FINANCIAL SECTOR ASSESSMENT PROGRAM PEOPLE’S REPUBLIC OF CHINA FINANCIAL INCLUSION TECHNICAL NOTE DECEMBER 2017 This Technical Note was written by Jennifer Chien and Luan Zhao. The note was prepared in the context of a joint World Bank/International Monetary Fund Financial Sector Assessment Program mission in the People’s Republic of China during May-June, 2017. The note contains technical analysis and detailed information underpinning the FSAP assessment’s findings and recommendations. Further information on the FSAP program can be found at www.worldbank.org/fsap. THE WORLD BANK GROUP FINANCE AND MARKETS GLOBAL PRACTICE CONTENTS Abbreviations and Acronyms ........................................................................................................................ i I. Executive Summary .............................................................................................................................. 1 II. Current State of Financial Inclusion ..................................................................................................... 7 A. Access to Finance for Individuals .................................................................................................... 7 B. Access to Finance for Firms .......................................................................................................... 14 III. Public and Private Sector Commitment .......................................................................................... 16 A. Conceptualization of Financial Inclusion ...................................................................................... 16 B. Appropriate Role of Government .................................................................................................. 17 C. Institutional Capacity and Coordination ........................................................................................ 19 D. Financial Inclusion M&E and Data Infrastructure ......................................................................... 20 E. Government Financial Inclusion Policies and Programs ................................................................ 22 IV. Provider Diversity and Sustainability ............................................................................................. 26 A. Banking Sector ............................................................................................................................... 28 B. Financial Cooperatives................................................................................................................... 29 C. “New-Type” Financial Institutions ................................................................................................ 31 V. Fintech................................................................................................................................................. 33 A. Leveraging Fintech to Reach the Underserved .............................................................................. 34 B. Addressing Risks ........................................................................................................................... 35 VI. Credit and Borrower Information ................................................................................................... 39 A. Credit Registry ............................................................................................................................... 39 B. Credit Bureaus and Data Analytics Services ................................................................................. 40 C. Access to Public Information ......................................................................................................... 41 D. Data Protection and Privacy Legal Framework ............................................................................. 42 VII. Provider Reach ................................................................................................................................ 43 A. Manned Outlets .............................................................................................................................. 45 B. Rural Cash Withdrawal Points ....................................................................................................... 46 C. AML/CFT ...................................................................................................................................... 48 VIII. Product Range and Appropriateness ............................................................................................... 49 A. Product Range ................................................................................................................................ 49 B. Product Appropriateness ................................................................................................................ 50 C. Level Playing Field ........................................................................................................................ 52 Figures Figure 1: Ownership of transaction accounts over time................................................................................ 7 Figure 2: Transaction accounts in China, by individual characteristics ........................................................ 7 Figure 3: Transaction accounts in China, by income group.......................................................................... 8 Figure 4: Usage of debit card, by income group ........................................................................................... 9 Figure 5: Digital payment instruments.......................................................................................................... 9 Figure 6: Digital payment instruments, by urban and rural ........................................................................ 10 Figure 7: Digital payments use ................................................................................................................... 10 Figure 8: Opportunity for increasing financial access and usage................................................................ 11 Figure 9: Savings ........................................................................................................................................ 11 Figure 10: Saving, by urban and rural......................................................................................................... 12 Figure 11: Borrowing, by channel .............................................................................................................. 12 Figure 12: Borrowing, by income group ..................................................................................................... 13 Figure 13: Borrowing, by urban and rural .................................................................................................. 13 Figure 14: Use of commercial insurance, by urban and rural ..................................................................... 13 Figure 15: Access to account for firms, by firm size .................................................................................. 14 Figure 16: The obstacles in doing business for firms, by firm size ............................................................ 15 Figure 17: Credit demands firms were met, by firm size ............................................................................ 15 Figure 18: Funding sources for investment, by firm size............................................................................ 16 Figure 19: Physical Access Points .............................................................................................................. 43 Figure 20. ATM and POS device penetration ............................................................................................. 44 Figure 21: Remote access points................................................................................................................. 45 Figure 22: Collateral requirement for firms, by firm size ........................................................................... 52 Figure 23: Value of collateral requirement for firms, by firm size ............................................................. 52 Tables Table 1. Summary of Main Policy Recommendations ................................................................................. 4 Table 2: Main Types of Financial Providers in China ................................................................................ 27 Table 3: Transaction data of rural cash withdrawal points ......................................................................... 47 Table 4: Reasons for Not Applying for a Bank Loan ................................................................................. 53 ABBREVIATIONS AND ACRONYMS ABC Agricultural Bank of China ADBC Agricultural Development Bank of China AML/CFT Anti-Money Laundering/Combating the Financing of Terrorism ASR Agricultural Supplementary Relending ATM Automated Teller Machine BRICS Brazil, Russia, India, China, and South Africa CBRC China Banking Regulatory Commission CCB China Construction Bank CDB China Development Bank CDF Community Development Funds CFCA China Financial Certification Authority CFPA China Foundation for Poverty Alleviation CGAP Consultative Group to Assist the Poor CGFS Committee on the Global Financial System CGC Credit Guarantee Companies CGS Credit Guarantee Scheme CHFS China Household Finance Survey CPCA China Payment and Clearing Association CIRC China Insurance Regulatory Commission CSRC China Securities Regulatory Commission CRC Credit Reference Center CRS Cash Recycling Systems DFS Digital Financial Provider EAP East Asia and Pacific ES Enterprise Survey EU European Union FIP Financial Inclusion Plan FSAP Financial Sector Assessment Program FSB Financial Stability Board G20 Group of Twenty G2P Government to Person GDP Gross Domestic Product GNI Gross National Income GPFI Global Partnership for Financial Inclusion ICBC Industrial and Commercial Bank of China ICT Information and Communications Technology IT Information Technology JSCB Joint-Stock Commercial Bank KYC Know Your Customer M&E Monitoring and Evaluating MCC Micro Credit Companies MFI Microfinance Institutions MIIT Ministry of Industry and Information Technology MNO Mobile Network Operator MoA Ministry of Agriculture i MoC Ministry of Commerce MoF Ministry of Finance MSE Micro and Small Enterprises MSME Micro, Small, and Medium Enterprises NDRC National Development and Reform Commission NFCS Net Financial Credit System NGO Non-Government Organization NIFA National Internet Finance Association of China NPL Non-Performing Loan OECD Organisation for Economic Cooperation and Development PBC People’s Bank of China POS Point of Sale PSBC Postal Savings Bank of China P2P Peer-to-Peer RCOMB Rural Commercial Bank RCOPB Rural Cooperative Bank RCC Rural Credit Cooperative RFI Rural Financial Institution SCB State-Owned Commercial Bank SCOPA State Council Office of Poverty Alleviation SME Small and Medium Enterprise VTB Village and Township Bank WB World Bank WBG World Bank Group ii I. EXECUTIVE SUMMARY1 1. China has made significant progress in financial inclusion. High levels of account penetration, savings, and usage of payments services have been achieved, largely due to extensive branches and access points (particularly a vast network of rural cash withdrawal points), innovations by non-bank payment providers, and expansion of government-to-person transfers and bankcard programs. Account penetration in China is quite high, with estimates ranging from over 80 to 90 percent, which compares well to the EAP regional average but is lower than the high-income country average. The rapid growth in fintech has led to millions of previously underserved mass retail consumers accessing lower cost and better tailored financial products and services. 2. Despite impressive progress, challenges remain in reaching the “last mile” and providing a wider range of financial services. In particular, access to credit and insurance products is lacking for individuals and micro and small enterprises (MSEs) that are technically “banked”, but still underserved. Actual usage of financial services such as transaction accounts could also be improved, especially among rural residents. 3. Expanding beyond the current achievements in financial inclusion requires a more modern conceptualization of financial inclusion. Many stakeholders still believe that financial inclusion means promoting credit to the rural poor via subsidized approaches and preferential policies. To date, the gains in financial inclusion in China, while substantial, have been primarily driven by government mandate and encouragements. To achieve high levels of financial inclusion over the long-term requires a commercially sustainable approach, where providers are internally motivated by market-based principles. 4. Heeding the principle of commercial sustainability requires a shift in the role of government from promoter to enabler. The new Plan for Advancing Financial Inclusion Development (“FIP”) refers to “commercial sustainability” as one of its main guiding principles, a significant step forward. However, while some stakeholders appreciate the importance of market-based approaches, there is a general lack of understanding regarding how these principles translate into action. Mixed messages from government, with quantitative targets on physical outlets in rural areas and the number of loans disbursed, contradict the principle of commercial sustainability. A market-based approach does not equate to the government instructing providers to serve underserved segments in a commercially sustainable manner. Rather, a market-based approach entails policymakers taking concrete action to improve the legal and regulatory framework and financial infrastructure (e.g. credit infrastructure, legal framework for agent- based models, rural financial institutions (RFIs), and “new-type” providers, ICT infrastructure) so that providers can serve the underserved in a commercially sustainable manner. 1 This technical note focuses primarily on financial inclusion with respect to individuals and microentrepreneurs. The note focuses on select areas that were identified as key remaining challenges, in order to provide deeper analyses and recommendations in these areas. This note is not a comprehensive review of all past and current efforts and issues related to financial inclusion in China. For example, many important developments in secured transactions and SME finance are not covered in this note, nor are insurance or securities products discussed, although many advances have been made in these areas with respect to financial inclusion. 1 5. Strong national leadership, inter-agency and national/local coordination, and collaboration with the private sector should be strengthened. Strong leadership is needed at the national level, particularly as the role of government shifts to focusing on improving the enabling environment for providers, which requires coordinated action and consensus across different agencies and suffers when too many key actions are decentralized. A high-level body empowered to make broad, strategic decisions should be established in order to take necessary action at an ecosystem-wide level. A formal entity or mechanism is needed that sits above financial sector regulators and other agencies, is chaired by a high-powered individual, can provide strategic vision and thought leadership, and has the authority to make decisions on issues that are cross-sectoral, related to broader infrastructure, or entail national/local coordination. Better and more active coordination and information sharing among stakeholders is also needed, both between national and local governments as well as between the public and private sectors. 6. Promotional government policies and programs should be employed in a limited, strategic manner, and regular assessments undertaken to ensure efficiency and impact. A wide range of fiscal, monetary, and supervisory policies and programs have been adopted across agencies to promote financial inclusion, but impact is rarely assessed. Results should be consistently tracked and analyzed to design more effective policies that are better targeted and less distortive, with the results made publicly available. Promotional policies should be limited to only those “last mile” segments for which government support is necessary, and should be well- coordinated and not duplicative. Supervisory or prudential tools should be careful assessed and their potential impact on financial stability taken into account when such tools are used to promote financial inclusion. The M&E system for financial inclusion should be enhanced to allow for performance monitoring and evaluation, in order to allow for evidence-based policymaking. More use of pilots, linked to rigorous evaluations, should be encouraged. 7. Improvements to the legal and regulatory environment for RFIs and “new type” providers are needed in order to create a competitive, vibrant sector. Further reforms should be undertaken to address the weak corporate governance of rural credit cooperatives (RCCs), with the objective of transforming RCCs into commercially-oriented providers without undue interference by local government. The regulatory framework for village and township banks (VTBs) and microcredit companies (MCCs) should be revised to address constraints on geographic expansion (such as limitations to operating within a single county), funding, and ownership requirements that currently prevent such providers from achieving economies of scale and reaching their full potential in serving rural clients and MSMEs, combined with higher standards of regulation to ensure that RFIs and “new-type” providers are subject to sufficient levels of supervision. 8. Fintech should be further leveraged to reach the remaining underserved. Policymakers should study how fintech can be more fully leveraged to reach the remaining underserved, particularly in rural areas. For example, partnerships between fintech providers, RFIs, and “new-type” providers or outsourcing arrangements between traditional providers and fintech service companies could be encouraged and facilitated, particularly where such partnerships enable offline/online models. Potential barriers to undertaking such partnerships should be identified and addressed, such as poor IT systems or governance at RCCs, VTBs, and MCCs. Further improvements in ICT infrastructure to reach the “last mile” could help to expand 2 the positive impact of fintech, in particular increasing internet and mobile coverage in remote and rural areas, improving quality (e.g. expanding broadband and 4G), and decreasing costs. Finally, developing a deeper understanding of the needs and behaviors of “last mile” consumers and improving the financial literacy of such consumers, particularly with respect to digital finance, may also be beneficial. 9. At the same time, a clear and comprehensive legal and regulatory framework should be developed to help ensure the healthy growth of responsible fintech companies while still allowing for innovation. Advances in fintech have been accompanied by numerous risks, including loss of funds, data privacy violations, poor disclosure, and false promotion. The conscious move away from the previous “wait and see” approach already begun by policymakers should continue to be pursued, replaced with a more comprehensive regulatory framework and proactive, risk-based supervision undertaken in a coordinated fashion across regulators. Tools such as regulatory sandboxes can be employed to allow for continued innovation. A tiered approach to regulation and supervision should be developed for large fintech providers operating nation-wide (i.e. P2P platforms and online MCCs) as opposed to the current system of dual regulation and supervision by local financial offices. The legal and regulatory framework for financial consumer protection should also be adapted to address issues related to fintech, including disclosure and transparency for digital channels, sales practices, product suitability and responsible finance, and dispute resolution mechanisms in a digital environment. 10. Credit and borrower information should be further expanded to address information asymmetries. Lack of sufficient information on individuals and MSEs remains a major challenge, preventing financial service providers from properly assessing potential borrowers, and hence limiting access to credit. Fintech providers are developing separate credit information platforms, indicating gaps in the system and leading to fragmentation of information. While the PBC’s Credit Reference Center has achieved impressive coverage of individuals and MSEs, efforts should be made to further expand coverage to regulated digital credit providers. Policymakers should encourage the growth of the credit bureau industry, including by establishing a clear framework for licensing and oversight. Credit scoring and data analytics companies should be distinguished from credit bureaus, and appropriate credit reporting principles applied to such companies. 