Method and system for establishing financial transaction risk feature library

Abstract

The invention provides a method and a system for establishing a financial transaction risk feature library, and relates to the technical field of financial risk assessment. Transaction data of financial products, income conditions of other financial products in the same industry and income conditions of markets to which the financial products belong are acquired; and a comprehensive risk assessment system is constructed from three levels of microcosmic, mesoscopic and macroscopic in sequence. Calculating a trading volume mean value, a return rate fluctuation rate and a trading price change rate of the to-be-analyzed financial product, and generating a microscopic characteristic index; performing section division according to the industry, calculating the average yield and internal correlation of the industry, and generating a mesoscopic feature index; markets are divided according to geographic areas, and macroscopic feature indexes are generated by combining the yield rate of financial products, the market yield rate and the volatility index. And finally, generating a comprehensive risk index by integrating the feature indexes, further generating a feature vector, evaluating the risk of the to-be-analyzed financial product, and establishing a financial transaction risk feature library.

Description

Technical Field

[0001] The present invention relates to the technical field of financial risk assessment, and specifically provides a method and system for establishing a financial transaction risk feature library. Background Art

[0002] In the financial trading market, risk assessment is one of the core topics. Traditional risk assessment methods usually focus on single-level analysis, such as evaluating individual trading objects, or analyzing from the industry level, or even relying on the overall judgment of the macroeconomic environment. However, these methods often lack systematicness and cannot comprehensively capture the multi-level and multi-dimensional risk characteristics in the financial market. Especially in the complex environment of increasing market volatility, enhanced industry correlation, and expanding regional economic impact, single-level risk assessment means may miss key information, resulting in inaccurate risk ratings and affecting the scientific nature of investment decisions. Therefore, there is an urgent need for a comprehensive method that integrates multi-dimensional analysis to more comprehensively and accurately identify and evaluate the risk characteristics in financial transactions and improve the risk management capabilities of market participants.

[0003] In the prior art, the publication number CN117974280A discloses a method and device for establishing a bank financial transaction risk feature library, which performs missing value processing and outlier processing on the historical financial transaction data of different bank customers to obtain the processed historical financial transaction data of different bank customers; performs feature generation processing and feature selection processing based on automatic feature engineering technology to obtain multiple bank financial transaction risk features corresponding to the historical financial transaction data; determines the association relationship between different bank financial transaction risk features; the association relationship is used to describe the subordinate relationship between features; according to the association relationship, determines the database level where different bank financial transaction risk features are located; and establishes a bank financial transaction risk feature library according to the bank financial transaction risk features located at different database levels.

[0004] The main problems existing in the above solution are: mainly processing the historical financial transaction data of different bank customers, focusing on feature extraction at the customer level, with a relatively single data source, making it difficult to fully capture the impact of the external market environment and industry dynamics on transaction risks; and having a weak adaptability to the current dynamic changes in the market, and there may be a situation where the risk assessment results lag behind the actual market.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for establishing a financial transaction risk feature library to solve the problems raised in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for establishing a financial transaction risk feature library, the specific steps include:

[0009] Step 1: During the acquisition time period, obtain the transaction data of the financial product to be analyzed, where the transaction data includes the trading volume, transaction price, and rate of return in each transaction record of the financial product to be analyzed;

[0010] Step 2: Based on the transaction data, calculate the micro-parameters of the financial product to be analyzed, which include the average trading volume, the rate of change of the transaction price, and the volatility of the rate of return, and generate a micro-feature index of the financial product to be analyzed based on the micro-parameters;

[0011] Step 3: Obtain the rates of return of other financial products in the same industry to calculate the average rate of return and internal correlation of the industry to which the financial product to be analyzed belongs, and then generate a meso-feature index of the financial product to be analyzed;

[0012] Step 4: Obtain the market rate of return and volatility index of the market to which the financial product to be analyzed belongs to generate a systematic risk index, and then generate a macro-feature index of the financial product to be analyzed;

[0013] Step 5: Generate a comprehensive risk index based on the micro-feature index, meso-feature index, and macro-feature index, and judge the risk level of the financial product to be analyzed based on the comprehensive risk index, and integrate the risk level, micro-feature index, meso-feature index, and macro-feature index of the financial product to be analyzed into a feature vector, and construct a financial transaction risk feature library with the feature vectors of multiple financial products to be analyzed.

