discount policies. Since green credit and green
bonds are the primary financing channels for GF, the implementation of
this policy may interfere with our results. To more accurately capture the
impact of CBMP on GF, we regress Eqs. (1) and (2) after excluding the
samples from the seven provinces mentioned above. The absolute values of
the estimated coefficients in Table 4 are slightly smaller than the bench-

Oeconomia Copernicana, 15(4), 1223–1262
1241
mark regression results, indicating that our fundamental conclusion re-
mains accurate after accounting for the potential interference of the green
credit discount policy.
To conclude, the evidence of diverse robustness checks supports the no-
tion that CBMP significantly improves GF.
Heterogeneity analysis
Heterogeneity of property rights
Based on property rights attributes, it is possible to categorize Chinese cor-
porate bond issuers into state-owned (SOEs) and private (non-SOEs) enter-
prises, which hold distinct economic and social statuses. The reform of
CBMP may affect both types of bond issuers mentioned above, but the
influencing magnitude and manner could be different. Non-SOEs lack im-
plicit guarantees from the government, and the reform of CBMP can im-
prove their financing conditions by providing them with more financing
opportunities. However, Chinese investors prefer bonds issued by SOEs
due to their high liquidity and government endorsement (Fang et al., 2023).
After the reform of CBMP, green bonds issued by SOEs are increasingly
attractive to investors as a combination of reliable issuers and high-quality
assets. Therefore, following the expansion of MLF collateral, green bonds
held by SOEs may have lower costs and a larger financing scale than those
held by non-SOEs.
To test whether there is ownership heterogeneity in the effect of MLF
collateral expansion, we compared the influence of CBMP on SOEs versus
non-SOEs, respectively. In Table 5, the first four columns indicate that
CBMPʹs impact on bond spreads and financing scale is significant only for
SOEs, but not for non-SOEs. According to the findings, CBMP has a strong-
er influence on SOEs.
Bond rating heterogeneity
Also, we consider bond rating heterogeneity. Prior to 2018, the collateral
scope of the CBMP included only local government bonds, policy financial
bonds, central bank bills, government bonds, AAA rated corporate credit
bonds. In 2018, the reform of CBMP included eligible bonds rated no lower

Oeconomia Copernicana, 15(4), 1223–1262
1242
than AA into the scope of collateral, improving the credit worthiness of AA
and AA+ bonds. In light of the differential impacts of MLF collateral expan-
sion on AAA rated bonds compared to AA and AA+ rated bonds, we pre-
dict that there is rating heterogeneity in the impact of CBMP on GF. To
verify this hypothesis, the sample is classified into two distinct categories
according to bond ratings: AAA rated bonds and non-AAA rated bonds. As
shown in columns (5) to (8) of Table 5, when  is the dependent variable,
the coefficient of    is significant only for non-AAA rated bonds,
while it is not significant for AAA bonds. When  is the dependent
variable, the coefficient of    is 0.812 with a significance level of
1% for non-AAA rated bonds, and 0.658 with a significance level of 10% for
AAA rated bonds. The bond rating heterogeneity results show that CBMP
has a stronger promoting effect on GF for AA and AA+ bonds.
Green attention heterogeneity
In addition to the inherent characteristics of bonds and issuers, local
government behavior has a significant impact on the implementation of
CBMP, which may provide more information to explain the conclusions of
this paper. In particular, the government, as a leader in resource allocation
and policy selection, not only has decision-making power over the direc-
tion of local development, but also has responsibility for creating strategic
orientation for the region. Under the pressure of multiple performance
appraisals, the government’s green attention reflects its intention to allocate
limited resources to the green fields, thereby guiding the development of
economic models towards sustainability. As a result, governments in re-
gions with high green attention are more likely to implement policy
measures to encourage the development of GF. This has led to increased
demand and supply for green financial products, thereby enhancing the
influence of CBMP on GF. In addition, differences in governments’ green
attention also reflect their awareness of risks. In areas with higher green
attention, both government and financial institutions tend to attach greater
importance to the prevention and control of environmental and climate
risks. Therefore, when the PBOC designates green bonds as MLF collateral,
green bonds become a relatively low-risk investment option and are more
favored by governments with high green attention, which increases the
scarcity of green bonds. Given that scarcity is one of the ways in which

