A financial post-loan risk early warning method

Abstract

The present invention relates to the technical field of post-loan risk assessment, and discloses a financial post-loan risk early warning method, comprising: obtaining the loan data of a target person to form preliminary assessment data, when the collection time is reached, obtaining the account data of the target person to form risk assessment data, forming grade adjustment data, forming risk early warning data, and performing corresponding risk early warning. The financial post-loan risk early warning method determines whether the lender has evaded or delayed the post-loan risk assessment of the institution according to the flow data submitted by the lender, determines whether the repayment funds of the lender are stable according to the income source of the repayment account of the lender, and when the lender fails to repay in time, determines the repayment willingness of the lender according to the lender's answering of the loan collection call of the financial institution, and promptly performs risk early warning for the lender who cannot repay the loan, so as to avoid the situation that when the lender suddenly loses contact, it causes property loss to the financial institution that is difficult to recover the funds.

Description

Technical Field

[0001] The present invention relates to the technical field of post-loan risk assessment, and in particular to a financial post-loan risk early warning method. Background Art

[0002] Financial post-loan risk early warning methods refer to real-time monitoring and early warning of loan risks through a series of technologies and methods after the loan is issued, so as to take timely measures to reduce or control risks. It involves multiple fields such as data technology, machine learning, credit assessment and financial technology. The integration and progress of these technologies not only improves the efficiency and accuracy of risk monitoring, but also promotes the innovation and reform of financial institutions in post-loan risk management. With the continuous development of technology, the methods and tools of post-loan risk early warning have become more refined and intelligent;

[0003] By analyzing the borrower's historical repayment data, etc., possible risk patterns can be identified, and the borrower's financial status can be monitored in real time. By building a credit scoring model, the borrower's credit status can be evaluated, and the credit score can be updated in real time according to the borrower's repayment behavior, financial status, etc., and the risk level can be adjusted in time. Machine learning algorithms can be used to train large amounts of data to identify potential risk characteristics and predict future risks. Through anomaly detection models, anomalies that are inconsistent with normal repayment behavior can be identified. The borrower's daily behavior is monitored to identify possible risk signals. The borrower's related parties are monitored to discover potential risk transmission, and a series of early warning indicators are set. When the indicator exceeds the threshold, an early warning is triggered. According to the market environment and risk changes, the early warning threshold is adjusted dynamically to ensure the accuracy and timeliness of the early warning. Through the automated collection system, the repayment situation is monitored in real time, and the collection process is automatically initiated when it is overdue to reduce risks. The collection effect is evaluated in real time, and the collection strategy is adjusted in time to improve risk control efficiency. Third-party credit data is integrated to supplement the lack of internal data and improve the comprehensiveness of risk identification. Combined with macroeconomic data, the impact of the overall economic environment on loan risks is evaluated. Risk reports are generated regularly, the current risk situation is analyzed, and decision support is provided to the management. Based on the data in the risk report, the decision model is used to assist in the formulation of risk control strategies;

[0004] The existing financial post-loan risk warning methods cannot judge whether the borrower has evaded or delayed the institution's post-loan risk assessment based on the borrower's latest bank flow data submitted by the borrower, and cannot judge whether the borrower's repayment funds are stable based on the income source of the borrower's repayment account. When the borrower fails to repay on time, it is impossible to judge the borrower's repayment willingness based on the borrower's response to the financial institution's loan collection calls. This can easily lead to the inability to provide timely risk warnings for borrowers who are unable to repay their loans, making it impossible for financial institutions to promptly determine the whereabouts of borrowers with high repayment risks, resulting in property losses that are difficult to recover when the borrower suddenly loses contact. Its practicality has certain limitations. Summary of the invention

[0005] The present invention provides a financial post-loan risk early warning method for promoting the solution of the problems raised in the background technology.

[0006] The present invention provides the following technical solution: a financial post-loan risk early warning method, comprising:

[0007] Obtain the loan data of the target person;

[0008] The loan data includes loan duration and loan risk data;

[0009] Based on the loan data of the target person, preliminary assessment data is formed through preliminary assessment strategies;

[0010] Get the target person's loan time;

[0011] Calculate the final time, final time = loan time + loan duration;

[0012] Determine the duration of the assessment based on the preliminary assessment data;

[0013] The target person's borrowing time is taken as the starting time, and the final time is taken as the ending time to form a collection period;

[0014] Taking the target person's loan time as the starting time, each interval evaluation duration forms a collection time within the collection period;

[0015] When the collection time arrives, obtain the target person's account data;

[0016] Based on the target person’s account data, risk assessment data is generated through risk assessment strategies;

[0017] Based on the preliminary assessment data and risk assessment data, the grade adjustment data is formed through the risk adjustment strategy;

[0018] Adjust the data according to the level and form risk warning data through risk warning strategy;

[0019] Issue corresponding risk warnings based on risk warning data.

[0020] As an optional solution of the financial post-loan risk early warning method of the present invention, the preliminary assessment strategy is specifically:

[0021] Obtain the target person’s lending risk data;

[0022] Setting a first risk value and a second risk value;

[0023] If the loan risk data is ≤ the first risk value, the target person's preliminary assessment risk data is determined to be the first risk level;

[0024] If the loan risk data is greater than the first risk value, and the loan risk data is less than or equal to the second risk value, the target person's preliminary assessment risk data is determined to be the second risk level;

[0025] If the loan risk data is greater than the second risk value, the preliminary assessment risk data of the target person is determined to be the third risk level.