[0014] Further, an acquisition time period is 24 hours.

[0015] Further, the principle for generating the micro-feature index is:

[0016] The formula for calculating the average trading volume of the financial product to be analyzed is:

[0017]

[0018] where, represents the average trading volume of the financial product to be analyzed, i represents the index of the transaction record during the acquisition time period, and i ∈ [1, N], N represents the total number of transaction records, V i represents the trading volume of the i-th transaction record;

[0019] The formula for calculating the volatility of the rate of return of the financial product to be analyzed is:

[0020]

[0021] Among them, represents the average return rate of the financial product to be analyzed, R i represents the return rate of the i-th transaction record, and σ represents the return rate volatility;

[0022] The formula for calculating the price change rate is:

[0023]

[0024] Among them, △P represents the price change rate of the financial product to be analyzed, P 1 represents the transaction price of the first transaction record within the collection time period, P end represents the transaction price of the N-th transaction record within the collection time period;

[0025] The formula for generating the micro feature index is:

[0026]

[0027] Among them, U represents the micro feature index, u 1 、u 2 、u 3 respectively represent the weight coefficients of the average trading volume of the financial product to be analyzed, the return rate volatility of the financial product to be analyzed, and the price change rate of the financial product to be analyzed, u 1 +u 2 +u 3 =1, and u 1 =u 2 =u 3 .

[0028] Furthermore, the principle for generating the meso feature index is:

[0029] The formula for calculating the average return rate of the industry to which the financial product to be analyzed belongs is:

[0030]

[0031] Among them, represents the average return rate of the industry to which the financial product to be analyzed belongs, j represents the index of other financial products in the industry to which the financial product to be analyzed belongs, and j ∈ [1, M], where M represents the number of other financial products in the industry to which the financial product to be analyzed belongs, S j represents the average return rate of the j-th financial product within the same collection time period in the industry to which the financial product to be analyzed belongs, excluding the financial product to be analyzed;

[0032] The formula for calculating the internal correlation is:

[0033]

[0034] S R = [S R,1 …S R,i …S R,N

[0035] S R′ = [S R′,1 … S R′,i … S R′,N

[0036]

[0037] Among them, ρ R represents the internal correlation between the financial product to be analyzed in the industry and other financial products in its industry. Cov(S R , S R′ ) represents the covariance of the returns of the financial product to be analyzed and the other financial products with the closest transaction records in time within its industry. Var(S R ) represents the variance of the returns of the financial product to be analyzed. Var(S R′ ) represents the variance of the returns of the other financial products in the industry of the financial product to be analyzed that are closest to the collection time. S R represents the return matrix of the financial product to be analyzed. S R,i represents the return of the i-th transaction record of the financial product to be analyzed. S R′ represents the return matrix of the other financial products in the industry that are closest in time to the financial product to be analyzed at each collection time. S R′,i represents the return of the other financial products in the industry that are closest in time to the i-th transaction record of the financial product to be analyzed, represents the average return of the financial product to be analyzed, represents the average return of the other financial products in the industry of the financial product to be analyzed that are closest in time to the transaction records of the financial product to be analyzed;

[0038] The formula for generating the meso-characteristic index is:

[0039]

[0040] Among them, X represents the meso-characteristic index, x 1 , x 2 respectively represent the weight coefficients of the average return and the internal correlation of the sector. x 1 + x 2 = 1, and x 1 < x 2 . ​​

[0041] Furthermore, the principle for generating the macro feature index is as follows:

[0042] The formula for generating the system risk index is as follows:

[0043]

[0044] Y = [Y 1 …Y i …Y N

[0045]

[0046] where β represents the system risk index, Cov(S R , Y) represents the covariance between the return of the financial product to be analyzed and the market return, Var(Y) represents the variance of the market return, Y represents the market return matrix, and Y i represents the market return corresponding to the i-th transaction record, represents the market average return;

[0047] The formula for generating the macro feature index is as follows:

[0048] W = w 1 ·β + w 2 ·D VIX

[0049] where W represents the macro feature index, D VIX represents the volatility index, and w 1 , w 2 represent the weight coefficients of the system risk index and the volatility index respectively, w 1 + w 2 = 1, and w 1 > w 2 .

