Green loan risk assessment and pricing system

Abstract

The invention discloses a green loan risk assessment and pricing system, and the system comprises a data collection module which collects financial data, industrial chain information and environmental data of a green loan application enterprise, historical data of green loan and bank green financial products, and related data of carbon emission and carbon absorption of a loan enterprise; the data processing module performs data preprocessing; the feature extraction module extracts features of the data and constructs feature vectors; the model training module trains a deep learning model of risk assessment and pricing through the feature vectors; and the evaluation and pricing module performs risk, carbon absorption potential and carbon balance risk evaluation on the new loan application by adopting the model, performs loan pricing in combination with key technical factors of environmental benefits and green productivity, and formulates a green loan business strategy according to an evaluation result and a loan pricing suggestion output by the model. According to the method, comprehensive, detailed, accurate and scientific risk assessment can be carried out on green loans.

Description

Technical Field

[0001] The present invention belongs to the field of fintech, especially the field of green finance, and specifically relates to a green loan risk assessment and pricing system. Background Art

[0002] With the global attention to climate change, green finance has become an important development direction in the financial industry. As an important part of green finance, the scientificity and accuracy of the risk assessment and pricing mechanism of green loans are crucial for the sustainable development of banks.

[0003] However, existing green loan risk assessment and pricing systems often rely on traditional credit assessment models, lacking comprehensive consideration of environmental factors and policy orientations, and unable to meet the rapid development needs of green loan business.

[0004] At the same time, in the current risk rating system of the banking industry, the traditional five-level risk loan classification mainly focuses on individual loan businesses, with a relatively narrow assessment scope. This single-perspective risk assessment method is difficult to comprehensively reflect the complexity and potential risks of green finance business. In addition, existing green loans in the banking industry only focus on the assessment of the carbon emission field and lack attention to the carbon absorption field. Summary of the Invention

[0005] Aiming at the deficiencies that the existing banking industry only conducts loan credit assessment based on the historical financial income and expenditure data of applying enterprises, and only focuses on individual loan businesses and lacks attention to the carbon absorption field during the assessment, the present invention proposes a green loan risk assessment and pricing system. By collecting the financial data, industrial chain information, and environmental data of green loan applying enterprises, collecting the historical data of green loans and bank green financial products, and collecting the carbon emissions and carbon absorption of loan enterprises, it can realize the comprehensive risk rating of green financial product portfolios based on multi-dimensional data, provide more comprehensive and detailed risk assessment results for banks, and expand its assessment scope to the entire green finance product chain to construct a more comprehensive and accurate risk assessment framework. At the same time, this system adds the carbon absorption field to the assessment model, details the newly added paired carbon absorption of green loan enterprises, and then actively pairs with carbon absorption enterprises in the region through a deep learning model, so as to achieve a more scientific artificial intelligence large model for green loan risk assessment and pricing.

[0006] The green loan risk assessment and pricing system described in the present invention includes: A data collection module: used to collect the financial data, industrial chain information, and environmental data of green loan applying enterprises, collect the historical data of green loans and bank green financial products, and collect the data related to the carbon emissions and carbon absorption of loan enterprises; Data Processing Module: It is used to preprocess the collected data. The data processing includes: cleaning, denoising, removing outliers and missing values, standardization processing, and normalization processing; Feature Extraction Module: It is used to extract features from the data after preprocessing and construct feature vectors for the input of subsequent deep learning models; Model Training Module: It is used to design a deep learning model for risk assessment and pricing that includes convolutional layers and recurrent layers, and train the model with the feature vectors; Evaluation and Pricing Module: It is used to conduct risk assessment on new loan applications, as well as carbon absorption potential assessment and carbon balance risk assessment using the trained model, and perform loan pricing by combining key technical factors of environmental benefits and green production capacity. According to the evaluation results and loan pricing suggestions output by the model, formulate green loan business strategies.

[0007] Furthermore, it includes a Model Optimization Module: It is used to dynamically update the model according to market changes and policy orientations, and adjust the loan credit limit and pricing using optimization algorithms based on the risk assessment results, and collect the carbon emissions and carbon absorption data of enterprises after loans for model feedback and iterative optimization.

[0008] Furthermore, the formula used in adjusting the loan credit limit and pricing using optimization algorithms based on the risk assessment results is shown as formula (1) below: (1) In the formula, is the loan interest rate, is the base interest rate, is the natural constant, is the risk score, is the market carbon credit price, and are adjustment parameters.

