What Should We Realistically Expect from Regulatory Sandboxes?

Regulatory sandboxes are garnering attention as a potential way to unleash financial sector innovation, and it would not be surprising to see a wave of new sandboxes appear. Ivo Jenik and Kate Lauer, who look at regulatory sandboxes and their possible benefits for financial inclusion, describe a sandbox as a “framework set up by a financial sector regulator to allow small-scale, live testing of innovations by private firms in a controlled environment under the regulator’s supervision.” They lay out how sandboxes set objectives, identify eligibility and set parameters for experiments.

As Ivo and Kate point out, most formal sandboxes are in developed economies and arose out of a post-2008 desire to promote innovation amid heightened demands for regulatory compliance. There is also enthusiasm for extending formal sandbox approaches to low-income countries to promote financial inclusion.

To help us consider pragmatically the potential of sandboxes, let’s look at three regulatory issues that remain critical to the expansion of digital finance:

1. Licensing nonbank store of value issuers. Key to many transformative digital finance markets is the licensing of nonbanks to issue small-value transaction accounts (such as bKash in Bangladesh, WeChat in China and Wave Money in Myanmar). Businesses like these are substantial enterprises with millions of dollars in capital, special-purpose technology platforms and entrepreneurial managers. It is possible that some elements of their services could be tested in a sandbox, but such businesses cannot launch without a lot of up-front capital and technological investments. Sandboxes that limit experiments to small-scale “tests” are not likely to attract new businesses that require large investments or technology installations from the outset.

2. Allowing networks of cash-in and cash-out agents. A sandbox might enable providers to test remote oversight and basic business incentives for cash-in and cash-out agents (say, a pilot of a few hundred agents). This should be useful for evaluating agent onboarding, contracts and oversight — a basic proof of concept. The bigger challenge for cash-in and cash-out networks is that they evolve and must eventually operate at a large scale if they are to make a dent in financial inclusion. Initially testing a small number of agents might be a useful, and a sandbox might allow existing agent networks to test new features. However, a sandbox does not remove the need to eventually adjust regulation and supervision for large-scale deployments.

3. Piloting established providers’ digital credit products. In many markets, new entrants and even large existing players are testing new digital credit services. These products offer credit in ways that do not fit legacy regulations. A sandbox that allows quick testing of a product (within scale and other limits) would be valuable for regulators and lenders. A sandbox might help central banks more quickly test new services without having to change regulations for each innovation. But if digital credit products are to eventually scale, regulations might need to be changed down the road.

These examples help illustrate the nuance needed to assess when and how sandboxes might be useful. A few initial observations emerge. One is that sandboxes help facilitate a safe place for dialogue between regulators and industry; they offer a way to legitimize and open up discussions. This dialogue is important for regulators to learn, and it gives businesses more certainty in the testing. The strength of sandboxes appears to be their speed of testing, but the limitation is that tests are usually incremental. Transformative change will nearly always require more complete regulatory reform (to say nothing of the supervisory capacities needed to oversee new products and players in the market).

Another observation is that sandboxes are only one tool available to regulators, who are often changing regulations and experimenting in other ways. If an innovation has already been tested in another country, for example, regulators often look outside their jurisdictions for lessons. In some contexts, regulators can issue letters of no objection to allow experiments, a less formal “test-and-learn” approach. The central bank in the Philippines has permitted time-limited pilot tests under letters of no objection for a number of years. The “wait-and-see” approach China applied to WeChat mobile wallets informally allowed experiments and introduced regulations over time.

What might these initial observations suggest about the value of sandboxes? First, the relative value of sandboxes is context specific. Sandboxes may be the first step in opening dialogue or changing organizational culture in places where there is an established, closed or conservative approach to regulation. In other contexts, we may find there is already plenty of dialogue and flexibility to accommodate innovation, leaving less need for a formal sandbox.

Simone di Castri and Ariadne Plaitakis make an important point that sandboxes’ ability to promote innovation will be limited unless they are a part of a wider ecosystem that supports FinTechs. The authors suggest nine measures beyond sandboxes that are critical to support innovation.

My intuition, at this early stage, is that sandboxes can be a useful tool for incremental innovation and changing institutional culture, especially when they are designed for a specific context and with a clear set of objectives. But sandboxes may not always be the best or most practical option, and they are unlikely to replace the need for bolder transformative changes to regulation and supervision.