Oeconomia Copernicana, 15(4), 1223–1262
1243
CBMP works, the effect of CBMP will be more significant in areas with
high green attention.
To test this hypothesis, we construct the government green attention in-
dex # based on the frequency of green terms mentioned in government
work reports. # is the logarithm of the count of keywords related to
green subjects. Next, we construct a triple-difference model (DDD) by re-
placing the core explanatory variable    in Eqs. (1) and (2) with
#    , to investigate the heterogeneous effects of CBMP
under different levels of government green attention. Results from the first
and second columns of Table 6 indicate that, compared to regions with low
government green attention, CBMP has a stronger influence on promoting
GF in regions with high government green attention.
Environmental regulatory heterogeneity
Differences in environmental legislation between regions reflect the
government’s support for green economic development and will influence
the articleʹs conclusions. Specifically, local governments that prioritize
green economic development may introduce supporting incentives to en-
sure that environmental legislation is enforced. These measures not only
include policies that strengthen supervision, establish environmental
standards, and restrict emissions to penalize polluting activities, but also
policies that provide positive incentives for green behavior, such as tax
incentives, subsidies, and preferential interest rates. In this context, local
governments are more likely to proactively support GF development and
incentivize financial entities to use green financial instruments such as
green bonds as collateral to obtain monetary policy assistance. However,
local governments in areas with looser environmental regulations may
attach less emphasis on the green economy and have a relative lack of ap-
plicable support measures, resulting in less impetus for collateral-based
monetary policy. In the year the Environmental Protection Tax Law was
implemented (January 1, 2018), 19 out of the 31 provinces in China raised
their taxation standards. By observing changes in tax rates across regions,
we are able to accurately capture the extent of local environmental regula-
tions. As a result, we create a dummy variable  for environmental regula-
tion, which has a value of 1 if the bond issuer is in one of the 19 provinces
listed above and 0 otherwise. Subsequently, we build a DDD model by
replacing the core explanatory variable    in Eqs. (1) and (2)

Oeconomia Copernicana, 15(4), 1223–1262
1244
with      to examine how differences in environmental regu-
lations affect the impact of CBMP on GF. According to columns 3–4 of Ta-
ble 6, applying CBMPs in areas with strict environmental regulations has
a greater impact on promoting GF than in areas with loose regulations.
Further analysis
Impact on the primary market
As the secondary market operates solely as a bond trading platform with-
out direct involvement in financing activities, it is necessary to examine
how monetary policy affects bond issuance to determine whether CBMP
can effectively stimulate the development of GF (Fang et al., 2023). In the
literature on asset pricing, Chen et al. (2023) discovered that a reduction in
bond collateral capacity lowers the price and increases the spread in the
secondary market. In addition, the expansion of MLF collateral has a mar-
ket transmission effect, which can simultaneously reduce the spread of new
collateral in the secondary and primary bond markets (Fang et al., 2020).
Therefore, based on models (5) and (6), we further investigate how CBMP
affects the issuance spread and financing scale of new collateral in the pri-
mary market.
As shown in Table 7, when 8 is the dependent variable, the regression
coefficient of    is -0.046, and when :(;<= is the dependent
variable, the regression coefficient of    is 0.133. Moreover, the
above conclusion has passed at least the 5% level of significance test. The
results show that the positive effects of CBMP on reducing bond spreads
and increasing bond financing scale also exist in the primary bond market.
There may be two reasons for this phenomenon. First, in the primary mar-
ket, bond issuers determine the final issuance rate of bonds by inviting
financial institutions to bid competitively (Fang et al., 2023). In such a con-
text, the interest rates of similar bonds are comparable in the two markets.
Therefore, it is highly likely that the comparability of similar bonds be-
tween the primary and secondary markets may lead to synchronized
changes in their bond spreads. Second, green bonds’ performance in trad-
ing markets can influence investors’ perceptions and demand for similar
instruments in the primary market. As a result of green bonds’ lower fi-
nancing costs and significant investor interest, it can create a positive mar-

Oeconomia Copernicana, 15(4), 1223–1262
1245
ket perception that encourages issuers to issue more green bonds in the
primary market.
Impact on the real economy
GF is a way of channeling social funding to environment-friendly pro-
jects in order to mitigate climate risks and protect the environment (Nawaz
et al., 2021). Our results prove that CBMP significantly promotes the devel-
opment of GF. However, we know quite little about the green behavior of
companies that get green funding. The key to determining whether CBMP
benefits the green economy, is to examine whether companies issuing
green bonds would increase green investments or indulge in ʺgreenwash-
ingʺ (Shi et al., 2023a). Research shows that green bonds increase green
economic recovery by around 17% each year (Zhao et al., 2022). Similarly,
Zhang (2023) and Ma et al. (2024) found that GF plays an effective role in
mitigating corporate ʺgreenwashingʺ practices and incentivizing their ESG
performance. In contrast, Shi et al. (2023a) discovered that when corpora-
tions utilize funds raised from green bonds, they intentionally build an
image of environmental awareness while not altering their environmental
behavior. Based on models (7)–(9), we evaluate the impact of CBMP on
corporate green performance.
The regression results are presented in Table 8. From column (1), we can
see that the coefficient of    is 0.146 at the 1% significance level,
indicating that after the expansion of MLF collateral, companies issuing
green bonds are more willing to disclose environmental information. In
column (2), the coefficient of    is 0.058, demonstrating that
CBMP significantly strengthens the green innovation capabilities of enter-
prises. In column (3), we find that companies issuing green bonds exhibited
a higher level of pollution governance after the expansion of MLF collateral
compared with companies that did not issue green bonds. Overall, Table 8
demonstrates that CBMP can actually improve corporate green behavior
rather than greenwashing.
Although the above conclusions have preliminarily verified that CBMP
contributes to promoting corporate green performance, there are still con-
cerns that the improvement of corporate green performance is achieved
through mergers and acquisitions of high environmental governance com-
panies, which is greenwashing behavior. This part will conduct an in-depth
analysis on this issue. Based on model (10), we verify whether M&A can