[0009] Furthermore, the Evaluation and Pricing Module specifically includes: performing loan pricing based on the comprehensive risk rating, market interest rate, and green project yield output by the model. The formula used is shown as formula (2) below: (2) In the formula, is the loan pricing, is Risk_Rating, i.e., the comprehensive risk rating, is Market_Rate, i.e., the market interest rate, is Green_Project_ROI, i.e., the green project yield, 、 、 、 is the regression coefficient, is the error term.

[0010] Furthermore, the carbon absorption potential assessment includes evaluating the carbon absorption potential increased by enterprises through biological carbon sequestration and CCUS technologies, and the carbon balance risk assessment includes using a deep learning model to conduct risk assessment based on the carbon emission potential and carbon absorption potential of enterprises.

[0011] Furthermore, the formula used in the carbon absorption potential assessment is shown as formula (3) below: T (3) In the formula, is the carbon absorption potential, is the forest coverage rate, T is the technical carbon sequestration capacity, , , are model parameters; The formula used in the carbon balance risk assessment is shown as formula (4) below: (4) In the formula, R is the carbon balance risk score, e is the natural constant, is the carbon emission, is the carbon absorption potential, is the policy impact factor, is the market sensitivity factor, , , , are weight parameters.

[0012] Furthermore, it also includes a carbon absorption enterprise pairing module, which is used to identify and pair carbon absorption enterprises in the bank's private information database using an artificial intelligence algorithm. The formula used is shown as formula (5) below: (5) In the formula, is the minimum value among all pairing schemes, that is, the value corresponding to the optimal pairing scheme, is the total number of carbon emission enterprises applying for loans, is the i-th carbon emission enterprise applying for a loan, is the j-th carbon absorption enterprise in the bank's private information database, is the distance function between the i-th carbon emission enterprise applying for a loan and the j-th carbon absorption enterprise, is the cost function of the j-th carbon absorption enterprise, is the trade-off parameter.

[0013] Furthermore, the model training module includes: introducing a Dropout layer to reduce overfitting, constructing a model in combination with an attention mechanism, and optimizing model parameters using a backpropagation algorithm.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention proposes a green loan risk assessment and pricing system. By collecting the financial data, industrial chain information, and environmental data of green loan application enterprises, as well as the historical data of green loans and bank green financial products, and collecting the carbon emissions and carbon absorption of loan enterprises, and then through data preprocessing and feature extraction of these data for the training of a deep learning model, finally, the trained model is used to conduct risk assessment, carbon absorption potential assessment, and carbon balance risk assessment on new loan applications, and loan pricing is carried out in combination with the key technical factors of environmental benefits and green production capacity. According to the evaluation results and loan pricing suggestions output by the model, a green loan business strategy is formulated. 2. The present invention can achieve a comprehensive risk rating of the green financial product portfolio based on data from multiple dimensions, providing a more comprehensive and detailed risk assessment result for banks; and can expand its evaluation scope to the entire green financial product chain to construct a more comprehensive and accurate risk assessment framework.

[0015] 3. The present invention adds the carbon absorption field to the evaluation model, details the newly added paired carbon absorption of green loan enterprises, and then actively pairs with carbon absorption enterprises in the region through a deep learning model, so as to achieve a more scientific artificial intelligence large model for green loan risk assessment and pricing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a functional module diagram of a green loan risk assessment and pricing system provided by the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS In order to make the technical means, creative features, achieved purposes, and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0017] Such as Figure 1As shown in the figure, a green loan risk assessment and pricing system provided by the present invention includes: a data collection module, a data processing module, a feature extraction module, a model training module, and an evaluation and pricing module. After the data collection module collects various data related to green loans, it sends the data to the data processing module for data preprocessing. After the data processing module completes the data preprocessing, it inputs the data into the feature extraction module. After the feature extraction module extracts features from the input data and constructs a feature vector, it sends the feature vector to the model training module to train the model. After the model training module trains the model, the evaluation and pricing module uses the trained model to conduct risk assessment, carbon absorption potential assessment, and carbon balance risk assessment on new loan applications, and combines key technical factors of environmental benefits and green production capacity for loan pricing. Then, according to the evaluation results and loan pricing suggestions output by the model, a green loan business strategy is formulated.

