Banking in the big data age

The big data boom has sparked a major change in thinking, production and lifestyles, and it can be said that a new era has been opened up. For the financial industry with natural data attribute, on the one hand, large data can provide sufficient information support for the operation and management of financial institutions. On the other hand, large data-breeding new financial forms pose serious challenges to traditional financial institutions. What the financial institutions will do in this great social revolution is very much expected. To this end, The Economist has invited a number of senior managers of financial institutions as well as industry experts to explore the financial sector in the big data era of change and development.

Over the past decade, the Reform and development of China's banking industry has made the world's most notable achievements. Five big commercial banks are among the world's top companies in this year's "banker" Forbes ' list of big firms and market capitalisation. With the rise of the information revolution, represented by mobile internet, cloud computing, "Big Data" and "IoT of Things", the banking industry is faced with new opportunities and challenges. Whether the Chinese banking industry can use large data to realize management, management and service innovation determines its future sustainable development capability.

The banking industry has a preliminary basis for using large data

Large data is the product of the development of information technology and Internet industry to a certain stage, from Internet to IoT, from cloud computing to large data, information technology is moving from industrial base to industrial core. The banking industry, as a combination of information technology, Internet thinking and decision-making data has begun to embed the whole process of management. The big data is "deep learning", which can provide comprehensive, accurate and real-time decision information support for banks. The bank's business transformation, product innovation and management upgrade need to be fully used in large data. At present, the Bank in the customer analysis, risk management of large data application has accumulated a certain amount of experience for the future transition to a comprehensive large data application has laid a good foundation.

In the the 1990s, with the development of information technology, the domestic banking industry is adapting to the trend and applying information technology to business processing and internal management to improve the efficiency of service management. Into the 21st century, the big banks took the lead in promoting system centralization and data centralization, integrating the original decentralized information system, adapting to accelerate product innovation, enhance customer experience and other market demand, set up data Warehouse and data platform, information level continuously improve. In recent years, the banking industry has vigorously developed a new generation of customer-oriented core business system, information system construction is becoming more and more complete, electronic banking and other online financial services growth, in the promotion of customer experience and risk management capabilities, meet regulatory requirements, the formation and storage of a large number of available data resources. The data resources of the banking industry include not only the structured data of the core business of deposit and loan, but also the unstructured data such as telephone voice, online transaction record and video.

Since 2011, China Construction Bank (CCB) has started to build a new generation of enterprise-class whole-line sharing core business system, with customer as the center, service-oriented design framework, realize business and it integration, product rapid innovation, has begun to take shape at present. Especially in the new generation system design, fully consider the importance of data storage and application, and specially set up a data Integration layer module, including data buffer, data recording system, historical data storage, analysis data Warehouse, real-time data warehouse, public data mart.

Banks are beginning to try to access and integrate external data resources. In the traditional data analysis mode, the banking industry, for market analysis, internal management, regulatory needs, produced and recorded a huge amount of text-structured data, involving customer account funds transactions, financial information, as well as network silver browsing, telephone, video and other unstructured data. However, the traditional bank can only grasp the customer and banking business-related financial behavior, can not get customers in social life to embody interests, habits, consumption tendencies of emotional or behavioral data, can not form a linkage with the business data. With the rapid development of electronic commerce and the deepening of mobile finance, the banking industry has stepped up its docking with external data sources, identifying effective information, integrating multi-channel data and enriching the customer atlas. At present, a number of banks have made useful attempts.

First, the bank and the Electronic Business platform to form strategic cooperation. The banking industry shares the business data of small micro enterprises on the electric business platform and the personal information of the operators, and recommends the high quality enterprises with the loan intention from the electric business platform to the bank, and determines the credit level of the enterprise through the transaction flow, the evaluation of the buyers and sellers, and gives the credit limit CCB has made a useful attempt in this respect. In addition, there are banks to participate in the electronic business, data cooperation cases.

Second, the bank to build the platform of the electric business. The bank builds the electric business platform, obtains the independent discourse right of the data resource. To provide value-added services to customers at the same time, access to customer dynamic business information for the development of small micro-credit to lay the foundation for the bank to build a platform for the power of the driving force. 2012, the construction bank took the lead on the line "good melts business", provide business-to-business and customer mode of operation, including wholesale, retail, housing transactions and other fields, to provide customers with information dissemination, transaction matchmaking, community services, online financial management, online customer service and other ancillary services, the provision of financial services from the settlement, custody and the guarantee extends to the whole process of the online financing service to the merchant and the consumer.

Third, the bank establishes the Third-party data analysis intermediary, specially excavates the financial data. For example, some banks extend their one-on-one cooperation with the electric business platform to "tripartite cooperation", in between the bank and the electricity merchant, joins the third party company to be responsible for the data docking, provides the bank and its subsidiary data analysis mining value-added service. Its core is to analyze the customer's transaction data, accurately forecast the customer's consumption and transaction demand in a short time, so as to have a precise grasp of the customer's credit demand and other financial service demand.

