Credit Financial Products Loan Process and Related Products Introduction

Products in the field of financial credit, except wealth management, are mostly P2P credit, small loans, consumer finance, auto finance, bank credit and so on. Before the loan application begins, it is generally inevitable to acquire customers through marketing, registration, real-name authentication and bank card binding.

Having finished the basic credit process, here are the details of each process and the corresponding risk control products.

One is accurate marketing by identifying users' willingness to borrow, and the other is anti-fraud to prevent bonus hunter, advertising fraud and other fraudulent acts.

Corresponding products:

1. Identification of cheating equipment: Through equipment fingerprint technology, various attributes of running equipment can be obtained, so as to analyze the frequency and associated accounts of equipment participating in marketing activities and effectively identify cheating equipment.

2. Proxy IP analysis: Through proxy detection technology, the behavior of hiding real IP address by proxy technology can be found in time, and whether the current user uses proxy to access the website can be accurately identified. Combined with the risk control model of the organization itself and user information, cheating users can be effectively identified.

3. Fraudulent user behavior: Through the modeling and analysis of historical fraudulent user data, accurately analyze the characteristics of user behavior and effectively identify fraudulent users.

4. Precision marketing: The key point is whether the user group and label dimensions are rich, which is conducive to subdividing industries and positioning customer groups; The way to attract customers is to reach users through the selected channel media, which generally includes short messages, phone calls, online customized advertisements, mobile advertisements and so on. However, due to policy reasons, SMS and call services have been tightened, and the cost performance of SMS and call in precision marketing has gradually weakened.

The main prevention and control point is also anti-fraud, and the possible risks at this stage are as follows:

Conduct identity verification in the same industry as lending. In the process of anti-fraud identification, risk-free users come to the identity verification link to confirm their identity through information such as ID number, mobile phone number and name.

1. ID number authentication: citizen's identity information, including ID number and name, plus stripes. This query interface generally comes from the first and third police stations and the public security authentication center. It is not easy to find real-time interfaces on the market now, and most of them are stock interfaces. The reason is that these public security departments generally do not release it to the public, and there is no formal agency policy.

2. I apply for verification: mainly through OCR+ in vivo detection. The main purpose of OCR is to obtain credible photos on the ID card. In vivo detection is to obtain your live photos, and then compare the credible photos with the live photos to determine whether the operator is me. Credible photos can be obtained not only by in vivo detection, but also by calling police network photos, but the cost is high.

3. Operator verification: three elements of operator verification, that is, by providing name, ID number and mobile phone number, verify whether the name ID number is consistent with the name ID number registered in real name when opening an account with mobile phone number, and judge whether it is the mobile phone number of the owner. This verification of mutual gold products is less used.

Bank card verification: According to the needs of different industries and product design, bank card verification is generally divided into two elements (name, bank card number and mobile phone number), four elements (name, ID number, bank card number and mobile phone number) and six elements (generally referring to credit card, name, ID number, bank card number, mobile phone number, expiration date and CVN2 code). The main source of the interface is the main branches of major banks, and UnionPay is relatively complete, but most of them are well-known third-party payment channels. When checking the bank card, you also need to verify the real-name authentication. Therefore, many mutual gold products do real-name verification directly through this link during product design, which saves the cost of name+ID verification.

1. Authorization information acquisition: for the authenticated users, obtain the necessary information with or without perception, and prepare data for subsequent model scoring. Non-perceptual access includes multi-head lending data, consumer finance portrait data, mobile phone number status and duration data. Perception (users need to provide relevant account passwords) data include: operator statements, social security provident fund, career information, academic information, central bank credit information, etc.

2. User filtering: after obtaining the information submitted by users, the information captured by the system and the third-party data, various risk control rules are run based on these data to filter out users who do not meet the rules.

Risk control rules mainly include entry rules, anti-fraud rules, strict rejection rules, variable rules, scoring rules and so on.

3. User credit evaluation: customers who pass the filtering rules enter the credit evaluation and credit granting link. This link usually has a scoring link, that is, credit score, also called personal comprehensive credit score. It refers to the use of scientific and rigorous analysis methods to comprehensively investigate the internal and external subjective and objective environment that affects individuals and their families, and comprehensively judge and evaluate their ability to fulfill various economic commitments.

4. Credit limit: After credit evaluation, risk pricing is conducted according to the credit evaluation information of users, including credit limit, rate and term.

