Previously, Qiao Yang served as the head of risk strategy and data modeling at Discover, as the CEO of ZRobot, a financial technology company jointly established by JD Digits and the American company ZestFinance, and as the general manager of the credit management department of JD Digits. . During his tenure at JD.com, he and his team built JD.com's internal personal user credit scoring system "Xiaobai Credit", which attracted the attention of the industry.
Recently, news has also spread that Cheng Jianbo, the former vice president of JD.com and general manager of the Personal Risk Management Center, will serve as the general manager of Pudao Credit Information Co., Ltd. (hereinafter referred to as "Pudao Credit Information"). Previously, the risk control team led by Cheng Jianbo had long-term "escorted" the operations of JD Baitiao and accumulated practical experience in the field of consumer finance big data risk control.
The world of financial loan assistance may be small, with acquaintances coming and going. But for the entire loan assistance industry, its long-standing traffic data, risk control, and even customer service in the entire business model may return to the hands of banks and other licensed financial institutions under regulatory guidance in the future. This "landing" "The process may be longer.
Industry ups and downs
Li Li made up his mind to leave the Internet industry at the end of 2020. In 2018, he switched jobs from a credit card center of a joint-stock bank to work as a product manager in the consumer finance business of a major Internet company. After bidding farewell to the bank with rigid management, for the first time, the Internet work atmosphere where he was surrounded by employees born in the 1990s made him feel more relaxed. Freedom, salary income is also higher than the previous level in the bank from a book perspective. Although some hidden benefits such as physical examinations and medical reimbursement have shrunk, for him, a healthy young man, these are not the top issues.
However, after the initial novelty wore off, Li Li found that this purely online personal consumer credit business with United Loan as the main model was a bit "boring". The entire business model and industry chain have been very stable. From BD (business) to products, operations, and customer service, everyone performs their own duties. Everyone is just a small screw on the assembly line that can be replaced at any time. As time goes by, I seem to For those who are just "skilled workers" at a certain node, it is difficult to see any hope of improving their personal abilities or getting promoted in the workplace.
However, in the first half of 2019, the entire industry is still on the rise. Banks, consumer finance companies, small loan companies and other financial parties have great demand for consumer credit, and a steady stream of low-cost funding sources are provided. , on the contrary, the platform side seems to be struggling to obtain traffic sources on the asset side, and the entire business and team are still immersed in a prosperous atmosphere.
From 2018 to 2019, public data shows that the loan balances of three financial technology companies listed on the US stock market, 360, Qudian and Lexin, have been growing rapidly. This may serve as a development trend in the joint lending and loan assistance industries. a reference. Among them, Qudian clearly proposed the loan assistance business based on the "open platform" model in 2019.
Li Li now recalls that the turning point in the industry atmosphere occurred in the second half of 2019. In Hangzhou, the crawler businesses of some relatively large third-party data companies in the industry were shut down, and the business leaders were taken away for investigation. Some cooperative banks began to require the source of user data to be explained when diverting traffic. Although the banks were more of a "formal" requirement at the time to meet regulatory requirements, it also sounded a warning to the industry.
In the first half of 2019, Chen Ming’s company had just shifted from P2P industry consolidation to cash loans. The company had an online small loan license and some self-operated loans. It also has joint loan business in cooperation with banks, consumer finance companies, etc., and its scale is considered to be a waist-high enterprise in the industry. Chen Ming thinks this is not safe enough. He hopes to change jobs to a consumer finance company or bank because he feels "it is safer within the system."
At the same time, Chen Ming’s company had just disposed of its P2P business with difficulty, and in accordance with regulatory intentions, it switched to a business focused on joint lending and loan assistance models.
But in the second half of the year, something happened to the third-party data company, and he was keenly aware that the regulatory storm on data information had begun.
But no one expected that the sudden COVID-19 epidemic in 2020 would press the pause button on the undercurrents between supervision and industry.
As an operator of an online consumer loan product, Zhang Yue saw too many "human realities" in the first half of 2020 - some users faked their diagnosis of COVID-19 in order to avoid repayment. Certificate, but in fact the data on new cases in the area where it is located can be found with a little verification. Some users drove from Tianjin to Beijing to forcibly "make an appointment" with the person in charge of the business because they owed several thousand yuan. Some users repeatedly applied for deferment of repayment for a few hundred yuan in arrears and threatened the customer service staff to jump off the building...
With users' economic and emotional fluctuations experiencing great fluctuations, the personal consumer credit market has grown almost While big data has been more widely used in the fight against the epidemic, the online contactless credit products of various banks have entered a stage of rapid development. In the second half of the year, when the impact of the epidemic on the economy gradually receded, the haze over the industry seemed to gradually dissipate. In the second half of the year, Zhang Yue felt that high-risk complaints from front-line customer service had slowed down, and business volume was recovering steadily.