[0050] Furthermore, the principle for integrating the feature vectors is as follows:

[0051] The formula for generating the comprehensive risk index is as follows:

[0052] Z = z 1 ·U + z 2 ·X + z 3 ·W

[0053] where Z represents the comprehensive risk index, U represents the micro feature index, X represents the meso feature index, W represents the macro feature index, and z 1 , z 2 , z 3 represent the weight coefficients of the micro feature index, the meso feature index, and the macro feature index respectively, and z​1 +z 2 +z 3 = 1, and z 1 > z 2 > z 3 ;

[0054] And the eigenvector obtained by integrating the risk level, micro feature index, meso feature index and macro feature index of the financial product to be analyzed is:

[0055] A = [U, X, W, Z]

[0056] Where A represents the eigenvector of a financial product to be analyzed;

[0057] Calculate the eigenvectors of all financial products to be analyzed, and combine them to generate a financial transaction risk feature library.

[0058] The present invention also provides a system for establishing a financial transaction risk feature library. The system is used to implement the above method for establishing a financial transaction risk feature library, and specifically includes:

[0059] A data acquisition module, which is used to obtain the transaction data of the financial product to be analyzed during the acquisition time period. The transaction data includes the trading volume, transaction price and yield in each transaction record of the financial product to be analyzed;

[0060] A micro analysis module, which is used to calculate the micro parameters of the financial product to be analyzed based on the transaction data, including the average trading volume, the change rate of the transaction price and the volatility of the yield, and generate the micro feature index of the financial product to be analyzed based on the micro parameters;

[0061] A meso analysis module, which is used to obtain the yields of other financial products in the same industry to calculate the average yield and internal correlation of the industry to which the financial product to be analyzed belongs, and then generate the meso feature index of the financial product to be analyzed;

[0062] A macro analysis module, which is used to obtain the market yield and volatility index of the market to which the financial product to be analyzed belongs to generate a system risk index, and then generate the macro feature index of the financial product to be analyzed;

[0063] A comprehensive analysis module, which is used to generate a comprehensive risk index based on the micro feature index, meso feature index and macro feature index, judge the risk level of the financial product to be analyzed based on the comprehensive risk index, and integrate the risk level, micro feature index, meso feature index and macro feature index of the financial product to be analyzed into an eigenvector, and construct a financial transaction risk feature library with the eigenvectors of multiple financial products to be analyzed.

[0064] Compared with the prior art, the beneficial effects of the present invention are:

[0065] By collecting historical transaction data at the micro level, collective risk characteristics of industry segments at the meso level, and market risk parameters at the macro level, the present invention achieves multi-level coverage of data dimensions, can timely capture the dynamic changes in the market, reflects the correlation and linkage effects among risk characteristics at the micro, meso, and macro levels, and provides more comprehensive basic information for subsequent risk characteristic analysis; by calculating the average trading volume, volatility of the rate of return, and price change rate of the financial product to be analyzed, it can comprehensively capture the trading activity, income stability, and price fluctuation trend of the financial product to be analyzed, and more comprehensively and three-dimensionally reflects the risk characteristics of the financial product to be analyzed; by dividing into segments according to the industry, reflecting the comprehensive performance of all trading objects within the segment, it can reflect the overall income level and development trend of the industry, and by measuring the income correlation among trading objects within the segment through internal correlation, it can reflect the degree of co-fluctuation within the industry, enhancing the depth of risk assessment; by dividing the market according to geographical regions, the risk assessment can comprehensively consider the characteristics of different regional markets, combine the market rate of return and the benchmark rate of return to measure the systematic risk, introduce the volatility index to capture the market's expectation of future fluctuations, and thus more comprehensively reflect the systematic risk status of the market.

[0066] The present invention also generates a comprehensive risk index through micro, meso, and macro characteristics, forming a comprehensive multi-level risk assessment system from the financial product to be analyzed to the industry segment and then to the overall market. The financial transaction risk characteristic library constructed in this way can accumulate and record the risk characteristics of different trading objects in the long term, form dynamic updates, and combine the risk profiles of trading individuals, industries, and markets, significantly enhancing the comprehensiveness and depth of risk assessment, being able to adapt to different types of trading objects and different market environments, and having a wider scope of use. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a schematic flowchart of the method in an embodiment of the present invention;

[0068] Figure 2 It is a schematic diagram of the system modules in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0069] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to specific embodiments.

