Financial credit risk identification method and model construction method

Abstract

The invention discloses a financial credit risk identification method and a model construction method, and relates to the technical field of financial credit risk identification and construction. Based on basic credit identification and basic risk identification, an identification model for evaluating a credit risk basic basis unit is generated through machine learning, and a basic evaluation value of a current credit risk of a borrower is updated; the method comprises the following steps: generating an identification model for providing a unit for accurately evaluating the credit risk of a borrower through machine learning, updating and adjusting the risk value of the current credit of the borrower, generating an identification model for providing a unit for evaluating the comprehensive credit risk through machine learning, and updating to obtain the comprehensive credit risk value of the borrower. According to the credit risk assessment method and system, the accuracy of credit risk assessment is improved, the flexibility of credit policies is enhanced, the credit risk management process is optimized, financial science and technology innovation is promoted, and then a more comprehensive, accurate and efficient credit risk assessment solution is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of financial credit risk identification and construction, and specifically to a financial credit risk identification method and a model construction method. Background Art

[0002] Identifying and assessing financial credit risks has always been one of the core tasks of financial institutions. In the field of financial credit, risk assessment is a key link in ensuring loan quality, preventing defaults and reducing credit losses.

[0003] Existing credit risk assessment models are often too simple and cannot fully consider the borrower's multiple risk factors. This makes the model prone to misjudgment and omission when assessing credit risk, reducing the accuracy of risk assessment. In addition, existing methods often have problems of data waste and data silos when processing data, which leads to a large amount of valuable data not being fully utilized, limiting the performance and effectiveness of credit risk assessment models. Existing methods often lack a mechanism for dynamic adjustment based on changes in the economic situation and credit environment, which makes it difficult for financial institutions to take effective and timely response measures when faced with market fluctuations and risk changes. In addition, existing methods often require a lot of manual intervention and judgment when dealing with credit risk assessment, which not only increases operating costs and time costs, but also easily introduces errors and uncertainties caused by human factors. Summary of the Invention

[0004] The purpose of the present invention is to provide a financial credit risk identification method to solve the problems raised in the above background technology.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions, including generating a basic credit identification of the borrower's current financial credit risk based on the borrower's basic information; Generate basic risk identification based on the borrower's current financial status; Based on basic credit identification and basic risk identification, and through machine learning, an identification model for the basis unit for assessing credit risk is generated. Based on basic credit identification and basic risk identification, the basic assessment value of the borrower's current credit risk is updated; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, machine learning is used to generate an identification model that accurately assesses the borrower's credit risk unit, and the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, machine learning is used to generate an identification model that provides comprehensive credit risk assessment results, and the borrower's comprehensive credit risk value is updated; Based on the borrower's current comprehensive credit risk value and the previous comprehensive credit risk value, the increase and decrease in risk are identified, and based on this, an adjustment model for basic credit identification is generated through machine learning to complete the update of basic credit identification.

[0006] Optionally, the calculation formula of the unit providing the basis for assessing credit risk is as follows: Q1=SQRT(XP×(1-FL))+SR×H; FL=FZ / ZZ; in: Q1 is the credit risk basic assessment value; XP is the credit score, which indicates the borrower's credit status and ranges from 0 to 100; FL is the debt ratio and ranges from 0 to 1; FZ is the total liabilities, ZZ is the total assets; SQRT is the symbol for square root. Here, the square root of the result of (XP×(1-FL)) is processed to smooth the result and then evaluated in combination with the result of SR×H. SR is the income level, SR represents the borrower’s annual income; H is the macroeconomic impact coefficient, which is adjusted according to the economic situation and has a value range of 0.5 to 1.5. It is close to 1.5 when the economic situation is good, that is, when the line graph of the borrower's enterprise profit shows an upward trend, and close to 0.5 when the economic situation is bad, that is, when the line graph of the borrower's enterprise profit shows a downward trend; SR×H reflects and evaluates the borrower’s repayment ability under different economic environments; SQRT (XP × (1-FL)) is a correction to the borrower's basic credit rating, taking into account the impact of liabilities on credit status; If the Q1 value is high, a low-risk identification is generated; A low Q1 value generates a high risk identification.

