I. The growing relevance of DFS supervision

The rapid digitization of the financial sector in Africa and globally has significantly increased the importance of digital financial services (DFS) regulation and supervision. The emergence of mobile money platforms, digital wallets, and innovative fintech solutions has led to unprecedented data that presents regulators with a wealth of untapped information for assessing and mitigating risks.

Central banks are responding by developing regulatory frameworks, enhancing data infrastructure, and leveraging supervisory technology (suptech) to monitor and assess risks arising from the burgeoning digital financial ecosystem.

Central to DFS supervision is the effective use of data, and to that end, central banks are investing in robust data infrastructure and advanced analytical techniques. Using suptech is a critical enabler, providing regulators with cutting-edge tools and techniques to oversee DFS efficiently.

Despite these efforts, the journey toward adequate DFS supervision remains particularly challenging in developing countries where supervisors face limited resources, fragmented systems, and gaps in technical expertise. This article draws on our experience working with central banks in Ethiopia, Fiji, Malawi, and Samoa to provide insights into the data infrastructure and suptech adoption aspects of DFS supervision. These observations offer valuable guidance for central banks in other developing countries looking to establish a solid foundation for adequate DFS supervision and foster financial inclusion.

II. Existing gaps in the current architecture

The United Nations Capital Development Fund (UNCDF), in collaboration with Talanta 10, has been working with partner central banks in Ethiopia, Fiji, Malawi, and Samoa to assess their current data architecture and determine their readiness to adopt suptech solutions in response to the growing relevance of DFS supervision. Our assessments have revealed gaps in the current architecture that restrict central banks from harnessing the full potential of data for supervision in the digital age.

One such gap is regulatory reporting, which remains the primary data source for DFS supervision. However, this approach has several limitations, such as template-based and aggregate data reporting that is fixed by use case, resulting in compliance burdens on regulated entities and limited opportunities to gain insights from newer sources of unstructured data such as information from the web, images, social media posts, etc. Additionally, infrequent reporting limits the effectiveness of supervision to a retrospective rather than proactive stance.

In addition to issues with regulatory reporting, the current data architecture suffers from several other weaknesses, such as the semi-automation of data processes, disjointed and fragmented systems, and legacy IT systems. These challenges result in significant inefficiencies, costs, and operational risks, requiring considerable time and resources dedicated to routine and tedious tasks rather than analytical work central to supervision and decision-making.

III. Some options to consider in addressing the gaps

Our engagements with central banks in Ethiopia, Fiji, Malawi, and Samoa have provided valuable insights into addressing the gaps in the current data architecture and harnessing the power of data in DFS supervision. Here are some options to consider:

1. Foster data harmonization and streamlined reporting

Central banks can work towards harmonizing data requirements and reporting templates across functional units. This collaborative approach ensures efficient data collection and analysis for multiple purposes, reducing compliance burdens for regulated entities. Exploring innovative data collection and management solutions, such as Application Programming Interface (APIs) and automated reporting systems, can also minimize manual intervention and streamline data processes.

2. Embrace advanced analytical techniques and suptech

Investing in advanced analytical techniques and suptech solutions empowers central banks to adopt data-driven supervision practices. Central banks can leverage tools like machine learning (ML), natural language processing (NLP), and data visualization to gain valuable insights into industry trends, potential financial consumer risks, and overall financial stability. By establishing robust data infrastructure, central banks can enhance their capabilities in data management, risk-based supervision, evidence-based policies, and promotes digital innovation and financial inclusion. With advanced tools and techniques, central banks can collect granular data, disaggregated by different metrics like sex, age, and geolocation, for a comprehensive understanding of risks and experiences faced by different consumer segments.

3. Cultivate collaboration and information sharing

Promoting collaboration and information-sharing across functional units and with external stakeholders, including FSPs and other regulators, is crucial. Coordination and cooperation among various internal and external stakeholders with a shared mandate for regulating and supervising DFS players are necessary for harmonized data requests and seamless data sharing. By sharing information, supervisors can streamline data management processes, reduce duplication of efforts, and ensure that all relevant parties can access the latest information on industry trends and emerging risks. Besides, information sharing with FSPs will help fulfill their desire for relevant market-relevant statistics to understand customers' demands, experience, and market dynamics and benchmark themselves against competitors.