1. Anti-fraud before lending: Anti-fraud mainly focuses on two aspects:

The whole anti-fraud process is roughly as follows:

5. Credit Scorecard: Building a credit scorecard model can accurately evaluate the repayment ability and willingness of borrowers. The credit score card includes:

Loan management refers to the loan management from the date of loan issuance to the time of loan principal and interest recovery.

Risk control in loans is mainly monitoring and risk early warning, emphasizing whether the loan users' funds are used normally, whether their qualifications have changed, whether their income is stable and other behavioral changes that will affect normal repayment in the later stage of customer service.

Monitoring in lending is the least noticeable link in lending, and there is a situation of "emphasizing lending and neglecting management". Except for a few institutions that have access to users' travel data, payment data and real-time credit information data, most credit information agencies are unable to monitor loans and can only cooperate with these few institutions, but the cost of monitoring will naturally rise.

If the risk control before lending determines whether the enterprise can survive, then the management decision in lending determines the sustainable development of the enterprise.

In-loan monitoring is mainly aimed at these two dimensions, which are generally reflected in repayment behavior, litigation information, special list change, multiple application records change, contact information change and so on. If there is any change in the general monitoring dimension, it is necessary to find it in time and carry out risk early warning.

The changes in these dimensions reflect customers' repayment ability and willingness.

1. Behavior Scoring: Using the historical data and behavior characteristics of loan customers, establish a behavior scoring model, classify customers into risk levels, and implement different risk control measures;

2. Customer number status verification: judge the status of the customer number in the loan, give the corresponding status label, and find the risk in time;

3. Abnormal behavior early warning: multi-head lending behavior monitoring, abnormal early warning of repayment ability indicators, and cross-recognition of repayment willingness.

After the loan, it includes invitation to repay the loan and collection.

Re-loan invitation is mainly to find out the customers who repay on time through historical behavior and ask whether it is necessary to re-loan.

Collection, mainly for customers with overdue signs or behaviors, for collection contact. Generally, users who have overdue behaviors are easy to lose contact, so they need to repair the lost contact. We need to repair other contact information under the user's real name through data cross-matching, and then make a collection plan, formulate an automatic collection strategy, overdue loss model, and select collection channels. In order to improve efficiency, intelligent customer service can be set up for early collection, and later for serious overdue, you can choose to outsource collection.

1. Lost connection repair: If the target user cannot be contacted when the loan needs to be collected, the lost connection repair can be carried out to gain more ways to reach users and improve the success rate of collection.

2. Portrait of overdue customers: used to clarify the situation of the collection object. Multi-dimensional portrait data accurately outlines the relevant situation of overdue customers' repayment ability and willingness, and accurately quantifies it.

3. Collect scores: Assess the difficulty of collecting objects through scores. A modeling and scoring system that integrates multi-dimensional information such as commission amount, age and region, takes into account the repayment ability and willingness of overdue customers, and comprehensively evaluates the debtor's repayment possibility. Accurately predict the repayment probability through the post-loan behavior scoring model, and stratify customers according to the repayment possibility to assist the post-loan strategy;

4. Order splitting strategy: Collectors with different abilities are responsible for collecting objects with different difficulties. According to the collection score, combined with the business ability index of the total collection staff in the system, the single-splitting strategy is formulated to rationalize the distribution of cases and improve the collection effect.

5. Circular dialing: friendly automatic dialing, asking you to answer. For high-frequency, simple collection cases, trial-touch circular dialing is adopted to reduce manual operation, shorten dialing interval and effectively improve collection efficiency.

6. Normal repayment reminder: for the repayment due soon, multiple repayment reminders will be made by SMS or telephone. For example, SMS reminder will be given 5 days, 2 days, 1 day and 0 days in advance, and telephone reminder will be given 2 days, 1 day and 0 days in advance. This is a very common but effective repayment strategy. At the same time, repayment generally supports third-party withholding. For example, the third-party payment platform authorized by the user will deduct the principal and interest from the designated bank card on the repayment date to prevent the user from forgetting to repay.

7. Credit Re-evaluation: Whether it is normal repayment, overdue repayment or overdue repayment, after the end of a period, users' credit needs to be re-evaluated and the credit limit and interest rate are given again to facilitate users to refinance or blacklist.

8. Activation of old customers: activate marketing for users with good reputation to improve the repayment rate.

Reference article:

100 Secret Series of Wind Control Company 5 | Research Report of "Network Celebrity" Tongdun Technology Company and Products

100 Secret Series of Risk Control Companies 6 | Investigation Report of Bai Rongjinfu Company and Products

Analysis of Consumer Finance Big Data Risk Control Business

The product manager of "cash loan" should also know some risk control.

PPT, Same Shield transaction anti-fraud solution