Big Data Credit Information and Bank Risk Control Innovation

Big Data Credit Information and Bank Risk Control Innovation

Data will be one of the core competitiveness of banks in the future, which has become the consensus of the banking industry. In the era of big data, the competition faced by banks comes not only from within the same industry, but also from external challenges. Emerging enterprises such as the Internet and e-commerce have obvious advantages in product innovation capability, market sensitivity and big data processing experience. In this situation, using big data to innovate and improve the risk control of banks has gradually become an important topic of concern and discussion in the industry.

The deficiency of risk control in banking industry

The Outlook of China Finance and Banking in 20 15 released by PricewaterhouseCoopers pointed out that by the end of the third quarter of 20 14, the total non-performing loans of commercial banks in China increased by 36%, reaching 767 billion yuan, a four-year high. It is predicted that the upward trend of non-performing loans will continue in 20 15 years. Behind the above data, in addition to the reasons for the rising overdue risk caused by the economic downturn, the loopholes and defects of bank risk control are also important reasons.

Information asymmetry and loan fraud

With the rise of private lending such as P2P and small loans, it is becoming easier and easier for borrowers to obtain loans through non-bank channels. However, private lending institutions do not need to report data to the People's Bank of China, and the contradiction of unclear, opaque and unpredictable information such as loan application, debt and overdue situation in non-banking system is becoming more and more prominent. It is often not until the borrower is overdue or even lost contact that the bank passively learns about some historical overdue loans or excessive debts of borrowers in the private lending field.

Loan fraud is another problem faced by banks, especially in the field of credit cards and some loan products operating in the mode of credit factory. It is no secret that the bank's solidified card issuance review process and the operation mode of the credit factory are no longer secret. At present, credit cards, loan packaging and group fraud are common occurrences, especially in the field of credit loans. About 60% of credit loans come from fraud, and more than half of them are due to identity fraud and data packaging. In the case of incomplete data dimension, banks and other lending institutions are easy to be used by group fraudsters due to the lack of support from third-party big data and the lack of sufficient and effective cross-validation means.

Information untimely and post-loan risk prevention

The untimely acquisition of information has also brought varying degrees of trouble to the post-loan risk management of banks. For example, banks often want to know for the first time whether a corporate customer will face new legal proceedings after obtaining a loan, but most banks only rely on the credit manager to manually check the local court website from time to time to obtain information, which is very uncertain. Once the credit manager forgets to inquire or make mistakes in operation, the monitoring of post-loan judicial procedures will be invalid. This does not include continuous monitoring of customers' applications, liabilities and overdue situations in private lending. The means and efficiency of banks in the process of post-loan risk prevention greatly restrict the effect of bank risk control.

Contradiction between cost and efficiency

In order to solve the problems of information asymmetry and untimely information acquisition, banks often need to collect a large amount of data to assist in judgment. However, in the process of data collection, the usual method is to ask borrowers or enterprises to provide a lot of supplementary information, which involves a lot of labor costs and time costs. In order to improve efficiency, IT is necessary to build a background management system that can automatically collect some data and has a high degree of automation, but for many small and medium-sized banks, it is also a heavy burden to set up a special team of engineers and carry out a lot of IT development work.

Big Data Credit Information and Loan Risk Control

The Rise of Big Data Credit Information Industry

From June 2065438 to June 2005, the People's Bank of China issued the Notice on Preparing for Personal Credit Information, requiring eight institutions including Sesame Credit Management Co., Ltd., Tencent Credit Information Management Co., Ltd. and Lacarra Credit Information Management Co., Ltd. to prepare for personal credit information for six months. This means that these eight institutions will become the first commercial personal credit reporting institutions in China. As a result, the prelude of the big data credit information industry was officially opened, and the growth space of the personal credit information market was opened. Based on the scale of the personal credit investigation market of 60 billion US dollars, and considering the huge population base of China, the future market space of personal credit investigation in China is likely to reach 654.38+000 billion yuan.

