Abstract: China Finance 40 Forum and the International Monetary Fund recently held a closed-door seminar on "COVID-19 Pandemic, Virtual Banking, and SME Financing in China", aiming to elaborate on the Chinese experience in virtual banks providing financing services to SMEs during the pandemic. China's virtual banks introduced a number of measures to track the operations of SMEs, offer credit support and step up risk management and control, thus mitigating the financing difficulties faced by SMEs and the pandemic's impact on the economy.
The experience of China's virtual banks in supporting SMEs mainly comprises six aspects: firstly, personal data and data gathered on e-commerce platforms and social media are important for effective risk-control; secondly, intelligent digital risk management systems have the advantage of being low cost; thirdly, virtual banks leverage the synergy between human intervention and machine learning; fourthly, virtual banks and traditional mainstream banks collaborate with and complement each another; fifthly, the pressure on the management of duration and interest risks is mild during the pandemic; and lastly, loan forbearance and restructuring have little impact on the cash flow of virtual banks. Finally, to improve virtual banks’ services to the SMEs, experts suggested the following: increasing sharing of related business information, exploring the remote opening of electronic accounts for SMEs, improving the regulatory framework for virtual banks and tightening the security of digital financing.
I. Chinese virtual banks supporting SMEs amid COVID-19: practice and results
Credit support, which is reflected in: firstly, intense tracking and collection of client information. Since the outbreak of the pandemic, China's virtual banks have used businesses' digital footprint to determine which enterprises could survive the lockdown and even grow; secondly, fully exploiting the advantage of virtual banks by allowing mobile phone-based applications, and loans are granted in a contact-free manner to SMEs involved in the e-commerce, logistics, food and beverage industries as well as farmers; thirdly, loan forbearances are offered to SMEs severely impacted by the pandemic, while struggling SMEs are exempt from interest and penalty fees; fourthly, special credit funds are in place to satisfy the urgent financial needs of SMEs relating to operations, consumption and medical care, and priority is given to clients in the procurement, distribution and sales of goods and materials needed to fight the pandemic, as well as areas severely impacted by the pandemic in Hubei Province; and fifthly, the introduction of measures such as loan forbearance and adverse-credit exemption for specific individuals and groups. Statistics show that MYbank alone collaborated with fellow banks and granted 366.7 billion yuan in a contact-free manner to 8.4 million clients between 5 March and 30 April.
Risk management. For one thing, MYbank created a general reporting-monitoring system within ten days following the outbreak of the pandemic. The system includes more than 10 statements and 1,000 real-time indices and can provide real-time information on the state of business operations and the progress of business resumption based on clients’ industries and regions. For another, virtual banks have achieved rapid deployment of new products and risk strategies during the pandemic driven by core technologies such as big data, cloud computing and quantitative models. WeBank, for example, launched new products in merely ten days, setting a new record speed for product launch. Meanwhile, given their advantages with modelling, virtual banks introduced the hierarchical management of clients during the pandemic, which raises loan limits and lowers costs for good clients, offers loan forbearance and restructuring to clients experiencing short-term difficulties, and controls the amount of loans that can be granted to high-risk clients.
Results. It has been shown that the operations of virtual banks face only mild challenges during the pandemic. Although defaults briefly went up, as the pandemic comes under control and the intelligent risk-control system plays its role, virtual banks have been able to effectively respond to the unusual challenges posed by the pandemic. Studies also show that the development of digital financing has mitigated the impact of the pandemic on small family businesses in China. The fully online and contact-free financial services offered by virtual banks during the pandemic has allowed fintech to play the role of a macroeconomic stabiliser.
II. Six pieces of experience
Experience 1: Personal data and data gathered on e-commerce platforms and social media are critical data for effective risk-control.
The key risk with providing financial services to SMEs is information asymmetry, while the highest risk that influences SME’s operational stability comes from the proprietors whose personal data are key factors that may affect the quality of loans. Such data may include the stability of the proprietors’ marital status, domicile and financial standing, as well as the occurrence of any adverse changes, e.g. whether they have picked up bad habits and the quality of their friends, etc. In addition, to learn more about the proprietors, virtual banks also gather other data such historical data from financial institutions and peer-to-peer lending; data from telecom service providers; contributions to social-security fund and housing provident-fund; social data such as air travel, education, online behaviour, consumption and travelling. For business entities, operational data gathered by industry and commerce administration bureaus, tax bureaus, judiciary agencies, and value-added tax invoices are included.
Experience 2: The intelligent digital risk-control system achieves fully automated processes: from client login, risk assessment, marketing pricing to real-time loan granting and post-loan monitoring, giving it the advantage of being low cost.
As the amounts of loans to SMEs are small, the unit cost of their risk assessment is comparatively higher. As a result, traditional banks may not be able to actively help out SMEs and the process may also take longer. Virtual banks to a great extent solved the issue with the help of digital technology. Take the “310 model” of MYbank for example, the model has achieved application submission in 3 minutes and loan granting in 1 minute, with 0 human intervention. According to MYbank’s estimate, it costs about 2.3 yuan to grant a loan, which would cost around 2,000 yuan if performed offline.