[0007] Optionally, the calculation formula for accurately assessing the borrower's credit risk unit is as follows: Q2=Q1×(1-HK / 100)-SQRT(XP×(1-HK / 100)); HK=(HC / ZC)×100; in: Q2 is the credit risk adjustment value; HK is the repayment history assessment value, which reflects the quality of the borrower's repayment history and ranges from 0 to 100; HC is the amount repaid, and ZC is the total amount to be repaid; Q1×(1-HK / 100) reflects the combined impact of the credit score XP and the repayment history assessment value HK in the credit risk assessment: When the repayment history assessment value HK is high, the contribution of the credit score XP to the credit risk adjustment value Q2 decreases; When the repayment history assessment value HK is low, the contribution of the credit score XP to the credit risk adjustment value Q2 increases; SQRT (XP × (1-HK / 100)) uses the ratio of the historical repayment assessment value HK to the maximum value 100 to reflect the historical repayment score: If it is equal to 1, the repayment history is good; If it is less than and close to 1, the repayment history performance is relatively poor compared to the case where it is equal to 1; If it is less than and far away from 1, the repayment history performance is poor; A high Q2 value indicates low risk after considering repayment history and credit score; A low Q2 value indicates a high risk after considering repayment history and credit score.

[0008] Optionally, the calculation formula for providing the comprehensive credit risk assessment result unit is as follows: Q3=Q2×ZZ×(1-FX)-SQRT(Q2×(HK / 100)); in: Q3 is the comprehensive credit risk value; FX is the industry risk factor, which is adjusted based on the risk level of the borrower's industry and ranges from 0 to 1. Low-risk industries are closer to 0, while high-risk industries are closer to 1. Q2×ZZ×(1-FX) reflects the impact on credit risk under the comprehensive asset and industry risk factors; The ratio of the repayment history assessment value HK to the maximum value 100 is reintroduced into SQRT (Q2×(HK / 100)) to reflect the importance attached to repayment credit and its impact on credit.

[0009] Optionally, specific adjustments based on the industry risk factor FX are as follows: Low-risk industries have stable income sources and low default risks, including education, healthcare, and technology. Based on the current industry conditions, the education, healthcare, and technology industries are ranked by risk, i.e., 0 < education < healthcare < technology < 0.5; High-risk industries are those that are greatly affected by economic cycles and policies, including real estate, mining, and energy. Based on the current industry conditions, the risks of real estate, mining, and energy are ranked as follows: 1 < real estate < mining < energy < 0.5; Among them, the education, medical, technology industries as well as the real estate, mining, and energy industries will be ranked and adjusted in real time based on the current risk environment.

[0010] Optionally, a method for identifying and adjusting the credit score XP based on the comprehensive credit risk value Q3 is as follows: First, compare the comprehensive credit risk value Q3 with the previous comprehensive credit risk value Q3 of the same borrower. prev Compare the results; If the comprehensive credit risk value Q3 is higher than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[(Q3 / Q3 prev ) / 100+1]; If the comprehensive credit risk value Q3 is lower than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[1-(Q3 / Q3 prev ) / 100]; Among them, XP new The next credit score is the replacement input value of the credit score XP in the basis unit for assessing the credit risk of the same borrower next time.

[0011] The present invention aims to provide a method for constructing a financial credit risk identification model, comprising a data identification and collection module for generating a basic credit identification and basic risk identification of the borrower's current financial credit risk based on the identification and collection of the borrower's basic information and current financial conditions; Based on basic credit identification and basic risk identification, and through the construction of the identification model of the data analysis and processing module, the basic assessment value of the borrower's current credit risk is updated according to the basic credit identification and basic risk identification; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, the borrower's comprehensive credit risk value is updated; Based on the borrower's current and last comprehensive credit risk values, identify increases and decreases in risk and update basic credit identification; Optionally, the equipment used by the data identification and collection module includes a database management system, which is used to store and manage basic information of the borrower; The equipment used by the data analysis and processing module includes a credit risk assessment system, which is a software system integrating the above modules and is used to automatically calculate the credit risk value. The credit risk assessment system is used to execute the method described in any one of claims 1 to 6.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention introduces a unit for providing a basis for assessing credit risk, a unit for accurately assessing the borrower's credit risk, a unit for providing comprehensive credit risk assessment results, and a cyclical impact system. This enables the method to comprehensively consider multiple risk factors of the borrower and quantify them using mathematical formulas. This improves the complexity and accuracy of the model and reduces the possibility of misjudgment and omission.

