The Impact of Climate Change On Financing Cost of Heavy-Pollution Firms in China

We examines the effects of climate change on the nancing cost of heavy-pollution rms using rm-level panel data analysis. The empirical results demonstrate that the annual temperature and precipitation changes can signicantly promote the nancing costs of heavy-pollution rms, the positive impacts of annual temperature and precipitation changes on the nancing costs of large, medium and small heavy-pollution rms has shown a gradual weakening trend with an increase of rm size, and the positive effects of annual temperature and precipitation changes on the nancing costs of younger and older heavy-pollution rms has shown a decline trend with an increase of rm age. The evidences conrms that the impact of climate change on the nancing costs of heavy-pollution rms exhibit a signicant rm size and age discrimination of nancing behaviors. Government decision-makers have to identify and optimize the transmission effect of climate change response on nancing behavior decisions of heavy-pollution industries, nancial institutions alleviate nancial conicts and credit discrimination behaviors and optimize the eciency of nancial resource allocation. Firms’ executives correct climate change strategy, optimize the climate- relevant operation, investment and nancing activities, and alleviate unfavorable inuences of climate changes for heavy-pollution rms.


Introduction
In recent decade, the continuous increase of greenhouse gases has brought about global warming, and high-frequency and high-intensity extreme climate changes have brought huge challenges and disasters to human society and natural systems. For example, the melting of Antarctic glaciers has accelerated, forest res in Australia have been burning for several months, Philippines has experienced continuous drought and the United States have experienced extremely cold temperatures. These extreme climate changes have brought about huge loss for human social and economic activities (Sequeira, et al.2018;Lu,et al.2019; Chen and Li,2020). Climate change is the most urgent long-term threat to human society. Global warming and extreme climate change directly or indirectly pose long-term risks to the regional economy and society development, which have received widespread attention from the international community. In 1992, the United Nations General Assembly adopted the "United Nations Framework Convention on Climate Change." Climate change is a global problem faced by global countries and regions. Global climate change is huge disasters to human society and ecosystems, which bring immeasurable economic losses to human society. On December 12, 2015, the Paris Climate Conference passed the "Paris Agreement", and on April 12, 2016, member states signed a climate change agreement in New York. The response to climate change requires stronger international cooperation and strengthens the response to climate change ability. In the World Economic Forum in Davos, Switzerland in January 2020, economists and business leaders from global politician, business leadership and economist recognized that addressing climate change is one of the most important challenges for human society. The Chinese government has issued annual reports on its policies and actions to address climate change for 11 consecutive years. China has achieved positive results in addressing climate change, its ability to respond to climate change has gradually increased, and industrial enterprises have continuously improved their awareness of addressing climate change. How to deal with climate change, coordinate the contradiction between economic development and greenhouse gas emission reduction, and achieve sustainable and inclusive growth are ever-increasing concerned problems for global countries and regions (Bakhsh,et al,2020;Mahmood, et al,2020;Lu,et al.2021).
The bene t-cost approach responding to climate change is a hot topic (Koomey,2013). In recent years, climate change may cause major damage to many rms, and eventually may induce rms boost their industrial activities to move away from those areas highly affected by climate change (Linnenluecke et al., 2011). Mitigating climate change and reducing related greenhouse gas emissions are challenges faced by a good many rms in the world. Firms and industries have a central role in the physical effects of climate change under supporting societal adaptation (Linnenluecke, et al.2013). Increasing climate change causes worktime and labor productivity reduction (Takakura,et al.2017). More and more rms are adopting climate-related emission reduction strategies and information disclosure to effectively respond to national climate change (Canevari,2020). The challenge of responding to global climate change is an important task for corporate strategic reforms, which force many rms to enhance their rationalization and strategic reforms, improves the utilization of their resources allocation, and thereby enhances rms' market competitiveness (Clark and Crawford, 2012;Tamošiūnas, 2014). Climate change is closely related to the daily operational process and nancial performance for many rms. External stakeholders are concerned that the response to climate change will affect the daily operations of the rms' climate riskrelated processes, supervision, reputation and litigation risks (Sullivan and Gouldson, 2017). The exposure of to climate change can re ect rms' social responsibility actions that affect corporate market value and stock market response (Ziegler et al., 2011; Hsu and Wang, 2013). Different government commitments to address climate change will lead to signi cant differences in rm's market value and stock market portfolio risk (Berkman et al., 2019; Secinaro,et al.2020). In