Author | Wu Jun "General Manager of the Inclusive Finance Department of CITIC Bank"
Article | "China Finance" Issue 8, 2022
Since the 18th National Congress of the Communist Party of China, the Party Central Committee and the State Council have attached great importance to the development of inclusive finance and strived to create new advantages in the digital economy. Digital inclusive finance has emerged and has formed with the trend, becoming an important tool for financial services to the real economy in the new era. As a state-owned commercial bank, CITIC Bank inherits the red gene, seizes development opportunities, elevates inclusive finance to the strategic height of the entire bank, takes digital transformation as the pillar of the bank's development, and accelerates the digital transformation and development of inclusive finance with data asset construction as the core.
"Thinking" and "action" of data asset construction
Inclusive finance digital transformation is a systematic project involving financial technology, regulatory constraints and other aspects. Data asset construction is the key and foundation. Talking about digitalization of inclusive finance without data assets is tantamount to "cooking without rice". In the new era, we should think carefully and practice hard in promoting the construction of data assets.
"Thinking": The overall concept of data asset construction
Data assets are data resources owned or controlled by enterprises and recorded in a certain way, which can bring economic benefits to enterprises, and are the key production factors in the digital economy era. How to build a good ecology of data asset construction, standardize the key links in data asset construction such as data precipitation, data sharing, and data application, and achieve the full life cycle leap from "standardized production" to "effective application" is a key issue that needs to be solved in carrying out data asset construction.
data precipitation is the prerequisite, with the focus on standards, specifications and ease of use. implements the "cloud-to-use data to empower intelligence" action, promotes the construction of key areas such as smart cities, intelligent manufacturing, and smart logistics, accelerates the digital transformation of industries, enriches the supply of data resources, unifies data standards and rules, standardizes data production behavior, and creates a standard, standardized and easy-to-use data generation ecosystem.
data sharing is the foundation, the focus is on security, openness and order. improves data sharing systems and mechanisms such as laws and regulations, user management, and technical applications, and on the basis of ensuring information security, breaks the data shaft and promotes orderly sharing of data in different regions, industries and different entities, and creates a safe, open and orderly data sharing ecosystem.
data application is the core, focusing on agility, efficiency and precision. improves the data development application platform, carries out financial technology application and model optimization, conducts correlation analysis of massive data resources and taps the potential of value, and creates an agile, efficient and accurate data application ecosystem.
"Bank": CITIC Practice for Data Asset Construction
CITIC Bank takes the construction of "Digital CITIC" as the guide, established a Digital Transformation Office led by the Party Secretary, established a first-level department of the Big Data Center, and created a data ecosystem covering the entire life cycle management of data accumulation, data sharing, data applications, etc., and regards the construction of inclusive financial data assets as an important base and innovative frontier for digital transformation.
first, data accumulation, integration of the CCP win-win situation. CITIC Bank provides information technology to support the digital transformation of small and micro enterprises, deeply integrates into the business scenarios of small and micro enterprises, widely connects the public service information of small and micro enterprises, and realizes comprehensive, full-process and full-time acquisition of small and micro enterprise data. First, integrate into enterprise digitalization. Taking the construction of small and micro enterprises as the "tenta", through low-cost code development, system interconnection and embedding, small and micro enterprises will accelerate their digital transformation, promote the applications of digital office, online operations, remote collaboration, etc., and achieve in-depth integration with the daily business management data of enterprises. The second is to integrate into scenario digitalization. Taking the business scenarios of small and micro enterprises as the "tenta", CITIC Bank's financial service system is upgraded and transformed "industry standardization + personalization between enterprises", embedded in the digital interconnection between small and micro enterprises, and achieving organic integration with channel data such as core enterprise supply chains, industrial parks and Internet platforms. The third is to integrate into government digitalization. Taking the government public service department as the "tenta", "co-construction" and "interconnection".On the one hand, we actively participate in the construction of smart cities and provide technical support for the construction of public service systems, data management, etc.; on the other hand, we will widely connect with government departments' information management systems such as industry and commerce, taxation, housing, and procurement, and realize enterprise information acquisition in supporting government digital construction.