[0013] 2. The present invention can make full use of the borrower's data in multiple dimensions, including credit score XP, debt ratio FL, income level SR, repayment history assessment value HK, total assets ZZ and industry risk factor FX. Through data cleaning, organization and analysis, it can then mine valuable information from the data, thereby improving the performance and effectiveness of the credit risk assessment model.

[0014] 3. The present invention can flexibly adjust the threshold of the current borrower's credit score XP according to the changes in the comprehensive credit risk value Q3. This dynamic adjustment mechanism helps to take effective response measures in a timely manner when facing market fluctuations and risk changes, maintaining risk control while promoting business development.

[0015] 4. The present invention introduces mathematical models and algorithms to automatically process large amounts of data, reducing the need for manual intervention and judgment. This not only reduces operating costs and time costs, but also improves the accuracy and efficiency of credit risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of the method for identifying financial credit risk; Figure 2 This is a flowchart of the data identification and collection module in the method for building a financial credit risk identification model; Figure 3 This is a flowchart of the data analysis and processing module in this financial credit risk identification model construction method. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Regarding this financial credit risk identification method and model building method, it is different from the existing credit risk assessment model. The existing credit risk assessment model is often too simple and cannot fully consider the borrower's multiple risk factors. There are problems of data waste and data silos, and there is a lack of a mechanism for dynamic adjustment according to changes in the economic situation and credit environment. This algorithm unit improves the accuracy of credit risk assessment, enhances the flexibility of credit policies, optimizes credit risk management processes, and promotes financial technology innovation. At the same time, this method also solves, optimizes and innovates the problems and shortcomings of existing technologies, providing a more comprehensive, accurate and efficient credit risk assessment solution for financial credit risk identification methods and model building methods.

[0019] For example 1, please refer to Figures 1 to 3 ,This implementation provides a financial credit risk identification method, including generating a basic credit identification of the borrower’s current financial credit risk based on the borrower’s basic information; Generate basic risk identification based on the borrower's current financial status; Based on basic credit identification and basic risk identification, and through machine learning, an identification model for the basis unit for assessing credit risk is generated. Based on basic credit identification and basic risk identification, the basic assessment value of the borrower's current credit risk is updated; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, machine learning is used to generate an identification model that accurately assesses the borrower's credit risk unit, and the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, machine learning is used to generate an identification model that provides comprehensive credit risk assessment results, and the borrower's comprehensive credit risk value is updated; Based on the borrower's current comprehensive credit risk value and the previous comprehensive credit risk value, the increase and decrease in risk are identified, and based on this, an adjustment model for basic credit identification is generated through machine learning to complete the update of basic credit identification.

[0020] In this embodiment, the system, through the mutual cooperation of three algorithm units, combines the three operation results of Q1, Q2 and Q3 to jointly construct a comprehensive, accurate and flexible credit risk assessment and management framework. Q1 is the basic credit risk assessment value, which provides an important reference for subsequent credit risk assessment. Among them, the credit score XP reflects the borrower's historical credit record, the debt ratio FL reveals the borrower's debt burden, the income level SR reflects the borrower's repayment ability, and the macroeconomic impact coefficient H takes into account the impact of the economic situation on credit risk. Q2 is the credit risk adjustment value. The calculation purpose of this value is to more accurately assess the borrower's Credit risk, Q3 is the comprehensive credit risk value. The purpose of this value is to provide a more comprehensive and integrated credit risk assessment result. By comprehensively considering the borrower's economic strength and repayment potential, and the risk level of the industry in which he is located, it can more accurately assess the borrower's credit risk and formulate more reasonable credit policies and risk control measures accordingly. The calculation results of Q3 can also affect the calculations fed back to Q1 and Q2, making the three algorithms of this system highly correlated and entangled. This framework not only takes into account the borrower's personal credit status, repayment history, and economic strength factors, but also considers the impact of economic conditions and industry risks on credit risk, providing financial institutions with more accurate and comprehensive credit risk assessment results.