Oeconomia Copernicana, 15(4), 1223–1262
1246
improve a company’s environmental governance score in ESG. The result
in column (1) of Table 9 shows that M&A can indeed improve enterprises’
environmental scores, suggesting the likelihood of enterprises’ greenwash-
ing via M&A. Then, we verify whether corporate green performance is the
result of M&A from three aspects. First, assuming this premise is true, cor-
porate M&A activity should grow dramatically after green bonds are in-
corporated as MLF collateral. The explained variables in columns (2)–(4) of
Table 9 are the M&A scale, the number of M&A, and the M&A probability,
respectively, and the coefficients of (: are not significant. This shows that
CBMP does not significantly promote the M&A behavior of enterprises,
and the green performance of enterprises is not the result of M&A. Second,
if the corporate green performance is the result of M&A, it should be ob-
served that the environmental scores of the target enterprises should be
significantly higher than those of the acquiring enterprises. In column (5) of
Table 12, the explanatory variable K denotes the disparity in environmental
scores between the acquirer and the target. The insignificant coefficient of
(: implies that the hypothesis of enterprises acquiring entities with supe-
rior environmental performance via M&A is invalid. Third, we add the
variable (: to models (7)–(9). If a company engages in greenwashing, it
should be observed that after adding the variable (:, the core explanatory
variables 'L' should be insignificant. Columns (6)–(8) of Table 9 reveal that
the coefficients of (: and 'L' are significantly positive, signifying that
corporate green performance is not a result of M&A. The above conclusions
imply that, while M&A contributes to the enhancement of enterprises’ en-
vironmental scores, such operations do not induce corporate green perfor-
mance, and thus enterprises are not involved in greenwashing behavior.
Discussion
As an innovation of this study, the role of unconventional monetary policy
on GF is worth exploring. In academia, a debate exists over whether mone-
tary policy should actively promote the development of green finance. The
essence of the controversy is whether this behavior exceeds the statutory
boundaries of the central bank’s duties. Promoting the development of GF
is a long-term task, which means that the central bank should assume
ʺgreen responsibilitiesʺ and adopt normalized structural monetary policies.
However, classical monetary economics theory holds that the central bank

Oeconomia Copernicana, 15(4), 1223–1262
1247
should adhere to the principle of ʺmarket neutralityʺ and guide the alloca-
tion of financial resources through price signals (Aruoba & Drechsel, 2024).
Consequently, major developed countries view structural monetary poli-
cies as temporary remedial measures against negative shocks and contend
that the central bank should not directly undertake the task of guiding the
transformation of the real economy (Nyborg, 2017; Choi et al., 2021). The
opposing view holds that monetary policies in practice are detrimental to
the development of GF. On the one hand, monetary policies that follow the
market neutrality principle is non-neutral, as manifested in the fact that the
central bank’s large-scale asset purchase programs tend to favor carbon-
intensive industries (Campiglio, 2016). On the other hand, there is a ʺmar-
ket failureʺ in the financial market for green investment. Due to infor-
mation asymmetry, banks are more willing to hold non-green assets for
asset security reasons, which squeezes the space for green investment and
creates a huge green financing gap (Campiglio, 2016). Considering these
factors, some scholars have initiated the proposal of a “green central bank,”
advocating that the transition to a green economy should be incorporated
into the functions of monetary policy (Chen et al., 2025). We found that
considering green bonds as qualified collateral for the MLF facilitates the
development of GF and propels the green transformation of enterprises.
This suggests that the execution of normalized structural monetary policies
is advantageous for the green economic transformation. Consequently, it
offers a reference for the research on topics associated with the central
bank’s responsibilities and structural monetary policies.
According to the heterogeneity analysis results, CBMP’s role in promot-
ing GF is more significant for state-owned bond issuers. This shows that
CBMP mainly alleviates the green financing difficulties of state-owned
enterprises. However, China’s state-owned enterprises benefit from gov-
ernment policy support and possess an edge over non-state-owned enter-
prises in procuring financial resources like bank credit (Shi et al., 2023b).
Conversely, due to the imperfect credit risk assessment system, private
enterprises face financing constraints because of their relatively small scale
and lack of collateral guarantees (Guo et al., 2023). Improving the financing
difficulties of private enterprises is an important task in China’s financial
market reform (Shi et al., 2023b). Based on this, the future reform direction
of the MLF should appropriately take into account the characteristics of
enterprise property rights. For example, the government can strengthen the
risk-sharing mechanism for private enterprises’ green financing by estab-

