How Fintech Partnerships Actually Work: Banks, BaaS, and Distribution
Chime has 22 million customers. It offers checking accounts, savings accounts, a debit card, and direct deposit. It looks like a bank. It markets like a bank. It is not a bank.
Chime is a fintech company whose banking services are provided by Bancorp Bank and Stride Bank — two banks most people have never heard of. Every Chime checking account is actually a demand deposit account held at one of these banks. Every Chime debit card is issued under these banks' charter. The FDIC insurance sticker on Chime's website refers to these banks, not to Chime.
This is the partnership model that built modern fintech. The vast majority of fintech companies that offer banking products — Chime, Mercury, Brex, Ramp, Current, Dave — do not hold bank charters. They partner with chartered banks that provide the regulatory license, and the fintech provides the technology and customer experience.
Understanding how these partnerships work is essential for anyone building, investing in, or competing with fintech products. The partnership model determines the economics, the regulatory exposure, the customer experience, and increasingly, the survival of the company.
The Four Partnership Models
Fintech-bank partnerships are not one-size-fits-all. They fall into four distinct models, each with different economics, different control dynamics, and different risk profiles.
| Model | Who Owns the Customer | Who Provides the License | Revenue Split | Example |
|---|---|---|---|---|
| Banking-as-a-Service (BaaS) | Fintech | Partner bank | Bank earns fees per account/transaction; fintech keeps interchange and customer revenue | Chime + Bancorp Bank, Mercury + Choice Financial |
| White-Label | Bank | Bank (its own) | Bank pays fintech SaaS fee or revenue share for technology | Q2 powering digital banking for regional banks |
| Referral / Affiliate | Shared | Bank (its own) | Per-lead or per-conversion fee | NerdWallet referring users to Marcus savings accounts |
| Strategic Investment | Each owns their own | Each uses their own | Equity upside + preferential commercial terms | Goldman Sachs investing in Stripe |
Each model reflects a different answer to the fundamental question: who brings the customers, and who brings the license?
Model 1: Banking-as-a-Service (BaaS)
BaaS is the model that created the modern fintech industry. The bank provides its charter, its regulatory compliance infrastructure, its FDIC insurance, and its access to payment networks. The fintech provides the customer-facing product, the technology, and the user acquisition.
The economics work roughly like this: the partner bank earns a per-account fee (typically $1-5 per active account per month), a share of interchange revenue (usually 10-30% of the interchange earned on debit or credit card transactions), and sometimes a percentage of net interest income on deposits or loans. The fintech keeps the remainder.
For the bank, BaaS is a growth strategy that requires minimal investment in technology or marketing. A community bank with $500 million in assets might support a fintech with 2 million accounts, generating deposit volume and fee income that would be impossible to achieve through its branch network alone.
For the fintech, BaaS provides the regulatory license without the 18-36 month process of obtaining a bank charter, the $20-50 million in capital required to start a bank, and the ongoing burden of bank-level regulatory examinations.
The middleware layer — companies like Unit, Treasury Prime, Synapse (until its collapse), and Synctera — emerged to make BaaS easier. These companies sit between the fintech and the partner bank, providing APIs that abstract away the bank's core banking system, compliance workflows, and ledger management. The fintech integrates with the middleware API; the middleware handles the integration with the bank.
How money flows in a BaaS relationship:
When a Chime customer receives a direct deposit, the employer's payroll provider sends an ACH credit to Bancorp Bank, which credits it to the customer's account on Bancorp's core system. When that customer spends money with their Chime debit card, Bancorp is the card issuer and earns interchange from the merchant's acquiring bank via the Visa network. Bancorp keeps its share and passes the remainder to Chime.
Chime earns the majority of its revenue from interchange — estimated at $1.3 billion in 2023 based on its transaction volume and typical interchange rates for signature debit transactions. The partner bank earns fee income with virtually no customer acquisition cost.
Model 2: White-Label Technology
White-label is the inverse of BaaS. Instead of the fintech owning the customer and the bank providing the license, the bank owns the customer and the fintech provides the technology.
