Research on the Strategy of Industrial Structure Optimization Driven by Green Credit Distribution

Abstract

Credit is an important means to promote economic development, while green credit is conducive to the sustainable development of industry. This paper aims to build a multiple linear regression model and a dynamic panel data GMM estimation model to analyze the important factors that affect the optimization of the industrial structure. We then use an analytic hierarchy process to explore the relationship between green credit and industrial optimization. We compare this with the optimization rate of the industrial structure according to the real data, and then obtain the effectiveness of the hierarchical analysis of the three major industries in the eastern, central and western regions. Finally, neural networks are used to forecast the total amount and distribution of green credit in 2021. The final results show that there are regional and industrial differences in the influence of green credit on industrial structure optimization, and in the process of using green credit to promote the optimization and upgrading of industrial structure.

1. Introduction

Since entering the 21st century, the optimization and upgrading of China’s industrial structure have always been hot topics of social concern. Adjusting the industrial structure is not only a fundamental requirement for China to achieve stable and sustainable economic development, but also a key factor for improving economic efficiency, achieving high-quality development and enhancing the transformation of the model of economic development [1]. At the same time, a healthy industrial structure has a crucial impact on reducing energy, material consumption, controlling environmental pollution and improving the ecological environment. At present, the growth of China’s primary industry is relatively slow, while the secondary industry is growing rather rapidly. Additionally, the tertiary industry is speeding up the development processes in industries such as finance, insurance and consulting, breaking through the old one-way pattern of development, and maintaining a stabilized upgrading speed. However, it is not difficult to recognize that there are still many problems in the current industrial structure of China from both static and dynamic perspectives, and there is still a long way to go for optimization [1,2,3,4].

Across the world, green credit is often called sustainable finance or environmental finance. The goal of green credit is to help and encourage enterprises to reduce energy consumption, save resources and bring ecological and environmental factors into the accounting and decision making of the financial industry. In recent decades, the eco-environmental problems caused by China’s rapid economic growth have given birth to a new pattern of social development—green development. Green development is not only the follow-up to sustainable development but also an innovation of sustainable development with Chinese characteristics. Based on the background of the continuous development of China’s green economy and the vigorous promotion of the optimization and upgrading of the industrial structure, investigating the impact of green credit policies on the upgrading of the industrial structure is undoubtedly of great practical significance and is also an inherent requirement for the sustainable development of Chinese society.

In China, the industrial structure encompasses a large share of industries with high pollution and high energy consumption such as steel, coal and cement production, whereas a small share of the industrial structure consists of environmental protection and green industries. Nonetheless, for optimization’s sake, the central government has proposed to improve the industrial structure by resolving financial constraints in the industrial sector instituted in its 13th 5-year plan. Even though numerous studies have attempted to fish out the connection between the green credit supply and the industrial structure in China [5,6,7,8,9,10,11], no study has taken into account the optimization of the industrial structure’s association with green credit distribution and the theoretical implication in the current theoretical arena.

This paper aims to build a multiple linear regression model and dynamic panel data GMM estimation model to analyze the important factors that affect the optimization of the industrial structure, and then through the analytic hierarchy process and entropy method to carry out the green credit distribution of the three major industries in the eastern, central and western regions. Using the BP neural network, the paper also predicts the total credit amount and distribution in 2021, and finally examines the optimization effect of green credit distribution on the industrial structure.

Lastly, the remaining part of the study is divided as follows: Section 2 explains the hypothesis. Section 3 explicates the research design including the materials and methods, and Section 4 discusses empirical findings. Finally, Section 5 provides a conclusion and policy implications.

2. Research Hypothesis

The green credit of different industries has different effects on the optimization of the industrial structure. While the green credit ratio of the secondary and tertiary industries has a positive effect on the optimization of the industrial structure, the primary industry has a significant inhibitory effect on the optimization of the industrial structure [9,10,11]. When the green credit ratio of primary industry gradually reaches saturation, industrial upgrading will be restrained to a large extent.

The best strategies for attracting foreign direct investment, such as investment subsidies and tax-rate reductions, are determined by the rate of growth and profitability volatility [12]. In the United States, investment tax credits are driving significant growth in the solar photovoltaic and fuel cell industries [13,14]. Diverse incentive policies, such as fiscal and financial incentives, market-based instruments, and other support policies, have been shown to expedite renewable energy capacity investments and technology adaptations [15]. The investment efficiency of new enterprises is influenced by macroeconomic conditions and firm-specific characteristics [16,17]. Therefore, we posit the first hypothesis:

Hypothesis 1.