Oeconomia Copernicana, 15(4), 1223–1262
1248
lishing special guarantee funds or risk compensation funds. This helps to
alleviate financial institutions’ risk concerns, thereby enhancing the role of
MLF in promoting green financing of private enterprises and promoting
private enterprises to play a greater role in the development of the green
economy. These proposals could be included in future studies and ana-
lyzed in combination with data from other countries or regions.
In addition, we find that strict environmental regulations and increased
government green attention significantly enhance the positive effect of
CBMP on GF, demonstrating the importance of policy coordination. First,
the government’s high level of green attention represents the importance
and value of green development, which helps to create a favorable policy
atmosphere and social consensus and guide social resources to the green
field (Chen et al., 2024). Second, strict environmental regulations provide
a solid institutional guarantee for the effect of CBMP on GF (Du et al., 2023).
Environmental regulations compel enterprises to adopt green practices by
establishing specific environmental standards, emission limitations, and
other criteria (Du & Li, 2020). In this context, CBMP can provide financial
support to green enterprises, thus forming a virtuous cycle where envi-
ronmental regulations and CBMP synergistically foster the development of
green finance. Therefore, politicians should attach importance to policy
synergy and appropriately adopt policies that combine positive incentives
with negative penalties to promote the development of GF and assist in the
green transformation of the economy. Our findings suggest that investigat-
ing the optimal strategy of multiple policy combinations for improving
green finance is a novel approach to extending GF research subjects.
Conclusions
GF helps direct funds towards environmental conservation and sustainable
development, playing a crucial role in promoting ecological civilization
and driving economic structural transformation. In the context of a huge
green financing gap in China, how to employ monetary policy tools to ad-
just the economic structure and reduce green financing costs has become an
important topic. However, the effects of monetary policy, especially un-
conventional CBMP on GF, has not yet been explored fully in the existing
studies. To address the existing research gap, we investigate the influence
of CBMP on GF and the real economy. Our findings show as follows: First,

Oeconomia Copernicana, 15(4), 1223–1262
1249
the CBMP effectively decreased the spread of green bonds and increased
their financing scale in the secondary market, thereby facilitating the ad-
vancement of GF. Second, the positive correlation between CBMP and GF is
more pronounced in green bonds with AA+ and AA ratings, issued by
SOEs, and in regions with stringent environmental regulations and high
government green attention. These results hold up to series of robustness
checks, including changing the sample matching method, the placebo tests,
and excluding other policy interferences. Third, the policy shock in the
secondary market has a spillover effect on the primary market. It means
that endowing green bonds with the function of serving as collateral for
MLF in the secondary market would also lead to a reduction in the spread
of green bonds and an increase in their financing scale in the primary bond
market. Fourth, CBMP effectively incentivizes corporate green behavior
rather than “greenwashing”.
Based on the studyʹs main findings, we provide the following policy
recommendations. First, it is essential for the central bank to further im-
prove the CBMP framework by expanding other green assets as eligible
collateral. This measure helps to accelerate the economy’s green transfor-
mation while also contributing to improving the quality and efficiency of
economic growth. Second, the government should take steps to address the
green financing dilemma of non-SOEs. Our findings suggest that the CBMP
offers significant green financing benefits to SOEs, but limited benefits to
non-SOEs. Since non-SOEs suffer more funding challenges than SOEs, the
central bank should enhance its assistance for green financing of non-SOEs
in the future. Third, local governments need to strengthen their green at-
tention and implementation capacity. Heterogeneity analysis results indi-
cate that higher levels of green awareness and environmental regulations
can better leverage the effectiveness of CBMP. Therefore, enhancing local
governments green attention and implementation capacity becomes a criti-
cal pathway for promoting green finance development in the future.
Although our study provides new insights into the role of CBMP for
green finance, it has the following limitations. First, this study only ana-
lyzed the bond market’s impact on GF, not the credit market’s influence. To
support the development of the green economy, the PBOC has regarded
green bonds and green credits as eligible collaterals for the MLF. This
study examines the impact of CBMP on GF from the perspective of the
bond. Future research in this topic could analyze the impact of CBMP from
the perspective of credit. Second, the quantification methods for the financ-

Oeconomia Copernicana, 15(4), 1223–1262
1250
ing scale of GF may be limited. Ideally, the actual amount of funds raised is
the most accurate indicator to measure the actual financing scale of GF.
Due to the limitations of data availability, we are unable to identify wheth-
er a bond trading activity is a resale of a particular bond. Consequently, we
failed to accurately measure the financing scale of GF. To overcome this
limitation, we employ the trading volume of green bonds in the secondary
market and the issuance volume of green bonds in the primary bond mar-
ket as alternative indicators for the financing scale of GF. Green bond trad-
ing volume represents the market’s expectations for GF. The higher the
expectations, the greater the market demand for green bonds. Therefore,
bond trading volume can indirectly measure the financing scale of GF. The
primary bond market involves direct financing. The more bonds issued, the
larger the financing scale of such bonds. Therefore, although the construc-
tion of indicators for the financing scale of GF has limitations, it is reasona-
ble. If bond data contain information on resale attributes, future research
could use such new data to expand and refine this topic.
References
Akomea-Frimpong, I., Adeabah, D., Ofosu, D., & Tenakwah, E. J. (2022). A review of
studies on green finance of banks, research gaps and future directions. Journal of
Sustainable Finance & Investment, 12(4), 1241–1264.
https://doi.org/10.1080/20430795.2020.1870202.
Aloui, D., Benkraiem, R., Guesmi, K., & Vigne, S. (2023). The European Central
Bank and green finance: How would the green quantitative easing affect the in-
vestors’ behavior during times of crisis? International Review of Financial Analysis,
85, 102464. https://doi.org/10.1016/j.irfa.2022.102464.
Aruoba, S. B., & Drechsel, T. (2024). The long and variable lags of monetary policy:
Evidence from disaggregated price indices. Journal of Monetary Economics, 148,
103635. https://doi.org/10.1016/j.jmoneco.2024.103635.
Ashcraft, A., Gârleanu, N., & Pedersen, L. H. (2011). Two monetary tools: Interest
rates and haircuts. NBER Macroeconomics Annual, 25(1), 143–180.
https://doi.org/10.1086/657530.
Bahaj, S., Foulis, A., Pinter, G., & Surico, P. (2022). Employment and the residential
collateral channel of monetary policy. Journal of Monetary Economics, 131, 26–44.
https://doi.org/10.1016/j.jmoneco.2022.07.002.
Batrancea, I., Batrancea, L., Maran Rathnaswamy, M., Tulai, H., Fatacean, G., & Rus,
M.-I. (2020). Greening the financial system in USA, Canada and Brazil: A panel
data analysis. Mathematics, 8(12), 2217. https://doi.org/10.3390/math8122217.