[0026] As an optional solution of the financial post-loan risk early warning method of the present invention, wherein: the evaluation time is determined based on the preliminary evaluation data, specifically:

[0027] Obtain preliminary assessment data of target personnel;

[0028] Setting the evaluation duration, wherein the evaluation duration includes a first evaluation duration, a second evaluation duration, and a third evaluation duration;

[0029] If the target person's preliminary assessment data is the first risk level, the assessment duration is determined to be the first assessment duration;

[0030] If the target person's preliminary assessment data is the second risk level, the assessment duration is determined to be the second assessment duration;

[0031] If the preliminary assessment data of the target person is the third risk level, the assessment duration is determined to be the third assessment duration.

[0032] As an optional solution of the financial post-loan risk early warning method of the present invention, the risk assessment strategy includes a data stability analysis strategy, specifically:

[0033] Acquiring loan data of a target person, wherein the loan data includes loan flow data;

[0034] Obtaining the target person's account data, wherein the account data includes account transaction data;

[0035] If the account flow data ≠ the loan flow data, it is determined that the target person has not yet committed any risky behavior;

[0036] If the account flow data = loan flow data, a data submission notification is sent to the target person;

[0037] Set the judgment time;

[0038] Get the current time;

[0039] Calculate the judgment time, judgment time = current time + judgment duration;

[0040] The current time is taken as the start time and the determination time is taken as the end time to form a determination period;

[0041] Obtaining the target person's account data within the determination period, the account data including submission transaction data;

[0042] If the target person does not have any submitted transaction data, it is determined that the target person has submitted risky behavior;

[0043] If the target person has submitted transaction data, and the submitted transaction data ≠ loan transaction data, it is determined that the target person has not submitted any risky behavior;

[0044] If the target person has submitted transaction data, but the submitted transaction data = loan transaction data, it is determined that the target person has submitted risky behavior.

[0045] As an optional solution of the financial post-loan risk early warning method of the present invention, the risk assessment strategy also includes a flow analysis strategy, specifically:

[0046] Acquiring account data of the target person, wherein the account data includes a loan repayment account;

[0047] Get the current time;

[0048] Set the duration of flow analysis;

[0049] Calculate the flow analysis time, flow analysis time = current time - flow analysis duration;

[0050] The flow analysis time is taken as the start time and the current time is taken as the end time to form a flow segmentation period;

[0051] Obtain all income funds of the loan repayment account during the flow period;

[0052] The transfer time of each income fund withdrawal;

[0053] Calculate the transfer-in duration, transfer-in duration = current time - transfer-in time;

[0054] All the income funds are collected according to the transfer time, from large to small, to form a transfer order set;

[0055] Taking the first element in the transfer-in sequence set as the starting element, extract two adjacent elements in the transfer-in sequence set in sequence and define them as analysis elements;

[0056] According to the order of the two analysis elements in the transfer sequence set, the analysis element in the previous order is defined as the previous analysis element, and the analysis element in the next order is defined as the next analysis element;

[0057] Calculate element difference, element difference = |later analysis element - former analysis element|;

[0058] Calculate the average element, average element = (pre-analysis element + post-analysis element) ÷ 2;

[0059] If the element difference is ≤ average element × 60%, the element difference is judged to be small;

[0060] If the element difference is greater than the average element × 60%, the element difference is judged to be large;

[0061] The two analysis elements corresponding to the smaller difference of the judging element are identified as stable elements;

[0062] Extract the number of stable elements and set it as the stable number;

[0063] Get the number of elements in the transfer order set and set it as the transfer quantity;

[0064] If the stable amount is less than the transferred amount × 80%, the target person is judged to have repayment risk behavior;

[0065] If the stable amount ≥ the transferred amount × 80%, it is determined that the target person has no repayment risk behavior for the time being.

[0066] As an optional solution of the financial post-loan risk early warning method of the present invention, the risk assessment strategy also includes a personnel analysis strategy, specifically:

[0067] Obtain the amount of money to be repaid by the target person;

[0068] Get the amount of money repaid by the target person;

[0069] If the amount repaid = the amount to be repaid, the target person is determined to be actively repaying;

[0070] If the amount repaid is less than the amount to be repaid, a repayment notice will be sent to the target person;

[0071] If the target person accepts the repayment notice, the current time is obtained;

[0072] Set the waiting time;

[0073] Calculate the waiting time, waiting time = current time + waiting time;

[0074] The current time is taken as the start time and the waiting time is taken as the end time to form a waiting period;

[0075] If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan;

[0076] If the target person does not repay the loan during the waiting period, it is determined that the target person is not active in repaying the loan;

[0077] If the target person does not accept the repayment notice, the habit analysis strategy is implemented.