Second, data sharing, open in the specification. CITIC Bank is based on data security and consumer rights protection, and promotes open data exchange by standardizing technology application, authorization management and system construction. First, standardize the application of technology. Through blockchain distributed ledgers, smart contracts, consensus mechanisms, etc., we can solve the security, integrity and credibility of data communication. Through multi-party secure computing, federated machine learning and other methods, we can realize the sharing of core sensitive data available and invisible, and promote the safe utilization of data assets. Through data desensitization, secure isolation and other methods, a data security management firewall is established to block the transmission of core sensitive data between the production network and the office network, the intranet and the Internet. The second is to standardize authorization management. By improving the authorization mechanism of information subjects, we will promote the standardization of authorization content and forms, and continuously standardize authorization management. Based on the principles of "must know", "minimum authorization" and "specialized and dedicated", an internal data sharing authorization mechanism across regions, institutions and departments will be established to ensure reasonable authorization, secure interaction and controllable scope. Strengthen post-verification of data exchange, carry out data usage trace analysis, and strictly eliminate violations of authorized data use. The third is to standardize system construction. By continuously carrying out the construction of "Safe CITIC", we will integrate responsibility culture throughout the business, strictly implement laws and regulations such as the Cybersecurity Law and the Implementation Measures for the Protection of Financial Consumer Rights and Interests, formulate unified data security management measures and data security hierarchical management implementation rules for the entire bank, and refine the data asset management process links to ensure that the purpose, processing methods and scope of use of data are in compliance with laws and regulations.
Third, data application, development in exploration. CITIC Bank has built a "CITIC Brain" with automated, visual and efficient modeling through the construction of the basic data system of "one lake, one library and one platform" (data lake, new generation data warehouse, data analysis cloud platform), and widely used data assets in the entire process of inclusive financial business such as product research and development, operation management, risk prevention and control.
focuses on agile delivery and explores the digitalization of products and services. takes "data foundation building, technology application, and encouragement of innovation" as the core, builds an inclusive financial innovation test field, product research and development intelligent credit factory, and builds a digital product system for inclusive finance. By designing nearly 100 standardized scenarios and more than 200 reusable components, the product R&D cycle has been shortened by three quarters, the project's parallel production capacity has been increased by three times, and the inclusive financial system has been rapidly improved. Relying on supply chain and big data, we will deepen product development and application, develop "order e-loans", and support the start-up capital needs of order production of upstream small and micro enterprises; relying on the interconnection of ERP system data and channel data of core enterprises, we will develop "distribution e-loans", and support the batch ordering of downstream small and micro enterprises; relying on "specialized, precise and innovative" patent data, we will respond to the "double innovation" development strategy, develop "science and innovation e-loans", and realize "science and innovation loans, borrowing and repaying as soon as possible". In just two months after the product was launched, loans were issued to more than 200 small and technological innovation enterprises.
focuses on efficient support and explores the digitalization of operation and management. has established a digital marketing terminal with "one enterprise, one policy, thousands of people and thousands of faces", deeply integrate into the operating ecological scenarios of small and micro enterprises, and create more customer contact points into customer demand information collection points, actively explore scenario-based, fission-based, and chain-based marketing models, relying on data analysis, demand mining and opportunity insights to capture customers' real financial needs in a timely manner. Build an open marketing platform, promote the introduction and cleaning of tens of millions of data, build a "four-degree evaluation model", improve the customer labeling system, and realize digital business opportunity management. The data analysis cloud platform and "Puhui Cockpit" are put into production, providing about 120 reports, promoting the construction of data assets in the entire process of business, and realizing multi-dimensional intelligent statistical analysis and targeted development strategy adjustments. Create a "cloud business hall" for inclusive finance, launch a real-time marketing map, break through the limitations of physical outlets, expand services to ecological scenarios, and extend them to customers.
