摘 要

21世纪以来，我国经济快速发展，然而粗放式增长带来的是不合理的产业结构和不平衡的行业发展，并伴随着生态环境的恶化与资源的浪费。我国为了践行可持续发展，发布了许多政策法规以推动绿色经济发展。绿色信贷作为实施绿色金融的工具，通过银行引导资金流向来调节产业结构，对经济发展有着重要影响。我国绿色信贷已发展十余年，其规模不断扩大，体系也逐渐完善，但是在商业银行实施绿色信贷过程中也存在一些问题，例如银行实施绿色信贷积极性不高、绿色信贷规模增速较低、相关政策在地方不够落实等。鉴于此，本文将从目标商业银行的角度研究绿色信贷规模的变动情况，并对发展商业银行绿色信贷提出相关建议。

本文基于绿色信贷相关理论，从两方面梳理出商业银行绿色信贷规模的影响因素，并选取了12个主要可量化指标衡量银行绿色信贷规模变动。在分析了lasso和支持向量回归的特点和适用范围后，以国有政策性商业银行代表中国农业银行和普通股份制商业银行代表兴业银行2011-2018年的绿色信贷余额为样本，先用lasso方法筛选出银行的绿色信贷规模主要影响因素并用模型进行预测，再用支持向量回归针对银行绿色信贷数据训练集进行学习得到预测值，最后利用lasso和支持向量回归的组合模型，将lasso得到的关键变量作为输入进行支持向量回归得到预测结果。

经过实证研究得出结论，兴业银行和中国农业银行的绿色信贷规模共同影响因素为国家补贴、第三产业发展情况和人才占比，而上期资本充足率、不良贷款率和业务转移率也对兴业银行的绿色信贷规模有着显著影响。上述差异源于兴业银行的绿色信贷比率更高、相关部门职能更健全、环境风险管控更完善，而农业银行更多的是政策激励实施绿色信贷。对比上述模型的预测结果误差，得到SVR和组合模型的预测精度更高，模型的MAPE在4%左右。最后基于对两家目标商业银行的实证研究结果，从政府和商业银行自身两个角度对扩大我国商业银行绿色信贷规模提出了几点建议。

关键词：绿色信贷规模；lasso；支持向量回归；组合预测

Abstract

Since the 21st century, China’s economy has developed rapidly. However, this extensive growth has brought about unreasonable industrail structure and unbalanced industrial development, along with the deterioration of ecological environment and waste of resources. In order to practice sustainable development, China has issued many policies and regulations to propel the development of green economy. As a tool of implementing green finance, green credit possesses an important influence on economic development by guiding capital flow through banks to adjust industrial structure. China's green credit has been developing for more than 10 years, its scale is continuously expanding and its system is gradually improving. However, there are still some issues in the course of implementing green credit in commercial banks, such as low enthusiasm of banks in implementing green credit, low growth rate of green credit scale, and insufficient implementation of relevant policies in local areas. In view of this, this paper will study the changes of the scale of green credit from the perspective of the target commercial banks, and put forward relevant suggestions for the development of green credit of commercial banks.

Based on the relevant theory of green credit, this paper combs out the influencing factors of green credit scale of commercial banks from two aspects, and selects 12 main quantifiable indicators to measure the change of green credit scale of banks. After analyzing the characteristics and application scope of lasso and support vector regression, the paper takes the green credit balance of state-owned policy commercial banks on behalf of Agricultural Bank of China and common joint-stock commercial banks on behalf of Industrial Bank in 2011-2018 as samples. First, the lasso method is used to screen out the major influencing factors of banks' green credit scale and the model is used to predict them, then support vector regression is used to learn the bank's green credit data training set to obtain the predicted value, finally, using the combination model of lasso and support vector regression, the key variables obtained by lasso are used as input to support vector regression to get the prediction results.