Oeconomia Copernicana, 15(4), 1223–1262
1251
Borio, C., & Zhu, H. (2012). Capital regulation, risk-taking and monetary policy:
A missing link in the transmission mechanism? Journal of Financial Stability, 8(4),
236–251. https://doi.org/10.1016/j.jfs.2011.12.003.
Broer, T., & Kero, A. (2021). Collateralization and asset price bubbles when inves-
tors disagree about risk. Journal of Banking & Finance, 128, 106137.
https://doi.org/10.1016/j.jbankfin.2021.106137.
Campiglio, E. (2016). Beyond carbon pricing: The role of banking and monetary
policy in financing the transition to a low-carbon economy. Ecological Economics,
121, 220–230. https://doi.org/10.1016/j.ecolecon.2015.03.020.
Campiglio, E., Dafermos, Y., Monnin, P., Ryan-Collins, J., Schotten, G., & Tanaka, M.
(2018). Climate change challenges for central banks and financial regulators.
Nature Climate Change, 8(6), 462–468. https://doi.org/10.1038/s41558-018-0175-0.
CBNEditor. Green Finance in China Sees Roaring Growth, Beijing Flags Further
Standardisation Efforts. China Banking News, May 20, 2019,
https://www.chinabankingnews.com/2019/05/20/green-finance-in-china-sees-
roaring-growth-beijing-flags-further-policy-supports/.
Chang, Y., Ji, Q., & Zhang, D. (2021). Green finance and energy policy: Obstacles,
opportunities, and options. Energy Policy, 157, 112497.
https://doi.org/10.1016/j.enpol.2021.112497.
Chen, H., Chen, Z., He, Z., Liu, J., & Xie, R. (2023). Pledgeability and asset prices:
Evidence from the Chinese corporate bond markets. Journal of Finance, 78(5),
2563–2620. https://doi.org/10.1111/jofi.13266.
Chen, H., Deng, J., Lu, M., Zhang, P., & Zhang, Q. (2024). Government environmen-
tal attention, credit supply and firms’ green investment. Energy Economics, 134,
107547. https://doi.org/10.1016/j.eneco.2024.107547.
Chen, R., Wang, G., Jamil, N., & Iqbal, N. (2025). The green premium of unconven-
tional monetary policy: Evidence from the enlarged collateral framework by the
People’s Bank of China. Research in International Business and Finance, 73, 102655.
https://doi.org/10.1016/j.ribaf.2024.102655.
Chen, S., Huang, Z., Drakeford, B. M., & Failler, P. (2019). Lending interest rate,
loaning scale, and government subsidy scale in green innovation. Energies,
12(23), 4431. https://doi.org/10.3390/en12234431.
Cheng, T., Qiu, L., Lv, W., Yang, X., & Yang, G. (2024). Economic policy uncertainty
and municipal corporate bonds credit spreads: Evidence from China. Finance
Research Letters, 69, 106170. https://doi.org/10.1016/j.frl.2024.106170.
Choi, D. B., Santos, J. A. C., & Yorulmazer, T. (2021). A theory of collateral for the
lender of last resort. Review of Finance, 25(4), 973–996.
https://doi.org/10.1093/rof/rfab002.
D’Amico, S., Fan, R., & Kitsul, Y. (2018). The scarcity value of treasury collateral:
Repo-market effects of security-specific supply and demand factors. Journal of
Financial and Quantitative Analysis, 53(5), 2103–2129.
https://doi.org/10.1017/S0022109018000790.

