Sina Finance News, on November 17, it was reported that the 2018 Global Financial Technology (Beijing) Summit jointly organized by the China Financial Forty Forum and the Financial City was held today, and the guests held a round table discussion at the special conference on "regulatory technology and financial risk prevention".
The following is the transcript of the round table:
Host Li Wenhong: Ask Director Wang Xin, who thinks your research on FinTech is very in-depth and extensive. Can you talk about the risks you think the development and application of FinTech will bring to us?
Wang Xin: The question is big. I'll try to answer it.
The first risk is the risk caused by insufficient supervision of new entrants to financial services. Because with the development of digitalization and financial technology, many technology enterprises can also provide financial services, which makes the boundaries between new entrants and traditional financial institutions such as banks more and more blurred. The new entrants' supervision department has not had time or enough knowledge to supervise, which is a relatively large risk. I think it is very important to apply the same regulatory rules to new entrants and existing financial institutions.
The second risk is that new entrants will have a relatively big impact on traditional market participants, whether banks or non bank financial institutions. In payment and other fields, traditional financial service providers are facing great challenges from new entrants. Some financial demerits that cannot be fully prepared, especially financial institutions with relatively poor capital strength, may be subject to relatively large impacts, thus bringing risks.
The third risk may come from network security and other aspects, because many will be carried out on the network in the future. Many countries and international organizations have stressed the need to pay close attention to network security and related possible risks.
The fourth risk may be universal, because we will be affected by emotionalism in the financial market, but the public or many market participants do not fully understand financial technology and emerging business forms and services. Once there is a disturbance, it may be more likely to fall into panic, leading to systemic risks.
Fifthly, from a macro perspective, because of the new entrants, there will be new business forms, and the traditional financial services, business models in the credit field, credit quality, etc. may change considerably. Whether this will also bring some uncertain and relatively large risks to the macro-economy needs our high attention.
Host Li Wenhong: Now international organizations are paying close attention to these new problems and risks, and are strengthening tracking and research. Director Wang Xin felt that the question I asked was too big just now, so I asked Director Sun a smaller question. Could you please systematically introduce what aspects of cross-border regulatory cooperation should be concerned about.
Sun Tianqi: Just now, in the context of Fintech digital finance, cross-border breakthrough of market barriers is becoming easier and easier, and cross-border cooperation in behavior regulation is urgently needed. Behavioral supervision is to crack down on insider trading, monopoly and fraud, promote information disclosure, standardize collection, advertising, anti money laundering, anti tax avoidance, personal information protection, regulate cross-border information, etc. We need to pay attention to Fintech's illegal behaviors and cross-border proliferation of risks, especially in emerging market countries and low-income countries, because their judicial and regulatory systems are not perfect, which is more likely to cause some impact on these countries. Recently, as you know, in Southeast Asia, some domestic enterprises have gone to Southeast Asia during the process of Internet financial regulation in China, which has attracted the attention of the media.
Specific behavior supervision cooperation includes the following aspects:
1、 Unification of regulatory rules and standards. From two perspectives, the first is access. We found in our work that the access standards of different countries are very different. For businesses prohibited in China, our enterprises and foreign enterprises can get licenses in the UK and Australia, and set up a website to provide services to Chinese people. After getting the British and Australian license plates, I found that the British and Australian regulations are very strict, and I went to some small countries such as Cyprus to get the license plates, because the daily regulation of those countries is very loose. So many enterprises have set up many subsidiaries, and different subsidiaries have different licenses. A group has many kinds of financial licenses from different countries, and the regulatory standards of various financial licenses are different. Then provide cross-border financial services through the website. Therefore, the access should be unified, otherwise, if you get a license in the country where the access threshold is, you will provide financial services to the world through the digital platform. In the access link, it should be emphasized that the online digital world seems to have no national boundaries, but the financial license must have national boundaries. In addition, the construction of prudential supervision and behavior supervision system should also be unified as soon as possible.
2、 Promote cooperation in combating cross-border illegal financial activities, regulatory cooperation and judicial cooperation. It can promote the MOU and MMOU of actual operation.
3、 International organizations, such as G20, IMF and BIS, can play a greater role in promoting cooperation and exchange in behavior regulation and introducing rules.
4、 For our country, it is very important that in the development process of Fintech, the gap in financial regulation will certainly exist in the short term, and we should try to fill this gap as early as possible. I can't say that there is a problem. I asked six departments. All six departments said that they had nothing to do with me. I didn't issue the license plate, so it's not my responsibility. Functional supervision should be implemented.
5、 Large multinational enterprises providing digital financial services must play a greater role in the formation of international rules. These enterprises must not only consider problems from the perspective of enterprises, but also from the perspective of operating the entire industry. They must not sit idly by when problems occur in the industry and the market, and can not finally be thoroughly rectified, Only then complained that a mouse dung had harmed a pot of soup.
Host Li Wenhong: Could you please tell us what obstacles you encountered in the experiment, what failure cases you had, and how to solve the problems you encountered?