The effectiveness of industrial structure upgrading is restricted when the green credit ratio of the primary industry reaches saturation.

Due to differences in China’s regional economic development, national policies and cultural legacy, there are also differences in the distribution ratio of green credit in regions of China. Progress and integration of low-carbon energy technology, capacity consolidation, legislative development and feed-in tariffs all help to make renewable energy investment practicable [18,19,20,21,22,23]. Green investment subsidies can be abused and overheated, putting the welfare of green investment projects at risk [24]. Cutting fossil fuel subsidies could boost Central Eastern and North African economies’ gross domestic growth per capita, as well as employment and labor force participation [25]. In China, a green car subsidy scheme was successful in boosting the development of the country’s new energy vehicle industry while also reducing vehicle emissions [26]. When the total amount of green credit is certain, green credit is distributed relatively smoothly in the eastern regions, which are more economically developed, thus resulting in a higher distribution ratio [9,10,11]. The level of economic development in the western regions is relatively low, and the effect of green credit on the optimization of the industrial structure will be somewhat restricted [2,3,4,5,6,7,8]. Therefore, the western regions have the lowest green credit ratio. The central regions rely on potential advantages such as regional resources to obtain a green credit ratio second only to the eastern regions. Therefore, we posit the second hypothesis:

Hypothesis 2.

Regional differences make the distribution of green credit different.

As a tool of green finance, green credit has a considerable positive effect on promoting sustainable economic development and industrial restructuring. Green credit provides preferential interest rates and higher credit ceilings for low-pollution and low-energy-consumption industries while implementing measures such as tightening credit limits and raising loan interest rates for those pollution-intensive and energy-intensive industries [9,10,11]. Currently, fiscal incentive policies, strict environmental regulations and improved investment consciousness in China are driving increased interest in renewable energy investment; as a result, the Chinese renewable energy market has been expanding with significant economic benefits, and China has emerged as a global leader in the renewable energy industry [27]. Using various techniques, many scholars have proved that subsidies and tax-incentive policies can promote renewable energy investment from a macro perspective. The most important elements in the decision-making process for renewable energy investments include higher investment prices, the availability of financing, market stability and subsidy levels, [28,29,30,31,32]. In this way, it contributes much to optimizing and upgrading industrial structures in the direction of greening. Therefore, we posit the third hypothesis:

Hypothesis 3.

Reasonable distribution of green credit will promote the upgrading of the industrial structure.

3. Research Design

3.1. Data Sources and Variable Selection

There are many factors affecting the adjustment of industrial structure. Based on existing research, this paper uses the rate of industrial-structure optimization (ISR) as the dependent variable, while using the amount of green credit and the proportion of the primary, secondary and tertiary industries as the explanatory variables. Among which, the rate of industrial structure optimization is measured by the ratio of added value to GDP of the secondary and tertiary industries to the total GDP; the amount of green credit is the total amount of green credit input of a certain year; the indicators of industrial structure mainly refer to index changes of structures of three industries. This paper is based on the general practice of domestic scholars for reference, which uses the proportion of the output value of each industry in GDP to measure the structure of each industry, namely the proportion of the primary industry’s added value in GDP (AGR), the secondary industry’s added value in GDP (IGR), and the proportion of the tertiary industry’s added value in GDP (SGR) [2]. In addition, considering the availability of data at the government level, this paper also adds two explanatory variables: the proportion of government energy conservation and environmental protection expenditures and the proportion of energy consumption (coal). Among them, the proportion of government energy conservation and environmental protection expenditure reflects the degree of government intervention, which is defined by the ratio of the total energy conservation and environmental protection expenditure to the general public budget expenditure of a certain year; energy consumption mainly takes coal as an example, measured by the proportion of coal consumption of the three major industries in the total energy consumption. Industrial scientific and technological innovation, government financial subsidies, industrial net exports and industrial credit interest expense are used as control variables. The industrial scientific and technological innovation is defined by the number of patents approved by the industry each year; government financial subsidies are defined by the financial subsidies received by the industry each year; industrial net exports are the total export value of the industry minus the total import value of the industry. Industrial credit interest expense = total amount of credit × base interest rate of a certain year [3]. The definition of the variables possibly used in this study can be found in

Table 1. Variable Definition.