[0078] As an optional solution of the financial post-loan risk early warning method of the present invention, the habit analysis strategy is specifically:

[0079] Obtain the target person’s communication records;

[0080] Extract all receiving records in the communication records, wherein the receiving records include received records and missed records;

[0081] Extract all missed records in the communication records;

[0082] Obtaining call information of each missed call record respectively, wherein the call information includes the call name and call duration;

[0083] The call name includes a note name and an unfamiliar name;

[0084] Extract all missed records with unfamiliar names and define them as judged missed records;

[0085] Acquire the historical record data of each missed call record, and define it as a historical missed call record, wherein the historical record data includes answering data and rejecting data;

[0086] The number of answering data of each missed call record is obtained respectively, and is defined as the number of historical answering calls;

[0087] The number of rejection data of each record determined to be unreachable is obtained respectively, and is defined as the number of historical rejections;

[0088] The missed call records with the historical rejection number ≥ (historical answering number + historical rejection number) × 80% are identified as missed call analysis records;

[0089] Get the number of missed analysis records and set it as the number of missed calls;

[0090] Extract all received records in the communication records and execute the received information analysis strategy;

[0091] Get strange quantity;

[0092] If the number of strangers ≥ (number of strangers + number of missed calls) × 80%, it is determined that the target person is not actively repaying the loan;

[0093] If the number of strangers is less than (number of strangers + number of missed calls) × 80%, a repayment notice will be sent to the target person again;

[0094] Get the current time and waiting period;

[0095] If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan;

[0096] If the target person does not repay the loan during the waiting period, it is determined that the target person is not actively repaying the loan.

[0097] As an optional solution of the financial post-loan risk early warning method of the present invention, the received information analysis strategy is specifically:

[0098] Respectively obtain the historical record data of each received record and define it as the historical received record;

[0099] Get the number of historical received records for each missed record respectively, and define it as the number of historical received records;

[0100] Extract all received records with a missed call count of 1 and define them as received analysis records;

[0101] Get call information for each received analysis record;

[0102] Extract all received analysis records with unfamiliar names and define them as judgment records;

[0103] Obtain the call duration of each determination record and set it as the determination call duration;

[0104] Set the analysis duration;

[0105] The records where the call duration is less than the analysis duration are defined as unfamiliar records;

[0106] Get the number of unfamiliar records and set it as the unfamiliar number.

[0107] As an optional solution of the financial post-loan risk early warning method of the present invention, the risk adjustment strategy is specifically:

[0108] Set adjustment points;

[0109] If the target person is determined to have submitted risky behavior, the first score is calculated, the first score = (adjusted score × 30%) × 80%;

[0110] If it is determined that the target person has not submitted any risky behavior, the first score is calculated, and the first score = (adjusted score × 30%) × 20%;

[0111] If the target person is determined to have repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 80%;

[0112] If it is determined that the target person has no repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 20%;

[0113] If it is determined that the target person is not active in repaying the loan, the third score is calculated, the third score = (adjusted score × 45%) × 80%;

[0114] If the target person is determined to be actively repaying the loan, the third score is calculated, which is (adjusted score × 45%) × 20%;

[0115] Obtain the target person’s lending risk data;

[0116] Calculate the adjusted risk data, adjusted risk data = (loan risk data × 40%) + [(first score + second score + third score) × 60%);

[0117] Obtaining a first risk value and a second risk value;

[0118] If the adjusted risk data is less than or equal to the first risk value, the target person's level adjustment data is determined to be the first risk level;

[0119] If the adjusted risk data is greater than the first risk value, and the adjusted risk data is less than or equal to the second risk value, the target person's level adjustment data is determined to be the second risk level;

[0120] If the adjusted risk data>the second risk value, the level adjustment data of the target person is determined to be the third risk level.

[0121] As an optional solution of the financial post-loan risk early warning method of the present invention, the risk early warning strategy is specifically:

[0122] Acquire level adjustment data of the target person, wherein the level adjustment data includes a first risk level, a second risk level, and a third risk level;

[0123] If the target person's level adjustment data is the first risk level, monitoring and risk warning will not be performed;

[0124] If the target person's level adjustment data is the second risk level, the target person will be monitored and warned;

[0125] If the target person's level adjustment data is the third risk level, a risk warning will be issued to the target person.

[0126] The present invention has the following beneficial effects:

[0127] 1. The financial post-loan risk warning method obtains the lender's preliminary risk assessment data, determines the data submission time of the lender during the post-loan assessment, and judges whether the lender has evaded or delayed the institution's post-loan risk assessment through the latest bank flow data submitted by the lender, determines the lender's enthusiasm for post-loan risk assessment, and promptly issues risk warnings to the lender, so that financial institutions can make more accurate assessments, reduce errors in post-loan risk assessment data, and thus reduce financial institutions' property losses.

[0128] 2. This financial post-loan risk warning method obtains the funds transferred into the borrower's loan repayment account, determines whether the borrower's income source at the time of repayment is fixed funds such as wages, and thus determines whether the borrower's repayment source is stable, and promptly issues risk warnings to the borrower, reducing the situation where the borrower fails to repay on time due to unstable repayment sources, thereby reducing the property losses of financial institutions.