Xu Shen: First of all, let's respond to the behavior monitoring just mentioned by Director Sun. Behavior monitoring is a very important topic. The knowledge map we have made is also a solution for behavior monitoring. In particular, the knowledge map is a very effective technical tool for anti money laundering in many Internet financial fields, including internationally renowned Internet financial fields.
Just now, the host asked us about the problems we encountered in the experiment. Our laboratory has done a lot of experiments. Maybe everyone has seen success, but not failure. It is likely that there are many failures and few successes. When a traditional organization embraces a new technology and new concept, the biggest obstacle and problem is the awareness of the industry. The last guest of Yixin CTO this morning said that when AI is applied in the traditional financial industry, it is very important to manage business expectations. In fact, this is two-way. The essence is to manage the business's understanding of technology. Maybe I should be more objective. I attended the forum yesterday and today. The theme of our conference is Fintech. When we talk about Fintech, we seldom talk about Tech, either about our own business or how to talk about Fintech and Regtech as a new model.
Why did HKEx create a laboratory? We consider that there will be many Fintech companies and Regtech listed in our company in the future, and it is possible that we will also assume some functions that need supervision in the future. However, the key point is that we should strengthen our understanding of Fintech and Regtech. When the innovation lab is doing this, it meets a lot of businesses, wrong perceptions or wrong expectations of new technologies.
For example, for AI, the business will first think that AI is omnipotent now, so basically when talking about a topic, we have such a problem, can AI help us realize it immediately? I'm more specific, such as a chat robot I hope to tell investors a very simple rule. When you encounter such problems as how to trade options, how to trade futures, and what the trading period is, there are very important issues involved. There are two aspects of wrong understanding. The first compromise is that I put a chat robot, and soon you can automatically understand all the questions asked by users, In fact, this is not correct and objective. The first thing we do is to train this chat robot, input some problems that may be encountered into artificial intelligence as sample data, and also distinguish his intention and the entities of the questions he asks. After introducing this concept, our business immediately opened its mouth and said that you are not AI? AI should be able to automatically understand all languages, but in fact, we need a training process. In this case, a technician or AI product manager should be able to effectively sit with business partners and carefully distinguish what is effective and what is invalid.
Where is the capability of AI? Through training 30 samples, AI may understand 300 different ways to ask questions, which is the main effect of AI, so it can not be separated from training.
What is the wrong understanding of another big trend? We are talking about chat robots. Of course, some technology companies have already done this before, and chat robots can assemble their own files and provide them to questioners. So some of our business colleagues think that my chat robot will go to the Internet to learn and find the corresponding answers by itself? In fact, frankly speaking, the degree of intelligence of chat robots at this stage is relatively limited, which can be divided into two aspects, one is intelligent understanding problems, the other is intelligent assembly answers. At present, AI has done a good job in intelligent understanding of problems, but there are very limited intelligent assembly files, and as a regulatory agency or commercial institution, it has to pay a very strong responsibility for any answer given. Basically, AI needs to use a classification scheme to classify many or even unlimited different ways of asking questions, and then provide users with a standard answer. So the key point is to manage the business's understanding of new technologies, manage its awareness of new technologies, and manage its expectations of new technologies, so as to effectively manage business expectations and management expectations. The most important thing is, I think this is an art, not only an art of communication, but also an art of management, more importantly, The art of bridging the gap between technology and business.
Host Li Wenhong: Thank you for your patience in answering my questions. The following is the question session.
put questions to: I want to ask Director Sun a question, because today's topic is regulation and financial risk, and my question is the financial crisis. Now there is a view that the financial crisis is inevitable, which is generated from the current financial model. From 1998 to 2008 and 2018, domestic and foreign countries experienced more or less financial turbulence or turmoil. In a sense, FinTech makes the current financial products more complex and makes the whole supervision more difficult. So there is a view that the frequency of financial crises may accelerate in the current form. What does Director Sun think of this view?
Sun Tianqi: The development of technology must have its two sides and double-edged sword. From the perspective of the school you just mentioned, we can see that it has some negative effects on risk amplification, regulatory impact and so on. On the other hand, the development of these technologies has also enriched the data of the regulatory authorities, making it possible to visualize, track and penetrate many financial products and activities, and improving the regulatory capacity. We should balance the relationship between innovation and risk. For us, the key is that regulation cannot lag behind financial innovation too much. Similar to the past, from the ox cart to the car, from the steam train to the electric train, and then to the high-speed railway, the high speed has brought risks different from the ox cart, and also brought benefits different from the ox cart. The speed of the aircraft is higher after the flight, and the risks also have new characteristics. The management rules are changing and evolving throughout the process.
put questions to: I'm from Tsinghua University. I asked Mr. Xu of Hong Kong Stock Exchange. You just introduced the application of financial knowledge mapping in financial risk. Can you briefly introduce the application of financial knowledge mapping in financial investment, such as the discovery of customer value, the value of listed companies, and other aspects in marketing and marketing?