11. A comprehensive legal framework for data protection and privacy should be developed that provides clear rules on data ownership, sharing, usage, and informed consent. Such a framework is currently lacking in China. Given the rapid innovation in the use of alternative data such as transactional and social data in China, there is a unique opportunity to develop a cutting-edge regulatory approach towards big data and alternative that appropriately balances innovation with data privacy and other consumer protection concerns. A legal framework for data protection would also facilitate access to public information on potential customers from various government agencies, which has the potential to provide a wealth of information for providers but is currently difficult to access. 12. The strategic approach towards rural cash withdrawal points should be reformulated to allow such points to become more functional, sustainable, and fully leveraged. China currently has the largest agent network in the work, with nearly one million POS devices located in retail stores and local commune offices in rural areas providing payment 3 services to rural residents. But many service points have very low transaction volumes. Policymakers should shift away from strongly encouraging universal coverage in all villages and instead provide space for providers to innovate to achieve greater commercial viability, for example by selectively expanding the functionality of service points and relaxing restrictions so that they can operate as full-fledged agents providing a broader range of services, complemented by necessary safeguards. 13. Finally, it should be noted that long-term progress in financial inclusion can only be achieved in the context of broader financial sector reform, and can also contribute to such reform. Market-based approaches to financial inclusion will only be effective if current distortions in the market that create biases towards serving large corporations and local governments are also addressed at the same time, freeing up savings to be allocated to the underserved. Efforts to improve financial inclusion will be ineffective if considered in isolation, and should be designed in a manner that takes into account broader market forces. On the flip side, improved financial inclusion can play a role in helping to further rebalance the economic growth model towards a more efficient allocation of China’s high levels of savings. 14. Table 1 provides a high-level summary of the main policy recommendations included in the Technical Note. Table 1. Summary of Main Policy Recommendations Responsible Time Priority Recommendation Authority Frame Public and Private Sector Commitment • Ensure modern conceptualization of financial inclusion PBC, CBRC MT High incorporating market-based, commercially sustainable principles held by all stakeholders • Shift role of government from “guiding” financial service providers to improving the enabling environment for providers • Make clearer distinctions between the promotional role versus the regulatory and supervisory role of financial sector regulators, and between poverty alleviation efforts and financial inclusion efforts • Establish a high-level entity to provide strategic vision and State Council, ST/MT High leadership on financial inclusion and empowered to make PBC, CBRC, broad decisions at an ecosystem-wide level CSRC, CIRC, • Enhance active coordination and joint decision-making MOF, NBS, local between national agencies government, • Set more detailed strategic framework at national level, other ensure consistent implementation at local levels, and stakeholders enhance capacity at both levels • Develop and operationalize a robust M&E system for PBC, CBRC, ST High financial inclusion, including strengthening the supporting MOF, NBS, local data infrastructure to improve the scope, consistency, governments frequency, and availability of data relevant to financial inclusion 4 • Enhance performance monitoring and evaluations of promotional policies and programs to better understand the precise impacts of these actions as well as remaining bottlenecks to financial inclusion Provider Diversity and Sustainability • Improve the corporate governance of ABC and PSBC to CBRC MT Medium ensure they have the needed incentives, institutions, and systems to service rural households and SMEs • Further reform RCCs to address the weak corporate CBRC, local ST Medium governance with the objective of transforming RCCs into financial offices commercially-oriented providers without undue interference by local government • Reform regulatory frameworks to ease constraints for well- CBRC ST Medium performing VTBs and MCCs on geographic expansion, funding, and ownership requirements that currently prevent them from achieving economies of scale and reaching their full potential in serving underserved clients, combined with higher standards of supervision for such providers Fintech • Encourage and facilitate partnerships between fintech PBC, private MT Medium companies and traditional providers sector • Expand and improve ICT infrastructure in rural and remote PBC, Ministry of MT Medium areas Industry and Information Technology • Undertake demand-side research to better understand PBC, NBS MT Medium behavior and needs of “last mile” consumers • Move from “wait and see” approach to deliberate, PBC, CBRC, ST High proactive regulation and supervision of fintech CSRC, CIRC, • Develop clear, comprehensive legal and regulatory National Internet framework for fintech that is aligned with the current Finance financial sector legal framework, applies consistent rules Association by type of activity, addresses cross-cutting issues related to fintech, and addresses issues specific to particular fintech models • Ensure balance between mitigating risks and allowing for healthy growth and innovation • Strengthen risk-based supervision of fintech and improve PBC, CBRC, ST High coordination across regulators CSRC, CIRC, • Consider tiered approach to largest P2P platforms and National Internet online MCCs with direct supervision by national regulators Finance • Clarify role of National Internet Finance Association of Association China in regulating and supervising fintech • Enhance financial consumer protection legal framework for PBC, CBRC MT Medium fintech-related issues, including developing rules on product suitability and responsible finance Credit and Borrower Information • Expand coverage of PBC’s Credit Reference Center to PBC MT Low include regulated digital credit providers 5 • Encourage growth of credit bureau industry, including by PBC ST High establishing clear framework for licensing and oversight • Distinguish data analytics companies from credit bureaus and clarify the general credit reporting principles applicable to such companies • Establish clear policies at a national level regarding what PBC, State MT Medium information can be shared by government agencies and Administration facilitate efficient and cost-effective access to such for Industry and information Commerce, local governments • Develop a comprehensive legal framework for data State Council MT High protection and privacy • Consider implications of big data and alternative data and appropriate application of data protection principles to such data, with adaptations for any particular risks posed by such data Provider Reach • Encourage and facilitate strategic use of low-cost, manned PBC ST High outlets in a commercially sustainable manner • Undertake comprehensive assessment of rural cash withdrawal points to gather granular data, better understand current practices and models, and provide for segmentation of such points by potential, with appropriate strategies developed by segment type • Selectively expand functionality and relax restrictions on PBC MT Medium rural cash withdrawal points to allow for operation as full- fledged agents providing broader range of services, complemented by necessary safeguards • Encourage innovative approaches to combine service points with non-financial services and increased shared use of service points by multiple providers, including non-bank payment providers • Assess whether AML/CFT framework poses undue PBC MT Low obstacles to remote identification Product Range and Appropriateness • Conduct targeted, demand-side research on accounts, PBC, CBRC, ST/MT Medium savings, insurance, and digital finance to better understand CIRC the needs and barriers of remaining underserved segments 6 II. CURRENT STATE OF FINANCIAL INCLUSION A. Access to Finance for Individuals 15. Over the past years, China has made significant and impressive progress in expanding account penetration. According to the World Bank’s Global Findex, 79 percent of Chinese adults (age 15 and above) had a store-of-value transaction account2 in 2014, representing an increase of 15 percentage points since 2011 (Figure 1). It is worth noting that the PBOC currently estimates that account ownership by adults China is above 90 percent. Nevertheless, Global Findex data is used in this report in order to undertake cross-country and time-series benchmarking analyses. Account penetration in China exceeds the 64.7 percent average in other BRICS economies and the 57.4 percent average in upper middle income economies, though it is lower than the 90.6 percent average in high-income economies. China is ahead of its peers with the same level of GNI per capita in terms of account penetration. Account penetration in China is significant thanks to simple documentation requirements, low account opening and maintenance costs, and easy access to physical branches and access points, combined with the expansion of government-to-person (G2P) transfers via bank accounts. Figure 1: Ownership of transaction accounts over time % adults (age 15+) reporting ownership of a store-of-value transaction account 100 78.9 84.8 90.6 80 63.8 70.5 69.0 64.7 57.4 55.1 61.5 60 48.2 50.6 40 20 0 China Upper middle High income East Asia and BRICS(excluding World income countries countries Pacific China) (developing only) 2011 2014 Source: Global Findex 2014, World Bank 16. Despite impressive progress, challenges remain regarding reaching the “last mile” - the poor and rural population that report particularly high levels of financial exclusion. Nation-wide, rural residents, the poor, the less educated, and young adults have lower levels of account ownership than the national average. Given that the share of the population that is rural remains higher in China than is typical compared to countries with similar levels of per capita income, achieving rural-urban financial inclusion parity continues to be significant challenge in China. 2 A “store-of-value transaction account” refers to an account with both saving and payment functionalities. This term is shortened to “transaction account” in the rest of the report. There are two types of transaction accounts in China: bank settlement accounts and payment accounts. 7 Figure 2: Transaction accounts in China, by individual characteristics % adults (age 15+) reporting ownership of a store-of-value transaction account Urban 82.7 Rural 74.3 Older adults (ages 25+) 79.8 Young adults (ages 15-24) 74.2 Secondary education or more 89.8 Primary education or less 72.8 Richest 60% 83.6 Poorest 40% 72.0 Male 81.4 Female 76.4 0 20 40 60 80 100 Source: Global Findex 2014, World Bank Figure 3: Transaction accounts in China, by income group % adults (age 15+) reporting ownership of a store-of-value transaction account 100% 80% 60% 40% Poorest (20%) Poorer (20%) Middle(20%) Richer (20%) Richest (20%) China Upper middle income countries High income countries East Asia and Pacific (developing only) BRICS(excluding China) World Source: Global Findex 2014, World Bank 17. Considerable gaps exist between access to and usage of financial services such as transaction accounts and debit cards. According to Global Findex 2014, 11 percent of account holders in China report not making any deposits or withdrawals from their account in the past year, similar to the average across upper middle income countries. The gap between access to and actual usage of accounts is higher for the poor in China. According to Global Findex 2014, 17 percent of adults report having used a debit card in the past year (Figure 4). Lower usage of debit cards may be partly attributable to the increased use of other digital payment instruments, as discussed below. 8 Figure 4: Usage of debit card, by income group % adults (age 15+) reporting use of debit card 60 50 40 30 23.1 17.3 20 8.6 10 0 Use debit card in the past 12 months National Richest 60% Poorest 40% Source: Global Findex 2014, World Bank 18. The Chinese are increasingly using digital instruments for transactions, especially in urban areas. Overall, 31 percent of Chinese adults reported using a credit card, debit card, internet platform, or mobile platform to make or receive payments in 2014. While on par with the corresponding rate in upper middle income countries and above that of East Asia and Pacific (EAP) developing countries, the prevalence of digital payments in China is significantly lower than that in high-income countries, where 75 percent of adults report having used a digital instrument to make a payment in the past year (Figure 5). With respect to payments using a mobile phone, Chinese adults are almost as likely as high-income economies to report a mobile- based transaction (14 and 16 percent, respectively). Innovative digital payment products via internet or mobile platforms have particularly grown in usage in recent years. Figure 5: Digital payment instruments % adults (age 15+) reporting ownership of selected electronic payment instruments 80 75.3 64.2 60 45.0 38.7 35.1 35.6 40 31.0 32.0 28.3 27.8 22.4 20.6 19.2 14.9 17.3 14.3 15.8 17.2 13.8 14.8 20 13.3 10.98.8 11.0 5.8 8.2 9.5 7.6 8.9 7.3 0 Digital payment Use debit card Use mobile phone Use internet payment Use credit card instruments China Upper middle income countries High income countries East Asia & Pacific (developing only) BRICS(excluding China) World Source: Global Findex 2014, World Bank 9 Figure 6: Digital payment instruments, by urban and rural % adults (age 15+) reporting ownership of selected electronic payment instruments 40 30.3 30 24.8 24.2 22.7 20 11.0 9.9 10 6.1 6.4 0 Use debit card Use mobile phone Use internet payment Use credit card Urban Rural Source: Global Findex 2014, World Bank 19. Digital payments have been actively used for social transfers from the government. Thanks to the expansion of government-to-person (G2P) transfers via bank accounts, a vast majority of social transfer recipients in China reported receiving these payments directly into an account. According to Global Findex 2014, 66 percent of adults reported receiving social transfer into their account. This is below the high-income countries average of 83 percent, but above the averages of upper middle income countries and developing countries in EAP (58 percent and 52 percent, respectively) (Figure 7). Figure 7: Digital payments use % adults (age 15+) reporting each use case, which do so via digital instrument 100 88.2 83.4 80 59.8 68.7 68.4 65.6 66.2 66.0 59.6 58.5 58.1 58.2 60 52.7 52.0 49.8 52.2 49.7 43.7 49.5 45.5 44.9 48.1 43.4 41.0 37.8 41.6 33.9 40 33.3 19.7 18.5 22.3 21.3 16.1 15.2 13.3 15.0 20 9.9 6.3 9.4 11.7 11.7 5.9 0 Receive Send domestic Receive Pay utility bill Pay school fees Receive wage Receive agricultural remittances domestic payment government payment remittances transfer China Upper middle income countries High income countries East Asia & Pacific (developing only) BRICS(excluding China) World Source: Global Findex 2014, World Bank 20. There are opportunities for further shifting from cash to digital channels to increase access and usage of financial products and services. According to Global Findex 2014, about 13 percent and 5 percent of adults in China that are not bank account holders still pay utility bills and school fees in cash, respectively (Figure 8). Since the choice of whether to pay digitally or in cash often resides with the utility companies and schools, encouraging them to provide incentives for customers to make electronic payments could advance the access and usage of financial services. Likewise, encouraging those 8 percent of adults who are not account holders but are receiving agricultural payments (or 3 percent of adults sending or receiving domestic 10 remittances) exclusively in cash to instead do so through digital transactions could also increase the use of financial products and service. Figure 8: Opportunity for increasing financial access and usage Increased in account penetration (% adults age 15+ has an account) if shifting these payments from cash to account-based 100 92.2 90 84.3 86.6 80.3 81.2 81.3 82.4 78.9 80 70 Status quo Send Receive Receive Receive Paying Receive Pay utility domestic domestic government wage school fees agricultural bill remittances remittances transfer payment payment Source: Global Findex 2014, World Bank 21. Savings behavior is relatively prevalent among Chinese adults, and the majority of savers use formal financial institutions to save. Approximately 72 percent of Chinese adults report having saved or set aside money in the past year, significantly higher than adults in high income- and upper middle income countries but in line with the average rates in develping countries in EAP, suggesting a common cultural trend across the region (Figure 9). Figure 9: Savings % adults (age 15+) saving 100 80 72.1 67.0 71.0 62.7 56.5 60 41.2 46.7 43.3 32.2 36.5 40 27.4 18.7 20 0 China Upper middle High income East Asia & BRICS(excluding World income countries countries Pacific China) (developing only) Saved any money in the past year Saved at a financial institution Source: Global Findex 2014, World Bank 22. However, savings is comparatively low for rural residents and the poor in China. Gaps remain in levels of formal savings between the richest and poorest adults. For the poorest 20 percent of adults, 21 percent saved at a financial institution. This is higher than the average of 10 percent in both upper middle income economies and other BRICS economies, but significantly lower than the average of 31 percent in high income economies. Consistent with this picture, levels of formal savings in rural areas is much lower than that in urban areas (Figure 10). This suggests the need to develop simple, accessible, and low-cost savings products, particularly for individuals with low and irregular incomes. 11 Figure 10: Saving, by urban and rural % adults (age 15+) saving 100 76.6 80 68.3 60 45.5 37.5 40 20 0 Saved any money in the past year Saved at a financial institution Urban Rural Source: Global Findex 2014, World Bank 23. Chinese adults are as likely as adults in high income countries to report having borrowed in the past year, but are less likely to report having done so from a regulated financial institution. Overall, 36 percent of adults in China reported having borrowed money in the past year, with less than a third of those adults borrowing from a regulated financial institution (Figure 11). This borrowing pattern is simliar to that in upper middle income countries, developing EAP countries, and other BRICS. High-income countries are much more advanced, with nearly half of borrowing demands being served through formal institutions. By contrast, family and friends remain the dominant source of borrowing for households in China. Figure 11: Borrowing, by channel % adults (age 15+) borrowing 50 46.1 41.2 42.4 37.7 36.3 38.3 40 31.6 28.3 26.2 30 25.1 24.0 17.3 15.0 20 12.0 10.4 11.0 9.6 10.7 10.2 8.2 9.2 7.9 10 1.12.60.9 2.5 4.6 2.75.0 3.8 0 Borrowed any money Borrowed from a Borrowed from Borrowed from a Borrowed from a in the past year financial institution family or friends private informal store by buying on lender credit China Upper middle income countries High income countries East Asia & Pacific (developing only) BRICS(excluding China) World Source: Global Findex 2014, World Bank 24. Use of formal credit is particularly less common for rural residents and the poor in China. Among rural residents and the poorest 40 percent of adults, formal borrowing is even less common. Just 6-8 percent of these individuals report borrowing from a regulated financial institution, compared to 12 percent for both urban residents and the richest 60 percent of adults. This is despite there being no difference between these groups and the population as a whole with respect to overall borrowing prevalence (Figure 12-13). 