[0129] 3. The financial post-loan risk warning method, when the borrower fails to repay on time, promptly calls the borrower for repayment, obtains the borrower's answering of the collection call, determines whether the borrower deliberately defaults on repayment when knowing that there is a repayment matter, and determines when the borrower does not answer the collection call, whether the borrower deliberately refuses to answer the collection call to default on repayment or misses the collection call because the borrower is not used to answering unfamiliar calls, thereby determining the borrower's repayment willingness and promptly issuing risk warnings to borrowers who are unable to repay the loan, so that financial institutions can promptly determine the whereabouts of borrowers with high repayment risks, and avoid property losses that are difficult to recover when the borrower suddenly loses contact. BRIEF DESCRIPTION OF THE DRAWINGS

[0130] Figure 1 This is a flow chart of the financial post-loan risk early warning method of the present invention. DETAILED DESCRIPTION

[0131] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0132] Embodiment 1, a financial post-loan risk early warning method, see Figure 1 ,include:

[0133] Obtain the loan data of the target person;

[0134] The loan data includes loan duration and loan risk data. The loan duration is the longest repayment time promised by the target person when signing the loan contract with the financial institution. For example, if the target person signs the contract on January 5 and the longest repayment time promised by the target person when signing the loan contract is 6 months, the loan duration is 6 months. The loan risk data is the repayment risk value assessed by the financial institution when the target person signs the loan contract with the financial institution.

[0135] Based on the loan data of the target person, preliminary assessment data is formed through preliminary assessment strategies;

[0136] Obtain the target person's loan time, where the loan time is the time when the target person signs the loan contract with the financial institution. For example, if the target person signs the contract on January 5, the loan time is January 5;

[0137] Calculate the final time, final time = loan time + loan duration;

[0138] Determine the duration of the assessment based on the preliminary assessment data;

[0139] The target person's borrowing time is taken as the starting time, and the final time is taken as the ending time to form a collection period;

[0140] Taking the target person's loan time as the starting time, each interval evaluation duration forms a collection time within the collection period;

[0141] When the collection time arrives, obtain the target person's account data;

[0142] Based on the target person’s account data, risk assessment data is generated through risk assessment strategies;

[0143] Based on the preliminary assessment data and risk assessment data, the grade adjustment data is formed through the risk adjustment strategy;

[0144] Adjust the data according to the level and form risk warning data through risk warning strategy;

[0145] Issue corresponding risk warnings based on risk warning data.

[0146] Through the above method, the stability of the borrower's repayment funds can be judged based on the income source of the borrower's repayment account. When the borrower fails to repay on time, the borrower's repayment willingness can be judged based on the borrower's answering of the financial institution's loan collection calls, and risk warnings can be issued in time to the borrower who is unable to repay the loan, so that financial institutions can make more accurate assessments, reduce errors in post-loan risk assessment data, and reduce the situation where borrowers fail to repay on time due to unstable repayment sources. Financial institutions can promptly determine the whereabouts of borrowers with high repayment risks to avoid property losses that are difficult to recover when the borrower suddenly loses contact, thereby reducing property losses of financial institutions.

[0147] Embodiment 2: This embodiment is an improvement made on the basis of embodiment 1. The financial post-loan risk early warning method, the preliminary assessment strategy, is specifically:

[0148] Obtain the target person’s lending risk data;

[0149] Set a first risk value and a second risk value, wherein the repayment risk value is 100, the smaller the repayment risk value, the smaller the risk of not being able to repay in time, and the larger the repayment risk value, the greater the risk of not being able to repay in time, the first risk value is 30, and the second risk value is 70;

[0150] If the loan risk data is less than or equal to the first risk value, the target person's preliminary risk assessment data is determined to be the first risk level, which is a relatively low risk level;

[0151] If the loan risk data is greater than the first risk value, and the loan risk data is less than or equal to the second risk value, the target person's preliminary assessment risk data is determined to be the second risk level, which is moderate risk;

[0152] If the loan risk data is greater than the second risk value, the preliminary assessment risk data of the target person is determined to be the third risk level, which is a relatively high risk.

[0153] This embodiment also provides that the determination of the evaluation duration according to the preliminary evaluation data is specifically as follows:

[0154] Obtain preliminary assessment data of target personnel;

[0155] Set the evaluation duration, which includes the first evaluation duration, the second evaluation duration and the third evaluation duration. The first evaluation duration is 45 days, the second evaluation duration is 30 days, and the third evaluation duration is 15 days.

[0156] If the target person's preliminary assessment data is the first risk level, the assessment duration is determined to be the first assessment duration;

[0157] If the target person's preliminary assessment data is the second risk level, the assessment duration is determined to be the second assessment duration;

[0158] If the preliminary assessment data of the target person is the third risk level, the assessment duration is determined to be the third assessment duration.