[0018] Data collection module: used to collect the financial data, industrial chain information, and environmental data of green loan application enterprises, collect the historical data of green loans and bank green financial products, and collect data related to carbon emissions and carbon absorption of loan enterprises; Specifically, in this module, first, the initial data of green loan application enterprises and the entire industrial chain are collected. Specifically, the financial data, industrial chain information, environmental data, etc. of green loan application enterprises are collected. These data specifically include operating income, net profit, asset-liability ratio, energy consumption, etc.

[0019] Secondly, the historical data of green loans and bank green financial products are collected. For example, the green loan data for the past five years are exported from the bank's internal system, specifically including basic data and information such as the loan amount, loan term, interest rate level, and repayment records of historical green loans.

[0020] Finally, data related to carbon emissions and carbon absorption of loan enterprises are collected, specifically including but not limited to data information such as energy consumption, production process, forest coverage rate, and CCUS facilities.

[0021] Data processing module: used to perform data preprocessing on the collected data. The data processing includes: cleaning, denoising, removing outliers and missing values, standardization processing, and normalization processing; Specifically, in this module, the collected data are sequentially cleaned, that is, duplicate, incorrect, or severely missing records are removed, denoised, outliers and missing values are removed, and then standardization processing or normalization processing is performed to ensure that data from different sources and with different dimensions can be compared and analyzed on the same scale. Through this series of data preprocessing, the data quality can be effectively improved.

[0022] Among them, the following formula is used in the standardization processing: In the formula, is the standardized data, is the original data, is the mean value, is the standard deviation.

[0023] In other embodiments, other different standardization processing methods can also be adopted, that is, the Min - Max standardization method. The formula used is as follows: By using different standardization processing methods, the influence of different standardization processing methods on the model performance can be tested.

[0024] Feature extraction module: used to extract features from the pre - processed data and construct feature vectors for the input of subsequent deep learning models; Specifically, the PCA method can be used to extract features from the pre - processed data. PCA (principal components analysis), that is, the principal component analysis technology, also known as the principal component analysis, aims to use the idea of dimensionality reduction to transform multiple indicators into a few comprehensive indicators. Therefore, feature selection and dimensionality reduction can be carried out through PCA. In actual operation, it can be to use the PCA method to reduce the dimension to 10 main features.

[0025] Among them, the following formula is used in the PCA method: In this formula, is the principal component score, is the original data, is the mean vector, is the eigenvector, is the eigenvalue.

[0026] Furthermore, when extracting features, key features are emphasized, such as loan term, green project type (such as renewable energy, energy - saving buildings, etc.), percentage reduction of carbon emissions, trend of carbon trading price changes, etc.

[0027] Model training module: used to design a deep learning model for risk assessment and pricing including a convolutional layer and a recurrent layer, and train the model through the feature vector; Specifically, in this embodiment, the deep learning model includes a convolutional layer and a recurrent layer, that is, a convolutional neural network (CNN) or a recurrent neural network (RNN). In the deep learning model, the input layer contains selected features, and the output layer is the risk score. After data pre - processing and feature extraction of the above - collected data, the deep learning model is trained. In this process, the Adam optimizer and the cross - entropy loss function can be adopted.

[0028] During the training process, by introducing relevant environmental data as input features, the model can learn the impacts of factors such as carbon emissions, changes in carbon emission policies, and fluctuations in carbon market prices on the risk of green loans. The specific formula used is as follows: In the formula, is the output, x is the input feature vector, is the weight matrix of the convolutional kernel, is the bias term, is the activation function.

[0029] During the training process, by continuously adjusting model parameters such as the weight matrix and the bias term , the prediction error can be minimized.

[0030] Through multiple iterations of training, the model gradually converges and exhibits good generalization ability on the validation set. The formula used for iterative training is as follows: In the formula, is the loss function, is the true label, is the probability predicted by the model.

[0031] In an alternative embodiment, the model training module includes: introducing a Dropout layer to reduce overfitting, constructing the model in combination with an attention mechanism, and optimizing model parameters using the backpropagation algorithm.

[0032] Specifically, in this embodiment, according to the data characteristics and problem complexity, we select a recurrent neural network combined with an attention mechanism (RNN-Attention) as the basic architecture of the risk assessment and pricing model. This architecture can effectively capture long-term dependencies in time series data and highlight important information through the attention mechanism.