The banking industry has experience and talent in handling data. Data analysis and measurement model technology has been used in the traditional data field, and a large number of people proficient in metrology and analysis technology have been developed. In the aspect of risk management, China's financial regulatory department, in line with international standards, introduces the new Basel Capital Accord and other international norms, providing a set of risk management tool system for the banking industry. Banks in this framework, the use of historical data to measure credit, market, operation, liquidity and other types of risks, internal rating-related technical tools have played an effective, widely used in loan assessment, customer access to exit, credit approval, product pricing, risk classification, economic Capital Management, performance appraisal and other important areas.

Banks have initially tried to apply large data. China's banking large-scale use of large data technology is not yet mature, but a number of banks from the key point, the specific business application of large data mining technology to solve the problem of efficiency improvement. For example, some banks offer telephone sets, network online,, micro-bo, micro-letter integration in one service platform, but also some bank credit card center to develop intelligent cloud voice, focusing on customer service voice information mining and analysis, through the massive language data on the continuous online and real-time processing, for service quality improvement, operational efficiency promotion, The service mode innovation provides the support, thus comprehensively promotes the operation management level. There are also some banks in personal customer marketing, focus on customer data analysis, explore customer behavior patterns and potential needs, leading to targeted precision sales. For example, by analyzing customer behavior data and financial data to lock potential customers, according to customer behavior rules, and combining their region, behavior content to determine consumption habits, to carry out targeted marketing; analyze transaction information to effectively identify small micro enterprise customers, and use remote banking and cloud lending to implement Cross-selling. In addition, some banks also have their internal customer number and microblog, QQ, mailbox and so on, will be the Internet data and traditional data storage, set up a database, not only to understand customer finance, fund purchase transactions such as the frequency of the degree, but also can find other dynamic information such as business trips, preferences and social circles.

Lessons learned from the application of large international data

International experience in the application of large data in the financial industry is mainly reflected in the rapid judgment of macroeconomic trends, analysis and prediction of customer and counterparty behavior, fraud prevention, improvement of internal efficiency and outsourcing of non-core business.

Quickly judge the macroeconomic situation. The Bank of England has begun using large numbers to make quick judgments about the UK's real estate and labour market trends. Previously, the Bank of England through the statistical Department of real Estate sales data, employment data and so on to judge the real estate market and labor market trends, but the statistics department's data generally have several days or even weeks of time lag, is not conducive to the rapid assessment of the situation. At present, the Bank of England has obtained the latest economic performance by monitoring some web search keywords, such as "mortgage", "house price" and "position".

Analyze and forecast customer and counterparty behavior. ZestFinance, a credit evaluation firm founded by Google's former chief information officer Douglas Merrill, integrates massive data technology into a complete customer jigsaw puzzle that restores the customer's real situation and actual credit status more accurately, It also supports cooperative companies in providing "wage-day loans" to Americans who are struggling to obtain loans from banks (payday loan). The Spanish foreign Bank (BBVA), which has a memory-enabled ATM machine, ABIL, not only remembers the amount and frequency of withdrawals, but also gives advice on how to make withdrawals according to its account. Some fund companies in the United States began a few years ago to use social media data to analyze market sentiment changes, and to determine whether future transactions will expand or shrink. Recently, these fund companies further through the analysis of financial transactions of large data, identify the trading characteristics of counterparties, pre-contract the trading trend of counterparties, and take appropriate action to obtain the difference.

Fraud prevention. Use large data analysis software to prevent credit and debit card fraud. Improve the ability of banks to defend against fraud in areas such as transactions, transfers and online payments by monitoring customers, accounts, and channels. When monitoring customer behavior, large data can identify potential offending customers, prompting bank staff to focus on them, thus saving the resources of anti-fraud monitoring.

Improve internal efficiency. Bank of America uses large data to analyze the behavior of employees at a call center in the bank, by placing sensors in the employee's name brand, monitoring the employee's walking line and conversational tone, and knowing the employee's social status in the workplace. Monitoring results suggest that employees who enjoy a break and communicate with each other are more productive and can share tips on how to deal with "difficult" customers in their day-to-day interactions. When Bank of America discovered the phenomenon, it turned to a group break, with staff performance up 23%, and the level of stress in the staff's tone of voice down 19%. In addition, some European and American banks use large data to evaluate the performance of branch offices and achieve significant results.