[0159] Embodiment 3: This embodiment is an improvement made on the basis of Embodiment 2. In this embodiment, the risk assessment strategy includes a data stability analysis strategy, which is specifically:

[0160] Obtaining the target person's loan data, wherein the loan data includes loan flow data, which is the bank statement provided by the borrower to the institution when applying for a loan;

[0161] Obtaining the target person's account data, the account data including account flow data, the account flow data being the bank statement provided by the borrower to the institution at the time of collection;

[0162] If the account flow data ≠ the loan flow data, it is determined that the target person has not submitted any risk behavior, that is, the target person has actively submitted his or her latest bank statement within the specified data submission time for the institution to conduct post-loan risk assessment. Based on his or her active data submission behavior, it is determined that his or her risk is low;

[0163] If the account flow data = loan flow data, a data submission notification is sent to the target person, that is, the borrower is notified to submit his latest bank statement as soon as possible for the institution to conduct post-loan risk assessment;

[0164] Set the judgment time, which is 3 days;

[0165] Get the current time;

[0166] Calculate the judgment time, judgment time = current time + judgment duration;

[0167] The current time is taken as the start time and the determination time is taken as the end time to form a determination period;

[0168] Acquire the target person's account data within the determination period, wherein the account data includes submitted transaction data, and the submitted transaction data is the bank statement submitted by the borrower within the determination period;

[0169] If the target person does not submit bank statement data, it is judged that the target person has submitted risky behavior, that is, the borrower has not submitted his latest bank statement within the prescribed data submission time, which may be to evade or delay the institution's post-loan risk assessment. Based on his failure to submit data, he is judged to have a high risk;

[0170] If the target person has submitted bank statement data, and the submitted bank statement data ≠ the loan statement data, then it is determined that the target person has not submitted risky behavior, that is, the borrower may have forgotten to submit data, resulting in not submitting his latest bank statement within the specified data submission time. Based on his subsequent voluntary data submission behavior, it is determined that his risk is low;

[0171] If the target person has submitted bank statement data, but the submitted bank statement data = loan statement data, then the target person is judged to have submitted risky behavior, that is, the bank statement submitted by the borrower within the specified data submission time is the bank statement submitted when he applied for a loan, which may be to evade or delay the institution's post-loan risk assessment. Based on his behavior of not submitting new data, he is judged to have a higher risk.

[0172] This embodiment also provides that the risk assessment strategy also includes a pipeline analysis strategy, specifically:

[0173] Obtaining the target person's account data, including the loan repayment account, which is the account from which the target person transfers funds when the target person repays the loan to a financial institution, and the bank statement data of the account is the bank statement submitted by the borrower;

[0174] Get the current time;

[0175] Set the duration of the water flow analysis, which is 3 months;

[0176] Calculate the flow analysis time, flow analysis time = current time - flow analysis duration;

[0177] The flow analysis time is taken as the start time and the current time is taken as the end time to form a flow segmentation period;

[0178] Obtain all income funds of the loan repayment account during the flow period;

[0179] Extract the transfer time of each income fund, which is the time when the loan repayment account receives the income fund;

[0180] Calculate the transfer-in duration, transfer-in duration = current time - transfer-in time;

[0181] All the income funds are collected according to the transfer time, from large to small, to form a transfer order set;

[0182] Taking the first element in the transfer-in sequence set as the starting element, extract two adjacent elements in the transfer-in sequence set in sequence and define them as analysis elements;

[0183] According to the order of the two analysis elements in the transfer sequence set, the analysis element in the previous order is defined as the previous analysis element, and the analysis element in the next order is defined as the next analysis element;

[0184] Calculate element difference, element difference = |later analysis element - former analysis element|;

[0185] Calculate the average element, average element = (pre-analysis element + post-analysis element) ÷ 2;

[0186] If the element difference is ≤ average element × 60%, the element difference is judged to be small;

[0187] If the element difference is greater than the average element × 60%, the element difference is judged to be large;

[0188] The two analysis elements corresponding to the smaller difference of the judging element are identified as stable elements;

[0189] Extract the number of stable elements and set it as the stable number;

[0190] Get the number of elements in the transfer order set and set it as the transfer quantity;

[0191] If the stable amount is less than the transferred amount × 80%, the target person is judged to have repayment risk behavior;

[0192] If the stable amount ≥ the transferred amount × 80%, it is determined that the target person has no repayment risk behavior for the time being.

[0193] Embodiment 4: This embodiment is an improvement made on the basis of embodiment 3. In this embodiment, the risk assessment strategy also includes a personnel analysis strategy, which is specifically:

[0194] Obtain the amount of money to be repaid by the target person;

[0195] Get the amount of money repaid by the target person;

[0196] If the amount repaid = the amount to be repaid, the target person is determined to be actively repaying;

[0197] If the amount repaid is less than the amount to be repaid, a repayment notice will be sent to the target person, where the repayment notice will be sent by phone;

[0198] If the target person accepts the repayment notification, the current time is obtained, that is, the target person answers the repayment notification call;

[0199] Set the waiting time, which is 7 days;

[0200] Calculate the waiting time, waiting time = current time + waiting time;

[0201] The current time is taken as the start time and the waiting time is taken as the end time to form a waiting period;

[0202] If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan, that is, the target person may have forgotten to repay the loan, not because he is unwilling to repay the loan;

[0203] If the target person does not repay the loan during the waiting period, it is determined that the target person is not active in repaying the loan, that is, the target person knows that he / she is in arrears but still delays and does not repay the loan;

[0204] If the target person does not accept the repayment notice, the habit analysis strategy is executed, that is, the target person does not answer the repayment notice call.

