Sanya Juneja (Principal Researcher, Fii), Pavle Avramovic (Fii), Paula da Cunha Duarte (CCAF), Hakan Eroglu (BIS), Jon Frost (BIS), Tanya Ghuman (Fii), Asli Ergun Konuk (CCAF), Teresa Lam (Fii), Benedicte Nolens (BIS), Jermy Prenio (BIS), Bill Roberts (CCAF), Friederike Rühmann (BIS/World Bank), Nipuna Varman (CCAF), and Bryan Zhang (CCAF).
This timely study examines the conditions that allow open finance ecosystems to take develop and achieve sustainable impact in Emerging Markets and Developing Economies (EMDEs). Drawing on around 30 semi-structured interviews and extensive desk-based research across 9 markets, Brazil, Egypt, Ghana, India, Indonesia, Nigeria, the Philippines, Saudi Arabia, and South Africa, it moves the analysis beyond adoption status towards the structural drivers of a resilient ecosystem. Building on the ‘Global and APAC State of Open Banking and Open Finance’ reports, it introduces a novel framework, applicable well beyond the markets studied, that treats incentives, liability, and performance measurement as 3interconnected pillars to be addressed together rather than in isolation.
Produced by the Cambridge Centre for Alternative Finance (CCAF) at Cambridge Judge Business School, University of Cambridge, together with Financial Innovation for Impact (Fii) and in collaboration with the Bank for International Settlements (BIS), and with the support of the UK Foreign, Commonwealth and Development Office (FCDO), the report offers regulators, financial authorities, and industry a practical and evidence-informed basis for the design choices that shape open finance. It is also clear that open finance may not be the right priority for every economy at every stage, and that those choices should be weighed against each country’s own objectives and market context.
Highlights from the report
1
The 3 pillars are interdependent, and a gap in one can be partly offset by another
The report’s central contribution is a novel framework that treats incentives, liability, and performance measurement as interdependent rather than separate questions. Because the pillars interact, a shortfall in one area can be at least partially compensated by another: predictable redress, for example, can act as a substitute incentive where direct incentives to participate are weak. This can allow regulators to sequence priorities and rebalance over time rather than resolve every question at once, which is particularly valuable for authorities operating under institutional and resource constraints.
2
Mandates set the floor for participation, but incentives determine the impact
Whether an ecosystem is regulation-led or market-driven, it may underperform if key actors do not perceive clear and sustainable value in taking part. Mandates can establish a baseline, but financial and non-financial incentives, from reciprocity and guaranteed data quality to grants, tax relief, and sandbox access, are often needed to foster deeper engagement. A commercial model is one tool among many, and choosing not to implement one can be a legitimate policy decision.
3
Incentives should be calibrated to local market structure, not assumed to revolve around banks
Banks are not always the principal data holders. In many EMDE contexts, mobile network operators and their mobile money subsidiaries dominate payments activity and control the transactional data that underpins credit decisioning, as in Kenya and Uganda. Understanding the competitive environment and the distribution of dominant data holders before designing incentives can therefore matter more than borrowing a bank-centric template from advanced economies.
4
Liability frameworks tend to build on existing legal architecture rather than replace it
Open finance does not create accountability from scratch – it draws on overlapping financial, data protection and consumer protection regimes while introducing genuinely novel multi-party features. No single liability model emerges as universally superior, with single-party, multiple-party, fault-based and hybrid approaches each offering different trade-offs that depend on institutional capacity, digital infrastructure, and policy priorities. Consent itself emerges as a regulatory tool and clear, predictable liability can act as an incentive in its own right.
5
Performance measurement must go beyond technical and adoption metrics to track policy objectives
A system that is fast, reliable, and widely used may still fall short if it mainly serves the already-served or delivers little measurable progress against the policy objectives that justified intervention in the first place. The report sets out five design principles, attribution, validity, outcome orientation, proportionality and granularity, and proposes indicators that move beyond aggregate counts towards meaningful outcomes across competition, innovation, consumer protection, and financial inclusion. Whether to publish metrics can be approached through a proportional, tiered framework rather than a binary choice.
6
Customers are too often the missing piece, and belong at the centre of ecosystem design
Customers do not need to understand the underlying infrastructure, but they do need to experience clear convenience and tangible value. Adoption follows more naturally when data sharing is embedded in moments of genuine financial need, as illustrated by the integration of open finance into everyday payment journeys in Brazil, rather than presented as a standalone consent exercise. It is ultimately the customer’s decision to share data that allows the ecosystem to function.