12 Figure 12: Borrowing, by income group % adults (age 15+) borrowing 40 35.9 37.0 28.2 30 22.9 20 12.1 10 5.9 3.4 1.1 1.0 1.6 0 Borrowed any money Borrowed from a Borrowed from Borrowed from a Borrowed from a in the past year financial institution family or friends private informal store by buying on lender credit Richest 60% Poorest 40% Source: Global Findex 2014 (Demirguc-Kunt and others, 2015). Figure 13: Borrowing, by urban and rural % adults (age 15+) borrowing 40 34.4 37.9 27.6 30 22.0 20 12.1 7.5 4.0 10 1.0 1.1 1.6 0 Borrowed any money Borrowed from a Borrowed from Borrowed from a Borrowed from a in the past year financial institution family or friends private informal store by buying on lender credit Urban Rural Source: Global Findex 2014, World Bank 25. Coverage of agricultural insurance is yet to be broadly expanded. Within the rural insurance sector, there is an observed tendency to subsidize insurance premiums. While subsidizing premiums may play a role in expanding coverage, persistent use of fiscal subsidies to promote the industry is not a long-term solution, as it tends to exacerbate moral hazards and expose the government to fiscal risks. Beyond agricultural insurance, the availability of a range of insurance products that are commercially sustainable and accessible to underserved segments of the population remains limited. According to the China Household Finance Survey (CHFS), 10 percent of urban residents and only 4 percent of rural residents reported use of any commercial insurance in 2013 (Figure 14). 13 Figure 14: Use of commercial insurance, by urban and rural % adults (age 15+) borrowing 10 9.8 7.5 8 6 3.7 4 2.6 1.4 1.3 2 0.5 0.6 0.1 0.1 0 Any commercial Life insurance Pension (commercial) Property Insurance Other commercial insrance (commercial) (commercial) insurance Urban Rural Source: China Household Finance Survey (CHFS) 2013 B. Access to Finance for Firms 26. Ownership of accounts is virtually universal for firms in China, and access is generally high for basic products and services such as payments services and accounts. Over 95 percent of firms reporting owning a checking or savings account. This compares favorably to high-income OECD and other BRICS economies (Figure 15). Figure 15: Access to account for firms, by firm size % of firms reported owning a checking or savings account 100 98.496.695.1 99.495.195.6 87.983.3 94.596.895.1 94.291.484.3 73.1 80 60 40 20 0 China (2012) BRICS(excluding East Asia & Pacific High income: World China) OECD Larger firms ( 100+ employees) Medium size firms (20-99 employees) Small firms (5-19 employees) Source: WBG Enterprise Surveys 27. However, access to finance is perceived by firms as a significant obstacle to doing business in China. Over 20 percent of Chinese firms identify access to finance as their biggest constraint to do business, significantly higher than reported by peer groups (Figure 16). The World Economic Forum’s Global Competitiveness Index for 2015-2016 also reported access to finance as the second most problematic factor for doing business, after insufficient capacity to innovate.3 3 The Global Competitiveness Report 2015-2016 assesses the competitiveness landscape of 140 economies, providing insight into the drivers of their productivity and prosperity. 14 Figure 16: The obstacles in doing business for firms, by firm size % of firms consider access to finance the biggest obstacle 30 24.4 20.1 21.7 20 16.5 12.8 12.7 11.9 13.5 9.4 11.1 11.0 11.6 10.7 8.8 8.6 10 0 China (2012) BRICS(excluding East Asia & Pacific High income: OECD World China) Large firms ( 100+ employees) Medium size firms (20-99 employees) Small firms (5-19 employees) Source: WBG Enterprise Surveys 28. There is a considerable gap between supply and demand for financial services for small enterprises in particular. While the number of Chinese firms that demonstrate demand for credit is on par with other countries, actual access to credit for SMEs is far more limited. For example, among the 49.6 percent of small firms with 5-19 employees that reported having financing needs, only 23.4 percent of such firms had a loan or line of credit (Figure 17). This proportion of small firms with unmet credit demand was higher than any of the comparison groups. In contrast, 68.2 percent of large firms with over 100 employees that reported financing needs had a loan or line of credit, a rate similar to other countries. Figure 17: Credit demands firms were met, by firm size % of firms with a bank loan/line of credit, among firms report needing a loan 100 86.3 68.2 71.6 72.2 68.3 67.3 80 54.9 60.4 58.7 49.0 50.8 44.346.7 50.9 60 38.2 36.8 40 23.4 29.4 20 0 China Upper middle High income Developing EAP BRICS excluding World income countries countries China Large firms (100+ employees) Middle-sized firms (20-99 employees) Small firms (5-19 employees) Source: WBG Enterprise Surveys 29. Data from WBG Enterprise Surveys indicates that business investments in China are mostly financed internally. Small enterprises in China are particularly reliant on internal finance. For purchasing fixed assets, 92 percent of capital came from internal funding, significantly higher than the average of 76 percent in other BRICS countries and 70 percent in high-income countries (Figure 18). Not surprisingly, the proportion of capital coming from banks, equity, and trade credit are all comparatively low. 15 Figure 18: Funding sources for investment, by firm size Proportion of investments financed by supplier credit (%) China China) Pacific OECD World Small firms Large firms excludi East incom High Medium size firms ng Asia & e: Small firms Large firms BRICS( Medium size firms Small firms Large firms 0% 20% 40% 60% 80% 100% Internal finance Banks Trade credit Equity Others Source: WBG Enterprise Surveys III. PUBLIC AND PRIVATE SECTOR COMMITMENT A. Conceptualization of Financial Inclusion 30. While financial inclusion has become an increasingly mainstream concept in China, not all stakeholders have a modern conceptualization of financial inclusion. For many stakeholders in China, financial inclusion remains synonymous with subsidies and preferential policies, directed lending, and charitable activities. Many stakeholders believe that only a narrow range of financial products are relevant to financial inclusion (e.g. credit), targeted at farmers and the rural poor. Significant variations in the understanding of financial inclusion across local and national government authorities, financial service providers, and non-governmental organizations has often led to confusion, conflicting messages, and counterproductive results. 31. Expanding beyond the current achievements in financial inclusion requires a more modern conceptualization of financial inclusion shared across all stakeholders. Financial inclusion can be defined as access to and usage of a range of appropriate financial products and services by the underserved, delivered in a responsible and sustainable manner. Financial inclusion involves a range of products, including current accounts, payments, savings, insurance, and credit. Underserved segments typically consist of low-income individuals (both urban and rural areas), micro and small enterprises (MSEs), and those in remote, “last mile” areas. 32. To achieve high levels of financial inclusion over the long-term requires shifting to a market-based, commercially sustainable approach. The gains in financial inclusion in China to date, while substantial, have primarily been driven by government mandate and encouragement. To achieve further gains requires internal motivation by the market (as well as counter-acting the historical bias towards channeling high savings at low costs to certain sectors deemed strategic). Indeed, the new Plan for Advancing Financial Inclusion Development (2016- 16 2020) (“FIP”) approved by the State Council in December 2015 refers to “commercial sustainability” as one of the main guiding principles of the FIP, a significant step forward in the thinking of policymakers. However, there appears to be misunderstanding regarding what commercial sustainability entails. Many stakeholders still believe that it is too costly to serve underserved segments, and that financial institutions seeking to do so must be heavily subsidized in order to be “commercially sustainable”. 33. This belief is contradicted by the reality in China, where underserved segments are increasingly being reached on a commercially sustainable basis. The fintech4 revolution in China has succeeded in reaching hundreds of millions of mass retail consumers with low-cost, better tailored products and services. The experience of fintech companies has shown that low- income consumers can be served on a commercially sustainable basis if adapted processes are used. The fintech boom also revealed significant long-standing, unmet demand among mass retail consumers in China. Similarly, the international experience with microfinance over the past decades has shown that small-value loans to the underserved can be provided on a commercially sustainable basis using adapted lending methodologies, without subsidies or preferential treatment. B. Appropriate Role of Government 34. Heeding the principle of commercial sustainability requires a substantial shift in the role of government with respect to financial inclusion. The Chinese government has traditionally taken a direct approach to financial inclusion, with national and local governments utilizing a variety of policies, regulations, and promotional measures to advance financial inclusion. While there has been welcome movement away from direct intervention in the market, the primary default approach of policymakers is to “guide” or encourage financial service providers to take a wide variety of actions, such as expanding physical outlets in rural areas, increasing the number of loans disbursed, developing innovative financial products and services, and setting up financial inclusion and MSE finance departments (to name just a few examples). 35. The approach of the Chinese policymakers in “guiding” financial service providers to achieve financial inclusion conflicts with the principle of commercial sustainability. It is also questionable how effective such an approach is. For example, while an impressive number of rural cash withdrawal points have been established throughout China, many of these points have very few transactions (e.g. one transaction every few days). Providers may develop a long list of “innovative” products ostensibly for financial inclusion purposes, but with limited uptake. This approach leads to an overemphasis on quantity over quality, on actions that do not have clear impacts on financial inclusion. Further, actions that are not commercially sustainable or market-based ultimately do not have long-term benefits for financial inclusion. 4 Definitions of fintech vary. For purposes of this technical note, fintech can be broadly considered to include technologically enabled financial innovation that could result in new business models, applications, processes, or products in financial markets. See “FinTech credit: Market structure, business models and financial stability implications,” Committee on the Global Financial System (CGFS) and the Financial S tability Board (FSB), May 2017. 17 36. The more appropriate role of government is to improve the enabling environment (as well as to correct market failures and distortions). This requires a critical shift in thinking, from the government “guiding” providers to serve underserved segments in a commercially viable manner, to the government actively working to improve the enabling environment so that providers can serve the underserved in a commercially viable manner. Improving the enabling environment entails developing an understanding of existing and potential business models in the market, identifying the main obstacles to providers and the key factors behind market failures to reach the underserved on a broad, sustainable basis, and addressing such obstacles and failures. The Chinese government has already made substantial improvements in the national payments infrastructure and credit reporting system. Key remaining obstacles in the enabling environment in China include the legal and regulatory frameworks for agent-based models, rural financial institutions (RFIs), and “new-type” providers, information and communication technology (ICT) infrastructure, and further enhancements in credit infrastructure.5 37. Focusing on the enabling environment will ultimately have greater long-term impact on advancing financial inclusion. Improving the enabling environment has macro level, market-wide effects, helping all providers more feasibly reach the underserved, rather than the micro effects of smaller scale and ad hoc government policies and programs. Improvements in the enabling environment also provide the necessary motivation and commercial incentives for providers to innovate and develop new processes and products to reach underserved of their own volition, not due to the encouragement of government. These market-based incentives will have greater tangible impact and scalability than efforts driven by a sense of corporate social responsibility, which will ultimately result in more of the underserved being reached. 38. Clearer distinction should be made between the promotional role and the regulatory and supervisory role of financial sector regulators. The regulatory and supervisory role is obviously the core role of the financial sector regulators. Ensuring the healthy operation of all financial service providers and the protection of consumers are complementary goals to financial inclusion. For example, where used for financial inclusion promotion, supervisory and prudential tools should be carefully and appropriately designed and calibrated (see Paragraph 66 below). 39. Clearer distinction should also be made between poverty alleviation and financial inclusion. Again, these two goals are related and complementary, but are distinct. Financial inclusion can be considered as one tool to help address poverty alleviation, but it should not be considered the equivalent of poverty alleviation. Financial inclusion primarily focuses on underserved segments that are economically active (i.e. “bankable”). Greater care should be paid to distinguishing when the primary objective is poverty alleviation for “unbankable” segments. In such particular, limited circumstances, more direct policy support for financial services may be warranted, along with other social programs. 5 For further examples, see the G20 Principles for Innovative Financial Inclusion, Global Partnership for Financial Inclusion (GPFI), 2010, which highlight key improvements in legal and regulatory frameworks and financial and ICT infrastructure that can help to advance financial inclusion. 18 C. Institutional Capacity and Coordination 40. Strong national leadership is needed as the role of government shifts to focus on improving the enabling environment. Improving the enabling environment will require high- level, strategic decisions and coordinated action and consensus across different agencies. Such efforts suffer when too many key actions are decentralized. 41. There is currently a lack of a clearly established lead in financial inclusion with a broad mandate. The Financial Consumer Protection Bureau of the PBC has a mandate to promote financial inclusion, while the CBRC also has a Financial Inclusion Affairs Department. Both the PBC and the CBRC have been designated as “co-leads” for implementation of the FIP, in conjunction with a range of government authorities. While this co-leadership arrangement may be conducive to drafting action plans and delegating tasks (which has already been undertaken), it appears to be less conducive to tackling larger systemic issues with respect to financial inclusion in China. 42. A high-level body that is empowered to make broad, strategic decisions should be established in order to take necessary action at an ecosystem-wide level. A formal entity or mechanism is needed that sits above financial sector regulators and other agencies, is chaired by a high-powered individual, can provide strategic vision and thought leadership, and has the authority to make decisions on issues that are cross-sectoral, related to broader infrastructure, or entail national/local coordination. 43. Inter-agency coordination should be strengthened. There are numerous scattered policies, programs, and initiatives with respect to financial inclusion across multiple agencies. These efforts should be better aligned towards achieving common objectives, and not duplicative or at cross-purposes. While a cross-agency supervisory coordination mechanism exists, it appears that such mechanism is used primarily for information sharing and communication purposes, for example for soliciting inputs on draft guidance or policies. Such a mechanism is less well-suited for active discussion, joint decision-making, and ensuring coordinated action. Policymakers should consider establishing a cross-agency working group that meets regularly to share updates and discuss joint coordinated approaches to cross-cutting issues. Alternatively, the cross-agency coordination mechanism to implement the FIP should continue to be strengthened, which is expected to occur given broader financial sector reforms planned for the near future. 44. Improved coordination and communication between national and local government authorities is also critical. Given the unique characteristics of China, it is inevitable that many decisions and actions will need to be delegated to the local level. However, with respect to financial inclusion, the extent to which decisions regarding strategic approach are delegated to local levels is potentially problematic, particularly given the fact that local authorities are more likely to have an outdated conceptualization of financial inclusion and objectives and approaches of national and local authorities may not be fully aligned. The FIP is a good example. The FIP covers a range of important topics, but at a high-level. There are few details on how topics will be approached and what actions are to be taken. The FIP calls for all local governments to formulate specific implementation plans based on local conditions, an effort which is currently underway. While national authorities provide some guidance to local governments in this 19 process, it appears that local authorities have considerable autonomy in developing local implementation plans. There is a potential concern with inconsistent approaches across local levels that do not follow best practices or market-based approaches. 45. National authorities should ensure that best practices and a modern conceptualization of financial inclusion are fully reflected in local financial inclusion efforts. The FIP is an important opportunity to implement more coordinated, impactful approaches nationwide. In order to achieve this, a greater degree of strategic decisions should be made at the national level, in order to create a fuller framework that can then be implemented at the local level. Certain topics that clearly call for national-level action, such as improvements in credit infrastructure, should be led by national level authorities and not duplicated at local levels. Topics that call for national and local level coordination, such as improvements in ICT infrastructure, should be more closely coordinated. Capacity at the national level should be enhanced to undertake these tasks. 46. Coordination should also be improved at the local level. There appears to be duplication of efforts across local level government agencies. Clearer, more detailed guidance and structured approaches from the national level may help in this respect, as well as closer communication and collaboration. Capacity at local levels should also be enhanced given the extent of responsibilities delegated to the local level. 