[0033] In addition, during the training process, in addition to introducing a Dropout layer to reduce overfitting, early stopping or adding regularization terms can also be used to prevent overfitting.

[0034] Evaluation and Pricing Module: Used to perform risk assessment, carbon absorption potential assessment, and carbon balance risk assessment on new loan applications using the trained model, and perform loan pricing by combining key technical factors of environmental benefits and green production capacity. According to the evaluation results and loan pricing suggestions output by the model, formulate green loan business strategies.

[0035] Specifically, after using the trained model to conduct a risk assessment on a new green loan application, loan pricing can be performed by combining the key technical factors of environmental benefits and green production capacity, and its expression is as follows: In the formula, is the loan pricing, is the risk score, is the environmental benefit, is the technical factor.

[0036] Furthermore, in an alternative embodiment, the evaluation and pricing module specifically includes: performing loan pricing based on the comprehensive risk rating, market interest rate, and green project yield rate output by the model, and the formula used is as shown in formula (2) below: (2) In the formula, is the loan pricing, is Risk_Rating, i.e., the comprehensive risk rating, is Market_Rate, i.e., the market interest rate, is Green_Project_ROI, i.e., the green project yield rate, , , , are the regression coefficients, is the error term.

[0037] Among them, , , , can be obtained by historical data regression, is the error term with a randomly selected value.

[0038] In this embodiment, when a new green loan application is submitted, the system will automatically extract relevant features and input them into the trained deep learning model. The model will output a comprehensive risk rating, which comprehensively considers the credit status of the loan applicant and the impact of the environmental benefits of the green project. The model will also perform pricing prediction and recommendation through a formula based on factors such as the current market interest rate and the expected yield rate (Green_Project_ROI) of the green project. Based on the risk assessment results and pricing recommendations output by the model, the bank's credit department can quickly make a loan approval decision in combination with the actual situation to formulate a personalized green loan business strategy.

[0039] Through this embodiment, the present invention can achieve accurate assessment and pricing of green loan risks, improve the efficiency and security of the bank's green loan business, and promote the sustainable development of green finance.

[0040] In an alternative embodiment, the carbon absorption potential assessment includes assessing the carbon absorption potential increased by an enterprise through biological carbon sequestration and CCUS technologies, and the carbon balance risk assessment includes using a deep learning model to conduct a risk assessment based on the enterprise's carbon emission potential and carbon absorption potential.

[0041] In an alternative embodiment, the formula used in the carbon absorption potential assessment is as shown in the following formula (3): T (3) In the formula, is the carbon absorption potential, is the forest coverage rate, T is the technical carbon sequestration capacity, , , are model parameters; Among them, , , can be preset in advance through an artificial intelligence model. For example, = 0.5, = 1.2, = 0.3.

[0042] The formula used in the carbon balance risk assessment is as shown in the following formula (4): (4) In the formula, R is the carbon balance risk score, e is the natural constant, is the carbon emission, is the carbon absorption potential, is the policy impact factor, is the market sensitivity factor, , , , are weight parameters.

[0043] Among them, , , , can be preset in advance through an artificial intelligence model. For example, = 0.4, = 0.3, = 0.2, = 0.1.

[0044] In an alternative embodiment, it further includes a carbon absorption enterprise pairing module, and the carbon absorption enterprise pairing module is used to identify and pair carbon absorption enterprises in the bank's private information database by using an artificial intelligence algorithm. The formula used is as shown in the following formula (5): (5) In the formula, is the minimum value among all pairing schemes, that is, the value corresponding to the optimal pairing scheme, is the total number of carbon emission enterprises applying for loans, is the i-th carbon emission enterprise applying for a loan, is the j-th carbon absorption enterprise in the bank's private information database, is the distance function between the i-th carbon emission enterprise applying for a loan and the j-th carbon absorption enterprise, is the cost function of the j-th carbon absorption enterprise, is the trade-off parameter. Among them, The value of can be 0.6.

[0045] Specifically, the pairing of carbon emission enterprises and carbon absorption enterprises can be achieved through the carbon trading market. The carbon trading market is an effective means to reduce carbon emissions through market mechanisms. The government sets an emission cap (quota) for enterprises. Enterprises that exceed the quota need to purchase additional carbon rights in the carbon market, while enterprises that have not used up their quotas can sell their remaining carbon rights, thus promoting emission reduction.

