With the wide application of financial technology in the financial industry, innovative financial service models continue to emerge, the financial industry is also undergoing profound changes, emerging formats in the industry to bring new variables at the same time, but also produced new risk scenarios and risk characteristics, traditional compliance management and risk prevention means are more and more difficult to meet the requirements of regulatory authorities and financial institutions, the trend of technology to supervise financial technology is becoming more and more obvious, regulatory technology is also widely concerned by the market. According to fintech Global, there were 317 regtech investment deals worldwide in 2019, raising $8.5 billion, up 89% year-over-year.
Feng Haitao, vice president of Qunzhi and partner of the financial industry, introduced to Yiou that at present, driven by the dual application needs of the regulatory end and the compliance end, regulatory technology has shown great commercial potential, and the market opportunity window is accelerating.
From the perspective of compliance, financial institutions are increasingly exposed due to financial losses, reputation losses and business losses caused by lagging risk control compliance capabilities, and compliance capacity building is imminent. The data shows that in the whole year of 2018, the CBIRC organs, the former CBRC and the former CBRC sub-bureau issued more than 3,800 fines to banking financial institutions and practitioners. At the same time, the punishment is getting stronger and stronger, and the amount of fines has begun to rise from the million level to the tens of millions level. In addition to the direct losses caused by business non-compliance, the indirect losses caused by insufficient risk control compliance capabilities of financial institutions are immeasurable. In reality, the risk control and compliance department adopts a conservative attitude towards the new exploration of the business unit because it is difficult to judge the financial risks in the innovative business, and financial institutions are likely to miss many innovation opportunities.
In addition, in the context of the tightening of financial supervision, the regulatory authorities' requirements for the compliance of financial institutions have changed from requiring process compliance to focusing on results and effectiveness. Financial institutions traditionally rely mainly on manpower, based on expert rules to meet regulatory requirements, so they began to change their thinking, explore the use of artificial intelligence, big data and other innovative technologies to actively identify and control risks, reduce compliance costs, and enhance compliance capabilities.
On the regulatory side, the regulatory authorities are changing the regulatory concept of grasping the big and letting go of the small and waiting for spot checks on risk events. Fintech model innovation is emerging in an endless stream, and regulators are gradually realizing that the risk supervision system with traditional large financial institutions as the "starting point" may lead to regulatory lag and regulatory vacuum, and have begun to actively interact with the regulated regulators to listen to the problems of financial institutions' business innovation. "We are now very happy to see that many regulators are beginning to be willing to try some more innovative ways, such as using artificial intelligence to supervise, so as to timely discover and prevent various risk points and loopholes." Feng Haitao said.
The application of regtech involves a large amount of financial data, systems and business rules, with distinct data-driven characteristics, so improving the quality of basic data is the premise and foundation for the promotion of regtech applications.
Feng Haitao said frankly that the current regulatory authorities and financial institutions are facing some difficulties in handling financial data. First, data is multi-source decentralized. For financial institutions, the data islands between branches and head offices and different departments are concerned; for regulators, the phenomenon of data dispersion and data fragmentation of different regulatory objects is common; secondly, the data magnitude is very large. Financial institutions need to form dozens of hundreds of statements, involving a very large amount of data, although the regulatory authorities get the data integrated by financial institutions, but all the data submitted by the regulatory authorities is still summarized in a large amount; in addition, multi-source data often has inconsistencies, and the timeliness of data processing of financial institutions is also under great pressure.
However, compared with other industries, data governance in the financial industry is already in a relatively leading state. Feng Haitao introduced that in the stage of bank information construction, banking business data has moved from dispersion to concentration, data governance engineering has continued to the present, and the data standardization construction of financial institutions has been relatively standardized, which has laid the foundation for the current big data application, and is also an important consideration for the selection of artificial intelligence to financial regulatory technology.
Feng Haitao divides the application of regTech into three levels:
The first layer is that in order to improve regulatory efficiency and reduce the cost of compliance manpower and infrastructure input, regulatory authorities and financial institutions apply cognitive intelligence technologies such as nlp and knowledge graph to some repetitive labor in supervision and compliance work, hoping to liberate manpower to do more active and challenging tasks; The second layer is for regulators and financial institutions to change their regulatory and business philosophy, seeking to use regtech to achieve proactive risk identification and control; The third layer is to really drive business development with regtech, which is a higher level of technology application.