	Variable	Definition	Sign
Dependent Variable	Rate of Industrial Structure Optimization	The added value to GDP of the secondary and tertiary industries/Total GDP	ISR
Explanatory Variable	Amount of green credit	The total amount of green credit of a certain year	AGC
	The proportion of primary industry	Green credit of primary industry/Total amount of green credit	AGR
	The proportion of secondary industry	Green credit of secondary industry/Total amount of green credit	IGR
	The proportion of tertiary industry	Green credit of tertiary industry/Total amount of green credit	SGR
	The proportion of government environmental expenditures	Energy conservation and environmental protection expenditure/general public budget expenditure	PEE
	The proportion of coal consumption	Coal consumption/Total energy consumption	CC
Control Variable	Industrial scientific and technological innovation	The number of patents received by the industry each year	ISATI
	The government financial subsidies	Financial subsidy received by the industry each year	FS
	Industrial net exports	Export value-import value	IE
	Industrial credit interest expense	The total amount of credit × Base rate	ICI

.

The data in this paper come from the “China Statistical Yearbook” (2012–2018), the information disclosure document of the China Banking and Insurance Regulatory Commission, which contains the inter-provincial panel data of 31 provinces in the eastern, central and western regions of China from 2012 to 2018. The fiscal expenditures, energy conservation, environmental protection expenditures and general public budget expenditures of provinces and municipalities directly under the Central Government, and minority autonomous regions, come from the fiscal part of the China Statistical Yearbook. The added value and GDP of the three major industries come from the national economic accounting sector of the “China Statistical Yearbook”.

3.2. Model Building

3.2.1. Multiple Regression Model to Determine the Rate of Industrial Structure Optimization

Based on the existing literature, evaluating the rate of industrial structure optimization includes three main criteria: advanced, rational and green. With reference to existing research [4], this paper adopts advanced at the macro level and rational and green at the micro level to determine the evaluation indicators that affect the optimization of the industrial structure. Regarding the optimization and upgrading of the industrial structure, the paper selects rationalization, advancement and greening industrial structures to measure the optimization of the industrial structure. Rationalization is calculated by the redefined Theil index, which reflects the coordinated development of regional industries and the coupling quality of industrial development and employment; advancement mainly considers the importance of the development of the service industry in the regional economic structure. It is measured by the ratio of industrial added value to GDP contribution rate; greenization is defined by GDP per unit energy consumption, reflecting the regional economy.

In order to test for endogeneity bias to ensure the robustness of the findings of our multiple regression method, we employed the Arellano–Bond dynamic panel GMM estimator. With respect to the use of this method, we tend to resolve the issues of endogeneity, serial correlation or autocorrelation (performing the AR(1) and (2) tests). Furthermore, the Sargan test is performed to examine the validity of instruments used in the process—on the other hand, heteroskedascity through the two-step approach. However, the model estimation for the Arellano–Bond dynamic panel data estimation is written as:

$$\begin{array}{ll}{\mathrm{ISR}}_{\mathrm{it}}=& {\displaystyle \sum }_{\mathrm{j}=1}^{\mathrm{p}}{\mathrm{a}}_{\mathrm{j}}{\mathrm{ISR}}_{\mathrm{i}.\mathrm{t}-\mathrm{j}}+{\mathsf{\beta }}_1{\mathrm{AGC}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{AGR}}_{\mathrm{it}}+{\mathsf{\beta }}_3{\mathrm{IGR}}_{\mathrm{it}}+{\mathsf{\beta }}_4{\mathrm{SGR}}_{\mathrm{it}}+{\mathsf{\beta }}_5{\mathrm{PEE}}_{\mathrm{it}}+{\mathsf{\beta }}_6{\mathrm{CC}}_{\mathrm{it}}\\ & +{\mathsf{\beta }}_7{\mathrm{ISATI}}_{\mathrm{it}}+{\mathsf{\beta }}_8{\mathrm{FS}}_{\mathrm{it}}+{\mathsf{\beta }}_9{\mathrm{IE}}_{\mathrm{it}}+{\mathsf{\beta }}_{10}{\mathrm{ICI}}_{\mathrm{it}}+{\mathrm{v}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{it}},\mathrm{i}=1,\dots ,\mathrm{N},\mathrm{t}=1,\dots ,{\mathrm{T}}_{\mathrm{i}}\end{array}$$

(1)

In the model, v stands for the panel level effect, t denotes the time (2012–2018), i represents the cross-section of the 31 provinces in China, ε_it represents the independent and identically distributed (i.i.d.) over the whole data sample with variance σ_ε² and j represents the time lag that will be determined by Arellano–Bond test for the serial correlation.

3.2.2. Advanced at the Macro Level

The advancement of the industrial structure is mainly affected by macro factors. A country’s scientific and technological innovation, green credit investment, overseas investment and factors of different dimensions will determine whether a country’s industrial structure is advanced at the macro level. Here, we choose the total amount of green credit to evaluate whether the degree of the optimization of the national industrial structure is advanced [5].

