Financial product recommendation method and system based on risk tolerance assessment

Abstract

The invention relates to the technical field of product recommendation, in particular to a financial product recommendation method and system based on risk tolerance evaluation, and the method comprises the steps: obtaining risk tolerance evaluation index data of a data object demander in response to a recommendation request of the data object demander for a financial product sent by a client; performing correlation analysis on the original input index data and the original output index data, and selecting mutually independent data as updated input index data and updated output index data; performing normalization processing on the index data; establishing a data envelope analysis model based on the updated index data; based on the data envelope analysis model, determining a risk tolerance value of the data object demander; determining the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and based on the risk tolerance level of the data object demander, recommending a financial product of a corresponding level to the client.

Description

Technical Field

[0001] The present invention relates to the technical field of product recommendation, and in particular to a method and system for recommending financial products based on risk tolerance assessment. Background Art

[0002] In the field of financial investment, staff need to recommend different financial products to different consumers. Currently, most financial product recommendations are made by experienced staff who interview users and make subjective recommendations based on their personal experience, or by asking users to fill out questionnaires first, and then the staff makes personalized recommendations based on the questionnaire results.

[0003] The flaws of this recommendation method are: when filling out the questionnaire, users may conceal their personal income information, or have misunderstandings about the questions in the questionnaire, resulting in inaccurate results. The staff has limited energy and will not verify the authenticity of the questionnaire one by one. The staff recommends investment products based on their subjective experience, which may lead to omissions in the recommendation process or inaccurate recommendation results. Summary of the invention

[0004] In order to address the deficiencies of the prior art, the present invention provides a financial product recommendation method and system based on risk tolerance assessment.

[0005] On the one hand, a financial product recommendation method based on risk tolerance assessment is provided, which is applied to a server and includes: in response to a data object demander's recommendation request for a financial product sent by a client, obtaining risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; performing correlation analysis on the original input index data and the original output index data, and selecting mutually independent data as updated input index data and updated output index data; normalizing the updated input index data and the updated output index data; establishing a data envelopment analysis model based on the updated input index data and the updated output index data; determining the risk tolerance value of the data object demander based on the data envelopment analysis model; determining the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and recommending financial products of corresponding levels to the client based on the risk tolerance level of the data object demander.

[0006] On the other hand, a financial product recommendation system based on risk tolerance assessment is provided, including: a response module, which is configured to: respond to a data object demander's recommendation request for a financial product sent by a client, and obtain risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; an analysis module, which is configured to: perform correlation analysis on the original input index data and the original output index data, select mutually independent data as updated input index data and updated output index data; normalize the updated input index data and the updated output index data; a determination module, which is configured to: establish a data envelopment analysis model based on the updated input index data and the updated output index data; determine the risk tolerance value of the data object demander based on the data envelopment analysis model; an output module, which is configured to: determine the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and recommend financial products of corresponding levels to the client based on the risk tolerance level of the data object demander.

[0007] The above technical solution has the following advantages or beneficial effects: when a user needs financial product recommendation services, the staff only needs to input the user's evaluation index data to get the most matching financial product. The entire recommendation process reduces costs, improves data processing efficiency, and achieves accurate and fast recommendations of financial products. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0009] Figure 1 This is a flow chart of the method of embodiment 1. DETAILED DESCRIPTION

[0010] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0011] In the financial field, the assessment of investors' risk tolerance is the core link in formulating effective investment strategies. As financial markets become more diversified and complex, investors are facing more and more risk factors. Accurately assessing their risk tolerance can help investors make wise choices under uncertain economic conditions, thereby achieving financial goals and enhancing their investment confidence.

[0012] Traditional assessment methods can no longer fully reflect investors' true risk tolerance. Existing technologies mostly rely on simple questionnaires, quantitative indicator analysis or empirical judgment, often ignoring investors' behavioral patterns and psychological factors in a dynamic market environment.

[0013] At present, although there are some risk assessment tools on the market, most of them only consider single-dimensional indicators and cannot comprehensively analyze the various characteristics of investors. In addition, existing methods are inefficient in processing complex data and are difficult to adapt to the needs of large-scale users. At the same time, many assessment systems lack the ability to update in real time and intelligence, and cannot promptly reflect the impact of market changes on risk tolerance.

