The Embedded Finance Infrastructure Second Wave

Banking-as-a-Service providers lost 60% of their market value in 2023 as compliance issues mounted. The embedded finance infrastructure supporting them continued growing beneath the regulatory chaos. A second wave is now emerging with different risk profiles that most investors are overlooking.

The fintech disruption thesis was clean and compelling, but largely wrong about where the value would end up.

For a decade, the story was about challengers displacing incumbents. Better apps, lower fees, and no branches meant challenger banks were going to capture retail banking. BNPL was going to eliminate credit cards. Neobrokers were democratizing investing for the masses. Incumbent banks were slow, fat targets.

The incumbents are mostly still there. Several of the challengers are working through credit quality problems, regulatory tightening, or profitability questions that got papered over during the zero-interest-rate era. The layer of the market that compounded value most consistently through all of it is the layer nobody spent much time celebrating: the infrastructure beneath.

What the First Wave Left Behind

The most important thing the first wave of fintech built was the toolkit for anyone to offer financial products without becoming a bank.

Stripe is the clearest example. The company processes over $1 trillion in payment volume annually, placing it among the largest financial infrastructure operators on the planet. Stripe Treasury provides banking-as-a-service for platforms. Stripe Capital deploys merchant lending. Stripe Issuing enables card programs. The product has expanded well beyond payments into the financial operating system for the internet economy. A company building on Stripe is assembling financial primitives from infrastructure the way it assembles compute resources from a cloud provider.

Adyen operated the same logic at enterprise scale, and its 2023 correction is instructive. A 39% single-day stock collapse followed a growth warning, and most market commentary treated it as an infrastructure failure. The company's gross margin sits at 43% through the correction: pricing power consistent with an infrastructure moat rather than an application-layer product. Adyen refocused on enterprise unified commerce and kept growing. The market read deceleration as a crisis. The financial structure said otherwise.

The Second Wave: Embedded Finance Infrastructure

The embedded finance infrastructure second wave is the systematic proliferation of financial products through non-financial operators, built on top of the API layer that the first wave established.

Shopify Capital has disbursed over $5 billion in cumulative merchant funding through a model that illustrates the data-advantage thesis in financial services. Shopify's underwriting uses live transaction data from merchants' own stores. A bank assessing the same merchant for working capital starts with externally sourced credit data. Shopify starts with verified revenue history. That gap in informational advantage directly compounds into pricing and risk control. Shopify's data advantage over external lenders is significant: it holds verified live revenue history that no outside institution can access through conventional underwriting channels.

The same mechanics are visible in Uber's financial products, Amazon's embedded merchant lending, and Walmart's One Financial Services platform. These are commerce operators who discovered that financial products are higher-margin and stickier than commerce alone when the right data infrastructure sits underneath them.

Three foundational shifts compose the second wave:

From Product Competition to Infrastructure Positioning

The first wave competed on product quality. Better UX, lower fees, faster onboarding. The second wave competes on infrastructure positioning: whose rails does everyone build on, and what does that relationship generate over time? This is why Stripe and Adyen have compounded better long-term economics than most of the challenger banks whose growth they enabled. Infrastructure positioning generates switching costs. What consumer product quality builds is preference. Over any ten-year horizon, those are fundamentally different kinds of assets.

From Direct Disruption to Embedding

The challenger bank model tried to take customers from incumbents through competitive positioning. The embedded finance model routes around the incumbent category entirely. A Shopify merchant managing business finances through Shopify Balance isn't choosing Shopify over their bank in any conscious competitive selection process. The banking layer isn't a deliberate choice at all. It's a feature of the commerce infrastructure they already use. That level of disintermediation is harder to reverse than any market share loss measured in switching rates.

The Synapse Correction

The Synapse BaaS collapse in May 2024 clarified what happens when the middleware layer fails. Synapse's bankruptcy left over 100,000 customers temporarily locked out of accounts because reconciliation between Synapse, its sponsor banks, and its client applications had no clean resolution mechanism. The correction was swift: BaaS middleware providers are now under significant FDIC and OCC scrutiny, and operators who had treated BaaS as a simple API integration are reassessing the concentration and reconciliation risk that was always present and wasn't being priced into partnership decisions.

The market is moving toward either direct infrastructure relationships with sponsor banks or middleware arrangements with explicit contractual accountability for fund reconciliation and operational continuity scenarios.

The UK's Open Banking Lead

The UK's open banking framework offers a useful window into what mature financial infrastructure adoption looks like at scale.