Xu Shen: HKEx itself is not an investment institution, but recently we were looking at an area that has relations with investors. One of our major functions as an innovation laboratory is to participate in some innovation communities, including start-ups. We have seen a lot of start-ups about investment research and investment consultants applying AI. First of all, knowledge maps are generally used in the investment research field, which involves the ownership structure and the background of investors. In particular, some mainland companies are relatively ahead of others and have different application directions. But my general summary is divided into two schools. One school is focused on data cleaning. The entrepreneurial companies of this school usually use knowledge maps. Their strong ability is that they have some background. The practitioners from Wande Data are very experienced in data sorting, so they can effectively use crawler technology first, not only the announced equity structure of listed companies, At the same time, it can also effectively summarize some industrial and commercial registration data. You will find that their knowledge map is very characteristic, not only related to listed companies, but also related to non listed companies, and even can be subdivided into different industries. Such companies belong to the field and direction of data cleaning. The other direction will be more technologically advanced. The founders of these companies have many backgrounds in technology companies, even Microsoft Research Institute. The mainland is similar to the AI Whampoa Military Academy, but also Google In the background of Facebook, they will apply more methods to intelligent mining, which is very similar to the direction we explore. They can dig out a large number of financial consulting, including structured information in listing announcements and news. At the same time, when searching, they apply knowledge maps to the search. For example, I search for a company today, Not only the content of the company can be searched, but also the information about the relevant personnel of the company's corresponding shareholding structure can be searched. We are also exploring this field ourselves.
Just now we talked about the application of AI in two directions. The first is public opinion monitoring, and now there is a knowledge map. In the next step, we will combine the two. At present, when monitoring public opinion, we only directly monitor CP's information. If we combine it with the knowledge map, we can not only monitor CP's information, but also monitor the industry information of CP related major shareholders. In this way, you will find that the information will have a strong joint effect once it appears, and the biggest financial risk is the avalanche effect. If one company goes wrong, there will be some joint effects. The direction and potential that AI can explore in this field in the future is very huge. In particular, there are many good start-ups in the mainland that have actively explored and used this aspect.
Host Li Wenhong: Time relation, the question is over! According to the requirements of the sponsor, I need to make a summary, and I want to talk about three points of experience, because I basically participated in one-and-a-half days of discussion from yesterday afternoon to today to talk about some of my experiences and new understanding in the one-and-a-half days.
First, I feel that in the process of integration of finance and technology, regulators and institutions are more closely connected, including both financial institutions and non-financial institutions, especially some sci-tech innovation enterprises. These two days, we learned that there are some differences in the regulatory practices of different countries in terms of sandbox regulation, but I feel that in this process, no matter what the essence of sandbox regulation is, the most important thing in common is to allow regulators, financial institutions and technology enterprises to have more communication, exchanges and in-depth discussions, It enables regulators to know more about the new business model of the institution and the current regulatory framework and requirements. In this process, the institution can decide whether to further apply for a financial license, and the regulators decide whether to issue a relevant license to the applicant in this process. So I feel that in this process, it is a country with a deeper understanding of each other. Regulators understand the business model and institutions understand the regulatory framework.
So we also thank CF40 Forum and Peking University for providing us with such opportunities.
Second, I feel that the principle of technology neutrality is a very important principle for regulators. No matter what kind of technology you use, you must apply the same regulatory standards to the same business, so as to ensure fair competition in the market and an orderly environment for survival of the fittest. My deepest experience in the past two days is that we have talked a lot about regulatory sandbox recently in China. Basically, every article on FinTech calls for the introduction of regulatory sandbox in China. I also want to ask the authors of each article whether the regulatory sandbox you call for is British style, Hong Kong style or Singapore style, Therefore, in the financial field, it is necessary to penetrate the essence of business and implement supervision. When we draw on international experience, we should also look at its business essence through a name. For example, the Swiss director introduced to us that there is a very essential difference between the Swiss regulatory sandbox and the British regulatory sandbox. The Swiss regulatory sandbox means that if an institution receives deposits of less than 1 million Swiss francs, it can be exempted from the regulatory requirements. The regulatory sandbox in the UK is a completely different concept. When people called for the introduction of regulatory sandbox in China, did they realize that our pilot is also a Chinese regulatory sandbox?
The lady also asked just now whether the supervision sandbox should be unified internationally? We believe that this is just a regulatory practice, and there is no need for everyone to have the same concept and practice in this field. But we need to know what is the essence of a country when discussing its regulatory sandbox. So I think that everything should be understood and grasped according to the essence.
Third, machines and technology are important in the development of financial technology, but for us humans, the role of human beings is always more than machines and technology. How can people understand these technologies? In understanding this technology, what kind of system, mechanism and culture do we have to give full play to the positive role of technology while preventing related risks. In this process, we need our judgment. Judgment is very important. Good Judgment needs long-term support and experience accumulation, mutual discussion and data support.
This is my experience in the past two days, not a summary. Thank you, all the speakers!
Host Wang Haiming: Thank you. See you next year at the Financial Technology (Beijing) Summit!
Sina stated that all the meeting transcripts were filed in on-site shorthand. Without the speaker's review, the posting of this article on Sina. com for the purpose of transmitting more information does not mean that it agrees with his views or confirms his description.
Editor in charge: Zhao Ziniu
