On a recent trip to Asia, I visited the offices of a federation of cooperatives. The backroom where the cooperatives’ files were kept had the musty smell of a library, with stacks of paper and folders with wilted edges teetering toward the ceiling. The scene was a stark reminder of how captive many financial institutions still are to a paper-based reality, with data on customers or operations often buried or hard to access.
This year, I also heard a representative of a central bank in East Africa share how the lack of up-to-date data in-house, such as rates of access or the gender gap, forced it to ask financial institutions it supervised for data. The bank might then have to wait days for an answer, even for a potentially straightforward type of information.
Despite two very different environments, these two offices share a common challenge: the need for data and better ways to collect, store, and analyze it. This, in a nutshell, is what prompted the U.S. Agency for International Development (USAID), Bill & Melinda Gates Foundation (BMGF), Omidyar Network, and BFA to launch the Regtech for Regulators Accelerator (R2A) in 2016.
What We Did
R2A was a bet on both a set of topics (regulatory technology for regulators, or “regtech,” and supervisory technology, “suptech”) and an approach to innovation (crowdsourcing ideas from start-ups to address challenges faced by financial authorities). The singular focus was to demonstrate, in collaboration with financial authorities and technology companies, how regtech could help overcome certain long-standing constraints to effective oversight and compliance, and in so doing, increase trust in formal financial services.
So after two years of innovation and collaboration, what did we learn?
In markets where millions of people are now using formal financial services through digital channels, the stakes for consumer protection and market integrity are getting higher. Yet this also means a commensurate increase in opportunity to use the data generated from these services to revisit standard methods for oversight and compliance. Tools like machine-learning, application programming interfaces (APIs), and natural language-processing are not merely buzzwords in this context. They are, in the right circumstances, viable engines for regtech applications in markets in the developing world.
Even Prototypes of Regtech Applications Can Drive Meaningful Progress
Looking over the last 20 months with R2A, our initial hypotheses seemed to hold true:
1. Financial authorities see value in adopting technology to the same degree as financial institutions;
2. Tech start-ups see commercial opportunity in serving financial authorities as customers; and
3. Initial testing of regtech applications developed jointly by startups and financial authorities with a competition model can demonstrate efficacy in resource-challenged environments.
Our lessons were drawn from three competitions held to prototype regtech solutions specifically designed to address challenges identified by financial authorities. Interestingly, dozens more regulators have since approached R2A expressing a desire for similar engagements. Our work with the Comisión Nacional Bancaria y de Valores (CNBV) in Mexico showed how a combination of artificial intelligence-powered analytic tools and better data dashboards might facilitate better insights for anti-money-laundering (AML) efforts. We anticipate that the tool prototyped by R2A and the CNBV is also anticipated to enable the CNBV to reduce the length of AML-related on-site inspections from 3 to 5 weeks to 3 to 5 days.
Likewise, our work with the Bangko Sentral ng Pilipinas (BSP) in the Philippines showed how the use of a chatbot might simplify and automate the receipt and resolution of customer complaints. This made technology a key access point for a fundamental consumer protection mechanism, not just a new way to access financial services. Finally, our work with the BSP on a second use case showed how an API for data-reporting might dramatically reduce the hours spent by both the financial authority and firms to validate data and correct errors each reporting period, resulting in higher data quality and timely compliance. By using the API, the BSP would be able to reduce the time to process data submissions from 30 minutes to 10 seconds. And similarly, it would enable the BSP to go from 29 reporting schema, many of which involve Excel-based reports, to one unified scheme.
All told, R2A received 31 applications from companies across 13 countries and 5 continents to collaborate with the BSP and CNBV, most of whom were new to engaging with financial authorities. This proved to us that even a field as staid as market oversight and regulatory compliance can pique the interest of technologists when public sector counterparts show a willingness to embrace innovation to overcome long-standing data challenges. Two notable factors contributed to this level of interest: concerted efforts to reach out to the tech community, and care taken by R2A and the financial authorities to clearly define the functional and technical specifications related to the underlying problem.