[0205] This embodiment also provides that the habit analysis strategy is specifically:

[0206] Obtain the target person’s communication records;

[0207] Extract all receiving records in the communication records, the receiving records include received records and missed records, the receiving records are records of other persons making calls to the target person, including all call records that the target person has answered and missed;

[0208] Extract all missed records in the communication records;

[0209] Obtaining call information of each missed call record respectively, wherein the call information includes the call name and call duration;

[0210] The call name includes a remark name and an unfamiliar name. The remark name is the name of the person who marked the phone number by the target person, and the unfamiliar name is the phone number, indicating that the call was made by a person who was not marked by the target person;

[0211] Extract all missed records with unfamiliar names and define them as judged missed records;

[0212] Acquire the historical record data of each missed call record, and define it as a historical missed call record, wherein the historical record data includes answering data and rejecting data;

[0213] The number of answering data of each missed call record is obtained respectively, and is defined as the number of historical answering calls;

[0214] The number of rejection data of each record determined to be unreachable is obtained respectively, and is defined as the number of historical rejections;

[0215] The missed call records with the number of historical rejections ≥ (the number of historical answers + the number of historical rejections) × 80% are identified as missed call analysis records, that is, missed call analysis records are the call records made by strangers who are completely unknown to the target person, such as advertisements, frauds, etc.;

[0216] Get the number of missed analysis records and set it as the number of missed calls;

[0217] Extract all received records in the communication records and execute the received information analysis strategy;

[0218] Get strange quantity;

[0219] If the number of unfamiliar calls ≥ (number of unfamiliar calls + number of missed calls) × 80%, it is determined that the target person is not active in repaying the loan at present, that is, the target person will answer unfamiliar calls but refuse to answer collection calls from financial institutions to delay repayment;

[0220] If the number of strangers is less than (number of strangers + number of missed calls) × 80%, a repayment notice will be sent to the target person again. This repayment notice will be a reminder notice executed simultaneously through multiple channels such as SMS notification and WeChat notification;

[0221] Get the current time and waiting period;

[0222] If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan, that is, the target person may have forgotten to repay the loan and has a habit of rarely answering unfamiliar phone calls, which leads to the failure to repay the loan on time, not because of unwillingness to repay the loan;

[0223] If the target person does not repay the loan within the waiting period, it is determined that the target person is not actively repaying the loan, that is, it is assumed that the target person knows that he / she has arrears but still delays and does not repay the loan.

[0224] The received information analysis strategy is specifically as follows:

[0225] Respectively obtain the historical record data of each received record and define it as the historical received record;

[0226] Get the number of historical received records for each missed record respectively, and define it as the number of historical received records;

[0227] Extract all received records with a missed call count of 1 and define them as received analysis records;

[0228] Get call information for each received analysis record;

[0229] Extract all received analysis records with unfamiliar names and define them as judgment records;

[0230] Obtain the call duration of each determination record and set it as the determination call duration;

[0231] Set the analysis time, which is 30 seconds;

[0232] The records with the judgment call duration less than the analysis duration are defined as strange records, that is, strange records are the records of calls made by strangers who are completely unknown to the target person, such as advertisements, frauds, etc.

[0233] Get the number of unfamiliar records and set it as the unfamiliar number.

[0234] Embodiment 5: This embodiment is an improvement made on the basis of Embodiment 4. In this embodiment, the risk adjustment strategy is specifically as follows:

[0235] Setting an adjustment score, wherein the adjustment score is 100;

[0236] If the target person is determined to have submitted risky behavior, the first score is calculated, the first score = (adjusted score × 30%) × 80%;

[0237] If it is determined that the target person has not submitted any risky behavior, the first score is calculated, and the first score = (adjusted score × 30%) × 20%;

[0238] If the target person is determined to have repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 80%;

[0239] If it is determined that the target person has no repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 20%;

[0240] If it is determined that the target person is not active in repaying the loan, the third score is calculated, the third score = (adjusted score × 45%) × 80%;

[0241] If the target person is determined to be actively repaying the loan, the third score is calculated, which is (adjusted score × 45%) × 20%;

[0242] Obtain the target person’s lending risk data;

[0243] Calculate the adjusted risk data, adjusted risk data = (loan risk data × 40%) + [(first score + second score + third score) × 60%);

[0244] Obtaining a first risk value and a second risk value;

[0245] If the adjusted risk data is less than or equal to the first risk value, the target person's level adjustment data is determined to be the first risk level, which is a relatively low risk level;

[0246] If the adjusted risk data is greater than the first risk value, and the adjusted risk data is less than or equal to the second risk value, the target person's level adjustment data is determined to be the second risk level, which is a moderate risk level;

[0247] If the adjusted risk data>the second risk value, the target person's level adjustment data is determined to be the third risk level, and the third risk level is a relatively high risk.

[0248] This embodiment also provides that the risk warning strategy is specifically:

[0249] Acquire level adjustment data of the target person, wherein the level adjustment data includes a first risk level, a second risk level, and a third risk level;

[0250] If the target person's level adjustment data is the first risk level, monitoring and risk warning will not be performed;

[0251] If the target person's level adjustment data is the second risk level, the target person will be monitored and warned, such as continuously monitoring the target person's fund transactions to prevent the target person from suddenly defaulting on the loan;

[0252] If the target person's level adjustment data is the third risk level, a risk warning will be issued to the target person, such as adjusting the target person's repayment period and repayment interest rate, and regularly determining the target person's residence information and the information of the people he or she frequently contacts, to prevent the target person from running away or losing contact.