Oeconomia Copernicana, 15(4), 1223–1262
1252
Debrah, C., Chan, A. P. C., & Darko, A. (2022). Green finance gap in green build-
ings: A scoping review and future research needs. Building and Environment, 207,
108443. https://doi.org/10.1016/j.buildenv.2021.108443.
Desalegn, G., Fekete-Farkas, M., & Tangl, A. (2022). The effect of monetary policy
and private investment on green finance: Evidence from Hungary. Journal of Risk
and Financial Management, 15(3), 1–18. https://doi.org/10.3390/jrfm15030117.
Dikau, S., & Volz, U. (2021). Central bank mandates, sustainability objectives and
the promotion of green finance. Ecological Economics, 184, 107022.
https://doi.org/10.1016/j.ecolecon.2021.107022.
Du, J., Shen, Z., Song, M., & Vardanyan, M. (2023). The role of green financing in
facilitating renewable energy transition in China: Perspectives from energy gov-
ernance, environmental regulation, and market reforms. Energy Economics, 120,
106595. https://doi.org/10.1016/j.eneco.2023.106595.
Du, W., & Li, M. (2020). Assessing the impact of environmental regulation on pollu-
tion abatement and collaborative emissions reduction: Micro-evidence from
Chinese industrial enterprises. Environmental Impact Assessment Review, 82,
106382–10. https://doi.org/10.1016/j.eiar.2020.106382.
Fang, F., Si, D.-K., & Hu, D. (2023). Green bond spread effect of unconventional
monetary policy: Evidence from China. Economic Analysis and Policy, 80, 398–413.
https://doi.org/10.1016/j.eap.2023.08.019.
Fang, H., Wang, Y., & Wu, X. (2020). The collateral channel of monetary policy:
Evidence from China. NBER Working Paper Series, 26792.
https://doi.org/10.3386/w26792.
Fecht, F., Nyborg, K. G., Rocholl, J., & Woschitz, J. (2016). Collateral, central bank
repos, and systemic arbitrage. Swiss Finance Institute Research Paper, 16–66.
https://doi.org/10.2139/ssrn.2871337.
Gârleanu, N., & Pedersen, L. H. (2011). Margin-based asset pricing and deviations
from the law of one price. Review of Financial Studies, 24(6), 1980–2022.
https://doi.org/10.1093/rfs/hhr027.
Geng, G., Han, Z., Wu, H., Cheng, M., WANG, R., & Liu, H. (2024). Collateral policy
of the central bank and corporate financing costs: Evidence from China. North
American Journal of Economics and Finance, 102042.
https://doi.org/10.1016/j.najef.2023.102042.
Green, T. C., & Jame, R. (2013). Company name fluency, investor recognition, and
firm value. Journal of Financial Economics, 109(3), 813–834.
https://doi.org/10.1016/j.jfineco.2013.04.007.
Grilli, R., Giri, F., & Gallegati, M. (2020). Collateral rehypothecation, safe asset scar-
city, and unconventional monetary policy. Economic Modelling, 91, 633–645.
https://doi.org/10.1016/j.econmod.2019.12.004.
Guo, K., Ke, B., & Tang, S. (2023). Private firms’ financial constraints and share
pledging by controlling shareholders of publicly listed firms: Evidence from
China. Journal of Corporate Finance, 80, 102393.
https://doi.org/10.1016/j.jcorpfin.2023.102393.

Oeconomia Copernicana, 15(4), 1223–1262
1253
Guo, S., & Zhang, Z. (2023). Green credit policy and total factor productivity: Evi-
dence from Chinese listed companies. Energy Economics, 128, 107115.
https://doi.org/10.1016/j.eneco.2023.107115.
Hafner, S., Jones, A., Anger-Kraavi, A., & Pohl, J. (2020). Closing the green finance
gap – A systems perspective. Environmental Innovation and Societal Transitions,
34, 26–60. https://doi.org/10.1016/j.eist.2019.11.007.
Liu, X., Wang, C., Zhang, X., Gao, L., & Zhu, J. (2022). Financing constraints change
of China’s green industries. AIMS Mathematics, 7(12), 20873–20890.
https://doi.org/10.3934/math.20221144.
Liu, X., Wu, Y., & Zhang, H. (2023). Collateral-based monetary policy and corporate
employment: Evidence from Medium-term Lending Facility in China. Journal of
Corporate Finance, 78, 102333. https://doi.org/10.1016/j.jcorpfin.2022.102333.
Ma, D., He, Y., & Zeng, L. (2024). Can green finance improve the ESG performance?
Evidence from green credit policy in China. Energy Economics, 137, 107772.
https://doi.org/10.1016/j.eneco.2024.107772.
McConnell, A., Yanovski, B., & Lessmann, K. (2022). Central bank collateral as
a green monetary policy instrument. Climate Policy, 22(3), 339–355.
https://doi.org/10.1080/14693062.2021.2012112.
Nawaz, M. A., Seshadri, U., Kumar, P., Aqdas, R., Patwary, A. K., & Riaz, M. (2021).
Nexus between green finance and climate change mitigation in N-11 and BRICS
countries: empirical estimation through difference in differences (DID) ap-
proach. Environmental Science and Pollution Research International, 28(6), 6504–
6519. https://doi.org/10.1007/s11356-020-10920-y.
Nyborg, K. G. (2017). Central bank collateral frameworks. Journal of Banking &
Finance, 76, 198–214. https://doi.org/10.1016/j.jbankfin.2016.12.010.
Palea, V., & Drogo, F. (2020). Carbon emissions and the cost of debt in the eurozone:
The role of public policies, climate-related disclosure and corporate governance.
Business Strategy and the Environment, 29(8), 2953–2972.
https://doi.org/10.1002/bse.2550.
Rosenbaum, P. R., & Rubin, D. B. (2023). Propensity scores in the design of observa-
tional studies for causal effects. Biometrika, 110(1), 1–13.
https://doi.org/10.1093/biomet/asac054.
Schmidt, J. (2020). Risk, asset pricing and monetary policy transmission in Europe:
Evidence from a threshold-VAR approach. Journal of International Money and
Finance, 109, 102235. https://doi.org/10.1016/j.jimonfin.2020.102235.
Shi, X., Ma, J., Jiang, A., Wei, S., & Yue, L. (2023a). Green bonds: Green investments
or greenwashing? International Review of Financial Analysis, 90, 102850.
https://doi.org/10.1016/j.irfa.2023.102850.
Shi, Y., Li, J. C., & Liu R. M. (2023 b). Financing constraints and share pledges: Evi-
dence from the share pledge reform in China. Journal of Corporate Finance, 78,
102337. https://doi.org/10.1016/j.jcorpfin.2022.102337.