[0014] Therefore, there is an urgent need for a more comprehensive, systematic and intelligent assessment method to accurately identify and analyze users' risk tolerance, help investors formulate personalized investment strategies, optimize asset allocation, and improve the scientificity and effectiveness of investment decisions.

[0015] Embodiment 1, as Figure 1 As shown, this embodiment provides a financial product recommendation method based on risk tolerance assessment, which is applied to a server and includes: S101: in response to a data object demander's recommendation request for a financial product sent by a client, obtaining risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; S102: performing correlation analysis on the original input index data and the original output index data, selecting mutually independent data as updated input index data and updated output index data; normalizing the updated input index data and the updated output index data; S103: establishing a data envelopment analysis model based on the updated input index data and the updated output index data; determining the risk tolerance value of the data object demander based on the data envelopment analysis model; S104: determining the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and recommending financial products of corresponding levels to the client based on the risk tolerance level of the data object demander.

[0016] Furthermore, the original input indicator data include: the age of the data object demander, the amount of debt, the proportion of expenditure, the duration of investment experience, the expected investment period, the types of risks experienced, the amount of risks experienced, the degree of risk aversion and the proportion of funds used for financial investment; the original output indicator data include: the annual income growth rate and the total asset growth rate.

[0017] The expenditure ratio is: (Annual total expenditure Total assets amount) 100%.

[0018] The duration of investment experience refers to: the length of investment experience the investor has (in years).

[0019] The expected investment period refers to the estimated length of time (years) that an investor will make a certain investment.

[0020] The types of risks experienced refer to the types of risks that investors have experienced in previous investment processes, such as purchasing power risk, financial risk, interest rate risk, market risk, liquidity risk, event risk, etc.

[0021] The amount of risk experienced refers to the amount of risk that investors have experienced in previous investment processes.

[0022] The proportion of funds used for financial investment refers to: (the proportion of funds used for financial investment Total assets amount) 100%.

[0023] The degree of risk aversion refers to the degree to which investors are averse to investment risks (can be expressed by a number between 1-10, with the larger the number, the higher the risk aversion).

[0024] The annual income growth rate is: (annual income growth amount Total income in the previous year) 100%, annual revenue growth Total income this year Total income from the previous year.

[0025] The total assets growth rate refers to: (total assets growth amount Total assets in the previous year) 100%, total asset growth Total assets this year Total assets in the previous year.

[0026] In this embodiment, when constructing evaluation indicators, we follow the principles of objectivity, comparability, and operability, select representative evaluation indicators, and form a systematic rating indicator system; based on the characteristics of the data envelopment analysis model, we divide the evaluation indicators into two categories: input indicators and output indicators.

[0027] The risk tolerance evaluation indicators are shown in Table 1.

[0028] Table 1 Risk tolerance evaluation indicators

[0029] The server is connected to the database of the financial trading platform through the API interface to collect relevant data on the target users' risk tolerance evaluation indicators in real time.

[0030] Furthermore, the step S102: performing correlation analysis on the original input indicator data and the original output indicator data, and selecting mutually independent data as updated input indicator data and updated output indicator data, includes: performing Spearman correlation analysis on the original input indicator data pairwise, and determining the correlation coefficient Is it greater than the set threshold? If it is greater than the set threshold, it means that the two original input indicators are independent of each other; if it is less than or equal to the set threshold, it means that the two original input indicators are not independent of each other, and the independent data are selected as the updated input indicator data.

[0031] Perform Spearman correlation analysis on the original output indicator data to determine the correlation coefficient Is it greater than the set threshold? If it is greater than the set threshold, it means that the two original output indicators are independent of each other; if it is less than or equal to the set threshold, it means that the two original output indicators are not independent of each other, and the independent data are selected as the updated output indicator data.

[0032] .

[0033] in, is the correlation coefficient, and They are Metrics for each user and indicators The ranking, and The indicators are and indicators The average ranking of Represents the total number of observed samples.

[0034] It should be understood that Metrics for each user The ranking It means: Metrics for each user Descending rank of the total number of sample observations (starting from 1).

[0035] index The average ranking Refers to: indicators of all users The sum of the ranks of divided by the total number of observed samples, that is, .