[0253] In this embodiment, whether the borrower's repayment funds are stable is judged based on the income source of the borrower's repayment account. When the borrower fails to repay on time, the borrower's repayment willingness is judged based on the borrower's answering of the financial institution's loan collection calls, and risk warnings are issued in time to borrowers who are unable to repay the loan, thereby enabling financial institutions to make more accurate assessments, reduce errors in post-loan risk assessment data, and reduce the situation where borrowers fail to repay on time due to unstable repayment sources. Financial institutions can promptly determine the whereabouts of borrowers with high repayment risks to avoid property losses that are difficult to recover when borrowers suddenly lose contact, thereby reducing property losses of financial institutions.

Claims

1. A financial post-loan risk early warning method, characterized by: include: Obtain the target person’s loan data; The loan data includes loan duration and loan risk data; Based on the loan data of the target person, preliminary assessment data is formed through preliminary assessment strategies; Get the target person's loan time; Calculate the final time, final time = loan time + loan duration; Determine the duration of the assessment based on the preliminary assessment data; The target person's borrowing time is taken as the starting time, and the final time is taken as the ending time to form a collection period; Taking the target person's loan time as the starting time, each interval evaluation duration forms a collection time within the collection period; When the collection time arrives, obtain the target person's account data; Based on the target person’s account data, risk assessment data is generated through risk assessment strategies; Based on the preliminary assessment data and risk assessment data, the grade adjustment data is formed through risk adjustment strategy; Adjust the data according to the level and form risk warning data through risk warning strategy; Issue corresponding risk warnings based on risk warning data.

2. The financial post-loan risk early warning method according to claim 1, characterized in that: The preliminary assessment strategy is specifically: Obtain the target person’s lending risk data; Setting a first risk value and a second risk value; If the loan risk data is ≤ the first risk value, the target person's preliminary assessment risk data is determined to be the first risk level; If the loan risk data > the first risk value, and the loan risk data ≤ the second risk value, the preliminary assessment risk data of the target person is determined to be the second risk level; If the loan risk data is greater than the second risk value, the preliminary assessment risk data of the target person is determined to be the third risk level.

3. The financial post-loan risk early warning method according to claim 1 is characterized by: The evaluation duration is determined based on the preliminary evaluation data, specifically: Obtain preliminary assessment data of target personnel; Setting the evaluation duration, wherein the evaluation duration includes a first evaluation duration, a second evaluation duration, and a third evaluation duration; If the preliminary assessment data of the target person is the first risk level, the assessment duration is determined to be the first assessment duration; If the target person's preliminary assessment data is the second risk level, the assessment duration is determined to be the second assessment duration; If the preliminary assessment data of the target person is the third risk level, the assessment duration is determined to be the third assessment duration.

4. The financial post-loan risk early warning method according to claim 1, characterized in that: The risk assessment strategy, including the data stability analysis strategy, is specifically: Acquiring loan data of a target person, wherein the loan data includes loan flow data; Obtaining the target person's account data, wherein the account data includes account transaction data; If the account flow data ≠ the loan flow data, it is determined that the target person has not yet committed any risky behavior; If the account flow data = loan flow data, a data submission notification is sent to the target person; Set the judgment time; Get the current time; Calculate the judgment time, judgment time = current time + judgment duration; The current time is taken as the start time and the determination time is taken as the end time to form a determination period; Obtaining the target person's account data within the determination period, the account data including submission transaction data; If the target person does not have any submitted transaction data, it is determined that the target person has submitted risky behavior; If the target person has submitted transaction data, and the submitted transaction data ≠ loan transaction data, it is determined that the target person has not submitted any risky behavior; If the target person has submitted transaction data, but the submitted transaction data = loan transaction data, it is determined that the target person has submitted risky behavior.

5. The financial post-loan risk early warning method according to claim 4 is characterized in that: The risk assessment strategy also includes a flow analysis strategy, specifically: Acquiring account data of the target person, wherein the account data includes a loan repayment account; Get the current time; Set the duration of flow analysis; Calculate the flow analysis time, flow analysis time = current time - flow analysis duration; The flow analysis time is taken as the start time and the current time is taken as the end time to form a flow segmentation period; Obtain all income funds of the loan repayment account during the flow period; The transfer time of each income fund withdrawal; Calculate the transfer-in duration, transfer-in duration = current time - transfer-in time; All the income funds are collected according to the transfer time, from large to small, to form a transfer order set; Taking the first element in the transfer-in sequence set as the starting element, extract two adjacent elements in the transfer-in sequence set in sequence and define them as analysis elements; According to the order of the two analysis elements in the transfer sequence set, the analysis element in the previous order is defined as the previous analysis element, and the analysis element in the next order is defined as the next analysis element; Calculate element difference, element difference = |later analysis element - former analysis element|; Calculate the average element, average element = (pre-analysis element + post-analysis element) ÷ 2; If the element difference is ≤ average element × 60%, the element difference is judged to be small; If the element difference is greater than the average element × 60%, the element difference is judged to be large; The two analysis elements corresponding to the smaller difference of the judging element are identified as stable elements; Extract the number of stable elements and set it as the stable number; Get the number of elements in the transfer order set and set it as the transfer quantity; If the stable amount is less than the transferred amount × 80%, the target person is judged to have repayment risk behavior; If the stable amount ≥ the transferred amount × 80%, it is determined that the target person has no repayment risk behavior for the time being.