Take the credit-adjustment approach of WXBank as another example, which offers clients customised loan limits and interest rates taking into consideration conditions such as the clients' personal credit during the pre-loan review. During the dynamic post-loan monitoring process, the loan limits are adjusted based on further analysis of the clients' needs. For instance, if a client is desperate for funds, has taken out multiple loans and has a poor repayment record, the bank will slowly freeze the client’s limits. In contrast, if a client is a frequent user with a positive borrowing record and greater funding needs, the client will be classified as a good client after taking into consideration the data feedback of other platforms, and the bank will gradually increase the limits.
Experience 3: With the help of machine learning, virtual banks could determine the amount, interest rate, term of every loan and whether to grant loan forbearance, and the process requires no involvement from a loan officer, thus reducing lending costs and bringing into play the synergy between human intervention and machine learning. Due to the uncertainty of the current circumstances, virtual banks are able to make timely risk judgments and lending decisions by incorporating the uncertainties in the risk-control models. For instance, with the help of machine learning, virtual banks set the interest on loans and determine whether to allow loan forbearance or restructuring. Meanwhile, risk management personnel and technicians can spend more time researching the overall risk performance as well as the willingness of business groups to take out, restructure and repay loans in different areas, industries and various state of operations. Take WXBank for example, it approves as many as 330,000 loan applications daily. With the help of fintech, this is performed easily by just 270 officers.
Experience 4: Virtual banks and traditional banks collaborate with and complement one another, and share the risks based on their respective capital contributions.
The two mostly collaborate in the form of syndicated loans. Although most traditional banks in China are capable of granting loans and assessing risks online, virtual banks still enjoy the unrivalled edge in data gathering and risk control, and may support traditional banks in risk identification, client appraisal, modelling and in- and post-loan management. Combined risk management by the two types of banks improves client screening.
During the pandemic, regulatory authorities have seen how virtual banks take advantage of fintech to expand their service channels and effectively serve long-tail clients. As a result, the regulators encourage close collaboration between virtual banks and traditional banks. The challenges in their collaboration arise from the effective connection of the fintech systems as well as accepting and agreeing with each other's ideas.
Experience 5: The duration and interest risks of virtual banks are manageable.
The assets and liabilities on virtual banks' balance sheets are mostly short-term, and the liabilities may be easily matched with assets, making liquidity management easier than traditional banks. In addition, risk-control technologies such as algorithms and big data allow better risk management. New technologies enable more accurate prediction of the asset duration of tens of millions of clients, which helps banks to appropriately source capital based on the term structures of assets.
Furthermore, the application of fintech allows virtual banks to pay greater attention to real-time data and the behavioural characteristics of business owners, thus allowing them to better assess changes of risks and manage interest rate risk.
Experience 6: Loan forbearance restructuring have little impact on the cash flow of virtual banks.
First of all, banks’ support for SMEs has been relatively reasonable. For example, measures such as blanket interest exemption and unconditional loan forbearance were introduced only in Hubei Province, whereas the extension applications are stringently assessed for other areas. As a result, the percentage of loan forbearance is very low across the country and has not led to cash-flow difficulties for the banks.
Secondly, virtual banks are relatively better equipped with big-data risk-control systems. Prior to the pandemic, they were already able to accurately screen their clients. During the pandemic, they are able to accurately predict cash-flow needs with varying risk models for different client groups, thus allowing them to effortlessly organise their cash flow.
Lastly, during the pandemic, consumption dropped and savings skyrocketed, positively affecting banks' liability end. In addition, despite the temporary rise in default rates at the onset of the pandemic, the resumption of business operations has mitigated the risk, and loan forbearance has posed little impact.
III. Policy suggestions on stepping up virtual banks' financial services for SMEs
First, enhance sharing of business data. Business data are indispensable for identifying the operational status of businesses. It is recommended to allow banking institutions greater access to information gathered by government agencies related to the operations of SMEs, including those gathered by the industry and commerce administration and taxation bureaus.
Second, explore the remote opening of electronic accounts for SMEs. The opening of electronic accounts is urgently needed to better serve SMEs, and the technologies needed for the remote opening of electronic accounts are mature, including identity identification and online authorisation. During the pandemic, the People's Bank of China introduced policies that encourage banks to offer electronic channels for corporate clients such as opening and updating accounts. It is recommended to select certain banks as trial points, with technological and operational support from relevant government departments, so as to make the policies a norm.
Third, improve the regulatory framework for virtual banks and expand their funding channels and sources. On the one hand, it is recommended to establish an independent and differentiated regulatory framework and business standards for virtual banks, which will give full play to their roles and allow regulated business development. On the other hand, within the boundary of risk control, the innovation of virtual banks' capital tools may be encouraged, and preferential consideration may be given to means to supplement capital and debt instruments, such as the issuance of preferred stocks, secondary capital bonds and perpetual bonds.
Fourth, emphasize the security of digital financing. Firstly: data security. Currently data used by virtual banks come from multiple sources, including e-commerce platforms, industry and commerce registration, and business operations, and greater emphasis needs to be placed on the data's authenticity, legality and security. Secondly: algorithm security. Flawed algorithm security may lead to risk resonance. Finally: channel security. Management of security of bank apps and APIs needs to be stepped up to protect the rights and interests of financial consumers.