[0046] Among them, the operation mechanism of the carbon trading market is as follows: 1. Quota management: The government sets a greenhouse gas emission cap (quota) for enterprises, and the amount of carbon that enterprises can emit each year is restricted.

[0047] 2. Carbon right trading: Enterprises that exceed the quota need to purchase additional carbon rights in the carbon market, while enterprises that have not used up their quotas can sell their remaining carbon rights.

[0048] 3. Carbon sink projects: Enterprises can increase carbon absorption by investing in carbon sink projects (such as afforestation) to offset part of their carbon emissions.

[0049] Through the pairing method in the pairing module designed in this embodiment, it is convenient for banks to quickly find the optimal pairing scheme in the carbon trading market, which is conducive to promoting the activity of the carbon trading market and promoting the development of green finance.

[0050] In an optional implementation manner, the system further includes a model optimization module: used to dynamically update the model according to market changes and policy orientations, and adjust the loan credit limit and pricing using an optimization algorithm based on the risk assessment results, and collect the carbon emissions and carbon absorption data of enterprises after loans for feedback and iterative optimization of the model.

[0051] Among them, the following iterative formula can be adopted: In the formula, are the updated model parameters, is the old model parameter, is the learning rate, is the gradient of the loss function. Among them, can take the value of 0.01.

[0052] In an optional implementation manner, the formula adopted for adjusting the loan credit limit and pricing according to the risk assessment result using the optimization algorithm is as shown in the following formula (1): (1) In the formula, is the loan interest rate, is the base interest rate, is the natural constant, is the risk score, is the market carbon credit price, and are adjustment parameters. Among them, can take the value of 0.05, can take the value of 0.03.

[0053] In summary, the present system can achieve accurate assessment and pricing of green loan risks, while promoting the activity of the carbon market and driving the development of green finance.

[0054] In summary, the green loan risk assessment and pricing system provided by the present invention has the following five beneficial effects compared with the prior art: 1. Full industrial chain analysis: Different from the existing banking industry that conducts loan credit assessment based on the historical financial income and expenditure data of applying enterprises, the present system provides a dedicated large model for green loans. Through the training of the bank's green loan private information database model, it conducts a full industrial chain green production capacity analysis on the industrial chain where the applying enterprise is located. Through the deep learning algorithm, the problem handling of the green loan information database, business processes, industry information of the industrial chain where the applying enterprise is located, and key technologies of green production capacity form the accumulation of green loan experience; the model architecture of green loans, financial knowledge concepts, industry market logic, and professional terms form dynamic knowledge input to achieve unsupervised learning and automatic update of artificial intelligence.

[0055] 2. Full product chain analysis: In the current risk rating system of the banking industry, the traditional five - level risk loan classification mainly focuses on individual loan business, with a relatively narrow evaluation scope, making it difficult to comprehensively reflect the complexity and potential risks of green finance business. Compared with this single - perspective risk assessment method, this system relies on advanced deep - learning algorithms, not only paying attention to the risk characteristics of green loans themselves, but also expanding its evaluation scope to the entire green finance product chain, covering diversified financial products such as green bonds, green funds, carbon finance special products, and green leases. Through the powerful learning ability of the deep - learning model, the system can deeply explore the internal connections and mutual influences among these financial products, thus constructing a more comprehensive and accurate risk assessment framework.

[0056] In specific implementation, the system will first collect and integrate relevant data of these green finance products, including historical performance, market environment, policy orientation and other dimensions. Subsequently, it uses deep - learning algorithms to deeply mine and analyze these data, extract key risk characteristics, and establish corresponding risk models. On this basis, the system can achieve a comprehensive risk rating of the green finance product portfolio, providing banks with more comprehensive and detailed risk assessment results. This cross - product - line comprehensive risk assessment method not only helps banks more accurately grasp the overall risk level of green finance business, but also effectively reduces the risk of individual green loan business. By optimizing resource allocation and strengthening risk management, banks can further improve the efficiency and safety of their green loan business, contributing to the sustainable development of green finance.

[0057] This invention is no longer limited to the risk rating of individual green loan business, but expands the evaluation scope to the entire green finance product chain including green bonds, green funds, carbon finance special products, green leases, etc. Through deep - learning algorithms, the system can comprehensively consider the risk correlation and mutual influence among various products, achieve a more accurate and comprehensive risk assessment, and provide banks with a more reliable basis for risk management.