In July 2020, the China Banking and Insurance Regulatory Commission issued the "Interim Measures for the Administration of Internet Loans of Commercial Banks". Some financial technology companies interpreted this normative document as positive. "The biggest highlight of the new regulations is that it fully affirms the The role of the loan market and institutions, and encouraging commercial banks to cooperate in absorbing new technologies to promote change and innovation in the credit industry, is a major benefit to leading financial technology platforms.”
But then, in August 2020. In the "Regulations on Several Issues Concerning the Application of Laws in the Trial of Private Lending Cases" promulgated by the Supreme People's Court in March, 4 times LPR (one-year loan market quoted interest rate) was set as the upper limit of judicial protection of private lending interest rates, and the interest rate red line was drawn even tighter. location. In the following months, a wave of "anti-collection alliances" emerged in various places. People who were trapped in the high-interest trap of black online loans and private loans regarded the provisions of the Supreme Law as a life-saving straw to get rid of debts. This situation It also once spread to lending disputes related to banks and consumer finance companies. Although the Supreme Court has regulations for non-financial licensed "private lending", licensed financial institutions still feel pressure from public opinion and moral aspects. , and the more an organization is at the top of the industry and attaches great importance to brand reputation, the greater the pressure.
For loan assistance institutions with consumer licenses, the upper limit of self-operated asset loan interest rates can be as close as possible to 24 (the previous "usury" red line), and in the loan assistance business In the model, some licensed financial parties have begun to require designs based on the upper limit of 4 times LPR.
Chen Ming saw that his company’s target customers were mostly blue-collar and lower-income people, and their personal financial risk resistance was poor. These characteristics have been amplified by the epidemic, and the market demand has shrunk under the epidemic. and depressed interest rate pricing space, which has greatly compressed the profit and profitability space of small and medium-sized loan assistance platform institutions. He began to actively look for his "next job", and his best goal was to enter a bank.
In October 2020, two leading platform institutions failed to go public one after another. The regulatory expectations and market reputation of the entire industry have plummeted.
In the fourth quarter of 2020, along with multiple public criticisms from senior officials of the Central Bank and China Banking and Insurance Regulatory Commission on issues such as disorderly innovation of Internet finance, driving up financing costs, and misuse of information and data, There is also the promulgation of the "Interim Measures for the Management of Online Small Loan Business (Draft for Comments)".
This heavy-handed regulatory document directly points to issues such as Internet platform institutions' excessive use of online small loan licenses, joint investment, and leverage ratios of 3:7 or even 1:99 to amplify credit funds.
After that, Li Li discovered that his cooperative banks began to shrink intensively, United Loans stopped adding new loans, and some banks even stopped lending altogether.
“The financial future of the Internet is worrying.
"At that time, Li Li made such a judgment and began to actively seek job-hopping opportunities in banks and consumer finance companies.
On the other hand, consumer finance companies and direct selling bank licenses that had been frozen for many years began to be released, "Supervision It also tends to allow licensed financial institutions to carry out financial innovation. "At the end of 2020, a foreign investment banking institution asked some experts from local mutual financial associations, the State Council Development Center, financial regulatory research institutions, etc. for their views on Internet lending and financial technology supervision through a symposium. The experts' core views include three Aspects: First, Internet antitrust and data strategies are the reasons behind China’s tightening of financial technology; second, the leverage ratio of the residential sector is high, and supervision will restrict consumer credit; third, supervision of online loans will be more stringent.
In the first half of 2021, while large and small platform institutions in the industry are still in the process of converting joint loans to loan assistance, Li Li and Chen Ming have left Internet platform companies and successively joined banks.
Loan assistance and the risk control game of small and medium-sized financial institutions
On July 7, 2021, some platform institutions received an email from the Central Bank’s Credit Bureau, requesting Further improve the rectification plan in accordance with the business cooperation flow chart of "Platform-Credit Agencies-Financial Institutions", which requires that "platform institutions shall not use information actively submitted by individuals when conducting business cooperation with financial institutions such as diversion, loan assistance, and joint loans." Information generated within the platform or information obtained from the outside is directly provided to financial institutions in the name of application information, identity information, basic information, personal portrait rating information, etc., and a comprehensive 'direct connection' between personal information and financial institutions must be achieved. ”
To understand this new regulation, it is necessary to clarify the cooperative relationship between the platform and financial institutions in the existing loan assistance business.
According to the previous China Popularization of Renmin University of China According to the "Research Report on Loan Assistance Business Innovation and Supervision" (hereinafter referred to as the "Loan Assistance Report") issued by the Hui Financial Research Institute, the cooperation processes of various participants in the loan assistance business include pre-loan cooperation, mid-loan cooperation, and post-loan cooperation.