Oeconomia Copernicana, 15(4), 1223–1262
1254
Sui, J., Liu, B., Li, Z., & Zhang, C. (2022). Monetary and macroprudential policies,
output, prices, and financial stability. International Review of Economics & Finance,
78, 212–233. https://doi.org/10.1016/j.iref.2021.11.010.
Tang, Y., Wang, B., Pan, N., & Li, Z. (2023). The impact of environmental infor-
mation disclosure on the cost of green bond: Evidence from China. Energy
Economics, 126, 107008. https://doi.org/10.1016/j.eneco.2023.107008.
Vespignani, J. L. (2015). On the differential impact of monetary policy across
states/territories and its determinants in Australia: Evidence and new methodol-
ogy from a small open economy. Journal of International Financial Markets,
Institutions & Money, 34, 1–13. https://doi.org/10.1016/j.intfin.2014.10.001.
Zhang, D. (2023). Does green finance really inhibit extreme hypocritical ESG risk?
A greenwashing perspective exploration. Energy Economics, 121, 106688.
https://doi.org/10.1016/j.eneco.2023.106688.
Zhang, R., Li, Y., & Liu, Y. (2021). Green bond issuance and corporate cost of capital.
Pacific-Basin Finance Journal, 69, 101626.
https://doi.org/10.1016/j.pacfin.2021.101626.
Zhao, L., Chau, K. Y., Tran, T. K., Sadiq, M., Xuyen, N. T. M., & Phan, T. T. H. (2022).
Enhancing green economic recovery through green bonds financing and energy
efficiency investments. Economic Analysis and Policy, 76, 488–501.
https://doi.org/10.1016/j.eap.2022.08.019.
Zou, J., Chen, P., Fu, X., & Gong, C. (2023). Does carbon trading affect the bond
spread of high-carbon enterprises? Evidence from China. Journal of Cleaner
Production, 417, 137882. https://doi.org/10.1016/j.jclepro.2023.137882.
Funding information
This study is funded by the Erasmus+ Programme Jean Monnet Chair of the Euro-
pean Union (project no. 101175462).
Compliance with ethical standards
This article does not contain any studies with human participants or animals per-
formed by the authors. Extracting and inspecting publicly accessible files (scholarly
sources) as evidence, before the research began no institutional ethics approval was
required.
Data availability statement
All data generated or analyzed are included in the published article. The raw data
supporting the conclusion of this article will be made available by the authors,
without undue reservation. The raw anonymized data can be provided by emailing
the primary author.

Oeconomia Copernicana, 15(4), 1223–1262
1255
Author contributions
All listed authors have made a substantial, direct and intellectual contribution to
the work, and approved it for publication. The authors take full responsibility for
the accuracy and the integrity of the source analysis.
Conflict of interest statement
The authors declare that the research was conducted in the absence of any commer-
cial or financial relationships that could be construed as a potential conflict of inter-
est.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not neces-
sarily represent those of their affiliated organizations, or those of the publisher, the
editors and the reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed or endorsed by the
publisher.

Annex
Table 1. Descriptive statistics of the sample










0.110 0.180 0 1.090 0.0200 0.0400 0.100 1345

12.90 2.970 5.970 19.69 10.87 13.11 15.40 1345

0.190 0.390 0 1 0 0 0 1345

2.970 1.470 1 12.27 1.970 2.730 3.630 1345

2.010 0.890 1 3 1 2 3 1345

0.560 0.140 0.140 0.910 0.460 0.570 0.650 1345

2.200 2.560 0.170 19.38 0.910 1.350 2.460 1345

24.98 1.890 22.19 29.04 23.58 24.51 25.95 1345
!"#
0.160 0.120 -0.160 0.560 0.100 0.130 0.210 1345
$
%
0.0800 0 0.0800 0.0900 0.0800 0.0800 0.0900 1345
Table 2. Benchmark regression results
Variables (1) (2) (3) (4)
&'
()*+,-
&'
()*+,-

.
#
-0.079
***
0.815
***
-0.079
***
0.823
***
(0.016) (0.258) (0.014) (0.250)

-0.030
***
-0.223
***
(0.003) (0.057)

-0.069
***
0.491
***
(0.007) (0.128)

0.056 0.566
(0.034) (0.614)

0.022
***
-0.133
***
(0.002) (0.035)

0.009
**
0.150
**
(0.003) (0.062)
!"#
0.025 -0.534
(0.039) (0.691)
$
%
-4.623
***
9.453
(1.530) (27.352)
/

0.121
***
12.815
***
0.430
***
8.019
***
(0.005) (0.085) (0.138) (2.460)
0
1345 1345 1345 1345
1
2
0.018 0.007 0.270 0.097
Note: *, * * and * * * respectively indicate that the significance is 10%, 5% and 1%. Robust standard errors are
shown in parentheses. The following tables are the same.