It is worth noting that big data credit reporting has become a battleground for Internet giants. In addition to Alibaba and Tencent, Internet companies such as Baidu, JD Finance, Xiaomi Finance and 360 Finance also indicated that they will build an Internet credit information system and intend to apply for the second batch of personal credit information licenses. Some institutions have submitted applications to the People's Bank of China. The high-profile involvement of internet companies shows that, on the one hand, the innovative characteristics and rapid expansion of internet companies have brought new vitality and opportunities to the traditional credit reporting field; On the other hand, the different advantages of Internet companies' big data and application scenarios will make the competition in the credit information market increasingly fierce.

Development Trend of Domestic Big Data Credit Information Industry

All kinds of big data companies are involved in the big data credit market, and the data dimensions and types have been greatly enriched compared with two years ago. Especially with the rise of the mobile Internet era, big data companies and big data services around mobile internet device information, geographical location information and operator information emerge in an endless stream, and they are used for P2P loan review and cross-validation. However, the source and validity of data still restrict the development of big data credit reporting industry. At present, the industry is still in the early stage of exploration, and there is no mature "killer" application tool.

Information islands still exist. Information island is an important factor restricting the development of domestic credit information industry at present. Information asymmetry and opacity bring a lot of long-term debt risks and fraud risks. When the domestic big data credit industry rises, the market has high hopes for eliminating information opacity and breaking information islands. Judging from the current development of the industry, it is impossible for information islands to disappear completely in a short period of time.

First of all, information closely related to loan risk control, such as public utility payment, fixed assets, social security, residence, etc., still belongs to relevant government departments. Although industrial and commercial, judicial and other information has been made public to the society, the openness of government information is still low, which will be a long and complicated process.

Secondly, it is difficult for internet companies with a large amount of citizen information to communicate with each other. At present, domestic Internet information such as social data, e-commerce data, geographical location data, search data and mobile device usage behavior data are concentrated in the hands of Internet giants such as Ali, Baidu, Tencent, JD.COM and 360. These companies have a lot of competitive relations in the process of staking, and the possibility of data exchange and information sharing is extremely low at present.

Finally, the information between credit reporting companies is also difficult to communicate. The core competitiveness of credit reporting companies lies in having their own unique information. As a direct competitor, it is impossible for credit reporting companies to use their own core data to enhance the competitiveness of competitors. It can be said that on the one hand, credit companies are committed to solving information asymmetry, on the other hand, credit companies are also building data barriers.

The application scenarios are gradually enriched, and portfolio credit evaluation may become the mainstream. Throughout the United States, where the credit information industry is developed, the application of credit information has long been not limited to the financial field, such as recruitment, renting, renting a car, making friends and other industries and fields all need to use personal credit information. With the promotion of "internet plus", the concept of big data and the development of P2P internet finance, domestic credit companies are also exploring and trying to enrich the application scenarios.

Judging from the development status of domestic big data credit reporting industry, because the status quo of information isolation and incomplete data sharing will exist for a long time, when the industry develops to a certain stage, it will produce combined credit evaluation. For example, the parties are required to issue credit reports from multiple institutions at the same time, and make holographic user portraits of the parties from different angles such as social, e-commerce, recruitment, browsing behavior and geographical location to judge their comprehensive situation. This is because unilateral credit evaluation has been unable to comprehensively evaluate a person, and it is necessary to give full play to the information advantages of major data credit companies in order to comprehensively evaluate.

Application case of big data credit reporting in the field of loan risk

Sesame credit that reflects the credit behavior of e-commerce. Based on Alibaba's e-commerce transaction data and Ant Financial's internet financial data, Sesame Credit has established data cooperation with public institutions and partners such as Public Security Network. The data covers credit card repayment, online shopping, transfer, wealth management, water, electricity and coal payment, rental information, address relocation history, social relations and so on. Sesame credit intuitively presents the credit rating by sesame score, which mainly includes five dimensions: user credit record, behavior preference, performance ability, identity characteristics and personal relationship. There are five grades from 950 to 350. The higher the score, the better the credit level and the lower the possibility of default. Sesame Credit Information also released a personal credit report, which was mainly provided by the central bank's credit information center, and recorded personal basic information, loan information, credit card information and credit report inquiry records.