3.2.3. Rational at the Meso Level

The rationalization of the industrial structure is mainly affected by the meso-level factors, and it reflects the degree of coordinated development between regions and industries and the degree of industrial coupling. Rationality at the level of regional synergy can be used as an evaluation criterion to reflect the optimization of the national industrial structure. Here we use the primary industry green credit ratio $X_1$, the secondary industry green credit ratio $X_2$ and the tertiary industry green credit ratio $X_3$ to reflect whether the degree of national industrial structure optimization is rational [6].

3.2.4. Greening at Micro Level

The greenness of the industrial structure is mainly affected by micro-level factors. According to the literature [7], the industrial structure optimization rate is defined as the dependent variable Y; the ratio of government expenditure on energy conservation and environmental protection to all fiscal expenditures, $X_4$, and the ratio of coal consumed by the three major industries in total energy consumption, $X_5$, are defined as explanatory variables to establish a multiple linear regression equation, which is:

$$Y_t={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+{\beta }_3{X}_3+{\beta }_4{X}_4+{\beta }_5{X}_5+\epsilon$$

(2)

By analyzing the absolute value of the coefficients and the degree of influence on the dependent variable, the optimization rate of China’s industrial structure has a strong correlation with the green credit ratio of the primary industry; the green credit ratios of the secondary industry and the tertiary industry have a relatively great impact on the optimization of the industrial structure; changes in the amount of green credit, the proportion of government environmental protection expenditures and the proportion of energy consumption (coal) hardly affect the optimization of the industrial structure, thus these variables can be eliminated in the analysis process [8]. In general, the optimization of China’s industrial structure has a certain correlation with the green credit ratio of three major industries, but specifically, it has a higher correlation with the green credit ratio of the primary industry. The impact of industrial structure optimization is the largest and negative.

According to the regression results in

Table 2. Regression Results.

Model	Unstandardized Coefficients		Standardized Coefficients	T-Stat	p-Value
	B	S.E.	Beta
Constant	−13.366	2.086		−6.408	0.099
Proportion of primary industry	−34.554	5.392	−1.953	−6.408	0.099
Proportion of secondary industry	7.139	0.908	1.281	7.862	0.081
Proportion of tertiary industry	14.434	1.925	3.185	7.498	0.084
Proportion of government environmental expenditures	0.757	2.059	0.011	0.368	0.776
Proportion of energy consumption (coal)	0.099	0.019	2.858	5.116	0.123

, China’s industrial structure optimization rate has a strong correlation with the primary industry’s green credit ratio. When other factors remain unchanged, if the green credit ratio of the primary industry increases by one unit, the industrial structure optimization rate will decrease by 34.554 units, indicating that when the primary industry’s green credit is saturated, the optimization of the industrial structure will be effectively restrained; the coefficients of the credit ratio of the secondary and tertiary industries are 7.139 and 14.434, for which adding one unit of green credit to the secondary and tertiary industries can increase the industrial structure optimization rate by 7.139 and 14.434 units, indicating that the two can effectively promote industrial structure optimization. Changes in the amount of green credit, the proportion of government environmental protection expenditures and the proportion of energy consumption (coal) have almost no impact on the optimization of the industrial structure, and these variables can be considered to be eliminated.

The findings presented in

Table 3. Robust Check Method—Dynamic Panel Data GMM Estimations.

Arellano-Bond DPD Estimations	Coefficients	T-Stat	p-Value
Constant	−11.386	−6.418	0.089
Proportion of primary industry	−24.854	−5.408	0.089
Proportion of secondary industry	7.189	7.262	0.081
Proportion of tertiary industry	14.434	7.898	0.084
Proportion of government environmental expenditures	0.787	0.378	0.676
Proportion of energy consumption (coal)	0.089	5.616	0.183
Lag of industrial structure optimization	0.023	6.325	0.013
Sargan test	11.623		0.852
AR(1)	−2.856		0.023
AR(2)	−0.478		0.523

outline the GMM dynamic estimations of our data to ensure the resolution of endogeneity, serial correlation and autocorrelation biases that are likely to occur in our main estimation model. From the outcome, we observed that our model was not effected by autocorrelation as the AR(2) test showed a p-value greater than 0.05. Additionally, the Sargan test implies that the instruments used were significant and strong enough to explain the relationship between the dependent and the independent variables because there were no over-identifying restrictions of the instruments.