[0070] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not represent any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0071] Embodiment:

[0072] Please refer to Figure 1 , the present invention provides a technical solution:

[0073] Step 1: During the acquisition time period, obtain the transaction data of the financial product to be analyzed. The transaction data includes the trading volume, transaction price and yield in each transaction record of the financial product to be analyzed;

[0074] In this embodiment, an acquisition time period is 24 hours.

[0075] Step 2: Based on the transaction data, calculate the microscopic parameters of the financial product to be analyzed, which include the average trading volume, the change rate of the transaction price and the volatility of the yield, and generate a microscopic characteristic index of the financial product to be analyzed based on the microscopic parameters;

[0076] In this embodiment, the principle for generating the microscopic characteristic index is:

[0077] The formula for calculating the average trading volume of the financial product to be analyzed is:

[0078]

[0079] Wherein, represents the average trading volume of the financial product to be analyzed, i represents the index of the transaction record within the acquisition time period, and i ∈ [1, N], N represents the total number of transaction records, and V i represents the trading volume of the i-th transaction record;

[0080] The average trading volume reflects the trading quantity of the financial product to be analyzed, is generated based on the number of transaction records within the acquisition time period and the trading volume of each transaction record, and is proportional to the trading volume of each transaction record.

[0081] The formula for calculating the return volatility of the financial product to be analyzed is as follows:

[0082]

[0083] Among them, represents the average return of the financial product to be analyzed, and R i represents the return of the i-th transaction record, and σ represents the return volatility;

[0084] The return is obtained by comprehensively considering the changes in the returns of the buyer and the seller within the collection period. That is, the time interval length between the first transaction and the last transaction within the collection period multiplied by the return per unit time is the return of the financial product to be analyzed within the collection period.

[0085] The return volatility reflects the stability of the financial product to be analyzed. The higher the volatility, the greater the change range of the return, the more unstable the return, and the higher the risk.

[0086] The formula for calculating the price change rate is as follows:

[0087]

[0088] Among them, △P represents the price change rate of the financial product to be analyzed, and P 1 represents the transaction price of the first transaction record within the collection time period, and P end represents the transaction price of the N-th transaction record within the collection time period;

[0089] The price change rate reflects the fluctuation range of the price of the financial product to be analyzed within the set time period. The larger the price change rate, the greater the price fluctuation range, and the higher the risk of the financial product to be analyzed.

[0090] The formula for generating the micro-characteristic index is as follows:

[0091]

[0092] Among them, U represents the micro-characteristic index, and u 1 , u 2 , u 3 respectively represent the weight coefficients of the average trading volume of the financial product to be analyzed, the return volatility of the financial product to be analyzed, and the price change rate of the financial product to be analyzed. u 1 + u 2 + u 3 = 1, and u 1 = u 2 = u 3 .

[0093] The micro - characteristic index reflects the risk situation corresponding to a specific financial product to be analyzed. The larger the micro - characteristic index, the higher the risk. At the same time, considering the trading volume, return situation, and price fluctuation situation of the financial product to be analyzed, these data need to be taken into account during risk assessment. Therefore, the weights of the three are the same. Moreover, the average trading volume is inversely proportional to the micro - characteristic index, while the return rate volatility and price change rate are directly proportional to the micro - characteristic index.

[0094] Step 3: Obtain the returns of other financial products in the same industry to calculate the average return and internal correlation of the industry to which the financial product to be analyzed belongs, and then generate the meso - characteristic index of the financial product to be analyzed.

[0095] In this embodiment, the industry sectors are classified according to the China Securities Regulatory Commission (CSRC) industry classification standard, and each industry sector contains several financial products.

[0096] In this embodiment, the principle for generating the meso - characteristic index is as follows:

[0097] The formula for calculating the average return of the industry to which the financial product to be analyzed belongs is:

[0098]

[0099] Among them, represents the average return of the industry to which the financial product to be analyzed belongs, j represents the index of other financial products in the industry to which the financial product to be analyzed belongs, and j ∈ [1, M], where M represents the number of other financial products in the industry to which the financial product to be analyzed belongs, and S j represents the average return of the j - th financial product in the same collection time period within the industry to which the financial product to be analyzed belongs, excluding the financial product to be analyzed.

