Break through the eight major pain points of banks and build intelligent risk control for banks

[51CTO.com original article] Breaking through the eight pain points of banks to build intelligent risk control for banks. The Intelligent Risk Control Center of Everbright Bank has been established for less than two years, but it has achieved considerable success in the bank. In 2019, the center developed hundreds of models for different scenarios, and launched more than 30 automated scenarios, which brought in a large number of new customers for the bank and accumulated a large number of high-quality assets. Currently, there are nearly 100 models under development. In the same year, the center won the Big Data Risk Control AI Innovation Award, the Intelligent Risk Control Best Practice Award and many other awards, and the "Intelligent Risk Control Service" was selected as the first batch of innovative cases in the financial "Regulatory Sandbox" of the National Financial Management Center.

The necessity of building intelligent risk control

Online finance is not simply the onlineization of offline finance, but requires a new intelligent risk control system to protect it.

What is intelligent risk control? It is the integration of technology, scenarios, customers, data and other aspects to provide new financial services.

What is new finance? With the continuous progress and innovation of cutting-edge technologies and new business forms, as well as the increasingly perfect combination with finance, financial technology has made great progress, gradually subdivided into "new technology" and "new finance", and then gave birth to the concept of "digital intelligence". Digital intelligence will take intelligent risk control as a starting point, integrate new technologies, and serve new finance.

The main difference between intelligent risk control and traditional risk control is the shift from causal relationship to correlation relationship based on big data, with more emphasis on data and customer experience.

Eight pain points in building intelligent risk control

What difficulties do traditional banks face in building intelligent risk control? Compared with Internet financial companies, traditional banks generally have the following eight pain points, which makes traditional banks lag behind in responding to market demand, mainly reflected in the fact that they are prone to miss the best promotion period for innovative products, are slightly passive in financial technology transformation, and lack the right to speak in the technology market.

This is how traditional banks should respond to pain points!

The introduction of new technical talents is very important. Currently, there are dozens of employees in the Everbright Intelligent Risk Control Center, one-third of whom have doctoral degrees, half of whom graduated from Tsinghua University and Peking University, and the rest are masters from well-known universities. The average age is only 29 years old. They have cutting-edge knowledge and professional capabilities in many fields such as finance, mathematics, and computers. They are a veritable team of "young algorithm scientists."

But this young team should not be underestimated. In terms of technology, the center insists on "independent research and development, open source and openness". In terms of service, it has completed the research and development of the cornerstone digital intelligent operation system and the design of the intelligent risk control SaaS service system, providing digital asset mining and intelligent risk control services for banks; in terms of algorithms, it has made breakthroughs in the research and development of multiple algorithms such as high-dimensional Logit distributed algorithms, graph convolution algorithms, and NLP improved algorithms, and has formed an implementable plan; in terms of platforms, it has completed the development of distributed data platforms, model development platforms, risk feature platforms, microservice platforms and other technology platforms. At present, it is gradually applying for invention patents to protect the above intellectual property rights.

How do traditional banks plan their business in a severe economic environment?

China's digital economy has entered a new stage of both quantitative and qualitative growth. As an important part of China's economy, digital transformation of banks is an inevitable trend. At the same time, market and industry trends require banks to vigorously develop retail and small and micro businesses. Therefore, in the process of implementing the transformation, Internet loan business has become the first area that most banks consider developing, as follows:

1. Focus on the mature self-operated Internet loan products already available in the market, such as provident fund loans, tax loans, etc., and combine the product characteristics and its own advantages to select some mature and self-operated Internet loan products with large traffic, good traffic qualifications or clear loan purposes as one of the bank's key development directions.

2. Through the analysis of existing customer groups and mining of existing customers, we can convert existing customers to Internet loan business customers. Through the analysis of existing customers, we can achieve customer segmentation, customer feature description, and product demand preference analysis. Based on the results of customer product demand and preference analysis, we can fully explore the association behavior between customers and banks and the funding demand status, and guide specific groups of customers to Internet loan business.

3. Through scenario-based construction and multi-scenario design, we can achieve all-round reach to customers and provide scenario-based customer Internet loan business needs. The most direct role of the scenario is to attract customers. When the customer's behavior is combined with the scenario, it has the characteristics of strong targeting and obvious direction. At the same time, scenario-based also has the advantage of risk control. Through the scenario, we can better determine the customer's loan intention and avoid the risk of cashing out and fraud caused by unclear borrowing intention.

《Expert Introduction》

Zhu Shihu, currently VP of the Intelligent Risk Control Center of China Everbright Bank, is one of the first batch of doctoral students majoring in artificial intelligence at the Department of Intelligent Science of Peking University. After graduation, he has been engaged in risk management for ten years. His main work areas are: Internet finance, intelligent risk control, big data models, artificial intelligence, etc. Many projects he presided over and developed won the People's Bank of China Science and Technology Progress Award and the China Banking and Insurance Regulatory Commission Innovation Award. He has also won many awards in the field of intelligent risk control. He has participated in or hosted domestic and foreign intelligent risk control forums as a keynote speaker many times, and has been reported by many media such as Toutiao many times.

[51CTO original article, please indicate the original author and source as 51CTO.com when reprinting on partner sites]