47. Finally, greater efforts should be made to engage with the private sector when developing financial inclusion strategies and approaches. In order to better understand the market and to develop appropriate regulatory frameworks, it will be important for policymakers to communicate more actively with providers. For example, as CBRC has quarterly meetings with providers, this forum could be used to listen and learn from market players. D. Financial Inclusion M&E and Data Infrastructure 48. The collection of financial inclusion data in China has been scattered across multiple agencies, though recent improvements have been initiated. In line with their 2014 Maya Declaration Commitment, PBC has designed indicators to closely monitor financial inclusion in selected areas, such as agriculture, farmers, and rural areas. CBRC has also started to collect information on coverage, accessibility, and satisfaction of MSE lending, based on data collected via the CBRC off-site surveillance information system. In addition, CBRC keeps track of the coverage ratio of financial service outlets to monitor the coverage of financial services in towns, townships, and villages, with data submitted by banks. As part of the FIP, PBC and CBRC have started to coordinate with CSRC, CIRC, and MoF, as well as with local governments, to consolidate and produce one single set of core financial inclusion indicators. A Financial Inclusion Indicators System was officially established at the end of 2016, with 51 indicators across three main dimensions (availability, usage, and quality of financial services). 49. While the range of data collection efforts by PBC, CBRC, and other agencies provide a useful picture of financial inclusion in China, further improvements can be made to enhance the comprehensive indicator system. More information on usage of financial services, such as usage of transaction accounts or bank cards, should be collected. Usage indicators to consider include: (1) percent of accounts with a zero balance, (2) percent of 20 accounts with no deposits or withdrawals in the past 30 days, or (3) percent of accounts with no deposits or withdrawals in the past 12 months.6 Reliable measurements on new fintech services are could be enhanced, particularly regarding the degree to which they are reaching underserved market segments. More information should also be collected on the productivity, efficiency, and financial sustainability of RFIs and “new-type” providers, credit guarantee systems (CGSs), and rural cash withdrawal points. Finally, while it is good that the authorities have started to collect data on satisfaction with financial services, current indicators related to “approval” should be further refined. 50. The measurement of MSE and rural lending should be improved to be in line with international best practices and consistent between agencies. The definition of MSME in China is still too broad, even though it was revised in 2011. For instance, thresholds for small and medium-sized enterprises in the manufacturing sector are 300 and 1,000 employees respectively, which are much higher than in comparison countries. This makes it difficult to track loans to actual small enterprises and leads to overestimation of lending to SMEs. Definitions for rural lending are broad as well, covering everything related to agriculture, farming, and rural development. These overly broad definitions of MSMEs and rural lending run the risk of misleading policymakers in strategic decision-making and the allocation of public resources towards the larger end of the scale. In addition, the indicator system for financial inclusion should be expanded beyond the traditionally narrow set of “agriculture, farmers, rural” segments to cover broader low-income categories, including urban consumers. 51. Greater emphasis should be put on improving the underlying data infrastructure, including regular collection of both supply-side and demand-side data. Efforts should focus on improving the scope, consistency, frequency, and availability of data relevant to access, uptake, and usage of financial products and services. However, efforts to collect data from financial institutions must also recognize the burden of reporting requirements, particularly for institutions with low technical and technological capacity, and balance these considerations with the value of the requested data. Regular collection of both supply-side and demand-side data are important. Until now, there have been few robust and comprehensive demand-side surveys to date on access to financial services of households and MSEs, including the China Household Finance Survey conducted by the Southwestern University of Finance and Economics. However, additional data collection and research efforts are required to better understand the needs and barriers of remaining underserved segments, as well as to monitor for potential over- indebtedness. 52. Standardization of data and indicators within and across regulatory agencies is needed. For example, PBC and CBRC use different classification systems for financial sector providers. The result is that financial sector authorities are speaking different languages when it comes to assessing progress or challenges, which makes product dialogue and consensus on cross-cutting issues difficult. Efforts to standardize data and indicators would potentially be a role well-suited for any higher-level entity established on financial inclusion, in order to ensure 6 For a further example, see G20 Financial Inclusion Indicators which includes an indicator on “high frequency of account use”, defined as taking money out of an account three or more time s in a typical month (at http://databank.worldbank.org/data/download/g20fidata/Indicators_note_formatted.pdf). 21 that all stakeholders involved in the collection of relevant data coordinate so that common definitions are used, data is not duplicative, and indicators jointly and comprehensively cover policy priorities and important market developments. 53. Authorities should improve the availability and dissemination of financial inclusion data. This can be achieved through either an online data portal or regularly published reports, or both. While both CBRC and PBC publish reports on financial inclusion, basic information such as size and numbers of household loans and branch penetration with breakdown by different types of provider remain unavailable to the public. More transparency can bring the needed monitoring from the public, helping in better designing policies. In this regard, policymakers should consider resuming the Rural Finance Service Map program previously initiated by CBRC. 54. China would benefit from a comprehensive monitoring and evaluation (M&E) system for financial inclusion to support evidence-based policymaking. Until now, impact assessments, results, and data on promotional government policies and programs has largely been missing. The M&E framework should be enhanced at the national, institutional, and program level to allow policymakers, regulators, and program managers to monitor progress over time through key performance indicators, associated with quantitative targets where possible. This would allow tracking progress of implementation and establishing a feedback loop through which modifications and improvements of policies can be made. The authorities should consider establishing a bureau or similar entity to coordinate and assess the impact of different financial inclusion programs across multiple agencies. The assessment should not only track inputs or outputs (such as the product volumes), but should be measured at the individual consumer-level and be linked to outcomes relevant to consumer well-being. Given the common use of “pilots” in China, there has been a particular missed opportunity to rigorously evaluate emerging policy/program approaches before scaling up. 55. Demand-side instruments should be more fully leveraged in M&E systems and policy discussions. Demand-side data gives a clearer picture of the distribution of financial services across consumer groups, shows how different financial and nonfinancial behaviors relate to one another, and also addresses double-counting and product dormancy issues common to supply-side data. Existing demand-side instruments, such as China Household Finance Survey, should be fully utilized in this regard. E. Government Financial Inclusion Policies and Programs 56. A wide range of fiscal, monetary, and supervisory policies and programs have been adopted across agencies to promote financial inclusion. Fiscal subsidies have been actively used at all levels of government to promote financial inclusion, ranging from exemptions of value-added taxes to fiscal rewards to financial institutions delivering financial services to rural households or MSEs. Monetary instruments, including lending facilities and differential reserve ratios, have also been used to encourage lending. The regulatory authorities have also adopted differentiated regulatory policies toward MSME lending. 22 57. Progress has been achieved in better use of promotional policies. Authorities have generally shifted from subsidizing interest rates to subsidizing transaction costs, as reflected in the fiscal support for setting up bank outlets in financially deprived locations and tax concessions based on actual performance in rural lending. Some fiscal subsidies target certain financial services rather than specific financial institutions, an improvement from earlier programs. The targeting of fiscal subsidies has also improved, as evidenced by the eligibility criteria of small loans under RMB 100,000 as the threshold for value-added tax exemption. Certain subsidy programs have also been consolidated into a special financial inclusion development fund. 58. However, in general, government policies and programs remain fragmented and piecemeal, and their impact is rarely assessed. Official statements tend to emphasize inputs and outputs, such as the number of programs and the amount of subsidies disbursed, rather than results, including the actual number of loans disbursed, customers benefited, and even more useful, welfare improvements for the targeted underserved segments. There is insufficient information on whether government programs have reached and benefited intended target beneficiaries and how various policies complement one another, both between line ministries as well as between national and local governments. As such, the real effectiveness and impact of various government policies and programs in promoting financial inclusion is quite unclear. 59. Promotional government policies and programs should be designed and implemented in an evidence-based manner. The government should consistently track and analyze impact in order to design more effective policies and programs that are better targeted and less distortive. For every subsidy program, quantitative results should be tracked at a minimum. Preferably, rigorous impact assessments should be conducted on select programs in order to provide reliable information on effectiveness, efficiency, and real impacts on consumers and the market. Robust monitoring systems should be embedded into all initiatives in order to generate quantitative, consumer-centric information on implementation progress and establish a feedback loop through which ongoing modifications and improvements can be made. 60. Promotional policies should be limited to only those “last mile” segments for which government support is necessary. The fiscal authorities at both the central and local government levels are faced with pressure from all parties asking for more. In the absence of an integrated and coherent strategy to promote financial inclusion, promotional policies are subject to the vagaries of various interest groups. Furthermore, the heavy focus on subsidized lending impairs the growth of operations that are commercially sustainable over the long-term, and deters the entrance of commercially-oriented providers.7 As noted above, policymakers should focus on improving the enabling environment to allow for market-based approaches, while fiscal policies should be used cautiously. 61. Where such policies are employed, policymakers should ensure that the policies are well-tailored, targeted, and complement market-based approaches. Many subsidies ultimately favor larger and corporate borrowers rather than rural households and MSEs. Failures in targeting bankable but excluded groups results in the inefficient use of public resources 7 Fiscal subsidies disbursed under the special financial inclusion fund and agriculture insurance alone amounted to 32.1 billion RMB in 2016. 23 without reaching those rural households and MSEs that are truly in need. Should the government wish to stimulate credit supply to rural households and MSEs, it would be better to base such awards on the number of loans or customers rather than the total size of the loan portfolio. 62. All promotional policies should be well-coordinated and not duplicative. For instance, local governments usually provide further premium subsidies in addition to that provided by the central government. This brings greater market viability, but also moral hazards such as fraud, breach of contract, and threats to fiscal stability. In this case, exit strategies should be introduced to avoid prolonged and unnecessary fiscal support that waste public resources. 63. Government interventions should be market-friendly and competition neutral. Many subsidy programs are still institution-specific. It is critical that subsidies be available to all financial institutions, or - in the case they are auctioned off - the auction should be accessible to all institutions. Authorities should also ensure that subsidies are accessible to both existing and new providers, such as rural financial institutions and fintech companies. 64. In determining appropriate subsidies for financial sector interventions, policymakers should make a clear cost-benefit analysis and comparison with other interventions that could help rural economies. Finance is only one policy area that helps foster rural development. Empirical analysis has shown that education and research are the most important areas of public investment in terms of effective poverty reductions.8 In China, a large share of government spending has been devoted to promotional functions, while gaps in core public services such as health and social protection remain significant and detrimental to the absorption capacity of any promotional efforts.9 65. Supervisory and prudential tools used for financial inclusion promotion should be carefully assessed and their potential impact on financial stability taken into consideration. Such tools should be utilized in a targeted and appropriate fashion, in adherence with international best practices. International guidance states that certain prudential and supervisory requirements should be stricter for microfinance activities, while other requirements can be lighter as appropriate.10 In particular, policies being considered to allow for greater tolerance of NPLs for MSEs and simplified write-off policies for MSE lending should be carefully assessed to ensure that they are in line with international guidelines. 66. Government subsidies should focus on subsidizing transaction costs rather than interest rates. While subsidies can be used positively to build critical financial infrastructure and to support the income of vulnerable or excluded households, their use for interest rate subsidies is less beneficial. MoF still subsidizes interest rates for poverty reduction loans. There are other 8 See for instance Mahmood, “Rural poverty in developing countries, implications for public policy”, IMF working paper, 2001. 9 See “China 2030: building a modem harmonious, and creative society”, World Bank and Development Research Center of the State Council, 2013. 10 For further details, see “Microfinance activities and the Core Principles for Effective Banking Supervision”, Basel Committee on Banking Supervision, 2010 and “Guidance on the application of the Core Principles for Effective Banking Supervision to the regulation and supervision of institutions relevant to financial inclusion”, Basel Committee on Banking Supervision, 2016. 24 types of microcredit programs still receiving fiscal subsidies on lending interest rates, including those targeting women, laid-off workers, and students. As well-documented in international literature, interest rate subsidies often prevent rural and micro borrowers from obtaining increased access to finance in the long-run, as such policies foster unsustainable business models on the part of existing providers and deter the entry of new players. The effectiveness of interest rate subsidies policies and programs should be evaluated.11 67. Finally, the design of CGSs should be strengthened to improve governance, targeting, and specialization. CGSs in China do not target SMEs effectively, as shown by the large average value of credit guarantees (about 523 million RMB in 2015). There were 7,340 credit guarantee companies in China in 2015. The industry is growing rapidly, now covering about 220,000 MSMEs. However, the loss ratio of CGSs in China jumped sharply from 0.5 percent in 2012 to 2.2 percent in 2015. Most CGSs suffer from poor governance, weak design, and insufficient institutional capacity. Few CGSs have sound credit assessment capacity, including the ability to build rating and scoring systems. Re-CGSs have been set up at both the central and provincial levels. but tend to perpetuate moral hazard in the credit market and generate contingent fiscal liabilities. It will be important to improve the design of CGSs, including improving targeting, tightening eligibility criteria, building up institutional capacity, and improving specialization, for instance by focusing on specific sectors or designated geographic areas. Authorities should undertake a stock-taking and evaluation of different schemes, as well as broader stock-taking of all promotional government interventions, in order to ensure effectiveness and impact. Summary of Main Policy Recommendations: • Ensure modern conceptualization of financial inclusion incorporating market-based, commercially sustainable principles held by all stakeholders • Shift role of government from “guiding” financial service providers to improving the enabling environment for providers • Make clearer distinctions between the promotional role versus the regulatory and supervisory role of financial sector regulators, and between poverty alleviation efforts and financial inclusion efforts • Establish a high-level entity to provide strategic vision and leadership on financial inclusion and empowered to make broad decisions at an ecosystem-wide level • Enhance active coordination and joint decision-making between national agencies • Set more detailed strategic framework at national level, ensure consistent implementation at local levels, and enhance capacity at both levels • Develop and operationalize a robust M&E system for financial inclusion, including strengthening the supporting data infrastructure to improve the scope, consistency, frequency, and availability of data relevant to financial inclusion • Enhance performance monitoring and evaluations of promotional policies and programs to better understand the precise impacts of these actions as well as remaining bottlenecks to financial inclusion 11 See for instance Helms, Brigit, “Interest rate ceilings and microfinance: the story so far”, September 2014, CGAP and Sinha, Sanjay, “Impact of restrictions on interest rates in microfinance”, September 2016, ADB. 25 IV. PROVIDER DIVERSITY AND SUSTAINABILITY 68. There are a range of different types of financial service providers relevant to financial inclusion in China. This includes state-owned commercial banks (SCBs) such as Postal Savings Bank of China (PSBC) and Agricultural Bank of China (ABC) and policy banks such as China Development Bank (CDB) and Agricultural Development Bank of China (ADBC). There are joint-stock commercial banks operating nationwide and regional banks such as city commercial banks (CCBs) and rural commercial banks (RCOMBs). Financial cooperatives include rural credit cooperatives (RCCs), rural cooperative banks (RCOPBs), and member-based cooperatives. New providers operating in rural areas include village and township banks (VTBs), microcredit companies (MCCs), rural village mutual funds, and loan companies. Fintech companies include non-bank payment providers, P2P platforms, internet-based lenders, internet banks, internet-based insurance, internet-based fund management, and internet-based crowdfunding. CBRC is the primary supervisory authority for commercial banks, financial cooperatives, and VTBs, while MCCs are supervised by local government agencies. Oversight for fintech companies appears to be led by PBC but shared among regulatory authorities depending on the type of provider or product, as outlined in the 2015 Guidance and Instructions on Promoting Sound Development of Internet Finance. 