[0021] See also Figures 1 to 3 The calculation formula for the unit providing the basis for assessing credit risk is as follows: Q1=SQRT(XP×(1-FL))+SR×H; FL=FZ / ZZ; in: Q1 is the credit risk basic assessment value; XP is the credit score, which indicates the borrower's credit status and ranges from 0 to 100; FL is the debt ratio and ranges from 0 to 1; FZ is the total liabilities, ZZ is the total assets; SQRT is the symbol for square root. Here, the square root of the result of (XP×(1-FL)) is processed to smooth the result and then evaluated in combination with the result of SR×H. SR is the income level, SR represents the borrower’s annual income; H is the macroeconomic impact coefficient, which is adjusted according to the economic situation and has a value range of 0.5 to 1.5. It is close to 1.5 when the economic situation is good, that is, when the line graph of the borrower's enterprise profit shows an upward trend, and close to 0.5 when the economic situation is bad, that is, when the line graph of the borrower's enterprise profit shows a downward trend; SR×H reflects and evaluates the borrower’s repayment ability under different economic environments; SQRT (XP × (1-FL)) is a correction to the borrower's basic credit rating, taking into account the impact of liabilities on credit status; If the Q1 value is high, a low-risk identification is generated; If the Q1 value is low, a high-risk identification is generated. In this embodiment: First, the "SR×H" calculation part of this algorithm unit is designed to combine the borrower's income level SR with the macroeconomic impact coefficient H to reflect the impact of the macroeconomic environment on the borrower's repayment ability. When the economic situation is good (H is close to 1.5), the borrower's repayment ability is relatively strong. When the economic situation is bad (H is close to 0.5), the borrower's repayment ability will be weakened. Through the calculation of this part, the borrower's repayment ability under different economic environments can be more accurately assessed, thereby providing more comprehensive information for the credit risk basic assessment value Q1; The "SQRT (XP × (1-FL))" calculation takes into account the borrower's credit score XP and debt ratio FL. A high credit score XP indicates a good credit standing, while a low debt ratio FL indicates a relatively light debt burden. This calculation reflects the borrower's credit standing and debt pressure. This calculation, combined with the product of the income level SR and the macroeconomic impact coefficient H, together constitutes the basic credit risk assessment value Q1, which reflects the borrower's basic credit and debt situation and provides a basis for subsequent credit risk adjustments. The credit score XP in this algorithm unit is an important indicator for assessing the borrower's credit status. In the unit that provides the basis for assessing credit risk, the level of the credit score XP directly affects the size of the credit risk base assessment value Q1, which in turn affects the initial judgment of credit risk. Therefore, ensuring the accuracy of the credit score XP is crucial. The accuracy and predictive ability of the credit score XP can be improved by continuously optimizing the credit score XP model. The debt ratio (FL) reflects a borrower's debt burden, while the income level (SR) reflects the borrower's repayment ability. In providing the basis for assessing credit risk, the weighting of the debt ratio (FL) and income level (SR) has a significant impact on the calculation of the credit risk basis assessment value (Q1). Financial institutions need to comprehensively consider the changing trends of the debt ratio (FL) and income level (SR), as well as the balance between them, to accurately assess a borrower's repayment ability. The macroeconomic impact coefficient H reflects the impact of the economic situation on credit risk. In the unit that provides the basis for assessing credit risk, by adjusting the macroeconomic impact coefficient H, we can flexibly respond to changes in the economic cycle and ensure the timeliness and accuracy of credit risk assessment. At the same time, the setting of the macroeconomic impact coefficient H also needs to take into account the economic differences between different industries and regions to achieve more refined credit risk management.

[0022] See also Figures 1 to 3 , the calculation formula for accurately assessing the borrower's credit risk unit is as follows: Q2=Q1×(1-HK / 100)-SQRT(XP×(1-HK / 100)); HK=(HC / ZC)×100; in: Q2 is the credit risk adjustment value; HK is the repayment history assessment value, which reflects the quality of the borrower's repayment history and ranges from 0 to 100; HC is the amount repaid, and ZC is the total amount to be repaid; Q1×(1-HK / 100) reflects the combined impact of the credit score XP and the repayment history assessment value HK in the credit risk assessment: When the repayment history assessment value HK is high, the contribution of the credit score XP to the credit risk adjustment value Q2 decreases; When the repayment history assessment value HK is low, the contribution of the credit score XP to the credit risk adjustment value Q2 increases; SQRT (XP × (1-HK / 100)) uses the ratio of the historical repayment assessment value HK to the maximum value 100 to reflect the historical repayment score: If it is equal to 1, the repayment history is good; If it is less than and close to 1, the repayment history performance is relatively poor compared to the case where it is equal to 1; If it is less than and far away from 1, the repayment history performance is poor; A high Q2 value indicates low risk after considering repayment history and credit score; A low Q2 value indicates a high risk after considering repayment history and credit score.