United States, industrial rms disclose that climate change-related performance did not signi cantly affect corporate operational e ciency and environmental e ciency, and carbon-intensive industrial rms perform better operational performance and higher response to climate change (Yu et al., 2016). Under the European Union carbon trading markets, the carbon-intensive rms widely adopt emission reduction strategies to address climate change (Sarasini, 2013;Cadez and Czerny, 2016;Wang and Sueyosh,2018). External agencies, stakeholders, and internal governance will pay much attention to the rms' strategy development to deal with climate change, the uncertainty of greenhouse gas regulation, and emission reduction strategies  (Nohara,et al.2013). The actions to address climate change and emission reduction strategies will directly affect rms' carbon emissions costs and their greenhouse gas emissions (Cadez and Guilding, 2017;Kennard,2020;Toft and Rüdiger,2020). Traditional knowledge, public policies and technological innovation may support nancial policy and sustainable development (Li,et al.2021 (Kuo, et al.2015). Firms' carbon information disclosure cause the negative market reaction (Lee,et al.2015). Indian rms' climate change information disclosure has a signi cant positive correlation with returns on equity, and those rms with higher returns on equity will disclose more information about climate change (Praveen and Mohammad, 2018). Climate change activists' public and private politics ethics may elicit different organizational internal and external responses (Hiatt, et al.2015;Toft and Rüdiger,2020). Climate change information disclosure lead to better environmental performance (Giannarakis, et al.2018). Firms can carry out heterogeneous mitigation technologies portfolios, including pollution control, green design, eco-e ciency, low-carbon energy (Wang, et al.2018).
Responding to climate change will bring signi cant differences in the nancial performance and risk impacts for rms in different industries. Extreme climate change events can lead to unsustainable business activities of industrial rms. The external environment and internal environment lead to large differences in climate change actions and development strategies of different rms, which affect the unfair resources allocation among different rms. Market investors gradually realize that addressing climate change may have a certain negative impact on the rm's nancial performance in heavily polluting industries, alleviating the increase of climate -change-related costs and addressing climate change strategies. As China's strategy for addressing climate change will be integrated into the "Fourteenth Five-Year" National Economic and Social Development Planning and the Ecological Environment Protection Planning, China will continue to take active measures to control greenhouse gas emissions and strengthen the response to climate change, pollution prevention, and ecological environment. The comprehensive integration of protection has enabled rms in different industries to update their strategies and enhance their comprehensive capabilities to adapt to climate change. Heavypolluting industries are the most affected industries by climate change strategies. Different heavypolluting industries will also enhance strategic actions and capacity establishment to address climate change. Currently, many rms in heavy-pollution industries may face certain nancing discrimination and nancial frictions in their strategic actions to address climate change. Climate change will also affect the capital structure of enterprises in heavy-pollution industries and their related nancing costs. Therefore, studying the impact of climate change on the nancing costs of heavily pollution rms is an important core scienti c issue in the process of tackling climate change. However, the current domestic and foreign literature rarely involves the relationship between climate change and rms' nancing costs, this article lls the gaps of their correlations from the perspective of rm's size and age.
Carbon information disclosure level reduce the cost of equity nancing (Li, et al.2017).The academic research on the relationship between climate change and nancing cost of heavily-polluting rms is a brand-new scienti c issue, and the statistical results of this paper have important scienti c research value for government decision-makers, nancial institutions and rm managers. There are three important innovative achievements in our study: First, this article selects 789 heavily polluting rms from 2008 to 2018 as the research objects, and analyzes the impact of annual temperature and precipitation changes on the nancing costs of heavily-polluting rms using panel data analysis. The empirical results con rm that an increase of annual temperature and precipitation can signi cantly increase the nancing costs of heavily polluting rms. Second, different rm sizes can signi cantly modiate the relationship between climate change and nancing costs of heavily-polluting rms, an increase of annual temperature and precipitation may gradually alleviate their positive impacts of nancing cost of heavily-polluting rms with the gradual increase of rm scale,, especially the changes of temperature and precipitation are more sensitive to the impact of nancing cost of small-and medium-sized heavily-polluting rms. Third, different rm ages can adjust the correlation between climate change and nancing costs of heavilypolluting rms with the gradual increase of rm ages, the increase of annual temperature and precipitation may gradually alleviate their positive impacts of nancing costs of heavily-polluting rms, especially climate changes are more sensitive to the impact of nancing costs of younger heavilypolluting rms.