Taking the marketing risk control solution launched by Lanyun Qunzhi to financial institutions as an example, the risk control compliance department can use the solution to identify the risks of existing customers, and then make suggestions for the marketing of business departments according to the risk preferences of different customers, thereby improving the pertinence and effectiveness of marketing. "When risk control and compliance departments can play an active role in the business unit, the business unit in turn will re-recognize their value."
According to reports, Qunzhi focuses on the research and application of artificial intelligence big data, and provides end-to-end application products and solutions such as elens' new generation of intelligent anti-money laundering, intelligent order review, intelligent risk control and credit system, and intelligent operation in the financial industry.
Specifically, elens' new generation of intelligent anti-money laundering system is applied to the anti-money laundering compliance business scenarios of financial institutions, which can help financial institutions to more accurately monitor and screen suspicious transactions and identify customers, thereby improving the overall efficiency of financial institutions' anti-money laundering compliance business and reducing compliance costs.
In terms of product system, it includes sanction list system, intelligent kyc, intelligent anti-money laundering suspicious transaction monitoring, and visual suspicious case screening platform. Specific to the business process, Qunzhi will be publicly released sanctions list and partner data combined, the use of machine learning, nlp technology to develop an anti-money laundering sanctions list system, in the form of artificial intelligence for sanctions list management, screening accuracy is about 60% higher than the rules-based screening system; in the suspicious transaction monitoring stage, through the cognitive intelligence training model system accuracy is higher, the system can reduce the amount of early warning submitted by the system by 60% to 70%; for suspicious cases, The system can visualize diagrams in an interactive and visual manner, automatically recommend possible crimes, and automatically generate large and suspicious transaction reports based on regulatory requirements.
Feng Haitao introduced that the traditional anti-money laundering requires three people to complete the workload in three days, and the compliance department may only need one person to complete it in one hour with the help of elens' new generation of intelligent anti-money laundering system, which greatly improves efficiency.
The intelligent single review expert system can be applied to specific business scenarios such as international settlement, foreign exchange, insurance, and investment banking, helping financial institutions deconstruct professional knowledge and liberating expert labor.
Taking the international settlement business as an example, in recent years, all commercial banks have vigorously developed intermediate business, and international settlement business has become an important source of profit for intermediate business. However, due to the complexity of the international settlement review process, the professional quality requirements of the review personnel are very high, the banks and their branches in the expansion of the field of international settlement services at the same time, have to support their own professional single review talent team, but the repetition of the audit business itself makes the high professional quality of talent mobility is larger, resulting in high labor costs.
Elens intelligent audit expert system uses ocr technology to electronically digitize the invoice information, extract the goods list in the invoice, invoice date and other elements through nlp technology, and precipitate the knowledge of business experts through the knowledge graph, automatically determine whether each audit point is passed, and automatically or assist the decision-making of the review personnel.
"The intelligent single review expert system entrusts the repetitive and boring work to the artificial intelligence, and the reviewer can do more professional things." Some artificial intelligence can not judge, still need the adaptability of the single reviewer, in a sense, artificial intelligence is not to replace people, but to turn people into superhuman. Feng Haitao said.
The intelligent risk control and automatic credit granting system is mainly used in the credit approval and credit granting business scenarios of small and micro enterprises, which can help financial institutions quickly build core risk control and credit granting capabilities in the credit field.
At present, there are many financial technology companies that provide risk control support for consumer credit business in the industry, and Feng Haitao introduced that the difference between the group wisdom at a glance lies in the deep understanding of the risk control and anti-fraud of small and micro enterprises in banks. At the same time, she also said that the bank itself is an institution that operates risks, intelligent risk control in addition to credit business, can also be applied to all aspects of the bank's internal risk management, and intelligent risk control in a broad sense is still a blue ocean worth digging deeper into.
At present, AI has entered the stage of extensive commercialization and exploration, and there are not a few companies in the market that provide solutions for vertical industries such as medical, financial, retail, travel, and education based on the underlying technology of artificial intelligence. Feng Haitao believes that the differentiated advantage of the group wisdom is to take cognitive intelligence as the technical foothold, have a deep understanding of the vertical industry, take the financial industry as an example, the company has a team of experts in the financial industry, can provide end-to-end application software products and solutions, and ultimately hopes to make artificial intelligence become productive and realize the industrialization of artificial intelligence. Specific to the field of financial regulatory technology, artificial intelligence technology will be more and more accepted by regulatory authorities and financial institutions, regulatory technology will develop in the direction of intelligence, dynamics, human-machine collaboration and multi-party interaction in the future, and regulatory technology will usher in the outlet.