[0023] In this embodiment, the calculation "Q1×(1-HK / 100)" is first used to adjust the basic credit risk assessment value Q1 based on the borrower's repayment history assessment value HK. Borrowers with good repayment history assessment values ​​HK (i.e., low HK) will receive a higher adjustment value, and vice versa. This calculation can further refine the credit risk assessment, taking the borrower's repayment history into consideration, and making the credit risk adjustment value Q2 more realistic. The calculation of "(XP × (1-HK / 100))" takes into account the interaction between the credit score XP and the repayment history assessment value HK. When a borrower has a poor repayment history (i.e., a high repayment history assessment value HK), the contribution of their credit score XP to the credit risk adjustment value will be reduced accordingly. This calculation component, as part of the credit risk adjustment value Q2, is subtracted from the adjusted value of the basic credit risk assessment value Q1 to jointly determine the final result of the credit risk adjustment value Q2. It reflects the combined impact of the credit score and repayment history in credit risk assessment. In the unit for accurately assessing the borrower's credit risk, the introduction of the repayment history assessment value HK makes the credit risk assessment more focused on the borrower's historical repayment behavior. The repayment history assessment value HK directly reflects the borrower's repayment willingness and credit record, which is of great significance for assessing the borrower's long-term credit status. In the unit for accurately assessing the borrower's credit risk, the synergistic effect of the credit score XP and the repayment history assessment value HK makes the credit risk assessment more comprehensive and accurate. The credit score XP reflects the borrower's current credit status, while the repayment history assessment value HK reflects the borrower's historical credit record. The combination of the two can more comprehensively assess the borrower's credit status and repayment ability. The credit risk adjustment value Q2 in the unit for accurately assessing the borrower's credit risk will be dynamically adjusted as the borrower's repayment history changes. This dynamic adjustment mechanism makes credit risk assessment more flexible and timely, helping to identify potential risks in a timely manner and take corresponding measures.

[0024] See also Figures 1 to 3 , the calculation formula for providing comprehensive credit risk assessment results unit is as follows: Q3=Q2×ZZ×(1-FX)-SQRT(Q2×(HK / 100)); in: Q3 is the comprehensive credit risk value; FX is the industry risk factor, which is adjusted based on the risk level of the borrower's industry and ranges from 0 to 1. Low-risk industries are closer to 0, while high-risk industries are closer to 1. Q2×ZZ×(1-FX) reflects the impact on credit risk under the comprehensive asset and industry risk factors; SQRT (Q2 × (HK / 100)) again introduces the ratio of the repayment history assessment value HK to the maximum value of 100 to reflect the emphasis on repayment credit and its impact on credit; The specific adjustments based on the industry risk factor FX are as follows: Low-risk industries have stable income sources and low default risks, including education, healthcare, and technology. Based on the current industry conditions, the education, healthcare, and technology industries are ranked by risk, i.e., 0 < education < healthcare < technology < 0.5; High-risk industries are those that are greatly affected by economic cycles and policies, including real estate, mining, and energy. Based on the current industry conditions, the risks of real estate, mining, and energy are ranked as follows: 1 < real estate < mining < energy < 0.5; Among them, the education, medical, technology industries as well as the real estate, mining, and energy industries will be ranked and adjusted in real time based on the current risk environment.