Mandated by the CMA in 2017 and overseen by the FCA, the UK open banking ecosystem has grown to over 8 million users actively sharing financial data with third-party apps. Variable recurring payments (open banking's equivalent of direct debit) are finding uptake in recurring billing use cases that previously required card or direct debit infrastructure. The trajectory was slow-then-compound: a decade of infrastructure investment generating an adoption inflection as use cases multiplied beyond the initial payment application.

The US equivalent is CFPB Rule 1033 (finalized October 2024), which requires financial institutions to share consumer financial data with authorized third parties on request. Implementation runs from 2026 to 2030, depending on institution size. The US open banking infrastructure layer is being established now, roughly ten years behind the UK.

What the KPIs Signal

Infrastructure positions are measurable in ways that disruption narratives aren't.

Payment volume compounding is the leading indicator. Stripe's $1 trillion-plus in annual TPV puts it in company with Visa and Mastercard by volume measure: organizations the market has always categorized as infrastructure rather than disruptors. Adyen's EBITDA margin of 53% (exceeding the 46% reported in 2023) through a growth correction suggests pricing power that product-dependent businesses rarely sustain under competitive pressure. Shopify Capital's $5.5 billion in cumulative merchant funding, deployed through data-advantage underwriting, shows financial products generating economics through operational data rather than customer acquisition spend.

Net Revenue Retention for infrastructure providers consistently runs above 115%, because the relationship between infrastructure and operator deepens as the operator grows. Each new product, geography, or customer segment the operator adds compounds the infrastructure's own embedded value. That's a growth model that runs on different mechanics from customer acquisition, and a more resilient one when markets tighten.

For a related analysis of how platform economics and data compounding create durable competitive advantages, read The Full Platform Bill.

Reading the Stack from Your Position

Operators sitting on one side or another of financial infrastructure decisions should ask three questions that clarify the strategic picture.

Which infrastructure layer do you depend on?

Most operators don't need to build financial products from scratch. They need embedded capabilities that their customers want bundled into an existing relationship. The post-Synapse recalibration means due diligence on middleware providers needs to be rigorous, with contractual accountability for fund reconciliation, sponsor bank relationships, and operational continuity scenarios, all spelled out before the contract is signed.

Where is your data advantage?

Shopify Capital works because Shopify has live merchant revenue data. Uber's financial products work because Uber has a transaction history. If your operational data gives you an informational advantage over external financial providers, embedded finance may be a genuine product opportunity rather than a distribution play. Where the data advantage is thin, distributing an external financial product through your interface is the sharper commercial choice.

What's your concentration exposure?

If your core operations run through a single payment provider, a single BaaS layer, or a single data aggregator, the Synapse case study is directly relevant to your strategic context. Infrastructure concentration risk in financial services is operational risk. It didn't get modeled carefully enough by most operators who discovered it exists only after something failed.

For European operators weighing these decisions against longer capital cycles, our analysis of The Mittelstand Long Game provides useful parallel context on build-versus-embed decisions in capital-constrained environments.

The first wave of fintech was loud and largely about competitive displacement. The second wave is structural, quieter, and compounding economics that application-layer operators are finding difficult to replicate. The operators who understand where the infrastructure value sits are making sharper decisions about what to build, what to embed, and where the relationship risk they haven't accounted for might be living.

Three Questions for Your Next Financial Services Strategy Review

1. Map your financial infrastructure dependencies. For every financial product or capability in your operations (payments, lending, data sharing, card issuance), identify the layer it sits on: direct sponsor bank relationship, middleware BaaS, or full-stack provider. Confirm contractual accountability for fund reconciliation and operational continuity scenarios in writing. If the contract doesn't address it, the risk sits with you.

2. Assess your data position before pursuing embedded finance. The product opportunity in embedded financial services is strongest where your operational data gives you an informational edge over external providers. Before building or sourcing an embedded finance product, audit what transaction or behavioral data you hold on your customers. Where the data advantage is thin, a distribution partnership with an established financial product provider typically makes more economic sense than building infrastructure.

3. If you're evaluating BaaS or embedded finance providers post-2024, ask the hard questions. Specifically: how does fund reconciliation work between sponsor banks and middleware? What is the contractual resolution mechanism in a disruption or insolvency scenario? Who bears operational liability if the middleware layer fails? These are questions many operators didn't think to ask before Synapse showed why they needed to.