The "Loan Assistance Report" points out that in pre-loan cooperation, each participating entity needs to complete customer acquisition, preliminary screening, customer drainage and other steps. In terms of customer acquisition and preliminary screening, loan assistance institutions can do so purely online. Or use a combination of online and offline methods to reach lending customers, and use big data, artificial intelligence and other technological means to conduct preliminary screening and risk control of lending customers before pushing them to the funding side, and select target customer groups that meet the funding side's prerequisites. “In terms of customer attraction, after loan assistance agencies screen out the target customer groups, they push the credit scores, credit recommendations and other information of the loan customers to banks and other funding parties. ”
According to a person from a financial technology company, the central bank’s email requesting rectification on July 7 mainly targeted this pre-loan customer acquisition link, requiring platform institutions and funding parties to strengthen this process. The completeness and detail of data reporting to the central bank.
In the "direct connection" mode, user information is directly transmitted between the platform and financial institutions. In the disconnected direct connection mode required by the central bank, it is introduced. However, the key to "cutting off direct connection" is that under the "direct connection" model, the quality of information data is directly related to the diversion service income of platform institutions. Linked.
Currently, the charging method for loan assistance business in the market is that the funder charges all fees and then returns the service fee and interest rebate to the loan assistance institution or credit enhancement institution. For example, in April this year. Xinye Technology, a U.S.-listed company, stated that it is fully shifting to a loan-assisted and profit-sharing business model. Previously, 360 Digits has also announced that it is continuously increasing the business proportion of its "capital-light" loan assistance model.
Accepted. After this cooperation model, the same goals of banks and financial technology companies will be closely tied: to achieve as much profit as possible as soon as possible, while controlling NPLs. In this model, the loan assistance platform will actually First, we conduct a risk control screening internally, use our own credit evaluation logic to match users to appropriate funding sources, and use some data that cannot pass the test to "fill up" the data, which will damage the mutual income of both parties.
In other words, when user data is “directly connected”, data quality is also implicitly linked to commercial revenue. Then, after "disconnecting the direct connection", financial institutions will not be able to directly verify the original data transmitted by the platform to the credit reporting agency, and they will be responsible for their lending income and non-performing assets.
Jin Tian believes that after the "disconnection of direct connections", the existence of credit reporting agencies has been directly strengthened in banks and platforms. The impact at least includes: the platform cannot directly output its own accumulated customer data, but can only provide to data processed by credit reporting agencies, thereby better protecting customers’ privacy data security; credit reporting agencies implement strict licensed operations, which raises the threshold for general market institutions to participate in loan assistance and joint lending; in view of the role of the platform weakening (it is no longer difficult to directly control customer quality), the original business model, especially the charging model, may face major changes. If the potential income is significantly reduced, it may further reduce the willingness of the platform to participate in loan assistance and joint lending, thus changing Market ecology.
In the "Interim Measures for the Administration of Internet Loans of Commercial Banks" issued in July last year, the loan assistance business was actually somewhat relaxed, clarifying that except for the core risk control links, commercial banks must be independent in all other aspects. Can cooperate with third-party companies. Specific cooperation contents include: marketing customer acquisition, joint loans, risk sharing (joint lenders and sharing parties must be licensed institutions), information technology, overdue collection, etc.
However, on February 20 this year, the China Banking and Insurance Regulatory Commission issued the "Notice on Further Regulating the Internet Loan Business of Commercial Banks", insisting on "implementing risk control requirements. Commercial banks should strengthen the main responsibilities of risk control and independently carry out Internet Loan risk management, and independently complete the risk control links that have an important impact on loan risk assessment and risk control, and it is strictly prohibited to outsource key links of pre-loan, loan, and post-loan management. ”
In this regard. The process from relying on external risk control to building its own core risk control will put the data technology and risk control management level of small and medium-sized banks to a great test.
Since larger loan-assisting platforms have access to multiple financial institutions, those with high approval rates and fast lending speeds are likely to be assigned to more users, and in the loan-assisted business Among them, financing guarantees and insurance institutions are introduced as credit enhancements, or through business agreements such as deposits on loan assistance platforms and cover-up repurchases. Funding institutions actually enjoy risk-free returns, and there is not much real risk control mechanism and level. "Practical" experience.
At an industry event in the first half of this year, the president of a private Internet bank spoke candidly about this difficult problem, the commonplace issue of independent risk control. Regulatory agencies have been very diligent about independent risk control for banking and other financial institutions. From the initial 2017 Document No. 141 to the promulgation of the latest new mutual lending regulations, independent risk control has been emphasized many times. But if you really think about it, in recent years For private banks, and for these smaller banks that later developed, small and medium-sized banks that do not have their own traffic, "How much ability do we have to achieve independent risk control, in terms of credit scoring, anti-fraud, and post-loan management?" To what extent, the regulatory authorities need to clarify that the platform belongs to the platform and the bank belongs to the counter. ”
So, are small and medium-sized financial institutions ready for the “independent risk control” that regulators have been talking about for many years?
(At the request of the interviewee, Chen Ming, Zhang Yue and Li Li are pseudonyms in the article)