[0058] 3. Carbon balance and carbon absorption assessment and carbon absorption enterprise pairing: This system provides a more comprehensive perspective on carbon balance by considering both the carbon emissions and carbon absorption of enterprises. This assessment method not only focuses on reducing carbon emissions, but also includes the potential of increasing carbon absorption through biological carbon sequestration and carbon capture and storage (CCUS) technologies. This enables banks to more accurately assess the green loan eligibility and risk level of enterprises.

[0059] Using large - scale artificial intelligence models, this system can actively identify and pair potential carbon absorption enterprises, which may increase carbon credits for enterprises applying for green loans through forest planting or other biological carbon sequestration methods, or through CCUS technologies. This pairing mechanism optimizes the loan approval process and promotes cooperation within the green finance ecosystem.

[0060] This system adopts deep learning algorithms to dynamically adjust loan pricing according to the carbon balance status of enterprises. This method takes into account the value of carbon credits and the impact of market and policy changes on the value of carbon credits, thus providing banks with a more flexible and market-responsive pricing strategy.

[0061] Combined with deep learning algorithms, this system can achieve professional evaluation and pricing of green loans for the analysis of the entire industrial chain. This pricing mechanism not only considers the financial status of enterprises but also incorporates key technical factors such as environmental benefits and green production capacity, making loan pricing more scientific and reasonable. Compared with existing technologies, this system can better motivate enterprises to make green investments and promote sustainable development.

[0062] 4. Deep learning model: This invention adopts deep learning algorithms, which can automatically extract complex and high-dimensional risk features from massive data. These features include not only traditional financial indicators but also multi-dimensional information such as environmental, social, and governance (ESG), making risk assessment more comprehensive and in-depth. At the same time, the deep learning model can handle non-linear relationships and implicit patterns between data, improving the accuracy of risk assessment.

[0063] After realizing the precipitation of private domain information of the existing green loan large model, the deep learning algorithm can achieve automatic interaction information and private domain management for enterprises applying for green loans, predict customer behavior, and then realize professional evaluation and pricing of green loans for the analysis of the entire industrial chain.

[0064] This system conducts a comprehensive green production capacity analysis of the industrial chain where the applying enterprise is located through deep learning algorithms. Compared with traditional evaluation models that only rely on the historical financial data of enterprises, this method can capture the potential value and risks of enterprises in the green transformation more comprehensively. This analysis can identify the position of enterprises in the green economy, predict their competitiveness and growth potential in the green industry, and thus provide more accurate loan decision-making support for banks.

[0065] 5. Model update and optimization: Based on the initial application, we continuously collect new loan data and project feedback to continuously optimize the deep learning model. This includes adjusting the network structure, adding regularization terms to prevent overfitting, and introducing richer features. In particular, we attempt to apply graph neural networks (GNNs) to the correlation analysis between green projects to further improve the accuracy of risk assessment.

[0066] After the model is optimized, it can significantly improve in terms of the accuracy of risk assessment and the reasonableness of pricing. Especially in identifying high-risk loan projects and reasonably setting loan interest rates, the prediction results of the model are highly consistent with the actual situation.

[0067] Meanwhile, we also evaluated the application effect of the system in actual business through indicators such as customer satisfaction surveys and loan default rates. This is conducive to improving the processing efficiency of the bank's green loan business, reducing the loan default risk, and promoting the sustainable development of green finance.

[0068] This system can input dynamic knowledge formed by the model architecture, financial knowledge concepts, industry market logic, and professional terms of green loans, and achieve unsupervised learning and automatic update of artificial intelligence. This advantage enables the system to continuously optimize itself, adapt to market changes, improve the timeliness and accuracy of risk assessment. Compared with traditional models that require manual regular updates and maintenance, this system greatly reduces labor costs and time delays.

[0069] This invention not only provides static risk ratings, but also can update risk assessment results in real time according to factors such as market changes and policy orientations, realizing dynamic risk management. This helps banks adjust the green financial product portfolio in a timely manner, optimize the decision-making process, reduce the risk of individual green loan businesses, and improve the overall business efficiency at the same time.

[0070] In summary, a green loan risk assessment and pricing system proposed by this invention realizes the accurate assessment of green loan risks and pricing prediction by comprehensively applying deep learning algorithms and relevant data, providing strong support for the healthy development of the bank's green loan business.