The application of large data has operational risk and operation risk, such as data loss, data leakage, data illegal tampering, information asymmetry in the process of data integration leads to wrong decision, the latter such as corporate reputation risk, data is the business risk after the opponent acquired. Therefore, data control must be strengthened. There are both successful experiences and lessons to be summed up. From the problems that have emerged, the biggest risk comes from cyber attacks and fraud: In 2011, cyber-bank fraud caused losses to 270 billion yen (about 22.5 billion yuan) in 53 Japanese banks; in 2012, fraud syndicates attacked the networks of at least 60 banks in Europe and the US, stealing bank funds; A domestic insurance company was attacked by hackers, which caused hundreds of thousands of of policy information to leak. To this end, the first is to attach great importance to and promote the unified data standards, and do a good job of data cleaning to ensure data quality. The second is to carefully delimit the data boundary, reasonably carry out internal and external data sharing and non-core data outsourcing. Third, the large data should pay more attention to privacy protection and information security, increase the investment in the attack against the network.

Strategies to drive large data applications

The 18 of the party put forward to adhere to the new industrialization of Chinese characteristics, information, urbanization, agricultural modernization Road, information has been upgraded to national strategy. China's banking industry to speed up the application of large data not only has the industry significance, and to promote China's information process, service "new four" development also plays an important role. China's banking industry should fully recognize the importance of large data analysis and application from the strategic height, and explore from the construction of management system and concrete application mode to build the core competitiveness of banking in the era of large data.

Set up a perfect management system of large data. The banking industry should fully understand the importance of large data, set up a large data work promotion mechanism at the head office level, make large data work plan, manage the data department to carry out the overall planning, organization and coordination, centralized management of the large data work, the Business Department undertakes the responsibility of data collection, analysis and application. Multi-mode integration of internal and external groups of data, the formation of management data, the use of data and dissemination of effective working mechanisms.

Enhance data mining and analysis application capabilities. To fully promote the decision making based on data within the bank, using information to create value concept, the introduction of data mining and large data use of professional methods and tools, training professional data mining analysis personnel, pay attention to the economic and financial, mathematical modeling, computer new algorithm and other complex skills, establish a forward-looking business analysis model, GRASP, Forecast market and customer behavior, apply data depth to business management processes, use data to guide work, design and formulate policies, systems and measures to achieve precision marketing and fine management.

To promote the construction of smart banks with large data technology. To promote the transformation of large data to productivity, accelerate the technological development of product innovation laboratories, apply the mature products of laboratories to the marketing and service of customers, promote the construction of intelligent banks, and transform the advantages of technological innovation into competitive advantages. Network services to use a good large data and other technical achievements, promote the popularization of intelligent vtm, remote banking, electronic banking service area, intelligent Interactive desktop, face recognition and other innovative services, the traditional banking service model and innovative technology organically combined, the use of intelligent equipment, digital media and human-computer interactive technology to bring customers "self-help, intelligent , wisdom "new feeling and experience. Intelligent Network in the construction of the promotion, should also fully adopt user interaction technology and experience equipment, to attract customers to browse, trial, compare various financial products, supplemented by staff recommendations, from the region, customers, products and other dimensions, mining customer needs, to achieve the right customers, at the right time, through the appropriate channels, recommend the appropriate products.

Establish a pricing system based on large data analysis. Current, changes in the frequency of transactions and liquidity accelerated, large data from a broader perspective, the volatility of the pre-debt, can be more flexible to calculate whether to meet regulatory requirements and loan demand changes, thus for the bank to deposit and loan, to lend the absorption strategy to provide quantitative support, can effectively reduce the cost of capital. Banks also need to use large data analysis, set up a comprehensive service and credit differential pricing system, so that different products, different industries, different regions of the implementation of differential pricing, and ultimately achieve a one-policy integrated, differentiated services to enhance the level of precision marketing. For example, the public and private customers will be gradually included in the pricing system, customer choice, different services to enjoy different credit concessions to achieve differentiated pricing and customer best experience dual purposes.

Rely on large data technology to enhance the level of risk management. Large data can better solve the problem of asymmetric information in traditional credit risk management, improve the pre credit risk judgment and credit risk early warning ability, and realize the precision and foresight of risk management. In the big data age, the banking industry can break the information island, fully integrate the customer's multi-channel transaction data, and the operator's personal finance, consumption, behavior and other information to credit, reduce the risk. If the construction bank relies on "good and financial business" to develop large data credit products "good and credit", banks can monitor social networking sites, search engines, things networking and E-commerce platforms, track and analyze customer relationships, emotions, hobbies, shopping habits and other aspects of their credit rating and repayment will change to the pre-sentence, In the first time credit business, lack of credit strong variables, timely use of educational background, past experience and other variables to combine analysis to establish a credit risk early warning mechanism. The shift from historical data analysis to behavioral analysis will make a great breakthrough in the current risk management model.

Large data is a very frontier and fast developing technology in the information revolution, the banking industry should solve the problem of internal data mining analysis and the security integration of external resources, quicken the transformation of talent team construction and technology achievement, accelerate the transformation and upgrade of the banking industry and sustainable development through the high efficient application of large data.