Theoretical Analysis And Hypothesis Design
Climate change will bring three related climate risks to rm's management: uncertainty of physical environment, uncertainty of regulatory policies and uncertainty of economic and nancial activities. Extreme climate change urges heavily-polluting rms to adjust their strategies and adaptation actions to deal with climate change in a timely manner, which are mainly re ected in climate risks related to climate change, strategic awareness and organizational vulnerability to adapt to climate change, and their management actions and investment activities to deal with climate change.
Organizational vulnerability and strategic awareness of coping with climate change are considered to have the signi cant impacts on the adaptation strategy of heavily-polluting rms to climate change. First, persistent climate change, especially, extreme climate change with high frequency and intensity will bring some potential climate-related risks to heavily-polluting rms. Recognizing extreme climate change will bring serious damage to investment activities, production and operation activities, rm managers realize the vulnerability of rm's management to deal with climate change (such as the defects or lack of ability to deal with oods) will bring about economic losses (Mele, et al,2021), such as tra c congestion, logistics delay, local rms' sales losses etc. Second, the continuous extreme hot or severe cold causes heavy-polluting rms to generate higher electricity consumption, soaring energy prices and greater productivity losses after electricity shortage. The continuous increase of climate-related risks will bring higher economic losses and nancial risks to heavy-polluting rms, which lead to an increase in debt nancing costs of heavy-polluting rms. Continuous extreme temperature leads to the decline of employee's productivity and the decrease of resource allocation e ciency, which causes the increase of climate-related risks of nancial friction and debt activities in heavy-polluting rms. These potential threats or adverse effects related to climate change urge heavy-polluting rms to enhance their strategic awareness of coping with climate change. Continuously increase business development and optimize operational processes related with climate change increase debt nancing activities, then organization's market adaptability to climate change and the exibility of its operation will be improved. Heavy-polluting rms continuously enhance investment activities and debt nancing related to climate risks, which may lead to an increase of rms' debt nancing costs. The Chinese governments have continuously supervise energy-saving and emission-reduction performance, and environmental regulation of heavy-polluting rms through multiple environmental regulatory policies, such as strict environmental laws and regulations, constraint targets for energy saving and emission reduction, environmental taxes and carbon trading scheme. Market investors tend to focus on those rms' nancing cost, extreme climate change and strict environmental regulations make market investors believe higher nancial risks of heavypolluting enterprises related to climate risks, which will reduce the investment opportunities and expected operating pro t margins of heavy-polluting rms. Meantime, market investors obtain higher return on investment and increase the nancing costs of heavy-polluting rms. From the analysis of agency theory, extreme climate change and stricter environmental regulation prompt rms' stakeholders to recognize higher climate risks and environmental regulation risks of heavy-polluting rms. As a result, nancial institutions may have credit discrimination and market friction, heavy-polluting rms face greater uncertainty of cash ow management, which credit rating entities reduce the quality of rm's credit rating. An increase of heavy-polluting rms' nancing costs enhance their awareness of environmental protection, timely adjustment of strategic actions to deal with climate change, improvement of rm's energy-saving and emission-reduction performance, and environmental pollutions (greenhouse gases, pollutants and sewage) improvement. Heavy-polluting rms increase the uncertainty of cash ow management and induce the increase of short-term debt activities and debt costs. Heavy-polluting rms try to reduce climate risks and environmental regulation risks related to climate change, then reduce agent problems and nancing costs.
Hypothesis 1: Climate change has a signi cant positive impact on the nancing cost of heavy-polluting rms.