[0100] The average return of each industry reflects the average of the returns of each financial product in the industry, which is the average of the returns of all financial products to be analyzed in the industry and reflects the overall return situation of the industry.

[0101] The formula for calculating the internal correlation is:

[0102]

[0103] S R =[S R,1 …S R,i …S R,N

[0104] S R′ =[S R ′ ,1 … S R′,i … S​R′,N

[0105]

[0106]

[0107] Among them, ρ R represents the internal correlation between the financial product to be analyzed in the industry and other financial products within its industry. Cov(S R , S R′ ) represents the covariance of the returns between the financial product to be analyzed and the other financial products with the closest transaction records in time within its industry. Var(S R ) represents the variance of the returns of the financial product to be analyzed. Var(S R′ ) represents the variance of the returns of the other financial products within the industry of the financial product to be analyzed that are closest in time to the collection moment. S R represents the return matrix of the financial product to be analyzed. S R,i represents the return of the i-th transaction record of the financial product to be analyzed. S R′ represents the return matrix of the other financial products within the industry that are closest in time to the financial product to be analyzed at each collection moment. S R′,i represents the return of the other financial products within the industry that are closest in time to the i-th transaction record of the financial product to be analyzed. represents the average return of the financial product to be analyzed. represents the average return of the other financial products within the industry of the financial product to be analyzed that are closest in time to the transaction records of the financial product to be analyzed;

[0108] The internal correlation of the sector reflects the tightness of the connection between the financial product to be analyzed within the sector and other financial products. The higher the internal correlation, the closer the connection between different financial products within the sector, and the lower the overall risk. Taking the collection moment of the financial product to be analyzed as the main reference, mark the time corresponding to each transaction record within the collection period, and query the return situation of other financial products in the same sector that are closest in time to each transaction record for calculating the internal correlation.

[0109] The formula for generating the meso-characteristic index is:

[0110]

[0111] Among them, X represents the meso-characteristic index, x 1 , x 2 respectively represent the weight coefficients of the average return and the internal correlation of the sector, x 1 + x 2 = 1, and x 1 < x​2 。

[0112] The meso - characteristic index reflects the risk situation at the sector level. The meso - characteristic index is directly proportional to the risk magnitude. The higher the average return rate, the lower the risk; the higher the internal correlation, the lower the risk. Therefore, the meso - characteristic index is inversely proportional to the average return rate and the internal correlation of the sector. In the micro - characteristic index, a part of the return rate has been considered. Therefore, here more attention is paid to the internal correlation of the sector. The weight coefficient of the average return rate is lower than that of the internal correlation, x 1 = 0.3, x 2 = 0.7.

[0113] Step 4: Obtain the market return rate and volatility index of the market to which the financial product to be analyzed belongs, so as to generate a system risk index, and then generate the macro - characteristic index of the financial product to be analyzed;

[0114] In this embodiment, the market is divided according to countries; the volatility index is used to measure the data of the expected volatility in the next 30 days in the S&P 500 index option market, which is calculated and released by the Chicago Board Options Exchange. A higher volatility index indicates an increase in market volatility, reflecting the uncertainty of the market, and a lower volatility index indicates lower market volatility and a more stable market.

[0115] In this embodiment, the principle for generating the macro - characteristic index is as follows:

[0116] The formula for generating the system risk index is as follows:

[0117]

[0118] Y = [Y 1 … Y i … Y N

[0119]

[0120] where β represents the system risk index, Cov(S R , Y) represents the covariance between the return rate of the financial product to be analyzed and the market return rate, Var(Y) represents the variance of the market return rate, Y represents the market return rate matrix, Y i represents the market return rate corresponding to the i - th transaction record, represents the market average return rate;

[0121] The system risk index, that is, the β - index, reflects the stability of the entire market, reflecting the relationship between the change in the return rate of the financial product to be analyzed and the change in the market return rate. When the β - index increases, it indicates that the financial product is more sensitive to market fluctuations, the change range increases, and the risk rises. ​

[0122] The formula for generating the macro - feature index is as follows:

[0123] W = w 1 ·β + w 2 ·D VIX

[0124] Wherein, W represents the macro - feature index, D VIX represents the volatility index, w 1 and w 2 respectively represent the weight coefficients of the system risk index and the volatility index. w 1 + w 2 = 1, and w 1 > w 2 .