According to the empirical study, the common influencing factors of the green credit scale of Industrial Bank and Agricultural Bank of China are state subsidies, the development of the tertiary industry and the proportion of talents. The capital adequacy ratio, non-performing loan ratio and business transfer ratio of the previous period also have a significant impact on the green credit scale of Industrial Bank. The above differences are due to the higher green credit ratio of Societe Generale Bank, better functions of relevant departments and better environmental risk control, while Agricultural Bank of China is more policy incentive to implement green credit. Comparing the prediction results of the above models, the SVR and the combined model have higher prediction precision, and the MAPE of the model are about 4%. In the end, in the light of the empirical results of two target commercial banks, this paper gives some advice on expanding the scale of green credit of commercial banks in China from the perspectives of government and commercial banks themselves.

Keywords: green credit scale; Lasso; support vector regression; portfolio forecasting

目 录

第1章 绪论 1

1.1 研究背景和意义 1

1.1.1 研究背景 1

1.1.2 研究意义 1

1.2 研究内容和方法 2

1.2.1 研究内容 2

1.2.2 研究方法 2

1.3 创新之处 3

第2章 文献综述 4

2.1 商业银行绿色信贷及其影响因素的相关研究 4

2.2 Lasso和支持向量回归的相关研究 4

第3章 绿色信贷的内涵及相关理论 6

3.1 绿色信贷的内涵 6

3.1.1 绿色信贷的含义 6

3.1.2 我国绿色信贷的相关政策 6

3.2 绿色信贷相关理论概述 7

3.2.1 可持续发展理论 7

3.2.2 企业社会责任理论 8

3.2.3 环境风险管理理论 8

第4章 我国商业银行绿色信贷实践分析 9

4.1 我国商业银行绿色信贷业务发展现状 9

4.1.1 绿色信贷规模 9

4.1.2 绿色信贷结构 10

4.2 目标银行的绿色信贷业务发展情况 11

4.2.1 中国农业银行的绿色信贷发展概况 11

4.2.2 兴业银行的绿色信贷发展概况 12

第5章 商业银行绿色信贷规模影响因素分析及模型构建 15

5.1 商业银行实施绿色信贷的影响因素 15

5.2 Lasso和支持向量回归以及组合模型的构造 16

5.2.1 Lasso方法 16

5.2.2 支持向量机回归 16

5.2.3 基于lasso和支持向量回归的组合模型 17

第6章 对商业银行绿色信贷规模预测的实证研究 19

6.1 变量选取 19

6.2 针对中国农业银行的绿色信贷规模的实证研究 20

6.2.1 基于lasso回归的预测 20

6.2.2 基于支持向量回归的预测 21

6.2.3 基于组合模型的预测 22

6.2.4 中国农业银行绿色信贷规模预测模型的结果对比分析 23

6.3 针对兴业银行的绿色信贷规模的实证研究 24

6.3.1 基于lasso回归的预测 24

6.3.2 基于支持向量回归的预测 26

6.3.3 基于组合模型的预测 27

6.3.4 兴业银行绿色信贷规模预测模型的结果对比分析 28

6.4 两家银行的对比分析 28

第7章 全文总结及政策建议 30

7.1 全文总结 30

7.2 政策建议 31

参考文献 32

致 谢 34

第1章 绪论

1.1 研究背景和意义

1.1.1 研究背景

我国经济高速发展的同时，环境问题日益凸显。为了实现生态文明和可持续发展，我国开始通过绿色金融大力促进绿色经济的发展。2007年，我国银监会、中国人民银行和环保总局共同颁布了《关于落实环保政策法规防范信贷风险的意见》，绿色信贷由此诞生，成为金融界的发展新课题。

银行业是影响我国经济增长和产业结构的重要行业之一，而股份制商业银行又是我国银行业的领头军，通过配置资金决定我国投融资活动，对我国经济的创新发展起到重要的作用。在我国主要绿色信贷业务银行中，股份制银行占多数，成为我国实施绿色信贷政策的主体。