Chinese banking system may have certain credit discrimination against heavy-polluting rm with different sizes when choosing credit loans. Extreme climate change brings different economic losses of heavy-polluting rms with different sizes, heavy-polluting enterprises enhance their ability to resist climate-related risks and improve their market adaptability to climate change with an increase of rm size. Meanwhile, climate change will bring heavy-polluting rms about the lower adverse impacts and effectively reduce their economic losses induced by extreme climate change. With an increase of rm size, if climate change bring about lower operating losses of larger heavy-polluting rms, nancial institutions provide more favorable external nancial resources, which are bene cial to reduce the nancing costs of the larger heavy-polluting rms. Extreme climate change will induce the local governments carry out stricter environmental regulation intensity, increase the operating costs and reduce the market competitive advantage of heavy-polluting rms (Hsu and Wang, 2013;Lu,et al.2020). Financial institutions are worried that climate change will bring about greater uncertainty of the business activities, investment activities and market demand of heavy-polluting rms. Climate changes(such as drought, ood, high temperature and severe cold) have a signi cant negative impact on the business activities of heavy-polluting rms, also causes interruption or adverse effects on upstream and downstream resource suppliers and buyers of the industrial chain. Unexpected climate change will bring small and mediumsized heavy-polluting rms about greater economic losses, and then small and medium-sized heavypolluting rm obtain higher nancing costs.
Hypothesis 2: Climate change may reduce the impact on the nancing cost of heavy-polluting rms with the gradual increase of rm scale.
Heavy-polluting rms with different rm age have different pressures from stakeholders, younger heavypolluting rms strive to improve the expected pro ts of heavy-polluting rms. To minimize the unfavorable factors induced by climate change, it is necessary heavy-polluting rm managers to raise greater external funds to optimize business the operational processes and market performance, to improve the market response speed related to climate change, which may lead to an increase of shortterm nancing costs. With the gradual increase of rm age, Heavy-polluting rms prefer to raise greater external funds through capital markets in order to improve business processes and market adaptability related to climate change and reduce the external dependence on debt funds. Therefore, older heavypolluting rms can alleviate the pressure on the rising nancing costs related to climate change. When nancial institutions provide heavy-polluting rms external debt funds, they must consider the operating losses of heavy-polluting rms caused by rm age and climate change, and then provide appropriately correctness of the heavy-polluting rms' nancing costs.
Hypothesis 3: Heavy-polluting rm age may moderate the relationship with climate change and nancing costs.

Econometric model
Generally, heavy-polluting rms constantly correct and adapt to the changing economic environment, but heavy-polluting rms seldom learn to adapt to the changes of climate environment. Heavy-polluting rms can endure seasonal climate and environmental changes, while extreme climate changes may cause heavy-polluting rms the vulnerability of coping with climate change, and the potential changes of extreme climate may exceed the market adaptability of heavy-polluting rms. Extreme climate changes will induce heavy-polluting rms their ability reduction of responding quickly to climate change, promote the mismatch between climate change and operation processes, and induce the decline of operation e ciency and expected pro ts of heavy-polluting rms. Rainfall changes will have a negative impact on the income of vineyard growers, while high temperature can make up the negative impact rainfall DFC it = c + α 1 CD it + α 2 AT it + β 1 Q it + β 2 size it + β 3 grow it + β 4 gs it + β 5 co it + β 6 bs it + β 7 age it + ξ it (1) In Equation (1), DFC it is the cost of debt nancing, CD it , AT it are the annual average temperature change rate and the annual precipitation change rate respectively, Q it, size it , grow it , gs it , co it , bs it , age it are rm market value, rm size, growth, liquid asset ratio, ownership concentration, board of director size and rm age respectively. α 1 , α 2 are the coe cient of the annual temperature change ratio and annual precipitation change ratio, β 1 ∼β 7 are the coe cients of relevant control variables, and ξ it is the residual error. In order to investigate whether rm size can moderate the relationship between climate change and the nancing cost of heavy-polluting rms according to Hypothesis 2, we construct the following model: DFC it = c + α 1 CD it × size it + α 2 AT it × size it + β 1 Q it + β 2 grow it + β 3 gs it + β 4 co it + β 5 bs it + β 6 age it + ξ it (2) According to the employees number employed in industrial enterprises, heavy-polluting rms are divided into larger-size enterprises, medium-sized enterprises and smaller-size enterprises, and whether the change of rm size can adjust the impact of climate change on the nancing cost of heavy-polluting rms. In order to investigate whether rm age can correct the impact of climate change on the nancing cost of heavy-polluting rms according to Hypothesis 3, we construct the following econometric model: DFC it = c + α 1 CD it × age it + α 2 AT it × age it + β 1 Q it + β 2 size it + β 3 grow it + β 4 gs it + β 5 co it + β 6 bs it + ξ it (3) Heavy-polluting rms are divided into younger rms and older rms, and whether rm age will correct the impact of climate change on the nancing costs of heavy-polluting rms.