[0125] When the volatility index increases, it reflects that the future market volatility intensifies and the market risk rises; when the volatility index decreases, it reflects that the future market volatility weakens and the market risk declines; the macro - feature index reflects the risk situation at the market level in different geographical regions, and is directly proportional to the risk magnitude. The higher the macro - feature index, the higher the market risk. The system risk index is directly related to financial products and reflects the sensitivity of financial products relative to the entire market. Therefore, the weight coefficient is relatively high, while the volatility index D VIX reflects the future trend change of the overall market, and the weight coefficient is relatively low. w 1 = 0.6, w 2 = 0.4.

[0126] Step 5: Generate a comprehensive risk index based on the micro - feature index, the meso - feature index, and the macro - feature index, and judge the risk level of the financial product to be analyzed based on the comprehensive risk index. Integrate the risk level, the micro - feature index, the meso - feature index, and the macro - feature index of the financial product to be analyzed into a feature vector, and construct a financial transaction risk feature library with the feature vectors of multiple financial products to be analyzed.

[0127] In this embodiment, the principle for integrating the feature vectors is as follows:

[0128] The formula for generating the comprehensive risk index is as follows:

[0129] Z = z 1 ·U + z 2 ·X + z 3 ·W

[0130] Wherein, Z represents the comprehensive risk index, U represents the micro - feature index, X represents the meso - feature index, W represents the macro - feature index, z 1 and z 2 and z 3Represent the weight coefficients of the micro - feature index, meso - feature index, and macro - feature index respectively, z 1 +z 2 +z 3 =1, and z 1 >z 2 >z 3 ;

[0131] Based on the combined effects of three levels: the financial product to be analyzed, the sector industry where the financial product to be analyzed is located, and the region where the financial product to be analyzed is located, a comprehensive risk index is generated, which reflects the risk situation of the financial product. Micro - features can directly reflect the dynamic risks of individual assets, being the most intuitive and having the highest weight coefficient; meso - features reveal the potential risks that may be hidden in the entire industry through the correlation between sectors, with the weight coefficient being the second; macro - features analyze the systematic risks and future trends of the entire market, with the weight coefficient being relatively low, z 1 =0.5, z 2 =0.3, z 3 =0.2.

[0132] And the eigenvector integrated from the risk level, micro - feature index, meso - feature index, and macro - feature index of the financial product to be analyzed is:

[0133] A = [U, X, W, Z]

[0134] Wherein, A represents the eigenvector of a financial product to be analyzed;

[0135] Calculate the eigenvectors of all financial products to be analyzed, and combine them to generate a financial transaction risk feature library.

[0136] Please refer to Figure 2 , the present invention also provides a system for establishing a financial transaction risk feature library. The system is used to implement the method for establishing the above - mentioned financial transaction risk feature library, and specifically includes:

[0137] A data acquisition module, which is used to obtain the transaction data of the financial product to be analyzed within the acquisition time period. The transaction data includes the trading volume, transaction price, and rate of return in each transaction record of the financial product to be analyzed;

[0138] A micro - analysis module, which is used to calculate the micro - parameters of the financial product to be analyzed based on the transaction data. The micro - parameters include the average trading volume, the change rate of the transaction price, and the volatility of the rate of return, and generate the micro - feature index of the financial product to be analyzed based on the micro - parameters;

[0139] A meso - analysis module, which is used to obtain the rates of return of other financial products in the same industry to calculate the average rate of return and internal correlation of the industry to which the financial product to be analyzed belongs, and then generate the meso - feature index of the financial product to be analyzed;

[0140] A macro-analysis module, which is used to obtain the market return rate and volatility index of the market to which the financial product to be analyzed belongs, generate a system risk index, and then generate a macro-characteristic index of the financial product to be analyzed;

[0141] A comprehensive analysis module, which is used to generate a comprehensive risk index based on the micro-characteristic index, meso-characteristic index and macro-characteristic index, judge the risk level of the financial product to be analyzed based on the comprehensive risk index, and integrate the risk level, micro-characteristic index, meso-characteristic index and macro-characteristic index of the financial product to be analyzed into a feature vector, and construct a financial transaction risk feature library with the feature vectors of multiple financial products to be analyzed.