4. Enhance capacity and expertise in data management and analysis

Central banks should prioritize building capacity and expertise in data management and analysis within their organizations. Strengthening the skills and expertise of staff with data roles on emerging data technologies such as big data, ML, and Artificial Intelligence (AI) is essential to enhancing their capacity to process and analyze complex and large volumes of data brought about by improvements in data architecture. This can be achieved by training staff, establishing partnerships with academic institutions, seeking technical assistance from external partners, and fostering peer learning and exchanges on best data management and analysis practices. By building a solid foundation of expertise in data-driven supervision, central banks can ensure they are well-equipped to adapt to the rapidly changing digital finance landscape.

5. Safeguard data privacy, cybersecurity, and the ethical use of data

As the digital financial ecosystem evolves, so do the risks associated with cyber threats. To maintain trust in the financial system and protect consumers, central banks must prioritize ethical use of data, cybersecurity and invest in robust measures to safeguard sensitive data. This can include implementing advanced encryption techniques, conducting regular vulnerability assessments, and promoting a culture of cybersecurity awareness among staff and stakeholders. Also, there is a need for a legal framework or public policy measures that ensure data is governed and used responsibly and securely and minimizes the potential for algorithmic bias in decision-making processes.

6. Supporting suptech adoption in developing markets

By leveraging advanced technologies such as APIs, AI, ML, etc., central banks can automate data management processes, enhance decision-making and improve overall efficiency for supervisors. Suptech solutions offer many advantages for supervisors, including real-time monitoring of financial service providers (FSPs), advanced risk assessment, and improved compliance management. By streamlining and automating the regulatory reporting processes and allowing FSPs to submit standardized data efficiently, suptech solutions reduce administrative burdens on regulated entities and enable supervisors to access timely and accurate data for decision-making.

In addition to improved data management, suptech allows for more proactive and targeted supervision, enabling supervisors to identify potential issues before they escalate into full-blown crises. By leveraging ML algorithms, supervisors can identify patterns and anomalies in transaction data that may indicate fraudulent activity. This allows for more effective fraud detection and prevention, ensuring the stability and security of the DFS ecosystem, which is important for building consumers’ trust, especially among vulnerable population segments, including women.

Although suptech adoption offers numerous benefits, its implementation often requires significant investments in technology, infrastructure, and capacity building, posing challenges for developing countries with limited resources. Despite these challenges, several countries, such as Rwanda, have successfully implemented suptech solutions to strengthen their financial sector supervision. The development of an electronic data warehouse by the National Bank of Rwanda (BNR) to automate and streamline data submission and analytics provides a prime example for other developing countries to learn from.¹ The UNCDF team closely follows Rwanda’s developments and believes the BNR journey provides valuable insights to guide prototype and full-stack solution developments in some of the assessed markets.

However, not every central bank has the capability to achieve this independently. Hence, UNCDF has undertaken the responsibility of supporting select countries to demonstrate the potential of technological advancements in addressing data management and use challenges under the Euorepean Union and Organisation of African, Caribbean and Pacific States supported DFS4Resilience Programme.

The primary objective of these engagements is to help partner central banks demonstrate their success in streamlining data management and analytical processes for prioritized use cases, in the hope that they can mobilize the necessary resources for full-scale solution production and establish themselves as data-driven central banks of the future. UNCDF aims to facilitate this process by collaborating with partner central banks to design, implement, test, and learn from these demonstrations.

Through these initiatives, UNCDF aspires to contribute to the body of knowledge and case studies on suptech implementations, especially from developing countries. By synthesizing insights gained from these engagements, UNCDF seeks to create a comprehensive knowledge base that can be used to guide future initiatives in this space.

IV. Conclusion: The road ahead for data-driven DFS supervision

UNCDF’s experiences with data architecture diagnostic assessments in Ethiopia, Fiji, Malawi, and Samoa, show that supervisors' data architectures in developing countries also adapt and mature to meet challenges introduced by rapidly growing digital financial ecosystems.

As these central banks embrace risk-based approaches, develop their robust data infrastructure, adopt and harness the power of suptech, and engage in multi-stakeholder collaborations, they gradually strengthen their capacity to supervise DFS effectively. While the road ahead may be challenging, better coordination and cooperation with private and public sector partners, such as UNCDF, can defray the costs of suptech demonstrations and implementations in some markets.

Alongside budgetary contributions, partners can help central banks stay up-to-date with the latest technological advancements and develop forward-thinking approaches to supervision. This can ensure that central banks remain well-equipped to safeguard financial stability, protect consumers, and promote inclusive growth in the ever-changing DFS ecosystem.

¹ Bank for International Settlements (2018) Innovative technology in financial supervision (suptech) - the experience of early users. FSI Insights 9. Basel: BIS. Available at: https://www.bis.org/fsi/publ/insights9.pdf