Barclays, HSBC, and Lloyds collectively spent an estimated £1 billion complying with the United Kingdom's open banking mandates, yet the direct return on that massive infrastructure investment remains anchored at zero. Open banking has been live in the UK for nearly eight years. The European Union's PSD2 framework has forced similar capital expenditures since 2019, which means the continent's largest financial institutions have funded a continental data grid without a mechanism to charge for the transmission. The United States is finally catching up to this dynamic, with the Consumer Financial Protection Bureau finalizing its Section 1033 rulemaking in late 2024. The result is a global regulatory push that mandates access but ignores economics. As of mid-2026, the open banking revenue model remains fundamentally broken for the institutions actually building the rails. No one in the ecosystem, not the source banks, not the middleware aggregators, and not the consumer fintechs riding on their rails, has built a revenue model that works at scale. The industry keeps talking about monetization as though it sits one product cycle away, treating the deficit as a mere adoption lag. It does not. The problem is structural, because the regulatory incentives that created the system are the exact same forces preventing anyone from fixing it.
Open banking created a giant infrastructure layer with no viable business model attached, and the participants most capable of fixing it have the least incentive to try.
Revenue Model Was: The Optimists Are Misreading The Adoption Metrics
The dominant narrative says open banking is still in the early adoption phase, and monetization will naturally follow once transaction volume scales. The proof points always sound impressive in isolation. The bull case relies on a cluster of massive valuations and user metrics, anchored by Visa paying $1.8 billion for Tink in 2022 and Mastercard acquiring Finicity for $825 million in 2020, alongside Plaid generating north of $300 million in annual revenue across 8,000 financial apps while Open Banking Limited tracks more than 11 million active UK users as of early 2025. The optimistic argument says the rails are being built, the passengers are arriving, and the ticket revenue is next.
That reading is incredibly shallow. Plaid's revenue is entirely real, but it sits almost exclusively on the aggregator layer, not on the balance sheets of the banks that actually built and maintain the underlying APIs. Tink's purchase price reflected Visa's strategic need to control the European data pipe, not a bank-level proof of monetization. Finicity gave Mastercard a defensive position in data connectivity, not a case study in how the source banks can charge for access. And those 11 million UK users? A massive share of them are using open banking entirely passively, routed through lending journeys and identity checks that they never framed as a separate product decision.
Two more outside views sharpen this exact same point. Juniper Research projected that open banking payments would reach $116 billion globally by 2027, but that specific forecast says absolutely nothing about who actually keeps the margin on those payments. Accenture has also repeatedly argued that API-led banking can raise institutional revenue, yet bank finance teams still struggle to connect that broad thesis to a live, functioning pricing schedule. Forecasts about total market size are not the same as forecasts about who gets paid, which leaves most of the market confusing transaction volume with value capture.
The fundamental disconnect is obvious. Plaid and TrueLayer sit directly between the source banks and the end customer, charging per-API-call fees that flow entirely away from the institutions bearing the heavy compliance cost. The banks built the core asset. A middleware firm built the cleaner interface. The middleware firm got paid.
Most analysts have this dynamic completely backwards. They track aggregator valuations and call it definitive proof of a healthy ecosystem. The banks, which actually built the thing, are running a high-maintenance public utility with no meter attached.
Four Structural Failures Breaking The Open Banking Revenue Model
The first failure is the severe cost-revenue inversion. PSD2 and the UK's CMA Order required banks to open their APIs but provided absolutely no mechanism for banks to charge third-party providers for that access. The policy logic was purely focused on consumer empowerment. The financial logic was never completed. A 2023 Open Banking Impact Report from the Open Banking Implementation Entity showed that UK banks' direct API call volumes exceeded 1 billion per month. Every single one of those billion requests requires server capacity, security monitoring, and compliance auditing, all served at zero marginal revenue to the institutions hosting the data. Regulators built an access mandate without a compensation model, prioritizing consumer rights while leaving the financial logic entirely incomplete, and they have shown no appetite to revisit this imbalance.
The second failure is aggregator capture. The companies that actually monetize open banking are not the banks that hold the deposits. They are Plaid, Yapily, TrueLayer, and Finicity. These firms act as middleware layers, translating fragmented and often clunky bank APIs into standardized, developer-friendly products. They then charge consumer and enterprise fintechs for access to that clean data feed. The value created by the bank's proprietary data asset is captured entirely one layer up by a company that simply built a cleaner label for it. The source banks get absolutely none of this margin. In markets where aggregation wins the developer experience battle, the source layer inevitably becomes a pure cost center while the interface layer becomes the actual business.
The third failure is payments adoption stalling out. Open banking payments were originally supposed to displace traditional card rails by offering merchants lower settlement costs. They have not achieved anything close to that displacement. Card-present and card-not-present transactions still dominate UK retail payments by a massive factor of roughly 20 to 1 in volume terms, according to Q1 2026 data from the Payment Systems Regulator. Variable recurring payments represent the mechanism most likely to compete with traditional direct debit, yet they remain an obscure niche rather than a default checkout option. Mastercard and Visa are not worried about this transition because the threat of account-to-account payment displacement has not materialized on any timeline the market once forecast.