3.3. Analysis of the Influencing Mechanism of Green Credit Ratio on Industrial Structure Optimization Rate

Based on relevant information [6], Zhang Yunhui and Zhao Jiahui found that there is a non-linear relationship among green credit, technological progress and industrial structure optimization [9]. Long Yunan and Chen Guoqing analyzed the current situation of China’s green finance development and the mechanism and system construction of China’s industrial-structure adjustment promoted by green finance, and established a grey correlation model to perform an empirical analysis of the correlation between green finance and industrial structure [10]. Yu Fei and Huang Ruiling explored the mechanism of the influence of bank credit on the optimization and upgrading of the industrial structure [11]. The green credit ratio of the three industries is mainly affected by the three major system factors of industrial finance fundamentals, scientific and technological innovation and industry environment. The specific influencing factors are presented in

Table 4. Influencing Factors of Green Credit Ratio.

System	Subsystem	Sign	Indicator	Unit
Finance fundamentals	The grey correlation between GDP of the three industries	α	Panel data of grey correlation degree in recent three years
	Loan interest expense	β	Total amount of credit × Base rate	Ten thousand yuan
Scientific and technological innovation	Overseas investment	Ψ	Net exports	Hundred million US dollars
	R&D investment	Φ	R&D expenditure/GDP
	R&D output	Ë	Patents received per million people	Piece
Industry environment	Energy consumption	Ω	Carbon emissions	Ten thousand tons
	Government financial subsidy	μ	Government environmental protection expenditures	Hundred million yuan

below:

With data collected from the statistical yearbook and the social responsibility report of the Chinese banking industry, an analytic hierarchy process (AHP) is adopted to calculate the green credit ratio of the three industries in the new model shown in Figure 1:

According to the related literature, financial fundamentals are usually assigned a weight of 0.5, scientific and technological innovation is assigned a weight of 0.3 and industrial innovation is assigned a weight of 0.2. Using the analytic hierarchy process, it can be concluded that the distribution of green credit is as follows: 10% to the primary industry, 60% to the secondary industry, and 30% to the tertiary industry.

4. Determination of Green Credit Distribution Strategy Based on Industrial Structure Optimization

4.1. Model Building

4.1.1. Analytic Hierarchy Process to Define Regions of the Eastern, Central and Western and Distribution Ratios of Three Industries

Taking into account the regional differences between the eastern, central and western regions, the regional characteristics of the three regions are taken into consideration. This paper selects the regional characteristics of three major regions $Q_{i\left(i=east,mid,west\right)}$ (annual GDP of each region) and local government’s expenditure on energy conservation and environmental protection $C_{i\left(i=east,mid,west\right)}$ as the first-level influencing factors to determine the three major regions, and then based on the results of the new model in “Figure 1: Analysis on the influencing mechanism of green credit ratio on industrial structure optimization rate”, the green credit ratio of three major industries $x_{i\left(i=1,2,3\right)}$ is used as the second-level influencing factor to determine the three major regions. As shown in Figure 2:

1. Entropy Method Determines the Weight of the First Level

Input the normalization matrix: $X=\left|\begin{array}{cc}{x}_{11}& x_{1m}\\ ⋮& ⋮\\ x_{n1}& x_{nm}\end{array}\right|$, where n represents the number of evaluated objects (n = 3). This refers to the three regions, namely the eastern, central and western. m represents the number of evaluation indicators (m = 2), which represents the regional characteristics and the expenditures of local governments on energy conservation and environmental protection. Since $x_{ij}$ in $X$ has all been standardized above, there is no need to standardize it here.

Calculate the proportion of the n-th sample under the m-th indicator and regard it as the probability used in the calculation of relative entropy. That is, calculate the value of each $p_{ij}$ in the probability matrix P:

$$p_{ij}=\frac{x_{ij}}{{{\displaystyle \sum }}_{i=1}^n{x}_{ij}}$$

(3)

Calculate the information entropy of each indicator. The smaller the value, the better, that is, for $j=\left(1,2,\dots ,m\right)$, the formula of the information entropy is:

$$e_j=-\frac{1}{lnn}{\displaystyle \sum }_{i=1}^n{p}_{ij}\ln \left(p_{ij}\right), j=\left(1,2,\dots ,m\right)$$

(4)

Calculate the utility value $d_j$ of information entropy:

$$d_j=1-e_j, j=\left(1,2,\dots ,m\right)$$

(5)

Normalize the information utility and get the entropy weight ${\omega }_{ij}$ of each indicator:

$${\omega }_{ij}=\frac{d_j}{{{\displaystyle \sum }}_{j=1}^m{d}_j}$$

(6)

which is,

$$F_{i\left(i=east,mid,west\right)}={\omega }_{i1}+{\omega }_{i2}$$

(7)

represents the weight coefficient of each of the three major regions in the first level.