de nition of relevant variables
Debt nancing cost is de ned by the difference both net capital expenditure and interest income divided by the sum of short-term and long-term debt, used as the measure of debt nancing costs of heavypolluting rms. Greater debt nancing costs imply the higher interest expenditure, conversely, lower debt nancing cost imply lower interest expenditure of heavy-polluting rm's debt.
Climate change is to measure regional climate change on the basis of the annual average temperature change ratio and annual precipitation change ratio in these cities that heavy-polluting rms are registered. The annual average temperature change ratios of heavy-polluting rms are calculated by the difference both the average temperature of the current year and the average temperature of the previous year and then divided by the average temperature of the previous year, which are used to measure the annual average temperature change. The annual precipitation change ratios of the heavy-polluting rms are calculated by the precipitation difference both current year and previous year and then divided the precipitation of the previous year, which are used to measure the annual precipitation change. The greater temperature change ratio and precipitation change ratios re ect greater climate change in these regions, conversely, the lower temperature change ratio and the precipitation change rate re ect the smaller climate change range in these regions.
The control variables mainly include market value, rm size, rm growth, tangible assets ratios, ownership concentration, board of director size and rm age. The rm market value is de ned as the market value of heavy-polluting rm divided by the replacement cost of assets, measure by Tobin Q value. The greater Tobin Q value imply higher market value and good investment opportunities of heavy-polluting rms, and lower debt nancing cost; On the contrary, the smaller Tobin Q value imply lower market value and poor investment opportunities, and higher debt nancing cost. The rm size is de ned as the natural logarithm of the number of employees in heavy-polluting rms; larger heavy-polluting rm size re ects the lower debt nancing cost, conversely smaller rm size re ects higher debt nancing cost. The heavy-polluting rm growth is calculated by the difference of sales revenue both current year and previous year, and then divided by the sales revenue in the previous year. The ratio of tangible assets is calculated by the ratio of tangible assets value to total assets value of heavy-polluting rms. The ownership concentration is calculated by the sum of shareholding ratios of the ve largest shareholders. The rm age is calculated by the difference both current date and initial public offering date, and then divided by 365.  According to the number of employees in heavy-polluting industrial rms, this paper divides total rms into large-size rms (with 1,000 employees), medium-sized rms (300~1000 employees) and small-size rms (less than 300 employees). According to rm's age, this paper divides heavy-polluting rms into younger rms ( rm age < 5 years) and older rms ( rm age> 5 years).    2.4400 and 2.6800 respectively, and their signi cance levels are signi cantly less than 1%. Therefore, this paper conducts empirical analysis on models 1, 2 and 3 using panel data analysis with the xed effects.  Table 4 shows the empirical results of the impacts of annual temperature and precipitation changes on the nancing costs of heavy-polluting rms. In Model 1, the coe cient of annual temperature change to nancing cost of heavy-polluting rms is 0.8396, Temperature change can have a signi cant positive impact on the nancing cost of heavy-polluting rm at the 5% signi cant level. It shows that the annual temperature change increases, the debt nancing costs of heavy-polluting rms increase, continuous hot or severe cold will lead to high-load power consumption of heavy-polluting rms, the soaring price of fossil energy and the shortage of electricity supply cause productivity losses and increase production costs of heavy-polluting rms. Bank's lenders are worried that the continuous hot or severe cold cause certain operating e ciency and expected pro ts losses of heavy-polluting rms decline, and the risk of capital loans will increase, which will raise the debt nancing costs of heavy-polluting rms. With the gradual deepening of China's carbon emission control and carbon trading mechanism, the continuous hot or severe cold prompt bank's lenders to pay more attention to the changes in environmental costs induced by fossil energy consumption and carbon emission control of heavy-polluting rms, which also raise the debt nancing costs of heavy-polluting rms. The coe cients between annual precipitation change and nancing cost of heavy-polluting rms is 0.1164, and annual precipitation change have a signi cant positive impact on the nancing cost of heavy-polluting rms at the 10%signi cant level. This shows that the sharp increase in annual precipitation may bring adverse impacts to heavy-polluting rms. For example, rainfalls in ow into warehouses and lead to accidental damage to inventory materials, and oods lead to tra c paralysis and production interruption, which bring short-term operating losses and pro t decline to heavy-polluting rms, and debtors will increase debt nancing costs because they bear the risk of accidental operating losses. Heavy polluting enterprises need to raise more external funds to improve the ability to market respond quickly and the protection of inventory materials, and then slow down the unexpected business losses induced by the sharp increase in precipitation.