Preliminary Insights for Future Engagements
Finally, while our compressed timeline for engagement meant that R2A could not develop full-fledged solutions, R2A’s sprint nevertheless yielded promising initial results and succeeded in developing functional prototypes for each of the three chosen challenges. By prototyping working solutions in close collaboration with the financial authorities, R2A reduced the difficulty in taking the next step, should the financial authorities choose to do so through their standard procurement processes. We are especially encouraged by the stated intent of our financial authority partners to sustain and build on the gains made with R2A, whether in the form of investing their own resources in additional regtech solutions, identifying additional use cases for regtech, or fully integrating the prototyped solutions.
As we look toward the evolution of regtech, there are several important insights we gained from our investment in R2A:
- Regtech should serve, not co-opt, a broader strategy for oversight and compliance: An excellent starting point for approaching regtech applications is to focus on ongoing pain points. But in the long run, greater opportunity exists to rethink long-standing approaches for the conduct of market oversight and craft a strategy within which regtech can operate. In other words, start by focusing on incremental gains while still scoping out more transformative changes. Likewise, a more holistic approach to regtech can reduce the risk of it being perceived as an easy fix for weak governance. Regtech can be a vehicle for more transparent, accountable oversight, but only if a baseline level of human capacity and organizational readiness exists (or can be strengthened) for introducing regtech applications in a phased manner over time.
- In an ocean of newly-available data, more is not always better: A temptation with the greater volumes of accessible, analyzable data is to default to collecting as much as possible. But the focus should remain on precision and on collecting only the data that serves a clear need for a financial authority with the skills to collect, process, and analyze it.
- Data-savvy authorities depend on data-savvy financial firms (and consumers): Many regtech applications focus specifically on improving the efficiency of information flows between firms and financial authorities. But in many markets, financial firms are equally early in their adoption of digital tools other than spreadsheets and email for operational and compliance purposes. This underscores the value of navigating the regtech landscape in a collaborative manner, where the adoption of new tools corresponds to the readiness of all market participants. This relates to another sector-wide concern: data protection and data security concerns that, if not accounted for, could impair the efficacy of regtech applications and trust in other digital interventions.
- Procurement practices may have an influence on the extent and type of technology solutions relied on by financial authorities: We think R2A’s success is in part due to how it facilitated productive engagement between two groups (financial authorities and tech companies) that do not often interact closely. These interactions were crucial to helping the participant authorities better understand what problems they wanted to solve and which regtech applications might offer viable solutions. Moreover, they occurred outside the traditional processing of defining “specs” and issuing a request for proposals. R2A’s focus was not on assessing how standard procurement processes used by either the BSP or CNBV select technology vendors. But the R2A experience hinted at how procurement processes can influence whether and how novel (and potentially better) technology solutions are integrated into the daily rhythm of oversight and compliance. Procurement aside, R2A demonstrated the value of “pre-procurement” exercises, like design-thinking workshops with industry, to uncover overlooked opportunities for improvement and bring new views to the table.
Our vision is to go beyond the prototypes that R2A pioneered. These preliminary insights, among many more, will help us do so. Whether we work with financial authorities to engage technologists through competitions, hackathons, bootcamps, or tried-and-true requests for information, we will continue to learn what works best for tackling challenges faced by partner financial authorities. In all cases, we will place a premium on tailoring the approach to regtrech to the capacity of the financial authority.
Without a doubt, the progress we have made with the BSP and CNBV was not merely a product of our assumption that regtech is here to stay. In fact, our initiative, R2A, had the benefit of working with highly dedicated financial authorities that were willing to take a chance on something new. And R2A itself has demonstrated the value of bringing to the table diligence, adaptability, and talent.
What does this mean for customers who rely on cooperatives in Asia or who live in a digitally vibrant Africa? We are not under any illusions that regtech applications will be a feasible reality tomorrow for every stakeholder. But we do believe that we can help chart out viable paths for incremental adoption of digital tools that are fit for purpose. Sooner rather than later, those stacks of disordered paper will give way to cloud-based storage of data, and those ad-hoc data requests from the central bank will be replaced by seamless flows of data and analysis.