The fourth failure is the consent architecture. Open banking's data sharing relies entirely on a user consent architecture that expires, degrades, and creates severe friction at every single renewal point. Plaid faced a class-action settlement in 2022 that was driven partly over these exact consent practices and how data was surfaced to the end user. This friction is not a superficial user experience problem waiting for a slightly better interface flow. It is a hard legal constraint that actively caps the data pipelines that any future monetization depends on. A predictable subscription revenue model simply cannot lean on a data access foundation that resets every 90 days. That regulatory reset forces repeated re-authentication at the exact moment a financial product needs smooth retention, guaranteeing a steady churn rate that degrades the lifetime value of the acquired customer.
A fifth evidence point makes this exact same case from a strictly transactional angle. The Payment Systems Regulator's Q2 2026 dashboard put open banking payments at a mere 0.6% of total UK retail transaction value. That is the sort of microscopic market share number that destroys the fantasy pricing models pitched by early-stage payment aggregators. It also confirms the obvious reality of the consumer market. The ecosystem has proved that early adopters will occasionally try the feature when prompted during a specialized checkout flow. It has absolutely not proved that the feature can generate enough consistent volume to earn back its massive compliance and engineering costs.
Each of these failures actively reinforces the others, creating a negative feedback loop. Together, they do not describe a difficult engineering problem waiting for a clever technical patch. They describe a fundamentally broken system.
The Application-Layer Counterargument Deserves Scrutiny
The strongest pushback against this pessimistic outlook runs as follows: open banking monetization is already happening, just not where the legacy incumbents expected to find it. Goldman Sachs's Marcus, before its strategic retreat, and Intuit's massive financial data business both built substantial product revenue directly on top of open banking infrastructure. The core argument holds that the ultimate financial value is not located in the transmission pipes. It is located entirely in the end-user applications. Banks should not expect to be the primary monetizers of this ecosystem, and they should instead expect to be the passive substrate that enables it. The smart money, under this specific view, is heavily concentrated on application-layer companies like Mono in Africa, TrueLayer in Europe, and MX Technologies in the United States, all of which have raised venture capital in the hundreds of millions on exactly this thesis.
It is a highly coherent argument. It is also a massive concession, not a genuine rebuttal. Saying that retail banks should not expect any revenue from their own proprietary data infrastructure simply because agile middleware players captured it does not actually defend the sustainability of the model. It merely explains why the model fails for the largest and most important actors in the system. Banks represent the single largest category of open banking infrastructure spend globally. If they cannot monetize that massive capital expenditure, the system as a whole operates in a permanent deficit. A business model that consistently enriches the connector while starving the source is not structurally stable over a long time horizon.
One specific data point would fundamentally change this analysis. If a major retail bank operating in a PSD2 jurisdiction, such as Santander or ING, announced a commercially validated premium API product generating more than €50 million annually in net new revenue by the end of 2026, that would provide a genuine signal of market viability. So far, no such product exists at that required scale. Santander has experimented heavily with API marketplaces to test developer appetite. BBVA has done the exact same thing to probe for willingness to pay. Neither institution has successfully turned the experiment into a permanent line item that can alter the bank's fundamental earnings model.
Strategic Fallout Across The Ecosystem
The broken revenue model has vastly different consequences depending on where the specific investment decision sits. The financial damage is not abstract. It is showing up in venture capital investment write-downs, enterprise procurement choices, and software product roadmaps right now.
Institutional Investors
Any venture capital or private equity fund still holding open banking aggregator equity at peak 2021 multiples needs to immediately revisit the core thesis of their investment. The Plaid valuation story, which once peaked at $13.4 billion after SoftBank's backing, has already been marked down significantly in secondary market estimates. The issue for late-stage investors is not whether open banking has inherent utility. It clearly does. The issue is that the enterprise value is concentrated in just two or three scale players globally, and at current capital efficiency ratios, the actual cash returns to late-stage investors look incredibly thin. The public and private markets consistently reward the software layer that charges the end user, not the infrastructure layer that merely complies with the regulation.
The critical trigger for investors to watch is the Consumer Financial Protection Bureau's Section 1033 implementation timeline. The rule requires readers financial institutions to make consumer-permissioned data available by strictly defined deadlines, with the largest institutions facing mandatory compliance as early as April 2026. If readers banks replicate the established European pattern of heavy compliance spending without any corresponding monetization strategy, the total addressable market for aggregators grows, but the underlying unit economics do not improve at all. Investors should ask aggregator management teams one blunt, uncomfortable question during their next board meeting: what is the exact gross margin per API call when the system reaches ten times its current volume? If the answer does not clear the standard software-as-a-service hurdle rate, then top-line growth is just a margin trap dressed up as an adoption curve.
The near-term action for asset managers is simple. Re-underwrite every single open banking position against strict, bank-grade economic realities, not inflated software-style revenue multiples. Demand granular disclosure of the average revenue per connected account, the specific re-authentication loss rates, and the total churn immediately following a consent renewal prompt. If a portfolio company cannot definitively show that the average bank connection lasts long enough to justify its initial customer acquisition cost, the fund must cut the exposure. The next quarter is not the time to sit back and wait for a miraculous policy fix that has already taken eight years to fail.
Related MarketIntel briefing: read The Open Banking Revenue Model Remains Broken for a connected view on this market signal.
Source context: readers can compare this market signal with broader data from S&P Global and Reuters.