BLOG

Do Regulatory Sandboxes Impact Financial Inclusion? A Look at the Data

In CGAP's blog series on regulatory sandboxes, we explore high-level lessons relevant to policy makers who are thinking about establishing regulatory sandboxes. There is one question, however, that we have not yet answered: Do regulatory sandboxes have a positive impact on the financial industry generally and on the inclusion of unserved and underserved customers, in particular?

Ugandan woman sewing. Photo: Mohammad Saiful Islam, 2013 CGAP Photo Contest
Photo: Mohammad Saiful Islam, 2013 CGAP Photo Contest

Beyond a report published by the Financial Conduct Authority almost two years ago, evidence demonstrating the impact of regulatory sandboxes remains scarce. To help grow this body of knowledge, we sought to find out how and why regulators are using sandboxes. CGAP and the World Bank surveyed financial sector regulatory authorities in almost 50 countries about their regulatory sandbox initiatives and other approaches to innovation. So far, 28 regulators have responded, and responses are still coming in. Next, we wanted to find out what kinds of players benefit from regulatory sandboxes and create a hypothesis about the actual impact sandboxes have on the market. To do this, we analyzed publicly available data on the experiences of nearly 100 firms participating in 12 regulatory sandboxes around the world.

The survey data point to the strengths and limitations of the sandbox concept and its potential impact. The following are some key insights on how and why sandboxes are gaining traction among regulators:

  1. Sandboxes are a response to an urgent need felt by regulators to understand fintech. Most respondents reported that their main motivation for launching a sandbox was either to deal with the growing fintech industry (92 percent) or to learn about emerging innovations (80 percent). This suggests that regulators want a deeper understanding of the increasing complexity of their work, and they see regulatory sandboxes as a way to gain this understanding. Interestingly, half of all respondents reported that they have adopted regulatory changes in relation to sandbox testing and similar initiatives, including in areas like crypto-assets, initial coin offerings (ICOs), crowdfunding, cybersecurity, cloud computing, e-money and biometric authentication. This suggests that regulatory sandboxes are already playing a role in evidence-based regulation.
     
  2. Regulators take consumer protection during sandbox testing seriously. Sandboxes are not a “wild west” for fintechs looking to exploit regulatory loopholes. Almost 70 percent of the 23 surveyed regulators with sandboxes had put in place safeguards to ensure consumer protection. While 52 percent of the surveyed regulators with a sandbox grant temporary waivers from full licensing regimes, most respondents require full authorization at the end of testing. 

    Consumer Protection Safeguards Adopted during Sandbox Testing
    Consumer Protection Safeguards Adopted During Regulatory Sandbox Testing
    Source: CGAP/World Bank Survey (2019). Preliminary results (16 responses).

  3. Sandboxes may be getting more attention than they deserve. Despite the attention they get from experts and the wider public, regulatory sandboxes attract fewer firms than complementary innovation facilitators, such as innovation hubs and accelerators. In our survey sample, innovation hubs accepted on average 170 applications and accelerators accepted 23. Meanwhile, regulatory sandboxes averaged just 13 applications. These results reflect the importance of nonsandbox innovation facilitators that can meet the needs of many more firms and complement the overall innovation facilitating ecosystem, as explained in a 2019 UNSGSA report. This is not to say that other innovation facilitators can substitute for sandboxes, but they can complement them.

When we complemented our survey data with publicly available data about firms using sandboxes, a few things stood out:

  1. Sandboxes are dominated by companies innovating in payments services, especially by those testing crypto-based solutions. About 30 percent of sandbox projects involve payments (including remittance or digital transaction accounts) and nearly 30 percent involve market infrastructure (exchanges, clearing and settlement, escrow services) and wholesale financial services. Perhaps not surprisingly, blockchain and crypto-asset projects collectively make up nearly 25 percent of projects accepted into the sandboxes covered by our survey. These business models range from digital asset exchanges to blockchain-enabled trade finance to settlement infrastructure for cross-border remittances.

    Companies Participating in Sandboxes (by Sector)
    Companies Participating in Regulatory Sandboxes (by Sector)
    Source: CGAP analysis of public information from 12 jurisdictions.

  2. Evidence of sandbox-driven regulatory change is weak. In contrast to the self-reported survey results, we see little evidence that sandbox programs have generated formal regulatory change or modernization. Of course, the impact from sandbox programs may be occurring at a more informal level (e.g., by helping regulators reconsider interpretation of existing rules) or may take longer to manifest.
     
  3. Most sandbox-tested innovations do not target excluded and underserved customers at the base of the pyramid. Less than 25 percent of sandbox tests focus on business models or technologies that explicitly address financial inclusion barriers or aim to address the financial needs of poor people. Aside from financial inclusion-focused sandboxes in Sierra Leone and Mozambique, only a handful of sandbox tests — including NOW Money (Abu Dhabi and Bahrain sandboxes) and Rahi Payment Systems (Rwanda) — overtly focus on projects for the unbanked. The number would be higher if we counted all services that theoretically may be relevant to the excluded and undeserved, regardless of the actual goals of the firms using sandboxes.

These findings raise the question as to whether sandboxes are living up to their potential and whether their impact can be improved with design tweaks. We believe that shifting to thematic sandboxes would be effective. Thematic sandboxes are designed around more specific policy goals than other types of sandboxes and could be used to advance financial inclusion policy. To do this, a regulator would need to design an inclusion-themed sandbox specifically to address a key regulatory barrier.

Creating a Financial Inclusion-Themed Regulatory Sandbox
 

Creating a Financial Inclusion-Themed Regulatory Sandbox

 

This type of sandbox could be used for a range of reforms, both basic (e.g., enabling nonbank e-money issuers) and advanced (e.g., regulating ICOs). The main benefit would be allowing regulators to focus their energies on regulatory changes affecting financial inclusion while collecting evidence to support those changes and de-risking the implementation process. We are keen to work with regulators willing to test this new approach.

This post is part of CGAP's "Regulatory Sandboxes: What Have We Learned So Far?" blog series. The series takes a critical look at the concept of the regulatory sandbox and how it has evolved in different parts of the world.

Comments

21 May 2019 Submitted by Jose Sanin (not verified)

Interesting article. A sandbox is a pretty young concept that is hard to sell and harder to implement in markets whit weak (or perceived as weak) institutions. I am not surprised the evidence doesn't show a big impact of sandboxes in financial inclusion yet.
I also hope, on the topic of thematic sandboxes, we start seeing international sandbox collaborations between regulators from both sides of international remittance corridors: this could vastly impact i) understanding of regulators of new International Remittance business and operational models; ii) homogenise regulation across corridors, making it easier to comply and oversight at both sides (and cheaper, and more impactful) ; and boost financial inclusion levels (especially on the receiving end).

Add new comment

CAPTCHA