The digital transformation of banks is accelerating, how to prevent potential risks? Industry knowledge: strengthen supervision!

The digital transformation of the financial industry is progressing rapidly.

The COVID-19 epidemic forced this road of transformation from uniform speed to acceleration. The urgency of market demand is beyond imagination: digital transformation has not only become an important topic in many financial forums recently, but also the financial technology leaders of some Internet banks and traditional financial institutions have become hot guests, or special lectures, online lectures or forum speeches. Financial technology giants have made great strides in the financial industry to provide technical support for financial institutions that urgently need digital transformation. Following the new strategy of "Digital Rural Credit Cooperatives" in Alibaba Cloud, more ecological resources such as intelligent risk control, intelligent data experience and digital life scenes will be opened to help small and medium-sized banks such as rural credit cooperatives and rural commercial banks form unique innovative service formats. On August 16, Tencent Cloud also formally signed a strategic cooperation agreement with Kunshan Rural Commercial Bank. The two sides will cooperate in new financial infrastructure areas such as bank private cloud platform construction, distributed architecture transformation, distributed database application and mobile terminal development. At the same time, relying on the superior resources of both parties, we will promote the digital construction of retail business and scene-based financial innovation, and build a new digital connection for future financial scenarios.

However, digital transformation has never moved business from offline to online as people imagined, once and for all; It's not the establishment of a financial technology department. It's so light to make several bank apps. Because it involves the all-round reconstruction of the entire internal structure, process and concept of financial institutions, it is not easy to realize the real digital transformation. Among them, in the process of transformation, how to effectively prevent and control the potential risks of smart finance in the future is particularly difficult.

As we all know, digital transformation is inseparable from big data, cloud computing and artificial intelligence; Only by giving full play to the power of technology can we realize the transformation from traditional finance to intelligent finance in the digital age. Using big data, financial institutions can build business and risk control models that meet their actual needs; Using cloud computing, efficient and fast operations can be carried out according to the model, and the results can be used in daily operations; Using artificial intelligence, we can realize the automatic distribution of high-frequency micro-loans and realize the real "second loan". It can be said that the intelligent finance in the digital age has subverted the traditional financial business model, saved a lot of manual operation processes, greatly improved customer experience and satisfaction, and made the customer-centered business philosophy a reality. This can be seen from the fact that the private internet bank in the head serves tens of millions or even hundreds of millions of customers every year and issues tens of millions of loans.

However, coins always have two sides. When financial institutions enjoy the convenient and efficient benefits brought by smart finance after digital transformation, potential risks also follow.

Xiao Gang, former chairman of the China Securities Regulatory Commission, recently warned that "the new format of deep integration of artificial intelligence and financial industry is the direction of financial model reform. While vigorously developing, we must also guard against possible systemic risks. " He said that because artificial intelligence mainly relies on models and algorithms, when technology is applied to the financial market, once the data quality is not high or there is deviation, it may have a butterfly effect and bring systemic risks. For example, in the capital market, the preset investment model is often implemented without human intervention, which may make the investment strategy highly consistent and impact the market at a certain point, thus causing systemic risks. There was a "lightning crash" in the American stock market, and the Dow Jones index plunged thousands of points in a very short time.

Although Xiao Gang's view has long been recognized by the industry, it is still of great significance in the context of financial institutions stepping up digital transformation and developing smart finance.

In fact, the industry has a deep understanding of the potential risks of smart finance. Huang Jinlao, chairman of Suning Bank, said in an interview that through digital transformation, financial services will penetrate into every scene and ecology like water, and financial services or financial products will be embedded in the production, trading and personal life of enterprises, but this "penetration" will also bring new risks and challenges. The first risk is data risk, which comes from the full automation of finance and relies on data to make decisions. Banking is the most abundant or comprehensive field of data application. How to integrate these data reasonably is a compulsory course for the digital transformation of banks. Improper response will lead to data fraud, data interruption, data leakage, data abuse and other risks. The second risk is technical risk, including the interpretability and evaluability of the algorithm. For example, build a model with 100 variables to evaluate whether loans can be issued, but who will measure whether the model is appropriate? It also covers the security risks brought by technology. In the case of high dependence on digital system, once the system is attacked or shut down, it may cause greater harm to financial security.

It is precisely because data collection comes from people and model building comes from people. Therefore, although smart finance is easy and intelligent, there are great risks due to data quality problems, algorithm parameter settings and other issues. If the person who built the model has bad intentions again, the hidden risk will be even greater. Once the risks of smart finance occur, they are often systematic. Therefore, it is very important to take precautions and do a good job in risk prevention and control.

How to guard against potential risks? The knowledge in the industry is to strengthen supervision.

In the era of intelligent finance, traditional regulatory concepts and means cannot be effectively matched. Therefore, financial institutions are accelerating the digital transformation, and the regulatory means of the regulatory authorities should also accelerate the transformation. Because the risk points of smart finance are hidden in data governance and algorithms, the supervised object should not only include the interpretable supervision of the model, but also let the supervised object explain clearly what logic the model is based on; It should also include the supervision of builders and designers of smart finance-related models and algorithms. To this end, it is necessary to improve the supervision system in time, plug the blind spot of supervision, and let the relevant employees have the qualifications recognized by the regulatory authorities through qualification identification. The most important thing is that the supervisory layer should have knowledgeable supervisors who can understand, manage properly and govern effectively.