Table 3. Robustness test based on changing the matching method
Variables





























(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


 !
-0.069
***
0.675
**
-0.073
***
0.842
***
-0.083
***
0.908
***
-0.084
***
0.911
***
-0.077
***
0.746
***
(0.015) (0.318) (0.014) (0.268) (0.015) (0.222) (0.015) (0.223) (0.013) (0.208)
"#
$
$
$
$
$
$
$
$
$
$
%#
$
$
$
$
$
$
$
$
$
$
&'
$
$
$
$
$
$
$
$
$
$
(
450 450 820 820 36137 36137 36893 36893 50800 50800
)
*
0.271 0.136 0.256 0.132 0.180 0.077 0.179 0.075 0.177 0.077

Table 4. Robustness test based on exclusion other policy interferences
Variables (1) (2)
&'
()*+,-

.
#
-0.070
***
0.697
***
(0.012) (0.255)
/

0.431
***
8.769
***
(0.126) (2.699)
AB
C
C
IC
C
C
ME
C
C
0
1126 1126
1
2
0.282 0.087
Table 5. Analysis of heterogeneity in individual characteristics
Variables
(1) (2) (3) (4) (5) (6) (7) (8)
&'
&'
()*+,-
()*+,-
&'
&'
()*+,-
()*+,-
FGH'
,*,
I
5
FGH'
FGH'
,*,
I
FGH'
,*,
I
JJJ
JJJ
,*,
I
JJJ
JJJ

.
#
-
0.116
***
-0.003 1.204
***
0.177 -0.138
***
-0.016 0.812
**
0.658
*
(0.017) (0.009) (0.272) (0.614) (0.022) (0.010) (0.343) (0.384)
/

0.456
**
0.333
***
11.131
***
-4.513 0.888
***
0.149 10.495
***
5.136
(0.183) (0.072) (2.839) (4.963) (0.226) (0.112) (3.517) (4.151)
AB
C
C
C
C
C
C
C
C
DB
C
C
C
C
C
C
C
C
$E
C
C
C
C
C
C
C
C
0
937 408 937 408 803 542 803 542
1
2
0.300 0.421 0.074 0.189 0.358 0.085 0.053 0.105
Table 6. Analysis of heterogeneity in government behavior
Variables (1) (2) (3) (4)
KL
MNOPQR
KL
MNOPQR
S
.

.
#
-0.016
***
0.168
***
(0.003) (0.050)

.

.
#
-0.074
***
0.702
**
(0.015) (0.274)
/
KOQL
0.426
***
8.017
***
0.411
***
8.333
***
(0.138) (2.458) (0.138) (2.464)
AB
C
C
C
C
DB
C
C
C
C
$E
C
C
C
C
0
1345 1345 1345 1345
1
2
0.270 0.097 0.266 0.094

Table 7. The spillover effect of the CBMP on the primary market
Variables (1) (2)
TU
VWXYZ[

.
#
-0.046
***
0.133
**
/

0.280
***
2.405
***
(0.076) (0.312)
AB
Y Y
$E
Y Y
0
2166 2166
1
2
0.222 0.092
Table 8. The impact of MLF collateral expansion on real economy
Variables (1) (2) (3)
S
\S
S

.
#
0.146
***
0.058
***
0.191
**
(0.016) (0.022) (0.097)
/

-1.148
***
-0.361
**
1.577
**
(0.116) (0.163) (0.723)
]S
5
]E
C
C
C
^
5
]E
C
C
C
0
38078 38078 38078
1
2
0.794 0.852 0.524
Table 9. Green governance or greenwashing? Based on the perspective of enterprise
M&A activities
(1) (2) (3) (4) (5) (6) (7) (8)
_`a
bcd
bce
fbc

dg
hae
`ad
$i
0.599
***
-0.092
***
0.048
***
0.004
(0.066) (0.027) (0.004) (0.006)

.
#
-0.008 -0.016 -0.010 -1.101
**
0.204
**
0.139
***
0.057
***
(0.419) (0.024) (0.023) (0.494) (0.097) (0.016) (0.022)
/

39.610
***
13.649
***
0.903
***
0.621
***
6.006 1.009 -0.857
***
-0.333
**
(1.812) (3.112) (0.179) (0.171) (4.315) (0.742) (0.119) (0.168)
0
36296 38078 38078 38078 3113 38078 38078 38078
1
2
0.805 0.154 0.156 0.157 0.360 0.524 0.795 0.852

Figure 1. Parallel trend results

Figure 2. Parallel trend results of the placebo event

Figure 3. Distribution of coefficients of the placebo test