London and Singapore both let fintech companies test novel services with regulatory support, but neither turns an experiment automatically into a licensed business. The better location is the one whose customers, payment infrastructure and licensing regime match the intended commercial product. Founders should plan the authorisation, capital, safeguarding, technology-risk and distribution work before entering a sandbox.

London's FCA sandbox is broad and continuously open to eligible firms serving the UK market. Singapore's MAS offers several sandbox routes, including customised testing and predefined options, alongside a detailed Payment Services Act licensing framework. The distinction matters because a useful test produces evidence for a specific market, not a portable regulatory approval.

A sandbox answers a testable question, not the whole business case

The FCA Regulatory Sandbox supports live tests with real consumers in a controlled environment. It accepts authorised firms, unauthorised firms seeking authorisation and technology companies that support regulated businesses. The FCA is explicit that the sandbox is not regulatory exemption. A firm conducting a regulated activity must have the appropriate permission or registration for the test.

The FCA's eligibility criteria require an in-scope proposition, genuine innovation, consumer benefit, readiness and a need for support. A well-developed testing plan, safeguards and adequate resources are positive indicators. This makes the programme useful when a founder has a focused uncertainty, such as how a novel payment method behaves with real users.

MAS describes its sandbox as a defined live environment in which selected legal or regulatory requirements may be relaxed for the experiment. Its official overview says that a firm leaving a successful test must comply fully with relevant requirements. That is the crucial boundary between experimentation and a durable commercial service.

London connects testing to a large UK buyer market

For a founder selling into UK banks, insurers or payment firms, London puts product teams near institutional buyers and the FCA. A test can help establish customer behaviour, operational controls and a route through rules that do not fit a conventional launch. It can also expose weaknesses before a large implementation.

Acceptance should not be marketed as proof that the FCA endorses the firm. The regulator publishes accepted propositions and may provide temporary permissions where appropriate, but the company remains responsible for compliance and customer safeguards. Procurement by a bank adds security, resilience, legal and third-party-risk reviews beyond the sandbox.

London's infrastructure context also matters. An open-banking payment, Faster Payments proposition and card service have different participants and rulebooks. A test should use the rails and counterparties expected in production. Otherwise it may validate an interface while leaving settlement, fraud allocation or reconciliation unresolved.

Singapore combines licensing detail with instant-payment reach

Singapore's Payment Services Act creates regulated categories and licence types for payment providers. The MAS licensing page distinguishes standard and major payment institutions, sets eligibility and capital conditions, and describes expectations for local presence, governance, compliance, audit and technology risk. A company should map its activities to those categories before designing a test.

The market also has established domestic payment infrastructure. PayNow allows customers of participating banks and major payment institutions to transfer Singapore dollars through FAST, using identifiers such as a mobile number or company number rather than the recipient's bank account details. For a payments founder, access to this infrastructure can be more commercially important than the availability of a sandbox.

MAS Sandbox Express is narrower than its name may suggest. The current official Sandbox Express page identifies predefined activities and notes that remittance ceased to be an eligible Express activity when the Payment Services Act began. Founders should confirm the applicable route rather than assume a fast-track test exists for every payment product.

Commercialisation begins when the temporary conditions end

A strong test plan starts with the production destination. It identifies the licence or partner required after testing, the capital and staffing needed, how customer funds will be protected, the evidence needed by banks and how users will be migrated or exited if the experiment stops.

The most useful measures are tied to the regulatory and commercial hypothesis. These might include completion and failure rates, fraud and complaints, manual intervention, reconciliation accuracy, control exceptions and support demand. Sign-ups alone say little about operational viability. A low-volume test may also fail to reveal performance under stress.

Founders should ask prospective customers what evidence they need before procurement. If a bank requires independent penetration testing, audited controls or a long production record, sandbox participation may help only one part of the case. The company should budget for the gap between a successful experiment and an enterprise rollout.

Choose by rails, customers and licence ownership

London is a logical first base for a product designed around UK regulation, UK bank customers or UK payment infrastructure. Singapore is logical for a product whose first users and partners sit there or whose Asian expansion depends on a Singapore-licensed entity. Neither city should be selected only because its regulator is known for fintech engagement.

A company can operate in both, but each market needs accountable ownership. Shared technology can reduce duplication, while local teams own regulatory reporting, partnerships, customer terms and incident response. Cross-border outsourcing and data transfers must be designed into the operating model.

The second market should follow a repeatable product and evidence of demand. Re-running a test can be justified when local rails or rules materially change the risk. It is wasteful when the company is using regulatory programmes as a substitute for customer discovery.

Limitations of the comparison

This article compares Greater London with the city-state of Singapore, so the geographic units are not equivalent. FCA and MAS programme pages show rules and accepted pathways, not the survival, revenue or authorisation outcomes of all participants. Public lists can also change after the research date.

Before publication, a cohort of comparable payments firms should be followed from application through test, licensing and commercial use. Interviews with customers and former participants, including firms that stopped, are necessary to test whether either route reduced time or cost.

Reporting by Shoreditch Talk