2. Determine the Second-Level Influencing Factors of the Three Regions Based on the Green Credit Ratio $x_{i\left(i=1,2,3\right)}$ in the New Model

The weight coefficients ${\beta }_{i\left(i=east,mid,west\right)}^{\prime }$ of three industries are shown in

Table 5. Weight Coefficient of Three Industries.

Industry	Primary Industry	Secondary Industry	Tertiary Industry
Weight	0.1	0.6	0.3

, as follows:

4.1.2. Determine the Comprehensive Weight $W_{i\left(i=east,mid,west\right)}$ of the Three Major Industries in the Three Major Regions

After the two steps in the above, the comprehensive weight coefficient of the three major regions is

$$W_{i\left(i=east,mid,west\right)}=F_{i\left(i=east,mid,west\right)}\times {\beta }_{i\left(i=east,mid,west\right)}^{\prime }$$

(8)

The final weight coefficients of the primary, secondary and tertiary industries in the three major regions are

$$T_{ij\left(i=east,mid,west j=1,2,3\right)}=W_{i\left(i=east,mid,west\right)}\times {\beta }_{i\left(i=east,mid,west\right)}^{\prime }$$

(9)

After programming, the final comprehensive weight coefficients are presented in

Table 6. Comprehensive Weight Coefficient.

Region	East			Central			West
Comprehensive Weight	0.43605			0.31895			0.245
Comprehensive weight of regional industry	East1	East2	East3	Central1	Central2	Central3	West1	West2	West3
Comprehensive weight of regional industry	0.043	0.2616	0.1308	0.0319	0.1914	0.0957	0.0245	0.147	0.0735

:

According to the estimation output in Table 6, when the total amount of green credit is certain, there are differences in the reasonable proportion of credit in the eastern, central and western regions, which should be related to factors of China’s economy, culture, politics and geographic location. The eastern region of China has obvious geographical advantages. It is an economically developed region in China and is at the forefront of opening up. It is a “test field” and “demonstration zone” for reforms in many fields, and has formed a good financial-development environment. Under this circumstance, the distribution of green credit is relatively smooth, and the role of green credit in optimizing the industrial structure can be better brought into play. Therefore, the proportion of green credit in the eastern region is relatively high, about 43.6%. The economic base of the central region is relatively poor, but it has absolute advantages in regional economic development in China and has potential advantages in agriculture, forestry and mineral resources, tourism resources and low labor costs. Therefore, green credit distribution is second only to the eastern region, about 31.9%. The development of the western region mainly relies on the support of national policies. From the economic development of the western region in recent years, it can be seen that the support of national policies has played a significant role, and because of the importance of the green credit policy as a tool of country’s sustainable development of green economy, appropriate green credit is conducive to the upgrading and optimization of the industrial structure in the western region. For three major industries, the proportion of green credit is almost the same in all regions. The secondary industry that has the greatest impact on the optimization of the industrial structure has the highest proportion of credit, primary industry has the lowest proportion of credit and the tertiary industry is in between.

4.2. Secondary Distribution of Industries under the Secondary Industry

Credit is an important means to promote economic development. Almost all industries need credit for development. Of course, different industries rely on credit to different degrees. According to relevant data, the industries that are actively supported by credit include electricity and tourism, which shows that the secondary industry relies heavily on credit. In order to simplify statistics and calculations, the secondary industry is divided into four industries: mining, energy (electricity, gas, and water production and supply), construction and manufacturing. The original green credit distribution ratio is defined as the ratio of each of the four industries’ annual GDP to the total GDP of the second industry. $R_{i\left(i=1,2,3,4\right)}$ is defined as the initial green credit distribution ratio, and then the ratio of the annual GDP of the above four industries is compared to the total GDP of the second industry $U_{i\left(i=1,2,3,4\right)}$ with the previous year $U_{i\left(i=1,2,3,4\right)}^{\prime }$; the variation range is:

$$∆U_{i\left(i=1,2,3,4\right)}=U_{i\left(i=1,2,3,4\right)}-U_{i\left(i=1,2,3,4\right)}^{\prime }$$

(10)

Define the adjusted distribution ratio as $R_{i\left(i=1,2,3,4\right)}^{\prime }$. The third distribution model for industries under the secondary industry is then as follows:

$$\left\{\begin{array}{c}{L}_{i\left(i=1,2,3,4\right)}=R_{i\left(i=1,2,3,4\right)}^{\prime }\times H\\ R_{i\left(i=1,2,3,4\right)}=R_{i\left(i=1,2,3,4\right)}\times \left(1+\delta \right)\\ -0.2\le \delta \le 0.2\end{array}\right.$$