[0025] In this embodiment, the algorithm unit first calculates the "Q2 × ZZ × (1-FX)" formula to combine the credit risk adjustment value Q2 with the borrower's total assets ZZ and the industry risk coefficient FX to comprehensively assess the borrower's credit risk. Borrowers with large asset sizes and low industry risks will receive higher comprehensive credit risk values. This calculation can more accurately reflect the borrower's comprehensive situation in terms of asset and industry risks, providing strong support for credit decision-making. The calculation of "(Q2 × (HK / 100))" takes into account the negative impact of the repayment history assessment value HK on the comprehensive credit risk value. When a borrower has a poor repayment history, that is, a high repayment history assessment value HK, the comprehensive credit risk value will be reduced accordingly. This calculation component, as part of the comprehensive credit risk value Q3, is subtracted from the result of the calculation of "Q2 × ZZ × (1-FX)" to determine the final result of the comprehensive credit risk value Q3. It reflects the important role of repayment history in comprehensive credit risk assessment. In the unit that provides comprehensive credit risk assessment results, the introduction of total assets ZZ and industry risk factor FX makes credit risk assessment more comprehensive and in-depth. Total assets ZZ reflects the borrower's financial strength and repayment potential, while industry risk factor FX reflects the differences in credit risk between different industries. The combination of the two can provide financial institutions with more accurate credit risk assessment results. The comprehensive credit risk value (Q3) in the unit providing comprehensive credit risk assessment results can be linked to and cyclically influence the credit score (XP) in the unit providing the basis for credit risk assessment, forming a dynamic credit risk assessment and adjustment mechanism. This mechanism makes credit risk management more flexible and efficient, and helps financial institutions adjust credit policies in a timely manner based on market changes and changes in the borrower's credit status. In addition, the comprehensive credit risk assessment result unit provides financial institutions with a comprehensive credit risk assessment result by comprehensively considering multiple dimensions of borrower information, including financial status, repayment history, asset size and industry risk. At the same time, by introducing the industry risk coefficient FX, differentiated credit policies can be formulated for different industries to achieve more refined credit risk management.

[0026] See also Figures 1 to 3 , based on the comprehensive credit risk value Q3, the method for identifying and adjusting the credit score XP is as follows: First, compare the comprehensive credit risk value Q3 with the previous comprehensive credit risk value Q3 of the same borrower. prev Compare the results; If the comprehensive credit risk value Q3 is higher than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[(Q3 / Q3 prev ) / 100+1]; If the comprehensive credit risk value Q3 is lower than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[1-(Q3 / Q3 prev ) / 100]; Among them, XP new The next credit score, that is, the replacement input value of the credit score XP in the basis unit for assessing the credit risk of the same borrower next time.

[0027] In this embodiment, based on the comprehensive credit risk value Q3 calculated by the comprehensive credit risk assessment result providing unit, this algorithm unit can timely understand changes in overall credit quality and adjust the threshold of the current borrower's credit score XP as needed, thereby achieving dynamic adjustment and optimization of credit policies. In addition, the cyclical impact mechanism can more quickly identify potential risks of borrowers' financial credit and take corresponding measures to intervene and manage them, thereby improving the efficiency and accuracy of credit risk management. By continuously optimizing credit risk management strategies, it can better support the development of the real economy and promote the healthy, stable and sustainable development of credit business. By dynamically adjusting the credit score XP based on real-time changes in the comprehensive credit risk value Q3, it is possible to adapt to market changes more quickly and improve the flexibility and efficiency of credit approval. When the comprehensive credit risk value Q3 increases, increasing the credit score XP can identify higher-quality borrowers and reduce potential risks. When the comprehensive credit risk value Q3 decreases, lowering the credit score XP helps maintain the credit approval rate, reduces the financial credit risk of a deteriorating borrower trend, and maintains business stability. Dynamic adjustment of the credit score XP helps better manage credit risk. When credit quality improves, increasing the credit score XP can further reduce risk. When credit quality declines, although lowering the credit score XP will increase risk to a certain extent, reasonable risk pricing and subsequent risk monitoring can control risk within an acceptable range. In addition, by continuously monitoring changes in the comprehensive credit risk value Q3 and the credit score XP, potential risk points can be identified in a timely manner and corresponding risk management measures can be taken. Through this flexible adjustment mechanism, it is possible to better balance risks and returns and achieve sustainable development. By dynamically adjusting the credit score XP, we can better adapt to market changes and enhance market competitiveness. When credit quality improves, increasing the credit score XP can demonstrate our attention and attraction to high-quality borrowers. When credit quality declines, lowering the credit score XP helps us stay in sync with the market and avoid losing market share due to being overly conservative. To sum up, the feedback adjustment mechanism of dynamically adjusting the credit score threshold Credit Score XP according to the changes in the comprehensive credit risk value Q3 can bring about beneficial effects in improving credit approval efficiency, timely identification, optimizing credit risk management and enhancing market competitiveness. This adjustment mechanism helps financial institutions achieve sustainable business development and enhanced market competitiveness while maintaining risk control.