Tencent Credit Information, which reflects the social behavior of the Internet. Tencent's credit data is more social data, and its credit products are divided into two categories: one is anti-fraud products, including face recognition and fraud evaluation; The second is credit rating products, including credit scores and credit reports. The main service targets of Tencent's credit investigation anti-fraud products include banks, securities, insurance, consumer finance, small loans, P2P and other commercial institutions. It can help enterprises identify users, prevent black accounts or organized fraud, find malicious or suspected fraudulent customers and avoid financial losses. For users such as blue-collar, students, self-employed and freelancers who have no personal credit report before, Tencent uses social services, portals, games, payment and other services to predict their risk performance and credit value through massive data mining and analysis technology, and establish personal credit scores for them.

Good loan cloud risk control that reflects the borrower's risk. Good Loan Cloud Risk Control is a big data risk control platform jointly built by Good Loan Network and FICO (the world's largest personal credit rating agency). Important data sources such as credit reporting companies, judicial data, industrial and commercial data and consumption data have been integrated, and risk databases in all fields of the whole industry required for risk control of financial lending institutions have been constructed, including anti-fraud risk list database, major risk identification list database and loan application record list database, totaling more than 70 million. The database with more than 6,000 dimensions can not only effectively supplement the local database of lending institutions, but also help them greatly improve their anti-fraud identification and credit risk identification capabilities. At the same time, it can provide services for credit institutions by combining FICO's credit decision engine. Financial institutions don't have to invest huge sums of money to build their own systems, and they don't have to spend huge energy and cost to find all kinds of risk control data.

The combination of bank risk control and big data credit reporting

Big data is difficult to solve all problems, but it can be used as an effective tool. What value can big data bring to the credit information industry? The author's judgment is that big data can't solve all the problems in credit risk control in the future; Or the types of loans that rely solely on big data for credit risk control and approval are still very limited.

However, big data can already solve some problems in the credit industry and will play an increasingly important role. For example, big data is increasingly used in anti-fraud identification, risk dynamic monitoring, user behavior analysis, user portrait and other fields. Banking institutions should embrace big data and dare and be good at using big data to assist risk control.

Make private lending information transparent to banks through big data. Through the data of big data credit reporting, banking institutions can understand the information of borrowers' private lending. At present, the information related to private lending provided by big data credit information companies mainly includes blacklist information, loan application information and inquired information. Take Good Loan Cloud Risk Control as an example, including blacklist information of various credit reporting companies and blacklist information of dozens of P2P platforms integrated by Good Loan Cloud Risk Control Platform. At the same time, it also includes the loan application record of 6,543,800,000 yuan and the query information doubled every week. This information reflects the borrower's private lending situation from the side. Through big data credit reporting, private lending information will become more and more transparent to banking institutions, identify more private lending risks, and better conduct loan review and anti-fraud identification.

Enrich the data dimension and improve the risk control ability of credit file customers. 20 14 the research by the policy and economic research Committee (PERC) on the role of non-financial information (which has also become alternative information) in credit decision-making shows that the incorporation of non-financial information such as water, electricity, coal, cable TV and mobile phones into the credit information system has significantly improved the credit acquisition ability of people with credit files.

At present, many banks gradually realize that the information that has been included in the traditional database of banks is not rich and complete, and begin to contact and cooperate with third-party big data credit companies frequently, such as customer information owned by banks and basic identity information of customers. But other information of customers, such as personality characteristics, hobbies, living habits, industry fields, living conditions, etc. It is difficult for banks to accurately grasp; On the other hand, it is difficult to analyze various heterogeneous data, such as customer's fund transaction information, web browsing behavior information, voice information of service call, video information of business hall and ATM, but other data can't be analyzed except structured data, let alone comprehensive analysis of various information, which can't break the pattern of "information island". Through cooperation with third-party big data credit reporting companies, we try our best to make up for our own shortcomings in obtaining information dimensions and data mining and analysis capabilities.

To sum up, the author believes that if banks want to further accelerate the pace of transformation, realize the vision of an honest society and inclusive finance, and shoulder the heavy responsibility of credit risk management, they must rely on the advantages of the Internet in risk control such as information use, pre-lending investigation and monitoring during lending, embrace big data credit reporting, make full use of all kinds of information at home and abroad to do a good job in customer credit reporting and credit enhancement, and further improve the level of risk control and management.

The above is Bian Xiao's sharing about big data credit reporting and bank risk control innovation. For more information, you can pay attention to Global Ivy and share more dry goods.