(11)

Among them, $L_{i\left(i=1,2,3,4\right)}$ is the green credit line allocated to the four industries, and $H$ is the total amount of green credit allocated to the secondary industry in a certain year. δ is the ratio of the annual GDP of the four industries to the total GDP of the second industry $U_{i\left(i=1,2,3,4\right)}$ compared to the previous year; the variation range is a random number generated by computer simulation. Its meaning is that if $∆U_{i\left(i=1,2,3,4\right)}>0$, then a random number $-0.2<\delta <0$ will be randomly generated by the computer, indicating the control green credit’s scale to an industry where GDP is of certain growth. The random number is drawn out and transferred to industries with $∆U_{i\left(i=1,2,3,4\right)}<0$; thus when $∆U_{i\left(i=1,2,3,4\right)}<0$, a random number $0<\delta <0.2$ is randomly generated by the computer to indicate support for industries with a decline in GDP. According to the above model, the distribution results are as follows:

In Figure 3 below, the red line represents the distribution results based on the model, and the blue line represents the distribution results based on the actual data. As can be seen from the figure, the credit scale of the construction industry and mining industry in the secondary industry occupies a small share in both actual distribution and model distribution, and the supply industry and the manufacturing industry together occupy more than half of shares. According to the green credit distribution share in the model, there should be somewhat more shares of the green credit scale allocated to electricity, gas and water each year. From the figure, we can see that there have always been large deviations between the scale of manufacturing loans in the model and the actual scale, and this is mainly due to the fact that the manufacturing industry is affected by external factors such as government guidance and the overall economic environment, which leads to the fluctuation of the model and the actual scale.

4.3. Neural Network Forecasts the Total Credit Amount and Distribution in 2021

This study selects a total of 8 years of data from 2011 to 2018 for the forecast. Based on the green credit of each of the three major industries from 2011 to 2016, $S_{t,i} \left(t=1,2,\dots ,6;i=1,2,3\right)$, the ratio of government expenditure on energy conservation and environmental protection to all fiscal expenditures, $A_t \left(t=1,2,\dots ,6\right)$, and the ratio of coal consumed by the three major industries in total energy consumption in 2011–2016, $U_t \left(t=1,2,\dots ,6\right)$, as input values; the total amount of green credit for each year from 2011 to 2016 is output as the expected value $E_j \left(j=1,2,\dots ,6\right)$ for neural network training; input $new{S}_{t,i}\left(t=7,8;i=1,2,3\right)$, $new{A}_t\left(t=7,8\right)$, $new{B}_t\left(t=7,8\right)$ by using the remaining two sets of data as the experimental group, and then the neural network is used to obtain the new expected value $new{E}_j\left(j=7,8\right)$, which is the predicted total amount of green credit $M$, which is evaluated by comparing it with the true value $E_j\left(j=7,8\right)$ from the social responsibility report of the Chinese banking industry, namely:

$$∆E_j=\frac{\left|new{E}_j-E_j\right|}{E_j}\left(j=7,8\right)$$

(12)

Because $∆E_j\le 5\%$, it is reasonable to use this method to predict the total amount of green credit in 2021.

According to the neural network, the prediction results of the analysis are presented in Figure 4:

The overall forecast results of the neural network are shown in Figure 5:

The forecasted green credit scale of 2021: 1,068,485.814 million. Through the neural network, green credit scale of China in 2021–2025 is predicted as shown in the following table.

It can be seen from

Table 7. Forecasted Green Credit Scale by Neural Network, 2020–2025.

2020	2021	2022	2023	2024	2025
901,891.8979	1,068,485.814	1,265,416.22	1,498,277.805	1,773,738.228	2,099,738

and Figure 6 that the green credit scale in 2021–2025 is steadily increasing, and it is expected that the scale will exceed 2 trillion in 2025, which means that the green credit scale in 2025 will be double that of 2020 and four times the green credit scale in 2016. It is important to take into comprehensive consideration the three dimensions: green credit policy at the macro level; the regional characteristics of the east, central and west regions; the local government’s investment in energy conservation and environmental protection at the macro level; and the industrial environment (energy consumption, government financial subsidy), scientific and technological innovation (overseas investment, R&D investment, R&D output) and financial fundamentals (grey correlation of GDP of the three major industries, loan interest expense) at the micro-level. These three dimensions suggest that we should focus on investing in green credit in the next five years, so as to ensure the amount of green credit investment matches and follows up the process of optimization and upgrading of the industrial structure, thereby promoting the steady development of green finance and achieving the green upgrading of other industrial structures. It is essential to promote the steady development of green finance and advanced, rational and green upgrading of the industrial structure.