[0028] For example 2, please refer to Figures 1 to 3 This implementation provides a method for constructing a financial credit risk identification model, including a data identification and collection module for generating basic credit identification and basic risk identification of the borrower's current financial credit risk based on the identification and collection of the borrower's basic information and current financial conditions; Based on basic credit identification and basic risk identification, and through the construction of the identification model of the data analysis and processing module, the basic assessment value of the borrower's current credit risk is updated according to the basic credit identification and basic risk identification; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, the borrower's comprehensive credit risk value is updated; Based on the borrower's current and last comprehensive credit risk values, identify increases and decreases in risk and update basic credit identification; The equipment used by the data identification and collection module includes a database management system, which is used to store and manage the basic information of the borrower; The equipment used by the data analysis and processing module includes a credit risk assessment system, which is a software system integrating the above modules and is used to automatically calculate the credit risk value. The credit risk assessment system is used to execute the method described in any one of claims 1 to 6.

[0029] In this embodiment, module 1: data identification and collection module Step 1: Identify and collect the borrower's credit score XP, total liabilities FZ, total assets ZZ, income level SR, amount repaid HC, and total amount to be repaid ZC; Step 2: Based on the current economic situation, identify the extent to which the borrower's macroeconomic environment affects its credit risk; Step 3: Identify the risk level of the borrower's enterprise based on the borrower's industry; Module 2: Data Analysis and Processing Module Step 1: Obtain data from Module 1 and construct a calculation model that provides a basis for assessing credit risk, accurately assesses the borrower's credit risk, and provides comprehensive credit risk assessment results; Step 2: Based on the total liabilities FZ and total assets ZZ, obtain the borrower's current debt ratio FL; Step 3: Based on the repaid amount HC and the total amount to be repaid ZC, obtain the borrower's current repayment history assessment value HK; Step 4: Input the corresponding acquired parameters into the calculation model of Step 1 in sequence, and calculate and update the basic credit risk assessment value Q1, the credit risk adjustment value Q2, and the comprehensive credit risk value Q3 in sequence; Step 5: Update the credit score XP based on the comprehensive credit risk value Q3.

[0030] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying financial credit risk, characterized in that: This includes generating a basic credit identification of the borrower's current financial credit risk based on the borrower's basic information; Generate basic risk identification based on the borrower's current financial status; Based on basic credit identification and basic risk identification, and through machine learning, an identification model for the basis unit for assessing credit risk is generated. Based on basic credit identification and basic risk identification, the basic assessment value of the borrower's current credit risk is updated; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, machine learning is used to generate an identification model that accurately assesses the borrower's credit risk unit, and the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, machine learning is used to generate an identification model that provides comprehensive credit risk assessment results, and the borrower's comprehensive credit risk value is updated; Based on the borrower's current comprehensive credit risk value and the previous comprehensive credit risk value, the increase and decrease in risk are identified, and based on this, an adjustment model for basic credit identification is generated through machine learning to complete the update of basic credit identification.

2. The financial credit risk identification method according to claim 1, characterized in that: The calculation formula for the unit providing the basis for assessing credit risk is as follows: Q1=SQRT(XP×(1-FL))+SR×H; FL=FZ / ZZ; in: Q1 is the credit risk basic assessment value; XP is the credit score, which indicates the borrower's credit status and ranges from 0 to 100; FL is the debt ratio and ranges from 0 to 1; FZ is the total liabilities, ZZ is the total assets; SQRT is the symbol for square root. Here, the square root of the result of (XP×(1-FL)) is processed to smooth the result and then evaluated in combination with the result of SR×H. SR is the income level, SR represents the borrower’s annual income; H is the macroeconomic impact coefficient, which is adjusted according to the economic situation and has a value range of 0.5 to 1.

5. It is close to 1.5 when the economic situation is good, that is, when the line graph of the borrower's enterprise profit shows an upward trend, and close to 0.5 when the economic situation is bad, that is, when the line graph of the borrower's enterprise profit shows a downward trend; SR×H reflects and evaluates the borrower’s repayment ability under different economic environments; SQRT (XP × (1-FL)) is a correction to the borrower's basic credit rating, taking into account the impact of liabilities on credit status; If the Q1 value is high, a low-risk identification is generated; A low Q1 value generates a high risk identification.