[0071] The above shows and describes the basic principles, main features, and advantages of this invention. Those skilled in the art should understand that this invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of this invention. Without departing from the spirit and scope of this invention, this invention will have various changes and improvements, and these changes and improvements all fall within the scope of this invention claimed. The scope of protection claimed by this invention is defined by the appended claims and their equivalents.

Claims

1. A green loan risk assessment and pricing system, characterized in that: include: Data collection module: used to collect financial data, industry chain information, and environmental data of green loan applicants, historical data of green loans and green financial products of banks, and data related to carbon emissions and carbon absorption of loan applicants; Data processing module: used to preprocess the collected data, including cleaning, denoising, removing outliers and missing values, standardization, and normalization. Feature extraction module: used to extract features from pre-processed data and construct feature vectors for input into subsequent deep learning models; Model training module: used to design a deep learning model for risk assessment and pricing including a convolutional layer and a recurrent layer, and to perform model training through the feature vector; Evaluation and pricing module: It is used to use the trained model to conduct risk assessment on new loan applications, as well as carbon absorption potential assessment and carbon balance risk assessment, and to price loans based on key technical factors of environmental benefits and green production capacity. Green loan business strategies are formulated based on the evaluation results and loan pricing recommendations output by the model.

2. A green loan risk assessment and pricing system according to claim 1, characterized in that: It includes a model optimization module: used to dynamically update the model according to market changes and policy guidelines, and use optimization algorithms to adjust loan credit limits and pricing based on risk assessment results, as well as collect carbon emission and carbon absorption data from enterprises after the loan for model feedback and iterative optimization.

3. A green loan risk assessment and pricing system according to claim 2, characterized in that: The formula used in adjusting the loan credit limit and pricing using the optimization algorithm based on the risk assessment results is as follows (1): （1） In the formula, is the loan interest rate, is the base rate, is a natural constant, Score the risk, is the market carbon credit price, and To adjust the parameters.

4. A green loan risk assessment and pricing system according to claim 1, characterized in that: The evaluation and pricing module specifically includes: performing loan pricing based on the comprehensive risk rating, market interest rate, and green project yield output by the model, wherein the formula used is shown in the following formula (2): （2） In the formula, To price the loan, Risk_Rating is the comprehensive risk rating. Market_Rate is the market interest rate, Green_Project_ROI is the green project rate of return. , , , is the regression coefficient, is the error term.

5. A green loan risk assessment and pricing system according to claim 1, characterized in that: The carbon absorption potential assessment includes assessing the carbon absorption potential increased by an enterprise through biological carbon fixation and CCUS technology, and the carbon balance risk assessment includes using a deep learning model to conduct a risk assessment based on the carbon emission potential and carbon absorption potential of the enterprise.

6. A green loan risk assessment and pricing system according to claim 5, characterized in that: The formula used in the carbon absorption potential assessment is shown in the following formula (3): T （3） In the formula, is the carbon absorption potential, is the forest coverage rate, T is the technical carbon sequestration capacity, , , are model parameters; The formula used in the carbon balance risk assessment is shown in the following formula (4): （4） In the formula, R is the carbon balance risk score, e is the natural constant, is carbon emissions, is the carbon absorption potential, is the policy influencing factor, is the market sensitivity factor, , , , is the weight parameter.

7. A green loan risk assessment and pricing system according to claim 6, characterized in that: It also includes a carbon absorption enterprise matching module, which is used to use an artificial intelligence algorithm to identify and match carbon absorption enterprises in the bank's private domain information database, wherein the formula used is shown in the following formula (5): （5） In the formula, is the minimum value among all pairing schemes, that is, the value corresponding to the optimal pairing scheme. is the total number of carbon emission enterprises applying for loans, For the i-th carbon emission enterprise applying for a loan, is the jth carbon absorption enterprise in the bank’s private domain information database, is the distance function between the i-th carbon emission enterprise applying for a loan and the j-th carbon absorption enterprise, is the cost function of the jth carbon absorption enterprise, is a trade-off parameter.

8. A green loan risk assessment and pricing system according to claim 1, characterized in that: The model training module includes: introducing a Dropout layer to reduce overfitting, building a model in combination with an attention mechanism, and using a back-propagation algorithm to optimize model parameters.

Images