69. The market potential for financial service providers to contribute to financial inclusion has yet to be fully realized. China has many financial institutions with enormous savings resources, but their involvement in channeling these resources to productive households and businesses remains limited. SCBs have a historical bias for lending to large corporations and local governments due to government priorities, slow progress in commercialization, and the perception of lower risks in these sectors due to implicit guarantees. More efficient allocation of capital towards the underserved, relying on market forces, can contribute to the broader rebalancing of the economy. Some banks are trying to downscale their lending operations, but they face internal capacity limitations and weaknesses in the enabling environment preventing them from reaching rural and MSE consumers on a commercially viable basis. 70. A multi-faceted approach to improve the institutional capacity of providers as well as the enabling environment will be required to leverage the full range of institutions to contribute to financial inclusion. SCBs remain the largest providers to households and MSEs. The corporate governance of PSBC and ABC should be improved further to ensure they have the necessary incentives, institutions, and systems to effectively serve rural households and MSEs. Further reforms should be undertaken to address weak corporate governance at RCCs, with the objective of transforming RCCs into commercially-oriented providers without undue interference by local government. Regulatory frameworks for VTBs and MCCs should be revised to address constraints on geographic expansion, funding, and ownership requirements that currently prevent such providers from achieving greater economies of scale and reaching their full potential in serving rural clients. 26 Table 2: Main Types of Financial Providers in China Types of Financial service Supervisory Description Number providers provider authority of physical branches Banks Agricultural Bank CBRC One of big-five commercial banks; 23,444 of China (ABC) formerly state-owned, now partly privatized and publicly listed Postal Saving Bank CBRC Originally the postal savings and 39,927 of China (PSBC) remittances bureau, now transformed into a large commercial bank that is partly privatized and publicly listed Agricultural CBRC Policy bank owned wholly by the 2,140 Development Bank government, offering policy-oriented of China (ADBC) agricultural lending Joint-Stock CBRC Second-tier commercial banks operated 15,216 Commercial Banks nation-wide (JSCBs) City Commercial CBRC Regional banks operated at the city level 14,264 Banks (CCBs) Rural Commercial CBRC Reformed from RCCs, usually operating 36,786 Banks (RCOMBs) in both urban and rural areas Financial Rural Credit CBRC Fully privatized and dispersed 38,969 Cooperatives Cooperatives ownership structure, but de facto (RCCs) controlled by the government through party appointment Rural Cooperative CBRC Fully privatized and dispersed 2,842 Banks (RCOPBs) ownership structure, but de facto controlled by the government through party appointment; reformed from RCCs New-Type Village and CBRC One of three “new type” small rural 3,961 Financial Township Banks financial institutions, usually sponsored Institutions (VTBs) by commercial banks Micro Credit Local finance Credit-only institutions that are 11,064 Companies office, guided by registered as companies (MCCs) regulation from CBRC Fintech Peer-to-Peer (P2P) CBRC Online platforms linking individual N/A Companies Platforms investors and borrowers. Some offline/online models. Non-Bank Payment PBC Mobile/electronic payments N/A Providers Internet Banks CBRC Focus on credit-only and online account N/A models Others (e.g. PBC, CBRC, Oversight for fintech companies appears N/A internet-based CSRC, CIRC to be led by PBC but shared among insurance, internet- regulatory authorities depending on the based fund type of provider or product management, crowdfunding) Source: Various government reports and own elaboration 27 A. Banking Sector 71. The banking sector dominates China’s financial sector, with large SCBs playing a major role in financial inclusion. The asset share of the large SCBs is decreasing, but such banks still account for over 45 percent of total banking assets at the end of 2016.12 The top five banks together with PSBC are all state-owned, providing about 34 percent and 28.4 percent of rural household loans and MSE loans in 2015, respectively.13 49 percent of branches nationwide and 41 percent of branches in rural areas were opened by large SCBs. The role of SCBs in terms of bank accounts and debit cards is even more significant. It is estimated that 72 percent of bank accounts and 67 percent of debit cards were distributed and maintained by large SCBs in 2015.14 72. ABC is well-positioned to serve rural markets due to proximity and its extensive network in rural areas, but needs new incentives, institutions, and systems. In recent years, the demand for financial service in rural areas grew rapidly. As a large commercial bank with a mandate for supporting the development of agriculture, rural areas and rural residents, ABC attaches great importance to the business in rural markets as it perceives a natural convergence between its commercial goals and the prospect of serving the rural market profitably. ABC has reoriented itself to focus on rural finance at the county level or below. However, market-based incentives for loan officers should be improved while institutional reforms and capacity-building should be strengthened to ensure that loan targets are met without undermining risk controls and commercial sustainability. Robust and strong risk management and credit risk management procedures are needed to ensure the commercial viability of rural lending. ABC’s experiments in branchless banking hold promise to expand outreach, but efforts should be made to ensure that their practices can be replicated in a sustainable fashion (as further discussed in Chapter VII). 73. PSBC should be further reformed through the establishment of effective corporate governance to ensure it has the needed institutional capacity to provide rural and MSE finance without exposing the fiscal budget to undue risks. With its extensive network, state- of-the art computer system, and ample savings resources, PSBC has great potential as a rural and MSE financial service provider, and has already reached about 17 million clients, with a total portfolio of over RMB 1.3 trillion by the end of 2016. But with its increasing scale and complexity of operations, PSBC has encountered internal constraints in institutional capacity, particularly in risk management and human resources, resulted in rising NPLs. Expanding the operation of PSBC in lending carries enormous risk if proper corporate governance structures, risk management, and human resources are not in place. Establishing stronger corporate governance at PSBC is a crucial pre-requisite before any major expansion in its operations can be contemplated. Further separation of PSBC from postal offices is needed for PSBC to improve corporate governance and avoid unnecessary costs and risks of associated operations. 74. Some joint stock and regional banks have made breakthroughs in downscaling to reach rural and MSE clients, but their impact and sustainability is unclear. The 12 JSCBs and 150 or so CCBs are the closest to being considered as China’s community banks, with the 12 China Rural Finance Service Report 2016. 13 China Rural Finance Service Report 2016. 14 China Payment and Settlement System Development Report, 2016. 28 largest networks and market share in their respective municipal jurisdictions. Some national JSCBs have started to develop small lending products by using a hybrid of credit-scoring and score-card techniques, while others have set up supply chain guarantee arrangements. It is too early to judge whether those practices will significantly expand the outreach of credit to rural households and MSEs in a sustainable manner. 75. Policy banks can potentially play an enhanced role in financial inclusion if their institutional capacity and governance can be strengthened. The role of ADBC has been limited to funding the procurement of grain, cotton, and edible oil, and serving the leading agricultural companies. CDB has sponsored some VTBs and MCCs, and has played a role in introducing the downscaling approach to MSE lending, but their current involvement on financial inclusion are still limited. Policy banks such as CDB and ADBC could potentially play an enhanced role such as apex financiers to further financial inclusion. However, their governance and institutional capacity needs to be strengthened before they are able to effectively fulfill such a strategic role, in order to effectively allocate resources to target groups and conduct appropriate monitoring and oversight of providers receiving funds. 76. While different institutional types of banks have varying levels of potential to contribute to financial inclusion that should be further enhanced, banks should not be given an explicit mandate to do so. The authorities currently mandate every bank to contribute to financial inclusion. For example, each of the large SCBs has established financial divisions dedicated to SME finance. However, many of the SME centers do not actually serve small businesses, as evidenced by the fact that average loan sizes of some SME centers are higher than that for typical MSE clients. Furthermore, some loans distributed are not market-driven but instead politically-driven, leading to unsustainable loans and market distortion. For long-term impacts on financial inclusion, serving underserved segments should be supported by market- based operational strategies and incentives and the necessary capacity and internally-generated motivations. The government should refrain from communicating unrealistic requests to all banks to focus on financial inclusion, as this approach is ultimately counter-productive. B. Financial Cooperatives 77. Given its large outreach, the RCC system remains one of the most important financial service providers in China for rural households, MSEs, and farmers, but is in urgent need of structural reform. By end of 2016, there were 1,125 RCCs, 40 RCOPBs, and 1,114 RCOMBs that were transformed from RCCs. The RCC system distributed 56 percent and 26 percent of rural household loans and MSE loans in China, respectively, and accounted for 59 percent of branches opened in rural areas. While most RCCs have reported improved earnings in recent years, their true financial conditions are mixed and their future prospects uncertain in an increasingly more competitive environment. RCCs have gone through rounds of restructuring with government resources to cover historical losses, but further reforms are needed for them to become resilient, commercially-oriented financial institutions. 78. RCCs should be freed from their ambiguous roles supporting policy objectives and become genuine commercial institutions without the label of “serving sannong”. RCCs have continued to suffer from the conflicting objectives of subsidizing rural credit while at the same 29 time achieving profit-oriented commercial goals. This conflict illustrates the moral hazard problems associated with soft budget constraints. Removing these policy objectives would help RCCs develop into true commercially-oriented financial institutions that provide commercially sustainable financial services to rural areas. Employing their comparative advantage of deep understanding of local consumers, and their relatively low-cost lending methodologies, RCCs have the potential to find their own niche and play an important role in future rural financial markets. 79. The most urgent reform needed for RCCs is to improve corporate governance and address interference by local government. Current practices of local government and regulatory agencies in vetting senior managers should be subordinated to shareholders’ nomination. The shareholders and management should be allowed to execute their responsibilities with full independence under a new governance framework. In return, shareholders and management will take on full responsibility and accountability for the success or failure of the businesses. 80. The reform effort requires repositioning and restructuring provincial RCC unions. There is a need to centralize IT systems and back-office operations to achieve economies of scale, a role that RCC unions could provide. However, the new operational structure of the RCC system should be based on efficiency, which means some form of top-down operations combined with corporate governance based on the bottom-up involvement of shareholders. 81. A “one size fits all” approach in RCC reform should be avoided. There will likely need to be more than one organizational model for RCCs in China, considering the diversified natural, social, and economic conditions in different areas of China and subsequent differences in demand for financial services. It is recommended that each provincial system make a strategic plan on how to transform their existing RCC system. In some cases, the direction might be to develop towards a merged and consolidated system, while in other cases, a two-tier organizational structure may be more appropriate. In some cases, it may be more appropriate to let some large urban-based RCBs, especially in Eastern China, leave the RCC system and operate independently from the provincial RCC union. Regardless of approach, these plans should all include a road map needed to achieve the plan, including capitalization, proposed corporate governance, and capacity-building. 82. Experiments with exit strategies for failing RCCs should be further advanced. Rules should be relaxed to enable weak institutions to be taken over and reorganized. This could include merger and acquisition by more successful neighboring RCC unions allowing successful RCC unions to operate outside of their immediate borders. Exit mechanisms for RCCs should also include liquidation. This is critically important because there is a tendency to over rely on mergers with "healthier" financial institutions, and perceptions of implicit guarantees lead to problems with moral hazards and feed notions that financial institutions cannot fail. As with all initiatives, these experiments should follow the internal logic of commercially sustainable development of the institution, rather than being unduly restricted by administrative boundaries. This would also encourage greater competition and innovation within the greater RCC system. 30 83. Broader experiments in member-based cooperative financial institutions at the grassroots level should also be undertaken to help fill the void in rural financial service availability. There are thousands of village-level, member-based, community development funds (CDFs) promoted by the State Council Office of Poverty Alleviation (SCOPA) and the MoF. There are also millions of specialized rural cooperatives with the potential to provide self-help financial services beginning with small savings and loans functions, though no official information is available on their activities and performance. However, caution is needed in promoting cooperative finance in view of the past setbacks of RCCs and rural credit foundations. The authorities are advised to avoid a top-down approach which may give rise to moral hazard and undermine governance of cooperatives, and instead allow for small-scale pilots and experimentation, with careful tracking of results. C. “New-Type” Financial Institutions 84. Some 10,000 new “new-type” financial institutions, including VTBs and MCCs, have been established in China in recent years, yet their role in broadening access to finance remains limited. More than 11,000 MCCs and 1,400 VTBs have been registered nationwide by the end of 2016. However, allowing new entrants into the rural and MSE financial markets introduced competition but has not brought about needed contestability. For instance, VTBs only account for 4.5 percent and 1.9 percent of total rural household and SME loans outstanding at the end of 2015, respectively. 85. A majority of “new-type” financial institutions do not reach intended target rural households and MSEs, nor are they operating commercial viability. The new-type financial institutions were designed to reach underserved segments, including rural households and MSEs. In reality, many of the new institutions favor large enterprise clients. They adopt credit methodologies that are not distinguishable from corporate banking approaches, nor adapted to new circumstances. The results of a survey of 279 MCCs in China conducted in 2015 show that the average loan amount for sampled MCCs is 1.90 million RMB. The number of loans less than 50,000 RMB account for only 1.5 percent of total loan portfolio.15 There are some poverty alleviation MCCs with average outstanding loan amounts around rural net income per capita, such as those of the China Foundation for Poverty Alleviation (CFPA), however, their outreach is very limited. Moreover, many of the new institutions are not operating on a commercial basis. It is estimated that about one-third of MCCs in China are undergoing financial difficulties.16 The business model for “new type” institutions needs to be examined more closely, and targeted, strategic regulatory reforms undertaken to improve the viability of the business model, as further described below. 86. Further easing geographic, ownership, and investment restrictions may potentially help “new-type” institutions to better achieve commercial sustainability. Most MCCs and VTBs are required to operate within a single county’s boundaries, and are therefore limited to small geographic areas with a small population and are unable to achieve scale. Geographic 15 See Zhang Rui, Chen Yehang and Yang Rubing (2015), “Survey analysis on loan to MCCs – based on sample survey of MCCs in China”, Journal of Finance and Economics, Au gust 2015 16 See speech by the head of the China MCC association as http://www.wdzj.com/news/hydongtai/78980.html. 31 restrictions also prevent sharing risks across regions and products. The existing limit on MCC borrowing to 50 percent of registered capital and fund-raising from no more than two banks makes it difficult for investors of MCCs to achieve an appealing return and for MCCs to grow to scale. VTBs must currently be sponsored by a bank whose share in equity capital must exceed 20 percent, while maximum equity held by any other single non-bank shareholder can’t exceed 10 percent. While the involvement of banks as the principal sponsor may help mitigate financial risks, this rule prevents many potential non-banking institutions from establishing a VTB without cooperating with an already existing banking institution. 87. A tiered approach could be considered to allow expansion of those “new-type” institutions that are in good financial condition and successfully reaching underserved segments. For instance, geographic restrictions could be partially lifted to allow well-performing institutions, especially those MCCs and VTBs successfully serving MSEs, to achieve improved economies of scale and reach their full potential in reaching underserved clients by expanding to multiple counties rather than being limited to one county’s boundaries. Leverage limitations could be eased for qualified MCCs as well, to help them achieve commercial sustainability and to expand their operations. More flexible funding sources for high-performing VTBs could be considered. For instance, private capital could be allowed to establish VTBs as the sponsor under certain conditions. Such appropriately targeted reforms can enable “new-type” institutions to operate more sustainably and efficiently and increase innovation and competition in better serving the underserved. 88. However, introducing more flexible operational and prudential rules should be done in a strategic and controlled manner to avoid potential risks. This easing of various restrictions should be taken in combination with higher standards of supervision for qualified “new-type” institutions. This easing could be introduced via a gradual approach. Geographic expansion would not necessarily be nation-wide, but simply beyond currently limitations on operating in neighboring counties. This tiered approach should be designed in such a manner that would provide proper incentives for other institutions to upgrade their tiers to better achieve policy objectives. 