[0036] Table 2 Description of the ranking of user indicators

[0037] Knowable , , obtained by Spearman correlation analysis formula If it is less than the set threshold, the indicator and indicators The correlation is not obvious, so it can be determined that the indicator and indicators Is independent.

[0038] If there is a high correlation between the input indicators, it may lead to redundant information, thus affecting the efficiency and accuracy of the data envelopment analysis (DEA) model. By identifying and eliminating indicators with strong correlation, the model can be simplified and its computational efficiency can be improved.

[0039] In practical applications, investors and financial institutions often hope to obtain the best evaluation results with limited resources. By determining independent indicators, the cost of data collection and processing can be effectively reduced and the efficiency of resource use can be improved.

[0040] Conducting correlation analysis on DEA indicators to find independent indicators can not only improve the effectiveness and interpretability of the model, but also optimize resource allocation and improve decision-making quality, thereby providing investors and financial institutions with more reliable risk tolerance assessments.

[0041] Furthermore, the S102: normalizing the updated input index data and the updated output index data includes: ;in, , , Indicates The decision-making unit Item data; express The decision-making unit The maximum value in the item data, express The decision-making unit The minimum value in the item data, Indicates the updated The decision-making unit Item data.

[0042] In a data envelopment analysis model, if the numerical range of some indicators is much larger than that of other indicators, it may have a disproportionate impact on the model results. Through normalization, the interference of this imbalance on the results can be reduced and the stability of the model can be improved.

[0043] Data envelopment analysis mainly consists of the following three parts: input data, output data and decision-making unit. Input data can be used as the consumption resource of the decision-making unit, and output data can be used as the output of the decision-making unit after resource consumption. The decision-making unit refers to the operation carrier that converts part of the input into the corresponding output. Each decision-making unit can be regarded as an entity of the same type, and each decision-making unit has the same input index and output index, and these index values ​​are required to be non-negative. The present invention regards each user as a decision-making unit.

[0044] Further, the S103: based on the updated input indicator data and the updated output indicator data, establish a data envelopment analysis model, and determine the risk tolerance value of the data object demander based on the data envelopment analysis model, including: taking each user as a decision-making unit, assuming a total of decision-making units, each of which has Input and The updated input index data and output index data of each user are used as the input and output of its decision-making unit respectively. express Any one of the decision-making units, For the The input vector of a decision unit, , Indicates The decision-making unit Type of input, For the The output vector of the decision unit, Indicates The decision-making unit Output is the weight of the input vector, is the weight of the output vector.

[0045] Build a data envelopment analysis model: ; Through Charnes-Cooper transformation, the fractional programming is transformed into a linear programming, , , , the model can become: ; The dual model of the linear programming model is as follows: ; Slack variables are introduced into the dual model. The model with slack variables is as follows: ; in, is the indicator weight of a decision-making unit, represents the negative slack variable, represents the positive slack variable, Represents the efficiency value of the decision-making unit.

[0046] Solve the model with slack variables and get the efficiency value of the decision-making unit The optimal solution , the efficiency value of the decision-making unit The optimal solution As the user's final risk tolerance value.

[0047] When the optimal solution and and When both are 0, the evaluation object is said to be effective in data envelopment analysis; when the optimal solution but and When it is not 0, the evaluation object is said to be weakly effective in data envelopment analysis; when the optimal solution When it is not equal to 1, the evaluation object is said to be invalid for data envelopment analysis.

[0048] When the optimal solution of the decision-making unit is And the slack variable , it means that the decision-making unit is judged as "strong DEA effectiveness" under the data envelopment analysis (DEA) model. This means that the quantity of any input cannot be further reduced without reducing output; at the same time, the quantity of any output cannot be increased without increasing input or reducing output. Therefore, the resource utilization of this decision-making unit has reached the optimal configuration state, and there is no redundant input or output.

[0049] However, if the optimal solution , but the slack variable or If it is not zero, it means that the decision-making unit is "weakly effective in DEA". In this case, although the resource allocation is close to the optimal in some aspects, there is still a certain degree of redundancy, and the quantity of all inputs cannot be reduced proportionally unless the output is reduced accordingly; similarly, the quantity of all outputs cannot be increased proportionally unless the input is increased. Therefore, the efficiency of the decision-making unit has not reached the theoretical optimal level and there is still room for improvement.