6. The financial post-loan risk early warning method according to claim 4 is characterized by: The risk assessment strategy also includes a personnel analysis strategy, specifically: Obtain the amount of money to be repaid by the target person; Get the amount of money that the target person has repaid; If the amount repaid = the amount to be repaid, the target person is determined to be actively repaying; If the amount repaid is less than the amount to be repaid, a repayment notice will be sent to the target person; If the target person accepts the repayment notice, the current time is obtained; Set the waiting time; Calculate the waiting time, waiting time = current time + waiting time; The current time is taken as the start time and the waiting time is taken as the end time to form a waiting period; If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan; If the target person does not repay the loan during the waiting period, it is determined that the target person is not active in repaying the loan; If the target person does not accept the repayment notice, the habit analysis strategy is implemented.

7. The financial post-loan risk early warning method according to claim 6 is characterized by: The habit analysis strategy is specifically as follows: Obtain the target person’s communication records; Extract all receiving records in the communication records, wherein the receiving records include received records and missed records; Extract all missed records in the communication records; Obtaining call information of each missed call record respectively, wherein the call information includes the call name and call duration; The call name includes a note name and an unfamiliar name; Extract all missed records with unfamiliar names and define them as judged missed records; Acquire the historical record data of each missed call record, and define it as a historical missed call record, wherein the historical record data includes answering data and rejecting data; The number of answering data of each missed call record is obtained respectively, and is defined as the number of historical answering calls; The number of rejection data of each record determined to be unreachable is obtained respectively, and is defined as the number of historical rejections; The missed call records with the historical rejection number ≥ (historical answering number + historical rejection number) × 80% are identified as missed call analysis records; Get the number of missed analysis records and set it as the number of missed calls; Extract all received records in the communication records and execute the received information analysis strategy; Get strange quantity; If the number of strangers ≥ (number of strangers + number of missed calls) × 80%, it is determined that the target person is not actively repaying the loan; If the number of strangers is less than (number of strangers + number of missed calls) × 80%, a repayment notice will be sent to the target person again; Get the current time and waiting period; If the target person repays the loan during the waiting period, it is determined that the target person is currently actively repaying the loan; If the target person does not repay the loan during the waiting period, it is determined that the target person is not actively repaying the loan.

8. The financial post-loan risk early warning method according to claim 7 is characterized by: The received information analysis strategy is specifically: Respectively obtain the historical record data of each received record and define it as the historical received record; Get the number of historical received records for each missed record respectively, and define it as the number of historical received records; Extract all received records with a missed call count of 1 and define them as received analysis records; Get call information for each received analysis record; Extract all received analysis records with unfamiliar names and define them as judgment records; Obtain the call duration of each determination record and set it as the determination call duration; Set the analysis duration; The records where the call duration is less than the analysis duration are defined as unfamiliar records; Get the number of unfamiliar records and set it as the unfamiliar number.

9. The financial post-loan risk early warning method according to claim 1, characterized in that: The risk adjustment strategy is specifically: Set adjustment points; If the target person is determined to have submitted risky behavior, the first score is calculated, the first score = (adjusted score × 30%) × 80%; If it is determined that the target person has not submitted any risky behavior, the first score is calculated, and the first score = (adjusted score × 30%) × 20%; If the target person is determined to have repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 80%; If it is determined that the target person has no repayment risk behavior, the second score is calculated, the second score = (adjustment score × 25%) × 20%; If it is determined that the target person is not active in repaying the loan, the third score is calculated, the third score = (adjusted score × 45%) × 80%; If the target person is determined to be actively repaying the loan, the third score is calculated, which is (adjusted score × 45%) × 20%; Obtain the target person’s lending risk data; Calculate the adjusted risk data, adjusted risk data = (loan risk data × 40%) + [(first score + second score + third score) × 60%); Obtaining a first risk value and a second risk value; If the adjusted risk data is less than or equal to the first risk value, the target person's level adjustment data is determined to be the first risk level; If the adjusted risk data is greater than the first risk value, and the adjusted risk data is less than or equal to the second risk value, the target person's level adjustment data is determined to be the second risk level; If the adjusted risk data>the second risk value, the level adjustment data of the target person is determined to be the third risk level.

10. The financial post-loan risk early warning method according to claim 1, characterized in that: The risk warning strategy is specifically as follows: Acquire level adjustment data of the target person, wherein the level adjustment data includes a first risk level, a second risk level, and a third risk level; If the target person's level adjustment data is the first risk level, monitoring and risk warning will not be performed; If the target person's level adjustment data is the second risk level, the target person will be monitored and warned; If the target person's level adjustment data is the third risk level, a risk warning will be issued to the target person.

Images