89. The entry of new institutions calls for a strengthening of regulatory and supervisory framework. CBRC has issued nation-wide regulations, but MCCs are registered and supervised by local financial offices. Local governments have regulatory and failure resolution functions over MCCs, but their capacity to carry out those functions is inadequate. There is a need to clarify the responsibilities of local and central governments in oversight over locally registered financial institutions to ensure their safety and soundness in the provision of financial services. A tiered system of MCCs should be considered to allow for risk-based supervision. Large MCCs that are operating nation-wide, specifically online MCCs (as further discussed in Chapter V), should be directly supervised by CBRC. Smaller MCCs may continue to be supervised by local governments, with increased guidance and clear rules from CBRC. Summary of Main Policy Recommendations: • Improve the corporate governance of ABC and PSBC to ensure they have the needed incentives, institutions, and systems to serve rural households and MSEs • Further reform RCCs to address weak corporate governance with the objective of 32 transforming RCCs into commercially-oriented providers without undue interference from local government • Reform regulatory frameworks to ease constraints for well-performing VTBs and MCCs on geographic expansion, funding, and ownership requirements that currently prevent such providers from achieving economies of scale and reaching their full potential in serving rural clients, combined with higher standards of supervision for such institutions V. FINTECH 90. China is at the global forefront of fintech; initially beginning with digital payment services provided by non-bank payment providers, the fintech industry has rapidly evolved to include a range of financial products and services. The growth of fintech in China reflects the increasing digitalization of Chinese society in general, particularly with respect to the pervasive use of mobile commercial applications and social networking platforms. For example, Alipay was launched to facilitate payment on Alibaba e-commerce platforms such as Taobao and Tmall, while Tenpay was used on Tencent’s social media platforms such as WeChat and QQ. Non-bank payment providers now provide both online and offline payments via e-wallet services and have hundreds of millions of customers. These huge sets of transactional data also serve as the basis for providing a broader set of financial products and services. Non-bank payment providers were soon followed by internet lending (e.g. Taobao-based loans, Ant Check Later), internet insurance, and internet fund management. 91. Digitally-enabled business models have also emerged that have further enabled reaching mass retail consumers. After increasing rapidly in recent years, the number of peer- to-peer (P2P) lending platforms in China has now declined to 3,156 as of March 2016.17 A more limited number of internet equity-based crowdfunding platforms have also been established. Online banks WeBank and MYBank have also emerged, leveraging big data analytics to serve customers with credit products at low costs via digital channels. 92. Fintech innovations share several common characteristics, including leveraging network effects, economies of scale, advanced information and communications technology (ICT), and big data analytics. Fintech companies have used big data analytics to provide diversified and customized financial services better tailored to the needs of customers, and cloud computing and other advanced ICT to efficiently expand operational systems while managing risks. For example, as online, network-based companies, both Alibaba and Tencent have taken advantage of the ability to access large existing customer networks and relationships; leverage big data produced via commercial or social interactions to inform product design and delivery and reduce information asymmetries; and realize lower transaction costs and efficient delivery by utilizing existing digital infrastructure and digital delivery channels. 93. Such characteristics have allowed fintech companies to reach millions of previously underserved consumers with new financial products and services better adapted to meet the needs of mass retail consumers, and at lower costs and faster speeds. Consumers 17 PBOC 33 traditionally had few options for investing their wealth, and individuals and MSEs had limited access to credit. Digital technologies have improved the accessibility of low-cost financial products and services, and the huge uptake of fintech products clearly demonstrates that long- standing, untapped demand is being met. For example, non-bank payment providers not only provide payment services to MSEs on e-commerce platforms, but also provide convenient payment channels to consumers in rural areas. Yuebao, a wealth management product that has experienced rapid uptake, allows consumers to earn interest on accounts that can be opened with just RMB 1. Farmers and MSEs can now access small loans via e-commerce platforms and internet banks, and low-income consumers in both rural and urban areas can access convenient financing channels such as Ant Check Later for small consumption loans. 94. Furthermore, the entrance of new fintech companies has increased competition and incentivized traditional providers to innovate and serve segments of the population previously considered commercially unviable. Traditional financial institutions have begun to pay more attention to underserved segments and to adapt their business models and product offerings, such as by developing their own wealth management products or providing online banking portals and e-commerce platforms. Fintech companies and traditional providers are also forming partnerships, which has significant potential to further benefit the underserved in China given the existing branch networks and trustworthy reputations of traditional providers. 95. Fintech has already contributed to substantial gains in financial inclusion, and should be further leveraged to reach the remaining underserved. Moving forward, policymakers will need to develop targeted strategies to more fully leverage fintech for financial inclusion, as further described below. The benefits of fintech need to be extended to a greater extent to reach rural and remote areas. Furthermore, fintech can be used to help further financial inclusion, but it is not the equivalent of financial inclusion, and not all fintech products or companies are beneficial to consumers. The fintech industry initially benefited from a loose regulatory environment and a “wait and see” approach employed by regulators. Guidelines and rules for certain products and services have been introduced after the fact. More comprehensive legal, regulatory, and supervisory frameworks are now needed in order to foster healthy growth while mitigating risks related to fintech. A. Leveraging Fintech to Reach the Underserved 96. While fintech has contributed to substantial gains in financial inclusion to date, such gains have been uneven, particularly in terms of concentration in urban areas (including smaller tier cities). Uptake of fintech products and services such as digital payments has been slower in rural areas, as has outreach of fintech companies. As noted by the Institute of Internet Finance, the level of development of digital finance is negatively correlated to the distance from Hangzhou. Fintech uptake has also been drive by young people born in the 80s and 90s. 97. Partnerships between fintech companies and traditional providers should be encouraged and facilitated. Such partnerships will be key to successfully reaching remote and rural areas, as they enable offline/online models that combine the technological advantages and efficiencies of fintech companies with the physical networks, local knowledge, and financial expertise of traditional providers. Such partnership arrangements and online/offline models are already emerging. For example, Taobao villages have been set up in rural areas, facilitating 34 access to e-commerce services and Alipay. In an effort to move into rural areas, MYBank has partnered with local entities such as agricultural supply chains, Taobao villages, and microfinance NGOs, applying its digitally-enabled models while leveraging local knowledge of customers. In addition, partnerships where traditional providers leverage fintech companies for services such as customer identification and assessment and risk management should be encouraged. 98. Potential barriers to undertaking such partnerships should be identified and addressed. Policymakers should work with the private sector to identify any obstacles (or uncertainties) that the current legal or regulatory framework may pose to innovative partnerships or arrangements. In addition, one barrier noted by fintech companies is the lack of appropriate partners, given the poor IT systems and governance of many VTBs, MCCs, and RCCs. As noted in Section IV, strengthening “new-type” institutions and RFIs would have the further benefit of helping to facilitate greater partnerships. 99. One of the biggest roles that the government can play in expanding fintech to rural and remote areas is by improving ICT infrastructure. Consumers outside of existing digital payment ecosystems/networks enjoy limited benefits from fintech. Rural households have similar numbers of mobile phones compared to urban households, and 83.3% of rural areas currently have broadband.18 While internet and mobile coverage is relatively high in China, further improvements can be made in increasing internet and mobile coverage to reach the “last mile”, improving quality (e.g. expanding broadband and 4G), and decreasing costs. Policymakers should continue efforts to further strengthen ICT infrastructure, which serves as the basis for increased use of digital payment transactions for “last mile” consumers, which then provides access to further products and services. Efforts to expand the retail payments infrastructure into rural and remote areas should also be continued, and access expanded. 100. Demand-side research should also be undertaken to better understand the behavior and needs of “last mile” consumers. To be able to design appropriate products and services for this segment requires a deeper understanding of their behavior and needs, in order to tailor features, functionality, and interfaces more appropriately for such consumers. B. Addressing Risks 101. The advances in fintech in China have been accompanied by numerous risks. These include risks of loss of funds, data privacy violations, poor disclosure, false promotion, and outright fraud. P2P platforms are the obvious example in this area. As the P2P industry expanded rapidly within a loose regulatory environment, certain P2P platforms engaged in illegal activities and Ponzi schemes, with platform operators absconding with customer funds. These risks arose due to low barriers to market entry and the lack of a clear legal and supervisory framework. While ad hoc rules existed for certain types of providers or certain issues, a comprehensive framework was lacking. 18 China Household Finance Survey, Southwestern University of Finance and Economics. 35 102. Policymakers should continue shifting from the previous “wait and see” approach to a more deliberate, proactive approach that allows for innovation while closely monitoring and mitigating risks. While the “wait and see” approach allowed for rapid growth and innovation in the fintech industry, many risks clearly arose due to this approach. In addition, legal uncertainty may have prevented the expansion of certain fintech companies, such as crowd- funding platforms. A clear legal framework and more proactive supervision would facilitate healthy growth of responsible players and help to crack down on unscrupulous providers. The Chinese government has clearly begun this process, launching a high-level initiative in the past two years to begin to address the risks of fintech and curb bad actors. As part of this initiative, the Guidelines on Promoting Sound Development of Internet Finance in July 2015 (“Guidelines”) and Implementation Plan for the Special Campaign to Address Risks of Internet- Based Finance in April 2016 (“Implementation Plan”) were issued by the PBOC in conjunction with a number of other regulatory agencies. The Guidelines provide high-level principles for internet finance. 103. A clear, comprehensive legal and regulatory framework should be developed that replaces the current piecemeal approach towards fintech. High-level principles should be supplemented with clear rules and standards that are comprehensive and enforceable. A variety of scattered rules may cause confusion and will be harder to enforce. The framework should establish basic licensing, governance, and reporting requirements, appropriate capital and liquidity requirements, and permissible scope of activities for fintech companies where such requirements are lacking. The framework should also aim to apply consistent rules for all similar types of activities, regardless of provider type. 104. The legal and regulatory framework for fintech should be aligned with the current legal and regulatory framework for the financial sector. The legal and regulatory framework should be harmonized with the existing legal and regulatory framework for traditional financial service providers, in order to provide a level playing field, avoid regulatory arbitrage, and ensure that consistent and strong standards apply for activities and risks that are the same across traditional providers and fintech companies. For example, rules regarding fraud should apply to all types of financial service providers, fintech or otherwise. Existing investor protection rules should be applied to fintech companies undertaking similar activities to securities firms, adapted as appropriate. 105. The legal and regulatory framework should also address fintech-related issues that are cross-cutting across all types of fintech companies. For example, consistent rules should be developed for the security of customer funds for P2P platforms, investment wealth management funds, and digital wallets. Certain risks that may be heightened by the digital nature of fintech companies and call for enhanced risk management to address cybersecurity, cyberfraud, and business continuity and an enhanced legal framework for data protection and privacy (see Section VII.D.). While the AML/CFT framework should apply consistently across fintech companies and traditional providers, certain adaptations may be required for the nature of fintech delivery channels and products. 106. Clear rules should also be developed to address issues that are specific to the unique risks caused by particular types of fintech models. For example, the nature of P2P and 36 crowdfunding platforms call for specialized rules. Policymakers should also consider certain specialized rules for digital credit. Given that such credit is issued very swiftly to large numbers of individuals with typically short terms, such loans may call for monitoring on a differentiated basis, for example tracking the percentage of loans that are 30 days past due rather than 90 days. Adapted rules should apply across all types of short-term digital credit, regardless of provider type. 107. The legal and regulatory framework should strike a careful balance between mitigating risks while still allowing for growth and innovation. Balanced and proportionate requirements should be put in place that allow for continued growth and stability. Policymakers should consider using tools such as regulatory sandboxes, which provide a safe space for fintech companies to experiment with innovative approaches in a relaxed regulatory environment, over a limited period of time, and in a structured manner with eligibility criteria and safeguards put in place. 108. In addition to strengthening the legal framework, supervision of fintech should be strengthened and better coordinated. Given the cross-cutting nature of many fintech-related issues, communication among financial sector regulators should be strengthened to ensure information sharing and coordinated approaches. Supervision should also follow a risk-based approach, for example by actively monitoring the market and focusing on those products, channels, providers, or business models that pose greater risks. Regulators should consider employing “regtech” solutions to address capacity constraints and effectively supervise large numbers of fintech companies. For example, technological tools can be developed to allow supervisors to monitor internet marketing, track transactions online, or directly link with providers’ systems in order to access real-time data. 109. A tiered approach should be considered where the largest P2P platforms and online MCCs are directly supervised by national regulators. P2P platforms and MCCs are currently subject to dual regulation and supervision by the CBRC and local financial offices, given that such institutions are credit-only. However, local financial offices have limited capacity to supervise such entities effectively. For the largest of such entities that have significant numbers of customers, supervision by CBRC would be more appropriate. Particularly for MCCs, the operations of online, nation-wide MCCs are quite different from locally based, offline MCCs. A differentiated licensing scheme should be developed for online MCCs more appropriate to their nature and scale, with consistent requirements nation-wide. For example, capital requirements and leverage ratios could be increased, although accompanied by more formal supervision, reporting, and risk management requirements. Where P2P platforms and MCCs are supervised by local financial offices, more support and capacity-building should be provided to such offices. 110. The role of the National Internet Finance Association of China (NIFA) in the regulation and supervision of fintech should be clarified. It appears that certain rule-making responsibilities have been delegated to NIFA, as NIFA has been tasked with developing detailed rules and standards, as well as a code of conduct, for fintech companies. NIFA is also developing a risk monitoring system. However, it is unclear how such standards and rules would be enforced, and what legal weight these rules have. Enforcing codes of conduct is often quite challenging as well. The role of NIFA should be clarified. If it is truly intended to act as a self- 37 regulatory body, its capacity will need to be improved and enforcement powers strengthened to ensure that it can effectively manage such a broad and important set of responsibilities. 111. Finally, the legal framework for financial consumer protection and financial literacy initiatives should be enhanced to address fintech-related issues. Rules should be put in place to ensure that consumers are receiving appropriate products suitable to their needs and that fintech products and services are being responsibly marketed. While fintech provides many benefits to the underserved, it also poses a variety of potential risks, including lack of transparency regarding product terms, conditions, and risks and aggressive marketing. Consumers face risks ranging from loss of funds to violations of data privacy to false or misleading promotions. These risks are heightened when dealing with underserved consumers who typically have lower levels of financial literacy (and digital literacy) and may be more vulnerable to fraud. Both financial consumer protection and financial literacy require further strengthening, including rules on disclosure and transparency (including disclosure rules adapted for digital channels), sales practices, data privacy, and dispute resolution adapted for fintech. 