3. The financial credit risk identification method according to claim 2, characterized in that: The calculation formula for accurately assessing the borrower's credit risk unit is as follows: Q2=Q1×(1-HK / 100)-SQRT(XP×(1-HK / 100)); HK=(HC / ZC)×100; in: Q2 is the credit risk adjustment value; HK is the repayment history assessment value, which reflects the quality of the borrower's repayment history and ranges from 0 to 100; HC is the amount repaid, and ZC is the total amount to be repaid; Q1×(1-HK / 100) reflects the combined impact of the credit score XP and the repayment history assessment value HK in the credit risk assessment: When the repayment history assessment value HK is high, the contribution of the credit score XP to the credit risk adjustment value Q2 decreases; When the repayment history assessment value HK is low, the contribution of the credit score XP to the credit risk adjustment value Q2 increases; SQRT (XP × (1-HK / 100)) uses the ratio of the historical repayment assessment value HK to the maximum value 100 to reflect the historical repayment score: If it is equal to 1, the repayment history performance is good; If it is less than and close to 1, the repayment history performance is relatively poor compared to the case where it is equal to 1; If it is less than and far away from 1, the repayment history performance is poor; A high Q2 value indicates low risk after considering repayment history and credit score; A low Q2 value indicates a high risk after considering repayment history and credit score.

4. The financial credit risk identification method according to claim 3, characterized in that: The calculation formula for providing comprehensive credit risk assessment results is as follows: Q3=Q2×ZZ×(1-FX)-SQRT(Q2×(HK / 100)); in: Q3 is the comprehensive credit risk value; FX is the industry risk factor, which is adjusted based on the risk level of the borrower's industry and ranges from 0 to 1. Low-risk industries are closer to 0, while high-risk industries are closer to 1. Q2×ZZ×(1-FX) reflects the impact on credit risk under the comprehensive asset and industry risk factors; The ratio of the repayment history assessment value HK to the maximum value 100 is reintroduced into SQRT (Q2×(HK / 100)) to reflect the importance attached to repayment credit and its impact on credit.

5. The financial credit risk identification method according to claim 4, characterized in that: The specific adjustments based on the industry risk factor FX are as follows: Low-risk industries have stable income sources and low default risks, including education, healthcare, and technology. Based on the current industry conditions, the education, healthcare, and technology industries are ranked by risk, i.e., 0 < education < healthcare < technology < 0.5; High-risk industries are those that are greatly affected by economic cycles and policies, including real estate, mining, and energy. Based on the current industry conditions, the risks of real estate, mining, and energy are ranked as follows: 1 < real estate < mining < energy < 0.5; Among them, the education, medical, technology industries as well as the real estate, mining, and energy industries will be ranked and adjusted in real time based on the current risk environment.

6. The financial credit risk identification method according to claim 4, characterized in that: Based on the comprehensive credit risk value Q3, the method for identifying and adjusting the credit score XP is as follows: First, compare the comprehensive credit risk value Q3 with the previous comprehensive credit risk value Q3 of the same borrower. prev Compare the results; If the comprehensive credit risk value Q3 is higher than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[(Q3 / Q3 prev ) / 100+1]; If the comprehensive credit risk value Q3 is lower than the previous comprehensive credit risk value Q3 prev , then the adjustment formula for credit score XP is as follows: XP new =XP×[1-(Q3 / Q3 prev ) / 100]; Among them, XP new The next credit score is the replacement input value of the credit score XP in the basis unit for assessing the credit risk of the same borrower next time.

7. A method for constructing a financial credit risk identification model, characterized in that: It includes a data identification and collection module, which is used to generate basic credit identification and basic risk identification of the borrower's current financial credit risk based on the identification and collection of the borrower's basic information and current financial status; Based on basic credit identification and basic risk identification, and through the construction of the identification model of the data analysis and processing module, the basic assessment value of the borrower's current credit risk is updated according to the basic credit identification and basic risk identification; Based on the basic assessment value, combined with basic credit identification and basic risk identification, as well as the borrower's historical repayment identification, the borrower's current credit risk value is updated and adjusted; Based on the risk value, combined with the borrower's historical repayment identification and industry risk identification, the borrower's comprehensive credit risk value is updated; Based on the borrower's current comprehensive credit risk value and the previous comprehensive credit risk value, the increase and decrease in risk are identified and the basic credit identification is updated.

8. The method for constructing a financial credit risk identification model according to claim 7, characterized in that: The equipment used by the data identification and collection module includes a database management system, which is used to store and manage the basic information of the borrower; The equipment used by the data analysis and processing module includes a credit risk assessment system, which is a software system integrating the above modules and is used to automatically calculate the credit risk value. The credit risk assessment system is used to execute the method described in any one of claims 1 to 6.

Images