Taking green credit scale of 1,068,485.814 million forecasts by the neural network in 2021 as an example, and according to the comprehensive weight distribution table (Table 7), the amount allocated to the first, second and third industries in eastern, central and western regions is shown in

Table 8. Green Credit Distribution Result, 2021 Unit: Million Yuan.

Region	East1	East2	East3	Central1	Central2	Central3	West1	West2	West3
Amount	36,006.15	216,036.92	108,018.46	26,336.80	158,020.81	79,010.41	20,230.5	12,138.3	60,691.49

below. For example, East1 means that the green credit line allocated to the primary industry in the eastern region in 2021 is 36,006.15 million, and the rest can be deduced from the above. Among them, the detailed distribution ratios of the secondary industry supported by green credit are shown in Figure 6. The distribution amount of green credit by industry in 2021 is shown in Table 8:

According to the forecast results, when the total amount of green credit is certain, the eastern region is distributed the most, the central region the second and the western region the least. Among them, the distribution of the secondary industry in each region is the most, followed by the tertiary industry, and the primary industry is the least. The forecast results are consistent with the results in the model in this paper. In summary, the development of China’s economy mainly relies on the eastern region. The eastern region has become the main force of China’s economic development with its unique regional, cultural and political advantages. It should have a certain tendency in the country’s support for economic development. At the same time, in response to the national green credit policy and the requirements of green credit to promote economic development, green credit should also occupy a major position in the eastern region.

5. Conclusions and Policy Recommendations

5.1. Conclusions

In conclusion, our research shows that green credit distribution in China can effectively affect the upgrading of the industrial structure. Among other things, green credit has a significant promotion effect on the secondary and tertiary industries, and when the green credit ratio of the primary industry reaches saturation, it will curb the optimization and upgrading of the industrial structure. When the total amount of green credit is certain, there are obvious regional differences in distribution. Among these, the eastern part has the highest distribution proportion, followed by the central part and the western part has the lowest proportion. In all regions, green credit distribution is almost the same in the three industries. The secondary industry that has the greatest impact on the optimization of the industrial structure has the highest proportion of green credit, the primary industry has the lowest proportion of credit and the tertiary industry is in the central part. Both the model and the actual data show that in the secondary industry, with the highest proportion of green credit, credit is distributed mainly in the supply industry and manufacturing industry, while credit distribution in the construction industry and mining industry is relatively small. The production and supply industries of electricity, gas and water should occupy more shares of green credit. We find our study to consistent with the findings of [33,34,35].

5.2. Policy Recommendations

Firstly, take measures tailored to local conditions, consider the regional structural characteristics of each region and the strength of government support in each region, and comprehensively analyze the regional credit distribution indicators from the external factors of environmental factors and carbon emissions. It is necessary to consider the regional characteristics thoroughly, in multi-level and multi-indicator fashion, and formulate a corresponding credit distribution ratio and implement the credit distribution policy based on regional differences [36,37].

Secondly, consider the characteristics of the industry and determine the distribution of green credit from the perspective of industry fundamentals, while at the same time combining the characteristics of the industry, examining the characteristics of the industry and the future innovation capabilities of the industry from the perspective of the current situation of the industry and the future development of the industry, so as to distribute credit quota from the macro and micro level [38,39,40]. Finally, continuously optimize the scientific and technological innovation environment, accelerate the establishment and optimization of a collaborative innovation mechanism for government, industry, university and R&D, strengthen the protection of intellectual property rights and promote scientific and technological innovation, which leads to industry development and innovation, and further industry optimization and upgrading.

Thirdly, promote high-quality economic development. Based on the development of the real economy, seize the core competitive sectors of high-end manufacturing in the secondary industry, such as new energy, and drive the optimization and adjustment of the industrial structure. Vigorously develop science and technology to promote the development of high-tech and high-value-added industries, optimize the industrial chain, and promote the transformation of the industrial structure to advanced. Based on the development principles of service of the secondary industry, economization of the tertiary industry and greenization of the economy, it is necessary to continue to promote the construction of ecological civilization for industrial development, and insist on energy conservation and emission reduction. On the one hand, we should eliminate backward production capacity and use new technologies to optimize high-energy-consuming industries. On the other hand, we will actively develop and utilize new energy resources to replace fossil fuels. While responding to the energy crisis, we will also promote green economic development and improve the greening of industrial structure. Implement a proactive opening-up policy to push forward a comprehensive, multi-level, and wide-ranging new pattern of opening-up, create a good international environment for all-round development and open up broad development space.