Empirical analysis of climate change and nancing costs of heavy-polluting rms
Model 2 shows that the interaction coe cient between annual temperature change and enterprise scale is 0.1027. Firm scale can adjust the positive relationship between temperature change and nancing cost of heavy-polluting rms at the 5% signi cant level, and rm scale can also signi cantly adjust the positive relationship between precipitation change and nancing cost of heavy-polluting rms at the 10% signi cant level. Compared with the annual temperature coe cient in model 1, The interaction coe cient between rm scale and temperature change decreased greatly, this also indirectly shows that the positive impact of temperature change on nancing cost of heavy-polluting rms tends to weaken with an increase of rm scale. The larger rm scale, the heavy-polluting rms with greater scale can resist the operational risks related to climate change, improve the market adaptability, mitigate the adverse effects of climate change on the operational risks of heavy -polluting rms, and reduce bank lenders' worries about the operational risks of heavy-polluting rms, and then obtain a certain degree of nancing loan concessions. meanwhile, the interaction coe cient between rm scale and precipitation also shows a slight downward trend, the positive impact of precipitation changes on nancing costs of heavy-polluting rms will be weakened with an increase of rm scale. The larger-scale heavy-polluting rms enhance their ability to resist the operational risks induced by the sharp increase in precipitation, and bank lenders' concerns about the operational risks, which is possible to reduce the nancing costs of heavy-polluting rms.
Model 3 indicates that the interaction coe cient between annual temperature change and rm age is 0.0679, and rm age can adjust the positive relationship between temperature change and nancing cost of heavy-polluting rms at the 10% signi cant level. Compared with model 1, the interaction coe cient between rm age and temperature change decreases signi cantly, the positive impact of temperature change on the nancing cost of heavy-polluting rms also shows a weakening trend with an increase of rm age. Moretime, the interaction coe cient between annual precipitation change and rm age is 0.0106, and rm age can adjust the positive relationship between precipitation change and nancing cost of heavy-polluting rms at the 10% signi cant level. Compared with Model 1, the impact of precipitation change on nancing costs of heavy-polluting rms slows down the rising trend with an increase of rm age. Older heavy-polluting rms can improve their market adaptability to climate change, reduce the adverse operational risks induced by climate change, alleviate bank lenders' worries about the expected operational risks of heavy-polluting rms, and reduce the pressure of climate change on the rising nancing costs of heavy-polluting rms.
Moreover, the tangible assets ratio and ownership concentration have a signi cant positive impact on the nancing cost of heavy-polluting rms at the 5%signi cant level, extreme climate change induces heavypolluting rms to enhance their strategic awareness and improve their ability coping with enterprise changes. More funds raised by heavy-polluting rms invest more physical assets to resist climate change, and then mitigate operational risks induced by climate change. Heavy -polluting rms with higher ownership concentration reduce the pressure from external stakeholders, improve the market response speed of organizations and reduce the operational risks induced by climate change. Therefore, tangible assets and ownership concentration reduce the nancing costs of heavy-polluting rms. The larger and older heavy-polluting rms can enhance their adaptability adapting to climate change, enhance their market response speed to climate risks, and slow down the pressure of climate change on the rising nancing costs of heavy-polluting rms.   According to the rm scale division standard, this paper divides 789 heavy-polluting rms into large-size, medium-size and small-size heavy-polluting rms, and makes an empirical analysis on Model 1. Table 5 indicates heavy-polluting rms also shows a gradual weakening trend with the gradual increase of rm scale. The larger heavy-polluting rms have the more diversi ed and multi-regional strategic pattern, the sharp increase of local precipitation have relatively less adverse impact on the nancing activities, investment activities and business activities of larger heavy-polluting rms. On the contrary, the sharp increase of local precipitation have a greater impact on the expected business pro ts of small-size rms.   Note: The values in the parentheses are T statistic value in GMM method; AR (2), Sargan Test statistics are p values in parentheses; ***, **, * indicate passing the test at the signi cant level of 1%, 5%, and 10%, respectively.