112. Policymakers may also wish to consider rules regarding product suitability and responsible finance. Much of the internet-based microcredit currently being offered is short- term consumption finance, oftentimes consisting of unsolicited, pre-approved credit offers sent directly to mobile phones and very easily obtained with one click. As access to credit to mass retail consumers has historically been limited in China, such models may not cause alarm at the moment, but over the long-term, such aggressive models pose a potential risk of overindebtedness, particularly when offers are targeted to lower literacy and more vulnerable consumers. Product suitability rules require providers to gather sufficient information about a consumer to render a product suitable to a particular consumer’s needs, level of sophistication, and financial capacity (e.g. ability to repay) before providing any product or service. Summary of Main Policy Recommendations: • Encourage and facilitate partnerships between fintech companies and traditional providers • Expand and improve ICT infrastructure in rural and remote areas • Undertake demand-side research to better understand behavior and needs of “last mile” consumers • Move from “wait and see” approach to deliberate, proactive regulation and supervision of fintech • Develop clear, comprehensive legal and regulatory framework for fintech that is aligned with the current financial sector legal framework, applies consistent rules by type of activity, addresses cross-cutting issues related to fintech, and addresses issues specific to particular fintech models • Ensure balance between mitigating risks and allowing for healthy growth and innovation • Strengthen risk-based supervision of fintech and improve coordination across regulators • Consider tiered approach to largest P2P platforms and online MCCs with direct supervision by national regulators • Clarify role of National Internet Finance Association of China in regulating and supervising fintech • Enhance financial consumer protection legal framework for fintech-related issues, 38 including developing rules on product suitability and responsible finance VI. CREDIT AND BORROWER INFORMATION 113. Lack of sufficient information on underserved individuals and MSEs remains a significant constraint. Inadequate and/or limited access to credit and payment histories and other relevant and predictive information (e.g. information from public records agencies such as tax and judicial authorities) was identified as a major challenge by all providers during the FSAP, preventing providers from properly identifying and assessing the creditworthiness of potential borrowers in a cost-effective manner and thereby limiting access to quality credit. Providers primarily rely on the PBC’s credit registry, the Credit Reference Center (CRC). Fully- functioning credit bureaus are lacking, as are value-added services such as credit scoring. 114. Providers are increasingly developing their own credit information systems, indicating gaps in the system and leading to inefficiencies and fragmentation of information. For example, due to lack of coverage by existing credit registries, particularly for rural consumers and MSEs, the strategy of several large commercial banks has been to gather data from a range of different sources, including local governments and third-party data companies. At the same time, multiple initiatives are underway to consolidate credit information in separate systems, such as the NIFA’s collection of credit information from its fintech-related members, the Shanghai Credit Information Service’s establishment of the Net Financial Credit System (NFCS) covering 300 P2P platforms, and individual online MCCs and P2P platforms’ efforts to create their own industry-specific platforms. Such fragmentation of information increases transaction costs for providers and prevents providers from easily obtaining a full, system-wide view of an individual’s total level of indebtedness and complete credit history. 115. Policymakers should work towards developing a multi-layered market structure to address existing gaps and needs and expand the coverage and depth of credit information. This multi-layered market structure should consist of a comprehensive credit registry, multiple credit bureaus and scoring companies, and third-party data companies. Such entities all play complementary roles necessary to fill the existing gaps in credit information in China, particularly for those individuals and MSEs without a formal or lengthy credit history. Achieving such a structure requires expansion of the CRC, development of a credit bureau industry as well as scoring and risk management services, and a comprehensive legal framework for data protection. A. Credit Registry 116. While the CRC has achieved impressive coverage of individuals and MSEs, efforts should be made to further expand coverage to additional sources of information. The stated objective of the CRC is to capture data from all regulated credit institutions. As of the end of 2016, the CRC covered 910 million individuals, including 430 million with a credit history, and 39 22 million enterprises, including 6.4 million with a credit history.19 However, data from new fintech companies such as P2P platforms is not collected. 117. The legal and regulatory framework for P2P platforms and other fintech companies should be strengthened to facilitate inclusion in and access to the CRC. Efforts should be undertaken to ensure that the CRC is fully capturing timely information from regulated online credit providers, particularly P2P platforms and online MCCs, in order to provide comprehensive coverage and address issues with fragmentation of data. Particularly as many fintech companies are serving previously underserved individuals and MSEs, inclusion in the CRC will both help to expand the credit files of such consumers, as well as to prevent overindebtedness. In order to expand coverage, policymakers should work to ensure that standards and requirements currently being developed for fintech companies sufficiently address policy concerns with data quality and data security, so that such providers can submit data to the CRC and be granted full access to CRC data (as opposed to the limited access currently available) once sufficient levels of data robustness are achieved. 118. Efforts should also be made to expand the CRC’s coverage of additional information. The CRC has been pursuing information collection agreements with the State Administration of Taxation and the State Administration of Industry and Commerce. Preliminary, limited efforts to obtain additional information from other government agencies, such as IDs for MSMEs and data from company registries, should continue to be pursued. B. Credit Bureaus and Data Analytics Services 119. Policymakers should promote the development of an efficient, safe, and reliable credit bureau industry, including developing a framework for licensing and oversight. In early 2015, the PBC granted permission to eight companies to operate as credit bureaus on a provisional basis. However, none of these eight companies (which provide data analytics) have received a license to date, due primarily to PBC’s concerns regarding the independent governance of such companies and lack of diverse ownership. Given the clear need to foster growth in credit bureaus in order to expand the availability of value-added services to lenders, the PBC should develop clear licensing procedures and guidelines and exercise proper oversight functions for credit bureaus. A framework should be developed to ensure credit bureaus observe the General Principles for Credit Reporting (e.g. data quality, security and efficiency, risk management and governance, legal and regulatory framework, consumer protection).20 Other measures to further encourage the development of credit bureaus should be examined further, such as enabling access of licensed credit bureaus to certain types of data included in the CRC. 120. Policy concerns regarding the current group of data analytics companies should be more clearly distinguished from the development of a credit bureau industry. Concerns with the operations of data analytics companies, which primarily utilize big data analytics and leverage data from within their own family of companies, should be separated from and not impede the development of credit bureaus. Both credit bureaus as well as data analytics 19 CRC 2016 Annual Report. 20 http://siteresources.worldbank.org/FINANCIALSECTOR/Resources/Credit_Reporting_text.pdf 40 companies can exist in parallel and be of use to providers as a supplemental data source, but require different regulatory treatment. Such companies are particularly useful in an environment where a large percentage of individuals and MSEs do not have a formal or extensive credit history. As noted by one of the new data analytics companies, 40-50 percent of consumers inquired about were not covered by the CRC; the data analytics company was able to create credit scores for 80 percent of this subset based on transactional data from e-commerce and payment platforms, thus addressing a critical gap. 121. Policymakers should consider what principles should be applied to data analytics companies. To address potential risks to consumers and providers, such companies should be subject to certain principles in the General Principles for Credit Reporting when developing tools that impact the provision of credit, such as transparency, data quality, ARCO rights,21 and other data protection and privacy principles. As discussed below, the adoption of a comprehensive data protection framework in China will help to address a number of these items. Further research may be required to better understand how new data analytics companies leveraging big data operate and what types of scoring models they use, in order to determine the appropriate application of credit reporting principles while still allowing for innovation. C. Access to Public Information 122. Providers also highlighted access to public information, such as information from the court system, tax authorities, and the State Administration of Industry and Commerce, as important sources of data on potential borrowers. Government agencies were noted to hold a wealth of valuable information, particularly regarding MSEs, which was difficult to access. Such information is scattered across various agencies at both the national and local levels and across provinces and can be time-consuming and costly to obtain, particularly where information is not digitized and physical visits to government offices are required. Clear guidelines are missing regarding what information can be shared, which was noted as a contributing factor behind government agencies’ reluctance to share such data. 123. Policymakers should work to ensure efficient and cost-effective access to public information. The State Administration of Industry and Commerce has been assigned to take the lead in integrating public information on MSEs. In Guiyang City, a pilot data-sharing platform was launched in January 2017 making data available from 51 government bodies, consolidating positive and negative information. Such efforts to facilitate easy access to data (particularly in electronic form) as well as to improve the quality of such data should be further encouraged. 124. Clear policies should first be established at a national level regarding what information is permissible to be shared. For example, in the case of Guiyang City, a new regulation was first issued establishing the limits and boundaries of such data sharing, which helped to address confusion among government agencies regarding what data can be shared. 21 Access, rectify, cancel, and oppose. 41 D. Data Protection and Privacy Legal Framework 125. At a broader level, there is an urgent need to establish a comprehensive legal framework on data protection and privacy in China. Though there are ad hoc rules that address certain aspects relating to data privacy and protection, a comprehensive framework on privacy, protection, ownership, sharing, and usage of any type of personal data is currently missing in China, a very timely issue with implications beyond the financial sector. For example, the Supreme People’s Court recently issued a ruling in May 2017 on what kind of information is illegal to share without a consumer’s consent. As a result of this lack of a comprehensive legal framework, some government agencies are disinclined to share any data, while conversely other private entities are obtaining and distributing data without clear boundaries or rules. Among other topics, a legal framework should be developed to address: (1) lawful collection and usage of personal data (e.g. how data is collected, for what purposes can data be collected, for what limited purposes can such data be used); (2) confidentiality and security of personal data; and (3) sharing of personal data.22 126. Authorities should take special consideration of the implications of big data and alternative data in the financial sector. When addressing the current data protection and privacy gap, authorities should consider the role that alternative data can play in the financial sector, and particularly for financial inclusion. Given the rapid innovation in the use of alternative data such as transactional and social data in China, there is a unique opportunity to develop a cutting-edge regulatory approach towards alternative data that appropriately balances innovation with data privacy and other consumer protection concerns. As noted earlier, this is an emerging issue globally, for which clear guidance has yet to be developed. As a first step, policymakers should consider when big data is and is not considered personal information. In such cases, appropriate data protection and privacy principles such as transparency, accountability, lawfulness of data collection, permissible use compatibility, ARCO rights, and data security measures should be applied to big data and alternative data. Particular issues that may arise with respect to big data and alternative data should also be further examined, such as obtaining informed and willing consent, how such data can be accessed and corrected by consumers, and rights regarding affirmative consent to data processing. Summary of Main Policy Recommendations: • Expand coverage of PBC’s Credit Reference Center to include regulated digital credit providers • Encourage growth of credit bureau industry, including by establishing clear framework for licensing and oversight • Distinguish data analytics companies from credit bureaus and clarify the general credit reporting principles applicable to such companies • Establish clear policies at a national level regarding what information can be shared by 22 For further information, see for example (1) UN Guidelines for the Regulation of Computerized Personal Data Files, adopted by the General Assembly Resolution 45/95 of 14 December 1990; (2) 1981 Council of Europe Convention for the Protection of Individuals with regard to the Automatic Processing of Personal Data; (3) OECD Guidelines Governing the Protection of Privacy and Transborder Flows of Personal Data, 2013; and (4) APEC Privacy Framework. 42 government agencies and facilitate efficient and cost-effective access to such information • Develop a comprehensive legal framework for data protection and privacy • Consider implications of big data and alternative data and appropriate application of data protection principles to such data, with adaptations for any particular risks posed by such data VII. PROVIDER REACH 127. The number of access points is quite high in China, driven largely by a vast network of rural cash withdrawal points. The density rate of China’s branch network (19 branches per 100,000 adults) lags behind that of upper middle income countries and other BRICS countries. In addition, in rural areas (defined as county level and below), branch penetration was 13.1 per 10,000 people in 2015, lower than urban areas where penetration was 22.1 per 10,000 people.23 However, China’s branch network is complemented by a large and far-reaching network of rural cash withdrawal points (similar though not the same as agents) which, at close to 1,000,000 points, is the largest agent-based network in the world by absolute size. When normalized by population, China’s agent density rate (88 agents per 100,000 adults) still compares favorably to other BRICS countries and developing EAP countries (Figure 19). However, rural cash withdrawal points have less functionality than agents, as these points cannot provide a full range of services to facilitate the uptake of financial products (i.e. account opening, loan repayments, insurance). Figure 19: Physical Access Points Number of access points per 100,000 adults 100 90.1 94.0 87.9 85.9 84.2 76.4 80 52.6 51.2 55.3 60 50.1 35.5 35.2 40 26.2 25.0 23.7 19.3 15.8 20 0 China Upper middle High income Developing EAP BRICS excluding World income countries countries China Branches per 100,000 adults ATMs per 100,000 adults Agents per 100,000 adults Source: IMF FAS, 2016; WB GPSS, 2017; national authorities. 128. The number of access points such as ATMs and point-of-sale (POS) terminals is also relatively high in China. The density rate of POS terminals (2,011 POS terminals per 100,000 23 PBOC 43 adults (age 15+) outperformed all peer groups except high income countries.24 The size of China’s ATM network is larger than that of average upper middle income countries, though smaller than high income and other BRICS economies. Overall, physical reach in China is on par with that of upper middle income countries and other BRICS, but significantly higher than that of developing countries in EAP. However, it is worthwhile to emphasize the distinction between access points that facilitate the full uptake of financial products (branches and potentilaly agents) and access points that are more geared towards facilitating transactions (branches, agents, ATMs, POS terminals, and personal mobile or internet devices). 129. However, similar to bank branches, the penetration of unmanned access points such as ATMs and POS devices is lower in rural areas compared to urban areas. The number of ATMs and POS devices per million people increased from 248 and 3,582 in 2011 to 668 and 17,742 in 2016. However, the number of ATMs and POS devices increased faster in urban areas than in rural areas. The penetration of POS devices in urban areas is about 2 times of that in rural areas in 2016 (Figure 20). Figure 20. ATM and POS device penetration Number of access points per million people ATMs, per million people POS devices, per million people 800 720 733 25,000 700 21,319 22,403 582 600 20,000 488 516 500 431 403 14,157 15,000 400 333 325 285 278 10,573 11,475 300 10,000 8,451 208 6,211 7,074 8,613 200 4,403 4,396 5,000 2,802 100 0 0 2011 2012 2013 2014 2015 2016 2011 2012 2013 2014 2015 2016 Rural Urban Rural Urban Source: China Payment and Settlement System Development Report (2012-16); PBC 130. High rates of mobile and internet access provide Chinese adults with additional channels through which to access and use financial products and services. Indeed, many emerging fintech models in China (e.g., non-bank payment providers, P2P platforms) rely on access to mobile or internet devices or build off data gleaned from the mass scale of digital transactions. Access to a mobile phone is virtually universal in China, with 97 percent of adults reporting that they or someone in their household has a mobile phone (Figure 21).25 By contrast, just over half of adults report that their household has access to the internet, which, while lower than the average in high-income countries, is far above the average in developing countries in EAP and other BRICS countries. In fact, the increasing prevalence of smartphones that facilitate 24 IMF FAS, 2016; WB GPSS, 2017; national authorities. 25 Gallup World Poll, 2014, www.gallup.com. 44 internet access may potentially reduce the significance and utility of traditional broadband internet connection. Figure 21: Remote access points % adults (age 15+) reporting household-level access to mobile phone or internet 100 97.097.0 93.995.7 88.989.9 86.286.3 85.886.8 84.485.5 79.0 80 60 52.0 42.5 41.6 36.8 40 26.3 20 0 China Upper middle High income Developing EAP BRICS excluding World income countries countries China have access to internet have access to mobile have access to either internet or mobile Source: Gallup World Poll, 2014. 131. As can be seen from the above, China enjoys a relatively high degree of physical access to financial service points, due in part to a concerted effort by policymakers to expand accessibility. Between physical branches (full branches as well as other types of sub- branches), self-service kiosks, ATMs, POS terminals, rural cash withdrawal points, and digital channels, many areas in China are covered. However, low usage of bankcards suggests that there is still a need to expand cost-effective service points beyond bank branches to remote and rural areas. Moving forward, the main challenges to be addressed include further expanding low-cost delivery channels to reach the “last mile” and working to improve the commercial sustainability and scalability of different low-cost delivery channels employed by providers. A. Manned Outlets 132. Providers have employed a variety of approaches to expand reach beyond full- service physical branches. Manned outlets beyond full-service branches are critical to allowing for the provision of a full suite of financial products and services to underserved segments, but at lower operational costs to providers. The approaches employed by providers include community sub-branches and small and micro sub-branches, mini-branches with self-service kiosks in villages, and leveraging post offices. The CBRC has issued lighter licensing requirements for community and small and micro sub-branches, specifically to allow for better servicing of underserved segments. Mini-branches with self-service kiosks may employ community volunteers, and can perform almost many of the functions of a normal branch. 