[0142] The above formulas are all dimensionless and take their numerical values for calculation. The formula is a formula obtained by collecting a large amount of data for software simulation to get the closest real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0143] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or by the combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0144] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0145] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application.

Claims

1. A method for establishing a financial transaction risk feature database, characterized in that: The specific steps include: Step 1: During the collection period, the transaction data of the financial product to be analyzed is obtained. The transaction data includes the transaction volume, transaction price and yield of each transaction record of the financial product to be analyzed; Step 2: Based on the transaction data, calculate the micro parameters of the financial product to be analyzed, including the average transaction volume, the transaction price change rate and the volatility of the yield, and generate the micro characteristic index of the financial product to be analyzed based on the micro parameters; Step 3: Obtain the yields of other financial products in the same industry to calculate the average yield and internal correlation of the industry to which the financial product to be analyzed belongs, and then generate the meso-level characteristic index of the financial product to be analyzed; Step 4: Obtain the market yield and volatility index of the market to which the financial product to be analyzed belongs to generate a systemic risk index, and then generate a macro characteristic index of the financial product to be analyzed; Step 5: Generate a comprehensive risk index based on the micro-characteristic index, the meso-characteristic index and the macro-characteristic index, and judge the risk level of the financial product to be analyzed based on the comprehensive risk index. In addition, integrate the risk level, micro-characteristic index, meso-characteristic index and macro-characteristic index of the financial product to be analyzed into a feature vector, and construct a financial transaction risk feature library with the feature vectors of multiple financial products to be analyzed.

2. The method for establishing a financial transaction risk feature database according to claim 1, characterized in that: In step 1, one collection time period is 24 hours.

3. The method for establishing a financial transaction risk feature database according to claim 1, characterized in that: The principle for generating the microscopic characteristic index in step 2 is: The formula used to calculate the average volume of the financial instrument to be analyzed is: in, represents the average transaction volume of the financial product to be analyzed, i represents the index of the transaction record within the collection time period, and i∈[1,N], N represents the total number of transaction records, V i Indicates the transaction volume of the i-th transaction record; The formula for calculating the volatility of the return rate of the financial product to be analyzed is: in, Represents the mean rate of return of the financial product to be analyzed, R i represents the rate of return of the i-th transaction record, and σ represents the volatility of the rate of return; The formula used to calculate the price change rate is: Among them, △P represents the price change rate of the financial product to be analyzed, P1 represents the transaction price of the first transaction record within the collection time period, and P end Indicates the transaction price of the Nth transaction record within the collection time period; The formula for generating the micro-characteristic index is: Among them, U represents the micro characteristic index, u1, u2, and u3 represent the weight coefficients of the average trading volume of the financial product to be analyzed, the yield volatility of the financial product to be analyzed, and the price change rate of the financial product to be analyzed, respectively, u1+u2+u3=1, and u1=u2=u3.

4. The method for establishing a financial transaction risk feature database according to claim 1, characterized in that: The principle for generating the meso-characteristic index in step 3 is: The formula for calculating the average rate of return of the industry to which the financial product to be analyzed belongs is: in, represents the average rate of return of the industry to which the financial product to be analyzed belongs, j represents the index of other financial products in the industry to which the financial product to be analyzed belongs, and j∈[1,M], M represents the number of other financial products in the industry to which the financial product to be analyzed belongs, S j It represents the average rate of return of the j-th financial product in the same collection time period, except for the financial product to be analyzed, in the industry to which the financial product to be analyzed belongs; The formula used to calculate the internal correlation is: S R =[S R,1 … S R,i … S R,N ] S R′ =[S R′,1 … S R′,i … S R′,N ] Among them, ρ R It indicates the internal correlation between the financial product to be analyzed in the industry and other financial products in the industry to which it belongs. Cov(S R ,S R′ ) represents the return covariance of the financial product to be analyzed and the other financial products in the industry to which it belongs that are closest in time to each transaction record. Var(S R ) represents the variance of the yield of the financial product to be analyzed, Var(S R′ ) represents the variance of the yield of other financial products in the industry to which the financial product to be analyzed belongs and closest to its collection time, S R represents the yield matrix of the financial product to be analyzed, S R,i represents the rate of return of the i-th transaction record of the financial product to be analyzed, S R′ represents the yield matrix of other financial products in the same industry that are closest to the financial product to be analyzed at each acquisition moment, S R′,i It represents the yield of other financial products in the same industry that are closest in time to the i-th transaction record of the financial product to be analyzed. represents the average rate of return of the financial product to be analyzed, It indicates the average rate of return of other financial products in the industry to which the financial product to be analyzed belongs, which are closest in transaction record time to the financial product to be analyzed; The formula for generating the meso-characteristic index is: Among them, X represents the meso-characteristic index, x1 and x2 represent the weight coefficients of the average return rate and internal correlation of the sector respectively, x1+x2=1, and x1<x2.