[0050] If the optimal solution , then the decision-making unit can be judged as "non-DEA effective". This situation usually indicates that the decision-making unit has a waste of resources, indicating that it has not effectively utilized resources in the input-output process and its efficiency is far from the optimal level.

[0051] It should be understood that the Charnes-Cooper transformation is a method of converting constraints containing proportional relationships into standard linear constraints by introducing new variables. In data envelopment analysis, the Charnes-Cooper transformation is often used to convert the efficiency problem of a multi-input, multi-output decision-making unit into a standardized linear programming model.

[0052] Further, the step S104: determining the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander specifically includes: when the user risk tolerance value is equal to 1 and the slack variable , the user's risk tolerance level is determined to be strong; when the user's risk tolerance value is equal to 1 and the relaxation variable and When it is not 0, the user's risk tolerance level is determined to be medium; when the user's risk tolerance value is less than 1, the user's risk tolerance level is determined to be weak.

[0053] For example, when and and When it is equal to 0, it means that the decision unit is effective; when but and When it is not 0, it means that the decision unit is weakly effective; when When it is not equal to 1, it means that the decision unit is invalid.

[0054] Further, the S104: based on the risk tolerance level of the data object demander, recommending financial products of corresponding levels to the client, specifically includes: if the risk tolerance level of the data object demander is a strong level, recommending financial products of high risk levels to the client; if the risk tolerance level of the data object demander is a medium level, recommending financial products of medium risk levels to the client; if the risk tolerance level of the data object demander is a weak level, recommending financial products of weak risk levels to the client.

[0055] The evaluation of users' risk tolerance needs to be described from different aspects using multiple indicators. These indicators often have inconsistent dimensions. Data envelopment analysis does not need to consider the issue of dimensions, which greatly improves the flexibility and accuracy of the evaluation. Data envelopment analysis can process multiple input variables and output variables at the same time, and can comprehensively consider the various characteristics of investors. This comprehensiveness makes risk assessment no longer limited to a single dimension, and can more accurately reflect the overall risk tolerance of investors.

[0056] Embodiment 2. This embodiment provides a financial product recommendation system based on risk tolerance assessment, including: a response module, which is configured to: respond to a data object demander's recommendation request for a financial product sent by a client, and obtain risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; an analysis module, which is configured to: perform correlation analysis on the original input index data and the original output index data, and select mutually independent data as updated input index data and updated output index data; normalize the updated input index data and the updated output index data; a determination module, which is configured to: establish a data envelopment analysis model based on the updated input index data and the updated output index data; determine the risk tolerance value of the data object demander based on the data envelopment analysis model; an output module, which is configured to: determine the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and recommend financial products of corresponding levels to the client based on the risk tolerance level of the data object demander.

[0057] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A financial product recommendation method based on risk tolerance assessment, characterized by: Applicable to servers, including: In response to a data object demander's recommendation request for a financial product sent by a client, obtaining risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; Performing correlation analysis on the original input indicator data and the original output indicator data, selecting mutually independent data as updated input indicator data and updated output indicator data; performing normalization processing on the updated input indicator data and the updated output indicator data; Based on the updated input indicator data and the updated output indicator data, a data envelopment analysis model is established; based on the data envelopment analysis model, the risk tolerance value of the data object demander is determined; Based on the risk tolerance value of the data object demander, determine the risk tolerance level of the data object demander; based on the risk tolerance level of the data object demander, recommend financial products of corresponding levels to the client.

2. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: The original input indicator data include: the age of the data object demander, debt amount, investment experience, expected investment period, types of risks experienced, amount of risks experienced and the proportion of funds used for financial investment; the original output indicator data include: annual income growth rate and total asset growth rate.

3. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: Perform correlation analysis on the original input indicator data and the original output indicator data, and select independent data as updated input indicator data and updated output indicator data, including: Perform Spearman correlation analysis on the original input indicator data to determine the correlation coefficient Whether it is greater than the set threshold. If it is greater than the set threshold, it means that the two original input indicators are independent of each other. If it is less than or equal to the set threshold, it means that the two original input indicators are not independent of each other, and the independent data are selected as the updated input indicator data; Perform Spearman correlation analysis on the original output indicator data to determine the correlation coefficient Is it greater than the set threshold? If it is greater than the set threshold, it means that the two original output indicators are independent of each other; if it is less than or equal to the set threshold, it means that the two original output indicators are not independent of each other, and the independent data are selected as the updated output indicator data.

4. The method for recommending financial products based on risk tolerance assessment according to claim 3, characterized in that: in, Correlation coefficient The calculation formula is: ； in, and They are Metrics for each user and indicators The ranking, and The indicators are and indicators The average ranking of Represents the total number of observed samples.

5. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: The updated input indicator data and the updated output indicator data are normalized, including: ； in, , , Indicates The decision-making unit Item data; express The decision-making unit The maximum value in the item data, express The decision-making unit The minimum value in the item data, Indicates the updated The decision-making unit Item data.

6. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: Based on the updated input indicator data and the updated output indicator data, a data envelopment analysis model is established. Based on the data envelopment analysis model, the risk tolerance value of the data object demander is determined, including: taking each user as a decision-making unit, assuming a total of decision-making units, each of which has Input and The updated input index data and output index data of each user are used as the input and output of its decision-making unit respectively. express Any one of the decision-making units, For the The input vector of a decision unit, , Indicates The decision-making unit Type of input, For the The output vector of the decision unit, Indicates The decision-making unit Output is the weight of the input vector, is the weight of the output vector; Build a data envelopment analysis model: ； Through Charnes-Cooper transformation, the fractional programming is transformed into a linear programming, , , , the model can be transformed into ； The dual model of the linear programming model is as follows: ； Slack variables are introduced into the dual model. The model with slack variables is as follows: ； in, is the indicator weight of a decision-making unit, represents the negative slack variable, represents the positive slack variable, It represents the efficiency value of the decision-making unit; Solve the model with slack variables and get the efficiency value of the decision-making unit The optimal solution , the efficiency value of the decision-making unit The optimal solution As the user's final risk tolerance value.

7. The method for recommending financial products based on risk tolerance assessment according to claim 6, characterized in that: When the optimal solution and and When both are 0, the evaluation object is said to be effective in data envelopment analysis; when the optimal solution but and When it is not 0, the evaluation object is said to be weakly effective in data envelopment analysis; when the optimal solution When it is not equal to 1, the evaluation object is said to be invalid for data envelopment analysis.

8. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: Based on the risk tolerance value of the data object demander, determine the risk tolerance level of the data object demander, including: When the user's risk tolerance value is equal to 1 and the slack variable When the user's risk tolerance level is determined to be strong; When the user's risk tolerance value is equal to 1 and the slack variable and When it is not 0, the user's risk tolerance level is determined to be medium; When the user risk tolerance value is less than 1, the user risk tolerance level is determined to be a weak level.

9. The method for recommending financial products based on risk tolerance assessment according to claim 1, characterized in that: Based on the risk tolerance level of the data object demander, recommend financial products of corresponding level to the client, including If the risk tolerance level of the data object demander is strong, then recommend financial products with high risk levels to the client; If the risk tolerance level of the data subject demander is medium, then recommend financial products with medium risk level to the client; If the risk tolerance level of the data object demander is a weak level, a financial product with a weak risk level is recommended to the client.

10. A financial product recommendation system based on risk tolerance assessment, characterized by: include: A response module is configured to: in response to a data object demander's recommendation request for a financial product sent by a client, obtain risk tolerance evaluation index data of the data object demander; the risk tolerance evaluation index data includes: original input index data and original output index data; An analysis module is configured to: perform correlation analysis on the original input indicator data and the original output indicator data, select mutually independent data as updated input indicator data and updated output indicator data; and perform normalization processing on the updated input indicator data and the updated output indicator data; A determination module is configured to: establish a data envelopment analysis model based on the updated input indicator data and the updated output indicator data; determine the risk tolerance value of the data object demander based on the data envelopment analysis model; The output module is configured to: determine the risk tolerance level of the data object demander based on the risk tolerance value of the data object demander; and recommend financial products of corresponding levels to the client based on the risk tolerance level of the data object demander.

Images