133. However, manned outlets face challenges with profitability and sustainability. A minimum level of active clients are needed for mini-branches to break even, indicating that it is not feasible to set up mini-branches in all rural and remote locations. Sub-branches have also suffered from difficulties in achieving profitability. 45 134. Low-cost, manned outlets should continue to be encouraged and facilitated in a strategic manner that follows the FIP principle of commercial sustainability. As an ongoing theme, the focus should shift to quality, and not only quantity. Policymakers have consistently encouraged providers to extend financial services to the village level. Such encouragement needs to be more strategic and take into account sustainability. Providers should not be mandated to increase the absolute numbers of such outlets, but should instead be encouraged to make strategic market-based decisions, for example by considering the optimal placement of such outlets to reach break-even levels of traffic and employing a tiered approach that combines manned outlets with unmanned access points in lower traffic areas. Partnerships with local governments to cover physical operating costs can be strategically employed in “last mile” scenarios. Policymakers should also ensure that the legal and regulatory environment does not impede innovations in self-service operations, for example in the usage of machines that can open accounts and issue bank cards (subject to appropriate KYC/CDD requirements, among other legal requirements). 135. As the network of postal offices is vast and reaches into the most remote and rural areas in China, policymakers may wish to consider its usefulness as a delivery channel for multiple providers. Postal offices appear to currently be leveraged primarily by the PSBC, but have the potential to serve as a low-cost delivery channel in remote areas for other providers as well. B. Rural Cash Withdrawal Points 136. Rural cash withdrawal points present a significant opportunity to be further leveraged for financial inclusion. Reaching the “last mile” in China requires a tiered system with increasingly low-cost delivery channels that can still provide necessary functionality beyond ATMs and unmanned devices. Given the huge network of rural cash withdrawal points in remote and rural areas, such points are well-positioned to serve this role. However, many such points currently have very low traffic and are not sustainable. A new strategy should be developed to better leverage rural cash withdrawal points and plan for their evolution into more fully functional agents. 137. Chinese policymakers have achieved considerable success in establishing a vast network of rural cash withdrawal points that reaches nearly every village in China. Rural cash withdrawal points typically consist of a POS device installed in local retail shops or local commune offices. Such points are set up by local banks, “new-type” institutions, and RFIs. As their name suggests, their initial purpose was to facilitate cash withdrawal, specifically of social transfers to rural residents. Since then, their functionality has expanded to include cash-in, including bill payments, remittances, and person-to-person transfers. There are ceilings for individual transactions and daily transactions, and caps on fees charged to consumers. At end of September 2016, there were 1,008,500 such points in China, with almost 531,800 administrative villages covered). These service points handled 495M transactions with a total value of 420B 46 RMB, with cash withdrawals comprising 50 percent of services provided.26 In some localities, service points enjoy subsidies on installation of POS devices and electricity. Table 3: Transaction data of rural cash withdrawal points 2010 2011 2012 2013 2014 2015 Rural cash withdrawal points 19,800 215,900 669,200 852,100 980,900 997,500 Number of withdrawals (millions) 4.85 22.5 72.0 92.7 157.0 236.0 Amount of withdrawals (billion RMB) 3.78 9.198 23.6 28.5 49.4 101.7 Number of total transactions1 (millions) 436 495 Amount of total transactions (billion RMB) 400.8 424.8 Source: PBC 1 Total transactions refers to withdrawals, remittances and bill payments. 138. However, the network of rural cash withdrawal points may actually be overextended, and not all points sustainable. While traffic varies substantially across service points, many service points have very low numbers of transactions. Providers have reported that some service points have only one transaction every few days, while some service points are dormant or have never even been used. Providers have clearly heeded policymaker calls, resulting in some small villages having a surplus of multiple service points established by different providers. Numerous providers indicated that such service points are established out of a sense of social responsibility and not due to market-based decisions, with the intention of each provider to cover all administrative villages. 139. The strategic approach towards rural cash withdrawal points should be reformulated to allow such points to become more functional, sustainable, and fully leveraged. There is already clear recognition by policymakers that the current model is not sustainable. While not every rural cash withdrawal point needs to be profitable per se as the channel plays a broader role for providers than simply making a profit (e.g. branding, customer stickiness, cross-selling), greater emphasis should be placed on sustainability. As with other financial inclusion-related issued, policymakers should refrain from directing the actions of providers (in this case, by pushing for universal coverage in all villages), and instead focus on improving the enabling environment so that providers have the space to innovate and improve commercial sustainability by tailoring service point operations to business needs and target markets. For example, greater commercial sustainability could be achieved by optimizing placement of service points, increasing functionality for financial and non-financial services to increase both foot traffic and revenue streams, encouraging shared use of agents by both traditional providers and non-bank payment providers, and loosening caps on agent fees and daily transaction limits. 140. Research should be undertaken to comprehensively assess current practices of rural cash withdrawal points, in order to segment such service points and to better understand the varying potential for different segments. As noted, traffic varies significantly across different service points. A one-size-fits-all approach for all rural cash withdrawal points is not suitable. More granular data is needed to assess different segments of service points, study gaps 26 PBOC. 47 in demand and supply, and design appropriate strategies to allow for and encourage different business cases and approaches. It is likely that for certain segments, a more enabling environment should be created to allow for expansion in services, while other segments in “last mile” locations may continue to have low levels of transactions and require targeted and limited support. 141. The functionality of rural cash withdrawal points should be increased so that they can operate as full-fledged agents, complemented by necessary safeguards. For certain rural cash withdrawal points that have the potential, there should be a major shift from thinking of these points as places for rural residents to withdraw cash and to conduct limited transactions to instead thinking of these points as full-fledged agents of financial service providers, with the possibility to provide a wider range of services as an extension of providers. For example, CGAP studies demonstrated a clear demand for increased deposit-taking and account opening services at service points, as well as increases in daily withdrawal maximums. Policymakers should allow flexibility for innovation and expansion in services, subject to stronger oversight, training, and management by providers. Based on careful study, limitations on transactions and agent fees could be relaxed as appropriate, and origination of a financial products and services expanded (for example, initial credit assessments appear to be allowed only in certain provinces). Deposit- taking should be more clearly allowed, subject to appropriate safeguards to prevent fraud and ensure safety of funds. Such an expansion in functionality would address unmet demand from consumers as well as make the agent-based model more appealing for agents and providers, thus helping to improve long-term sustainability. 142. Policymakers should also encourage innovative approaches to combining service points with other non-financial services. In some provinces, rural cash withdrawal points are being upgraded into broader centers providing integrated services, such as e-commerce, utilities payments, and purchase of inputs. Integrated service centers at village levels would help address issues with insufficient traffic, and their potential should be further explored. 143. Improving the shared use of agents should be studied. While rural cash withdrawal points are not intended to be exclusive, it appears that most points operate on the behalf of one provider. There appears to be a significant missed opportunity for shared use of agents, in particular by non-bank payment providers who have stated their intention of moving further into rural areas. Any obstacles or uncertainties in the legal framework should be examined and addressed, for example obstacles to employing different technology at rural cash withdrawal points or for non-bank institutions to employ agents. C. AML/CFT 144. The AML/CFT framework should be assessed to determine if there are undue obstacles, particularly with respect to remote identification. There have already been adaptions to introduce risk-based measures, such as simplified CDD and tiered accounts that allow for remote and online identification. However, such tiered accounts were noted to pose constraints to reaching certain “last mile” consumers. In particular, caps on Type I payment accounts are quite restrictive, and the ID requirements to upgrade to higher levels accounts can be burdensome for some consumers. As a result, payment accounts opened online by the unbanked cannot function as a substitute for a full-fledged bank account. As a more risk-based 48 framework is developed, policymakers should consider whether further flexibility can be introduced. Technologies for online and remote verification should also be further strengthened. Summary of Main Policy Recommendations: • Encourage and facilitate strategic use of low-cost, manned outlets in a commercially sustainable manner • Undertake research to gather more granular data on rural cash withdrawal points in order to segment such points into different types and develop appropriate strategies by type • Selectively expand functionality and relax restrictions on rural cash withdrawal points to allow for operation as full-fledged agents providing broader range of services, complemented by necessary safeguards • Encourage innovative approaches to combine service points with non-financial services and increased shared use of service points by multiple providers, including non-bank payment providers • Assess whether AML/CFT framework poses undue obstacles to remote identification VIII. PRODUCT RANGE AND APPROPRIATENESS 145. Despite forward progress, there is room to improve the range and appropriateness of financial products and services for the underserved in China. For example, there is low uptake and usage of transaction accounts and bank cards among the underserved. Digital payments are relatively prevalent in China, but mainly targeted to and used by urban consumers. There is a relative under-emphasis by policymakers on basic account, savings, and insurance services for the underserved. Collateral requirements and information asymmetries are important barriers for household and MSEs to access to bank credit. 146. It should be noted that the recommendations discussed in the previous sections of this Technical Note are all intended to improve the enabling environment so that providers are internally motivated to develop more appropriate products and services. For example, as noted, collateral requirements and information asymmetries are important barriers for household and MSEs to access to bank credit. This Section VIII will highlight a few targeted areas where further actions can be taken to facilitate a broader range of appropriate product offerings. A. Product Range 147. There is a relative under-emphasis on basic account, savings, and insurance services, as well as on the sustainability of provision of these services. While the government emphasizes credit, there is unmet demand for other financial services. About 240 million Chinese adults still lack a transaction account, a basic and essential financial tool which addresses the need to make day-to-day payments and store value. About 30 percent of adults in China reported saving, but not at a formal financial institution. Formalizing the “hidden” savings 49 of these 350 million informal savers represents a significant opportunity, but also suggests there is a need to develop simple, accessible, and low-cost savings products, particularly for individuals with low and irregular incomes and who live far away from physical branches. Further efforts should be undertaken to promote a broader range of insurance products (beyond agricultural insurance) that are commercially sustainable and accessible to the underserved. 148. A broad range of appropriate products should be promoted for a variety of underserved segments and groups. The government emphasizes promoting financial inclusion for the “last mile”, including the MSEs, the remaining poor, farmers, the disabled, and the elderly. However, financial inclusion efforts should cover not only the unbanked, but also segments that already banked, but still underserved. As heterogeneity of the Chinese economy increases, the policy agenda should evolve as well to address all underserved areas, including formal and informal SMEs, rural households in both agricultural and non-agricultural business, migrant workers, and low-income urban residents. 149. Various issues that may be contributing to low levels of usage of financial products among the underserved should be addressed. For instance, efforts to move from cash to card payments for social transfers has led to increased ownership of bankcards and bank accounts in rural areas. However, the actual usage of bankcards under some of these programs can be quite low. According to Global Findex 2014, among the 35.7 percent of rural residents who owned a debit card, only 11.0 percent actually used the debit card during the past year. There are multiple reasons why use of bank cards may be low. Some rural households may not be aware that the cards on which they receive government transfers are linked to a bank account and can be used to receive other financial services. In this regard, increasing financial literacy and digital literacy of “last mile” consumers could be useful to enhance usage of financial services. B. Product Appropriateness 150. Collateral requirements remain an important barrier for household and MSEs to access to credit. Banks still rely heavily on collateral and guarantees due to lack of ability to process soft information characteristic of small households and MSEs. According to the Enterprise Surveys, lack of sufficient collateral was one of the most frequently cited reasons why SMEs don’t not apply for a bank loan. 73 percent of loans to small enterprises (5-19 employees) and 80 percent of loans to medium enterprises (20-99 employees) required collateral, which is higher than other BRICS countries and high-income OECD economies ( 50 151. Figure 22). Collateral requirements measured by value of collateral needed for a loan are higher for small enterprises at 219 percent, compared to an average of 155 percent for other BRICS countries and 176 percent for high-income OECD economies (Figure 23). Land and building, and machinery and equipment (including real restate) are two major types of collateral required for bank loan; only 11 percent of banks use accounts receivables and inventory as collateral from small enterprises. 51 Figure 22: Collateral requirement for firms, by firm size % of loans requiring collateral 100 79.179.772.9 85.184.881.8 81.7 71.570.2 73.7 81.1 77.5 80 60.9 65.0 64.7 60 40 20 0 China (2012) BRICS(excluding East Asia & Pacific High income: World China) OECD Large firms ( 100+ employees) Medium size firms (20-99 employees) Small firms (5-19 employees) Source: WBG Enterprise Surveys Figure 23: Value of collateral requirement for firms, by firm size Value of collateral needed for a loan (% of the loan amount) 300 239 218 222 221 198 219 176 189 195 171 200 142 151 155 125 149 100 0 China (2012) BRICS(excluding East Asia & Pacific High income: OECD World China) Large firms ( 100+ employees) Medium size firms (20-99 employees) Small firms (5-19 employees) Source: WBG Enterprise Surveys 152. The government is actively promoting product innovation in China, though to what extent those innovations are beneficial to financial inclusion are unclear. For example, local governments are promoting tailored rural finance products, including loans secured by management rights on contracted rural land. In addition, the government encourages providers to design and deliver “innovative” products to the underserved. While there are many attempts, it is unclear how many customers are truly benefiting from such products, how efficient and effective these innovations are, and to what extent these business models are scalable and commercially sustainable. Products that are driven by a top-down, government- mandated approach and/or are heavily subsidized often do not meet the needs of the underserved or benefit a limited few and are not sustainable in the long-run. The government should instead focus on addressing obstacles that impede market-driven innovation. C. Level Playing Field 153. RFIs and “new-type” institutions continue to feel informal pressure to charge low interest rates, impeding the ability to achieve commercial sustainability. Serving underserved segments requires higher costs and adapted techniques to be accomplished effectively. However, charging higher interest rates is still discouraged and deemed not politically feasible, particularly in rural areas. Informal pressure on RFIs, “new-type” institutions, and commercial lenders to lend at interest rates that do not cover their costs is 52 contrary to a market-based approach. Artificially low interest rates ultimately act as a disincentive for providers to serve MSEs and rural consumers on a large scale and over the long- term. As indicated in Table 4, high interest rates are actually not a major reason for small firms not applying for loans in China. Only 4 percent of small enterprises (5-19 employees) cited high interest rates as a reason for not applying for a bank loan. Rather, the most frequently reported barriers were lack of sufficient collateral, insufficient size of loan and maturity, and complex application procedures. Table 4: Reasons for Not Applying for a Bank Loan % of firms that cite each reason No need Application Interest Collateral Size of loan Did not Other for a loan procedures rates requirements and think it were were not were too maturity would be complex favorable high were approved insufficient Small firms China(2012) 57.2 9.5 3.9 10.6 9.6 6.9 0.9 Brazil(2009) 77.3 0.9 13.3 1.7 0.2 0.0 6.4 India(2014) 55.5 4.5 12.7 6.4 3.8 5.0 10.9 Mexico(2010) 68.4 11.5 10.4 0.5 0.7 2.4 3.6 Russia(2012) 53.7 10.3 25.5 6.2 0.3 0.0 2.0 SouthAfrica(2007) 53.2 13.5 18.0 4.5 1.5 6.6 2.5 Medium size firms China(2012) 54.2 10.8 10.1 10.3 5.6 3.5 2.6 Brazil(2009) 74.4 1.7 17.6 1.4 0.5 0.0 3.8 India(2014) 51.2 8.5 11.5 6.3 5.1 6.5 9.7 Mexico(2010) 71.1 8.5 2.3 3.3 0.8 2.1 11.9 Russia(2012) 54.7 8.2 20.9 10.0 0.8 0.0 1.0 SouthAfrica(2007) 70.0 7.1 12.7 2.2 0.3 3.4 3.9 Large firms China(2012) 66.6 7.5 11.6 7.2 2.6 1.0 0.5 Brazil(2009) 95.6 1.4 2.3 0.0 0.0 0.0 0.4 India(2014) 51.3 6.1 15.3 8.0 6.8 2.7 8.5 Mexico(2010) 84.9 1.4 2.5 0.1 0.0 1.4 9.7 Russia(2012) 81.7 7.1 6.1 0.6 0.0 0.0 0.3 SouthAfrica(2007) 85.2 3.4 4.8 2.0 0.0 0.8 3.9 Source: WBG Enterprise Surveys Summary of Main Policy Recommendations: • Conduct targeted, demand-side research on accounts, savings, insurance, and digital finance to better understand the needs and barriers of remaining underserved segments • Informal pressures on RFIs and “new-type” institutions to charge low interest rates in rural areas should be addressed 53