In Table 6, Cccording to rm age of heavy-polluting rms, this paper divides heavy-polluting rms into younger and older rms, and makes an empirical analysis of model 1, and the empirical impacts of climate change on the nancing costs of heavy-polluting enterprises under different rm age. The  The empirical results show that the annual temperature and precipitation changes have signi cant positive impacts on the nancing cost of heavy-polluting rms at the 10% signi cant level. The increase of temperature and precipitation signi cantly increase the nancing cost of heavy-polluting rms, and the rm size and rm age also signi cantly moderate the relationship between climate change and the nancing cost of heavy-polluting rms. With the increase of rm scale, the impacts of annual temperature and precipitation changes on the nancing costs of large-size, medium-size and small-size heavy-polluting rms gradually boost, especially the nancing costs of small-size heavy polluting rms induced by climate change greatly increase. With the growth of rm age, the annual climate and precipitation changes gradually increase the nancing costs of older and younger heavy-polluting rms, and climate change can signi cantly increase the nancing costs of younger heavy-polluting rms. Moreover, higher tangible asset ratio and ownership concentration may increase the nancing cost of heavy-polluting rms, while the rm size and rm age may reduce the nancing cost of heavy-polluting rms.
The above empirical results con rm that climate changes have signi cant positive impacts on the nancing cost of heavy-polluting rms, which are helpful for government decision-makers to identify the signi cant impacts of climate changes on the nancing costs of heavy-polluting rms and for rm managers to identify the optimal impacts of climate changes on their nancing costs. This paper puts forward the following relevant policy recommendations: (1) Government decision-makers identify and optimize the transmission effect of climate changes on the nancing behavior of heavy-polluting rms. Policies and governance actions to deal with climate change (such as carbon trading mechanism, energy-saving and emission-reduction policies, compensation system for ecological environment damage, etc.) directly affect the behavior of nancing decisions, investment activities and business activities of heavy-polluting rms. Climate changes signi cantly increase the nancing cost of heavy-polluting rms, Government decision-makers need to identify the transmission effects of climate changes on the nancing decisions of heavy-polluting enterprises in different heavy industries, different types and different environmental regulation ways and intensities. We should optimize policies to deal with climate changes and their governance actions, reduce the adverse effects of climate changes and environmental regulations on the nancing costs of heavy-polluting rms, and adopt regional differentiation strategies to promote the successful governance actions of heavypolluting rms to deal with climate changes.
(2) Local governments should Reduce nancial market friction and nancing discrimination, optimize the e ciency of nancial resource allocation, and reduce the impacts of climate changes on nancial resource allocation. The changes of temperature and precipitation have signi cant differentiated impacts on the nancing costs of heavy-polluting rms with different rm size and rm age. These policies to deal with climate change have important impacts on the nancing, investment, production and operation activities of heavy-polluting rms. Financial institutions may have certain market friction and nancing discrimination in order to effectively control loan risks. The government promotes market competition of nancial institutions, enhance the internal competitive structure and optimization of the banking system, establish nancial resource allocation. Heavy-polluting rms with the response to climate change and climate risks may reduce nancial market friction, reduce nancing discrimination under rm scale, rm age and ownership types of heavy-polluting rms, promote the fairness of nancial resource and reduce the nancing costs of small-size and medium-sized heavy-polluting rms.
(3) Firm managers identify the signi cant impacts of climate changes on the nancing, investment and production and operation activities of heavy-polluting rms, understanding the climate risks and operation risks may result in strategic actions to deal with climate changes, optimize the organizational structure, improve the strategic awareness and rapid response ability to adapt with climate change.
Reducing the adverse effects of extreme climate change and governance actions to deal with climate change may bring to the production, operation, nancing activities and expected operating pro ts of heavy-polluting rms, and comprehensively improve the strategic layout, organizational exibility and rapid response capability of heavy-polluting rms to quickly respond to climate change.
Declaration of interest statement: The authors have no relevant nancial or non-nancial interests to disclose. No con ict of interest exists in the submission of this manuscript, all authors accept that this manuscript is submitted to Journal of Cleaner Production. I would like to declare that this paper is our original unpublished work and it has not been submitted to any other journal.