focuses on precise prevention and control and explores the digitalization of risk management. adheres to the business orientation of "real small and micro, real business", and builds a digital risk control system covering before, during and after loans, and on the basis of promoting the implementation of the risk embedded management mechanism and improving the risk management system. Before lending, with the principle of "more dimensions, higher timeliness, and more accurate information", deepen the use of data assets, expand the risk information analysis and monitoring dimensions, establish a unified risk view and identity authentication system, conduct comprehensive and accurate holographic portraits, and intelligently identify potential risks. In loans, relying on real trading scenarios, big data, rule engines, biometrics and other technologies are used to strengthen risk measurement assessment, intelligent model research and development, improve digital and professional credit review and approval mechanisms, and gradually realize "one-click application and one-click approval". After the loan, we will accelerate the construction of sample databases, continuously correct and optimize risk control models and risk control strategies, promote the integration of post-loan early warning, set up red, orange and yellow, and pay attention to the fourth floor, nearly 100 active warning indicators, and build a risk control rule system with more than 30,000 combinations, and transform from "civil defense" to "intelligent control".
"Stability" and "Progress" in the construction of data assets
data assets is a long-term and continuous task that requires the joint efforts of the government, banks and enterprises and the long-term efforts. The core position of the financial industry in the modern economy requires us to always put the prevention and resolution of financial risks first. In the new era, data asset construction work should adhere to the principle of stability and seek progress while maintaining stability.
seek progress in stability, consolidate the foundation and make up for shortcomings
In the process of data asset construction, "stability" is the main theme and the overall situation, and is the foundation and prerequisite for the digitalization of inclusive finance. Only by taking steady steps can we achieve a long-term goal. We need to continue to consolidate the foundation and make up for shortcomings on the basis of consolidating existing development achievements.
Build a solid foundation for data quality. In recent years, with the popularization of technologies such as the Internet, smartphones, and artificial intelligence, data scale has increased explosively, but the data foundations of various industries vary and the quality of data is of varying quality. The problem of the "digital divide" among different groups is becoming increasingly prominent, the data barriers between different institutions have not been completely broken, the problem of imbalance and inadequate data generation in different regions still exists, and the problem of authenticity and validity of different types of data has not been solved. It is necessary for the government, banks and enterprises to form a joint force, increase the popularization of digital technology, promote active transformation of all industries and industries, clarify data standards, optimize data catalogs, and realize massive data accumulation; accelerate the establishment of an integrated national basic data platform with wide coverage, good accuracy, strong availability and high standards, and realize unified data collection, unified storage, unified processing and unified display.
fills the gap in information security. information security issues are a companion of big data technology, including problems caused by potential problems and security vulnerabilities of the system, improper system operation, external network attacks, and personnel moral risks. It is necessary to correctly understand the dialectical relationship between information security and big data technology, improve the data life cycle security protection mechanism, improve the data subject rights recognition mechanism and data comprehensive management agreement, establish a cross-subject data security sharing privacy computer system, and effectively prevent possible security risks.
promotes stability through progress and seeks development through innovation and innovation
In the process of data asset construction, "progress" is the direction and the purpose. "Seeking progress while stability" is by no means "wait, rely on, and want", but "seeking speed in good condition" and achieve high-quality and sustainable development in the digital exploration and innovation of inclusive finance.
innovates in controllable. should adhere to the principle of giving equal importance to safety, controllability and openness and innovation. On the premise that risks are generally controllable, large and medium-sized banks with strong risk management capabilities and key areas with controllable risks should be selected, and priority should be given to the application of national basic data platforms. Through the "negative list" and "sandbox supervision" models, we encourage the experiment of new data open source, data sharing and data application methods, encourage data innovation products to be first tried within a certain range, and steadily promote application based on summarizing pilot experience.
is being improved in development. The development of is the basis and key to solving all problems.At present, data asset management is still in its early stages, lacking a formed management mechanism, and there is no ready-made experience to follow, and it needs to be continuously improved in business practice. Taking data application as an example, the scientific nature of the bank's inclusive finance big data model, the algorithmic decision-making mechanism, etc. have not yet experienced a complete economic cycle. Various problems may arise in the application, but they cannot be discarded. Instead, in practice, they should go through the process of "development-verification-redevelopment-reverification" through cross-verification with expert judgments, regular special stress tests, and strengthen the construction of bad sample databases, and continue to correct and improve the accuracy of the model and ensure the scientificity of the algorithmic decision-making mechanism. ■
(Editor Zhang Xiaozhe)