5. The method for establishing a financial transaction risk feature database according to claim 4, characterized in that: The principle for generating the macro characteristic index in step 4 is: The formula for generating the system risk index is: Y=[Y1…Y i …Y N ] Among them, β represents the system risk index, Cov(S R ,Y) represents the covariance between the yield of the financial product to be analyzed and the market yield, Var(Y) represents the variance of the market yield, Y represents the market yield matrix, Y i represents the market rate of return corresponding to the i-th transaction record, represents the average market rate of return; The formula used to generate the macro characteristic index is: W=w1·β+w2·D VIX Where W represents the macro characteristic index, D VIX represents the volatility index, w1 and w2 represent the weight coefficients of the systematic risk index and the volatility index respectively, w1+w2=1, and w1>w2.

6. The method for establishing a financial transaction risk feature database according to claim 1, characterized in that: The principle for integrating the feature vectors in step 5 is: The formula used to generate the composite risk index is: Z=z1·U+z2·X+z3·W Wherein, Z represents the comprehensive risk index, U represents the micro characteristic index, X represents the meso characteristic index, W represents the macro characteristic index, z1, z2, z3 represent the weight coefficients of the micro characteristic index, meso characteristic index and macro characteristic index respectively, z1+z2+z3=1, and z1>z2>z3; The characteristic vector obtained by integrating the risk level, micro characteristic index, meso characteristic index and macro characteristic index of the financial product to be analyzed is: A=[U,X,W,Z] Where A represents a characteristic vector of a financial product to be analyzed; Calculate the characteristic vectors of all financial products that need to be analyzed and combine them to generate a financial transaction risk feature library.

7. A system for establishing a financial transaction risk feature database, characterized in that: The system is used to implement the method for establishing a financial transaction risk feature library as described in any one of claims 1 to 6, specifically comprising: A data collection module is used to collect historical transaction data, collective risk characteristics of industry sectors and market risk parameters of macro-regions. The historical transaction data includes the transaction volume, transaction price and yield of the financial products to be analyzed within a collection time period. The collective risk characteristics include the yield of each financial product to be analyzed in the sector within a collection time period. The market risk parameters include the market yield, benchmark yield and volatility index within a collection time period. The micro-analysis module is used to calculate the average transaction volume, volatility of yield and price change rate of each financial product to be analyzed, and generate a micro-characteristic index based on the average transaction volume, volatility of yield and price change rate; The meso-analysis module is used to divide sectors into sectors according to industries. One sector corresponds to one industry. The average rate of return and internal correlation of each sector are calculated. The meso-characteristic index is generated based on the average rate of return and internal correlation of the sectors. The macro analysis module is used to divide the market by geographical region, generate systemic risk index based on market yield and benchmark yield, and generate macro characteristic index based on market systemic risk index and volatility index; The comprehensive analysis module is used to generate a comprehensive risk index based on the micro-characteristic index, the meso-characteristic index and the macro-characteristic index, judge the risk level of the financial product to be analyzed according to the comprehensive risk index, and build a financial transaction risk feature library based on the micro-characteristic index of the financial product to be analyzed, as well as the meso-characteristic index and macro-characteristic index of the sectors and geographical regions corresponding to the financial product to be analyzed.

Images