A regional bank with 200,000 customers wants to offer a modern mobile banking app. It does not have the engineering team to build one from scratch. Instead, it licenses a white-label digital banking platform from a company like Q2, Alkami, or Finastra. The platform is branded as the bank's own product. Customers never know a third party built it.
The economics are typically SaaS-based: the bank pays a per-user monthly fee (often $1-5 per active digital banking user) plus implementation fees. Some white-label providers charge a percentage of revenue generated through the platform.
For the fintech technology provider, white-label is a B2B SaaS business with predictable recurring revenue. For the bank, it provides competitive digital capabilities without building an engineering organization.
The limitation is speed and differentiation. When 500 community banks all run the same Q2 platform, their digital experiences converge. The bank that wants a truly distinctive product — different workflows, unique features, novel integrations — often finds white-label platforms too rigid.
Model 3: Referral and Affiliate Partnerships
The simplest partnership model: one party refers customers to the other and earns a fee for each conversion.
NerdWallet, Bankrate, and Credit Karma built billion-dollar businesses largely on referral economics. When a user clicks through from NerdWallet to open a Marcus savings account at Goldman Sachs, NerdWallet earns a referral fee — typically $50-200 per funded account, depending on the product and the competitive intensity of the category.
Credit card referrals are particularly lucrative. The cost to acquire a credit card customer through referral channels can be $150-400, which card issuers are willing to pay because a single active credit card customer can generate $500+ in annual revenue through interchange and interest.
For banks, referral partnerships provide customer acquisition at a predictable cost per acquisition. For fintechs and media companies, referral fees are high-margin revenue that scales with traffic.
The risk is dependency. When Google changes its search algorithm, referral businesses that depend on organic traffic can see their revenue evaporate overnight. When banks cut referral budgets during downturns, the income disappears. There is no recurring contractual obligation — it is pay-per-lead.
Model 4: Strategic Investment
The fourth model is not a product partnership at all — it is a financial one. Banks invest in fintech companies for equity upside and strategic positioning.
Goldman Sachs, JPMorgan, and Citi all operate venture arms that invest in fintech startups. Citigroup has invested in over 100 fintech companies. These investments serve multiple purposes: financial returns, early access to innovative technology, potential acquisition pipeline, and defensive positioning (better to invest in a potential disruptor than be disrupted by it).
Strategic investments often come with commercial agreements. A bank that invests in a payments startup might also become that startup's first enterprise customer or provide access to its merchant network.
The tension is real: the bank wants the fintech to succeed enough to justify the investment, but not so much that it threatens the bank's own business lines. This is why strategic investments in directly competitive products are rare — banks typically invest in companies that are complementary rather than substitutional.
Why Partner Banks Are Under Regulatory Pressure
The BaaS model worked smoothly for nearly a decade. Then regulators noticed.
Starting in 2022, federal regulators — the OCC, FDIC, and Federal Reserve — began issuing consent orders and enforcement actions against partner banks, citing failures to properly oversee their fintech relationships.
Blue Ridge Bank received a consent order from the OCC in 2022 requiring it to improve oversight of its fintech partnerships, including enhanced due diligence, transaction monitoring, and compliance controls. The bank was required to submit a plan for managing third-party risk.
Cross River Bank, one of the most prolific BaaS banks, received a consent order from the FDIC in 2023 related to fair lending practices in its fintech lending partnerships. The order required enhanced compliance monitoring and reporting.
Piermont Bank and several other community banks involved in BaaS received supervisory letters requiring them to slow fintech partner onboarding until compliance infrastructure caught up with growth.
The regulatory concern is straightforward: when a bank outsources customer-facing functions to a fintech, the bank remains legally responsible for compliance with all banking regulations — BSA/AML (Bank Secrecy Act / Anti-Money Laundering), fair lending, consumer protection, and safety and soundness. If the fintech fails to collect proper KYC (Know Your Customer) information or processes transactions that violate sanctions, it is the bank — not the fintech — that faces regulatory consequences.
Many partner banks grew their fintech programs far faster than their compliance teams could support. A community bank with 50 employees might have one compliance officer overseeing 15 fintech partners with millions of combined customers. Regulators concluded this was unsustainable.
The Synapse Lesson
The most dramatic illustration of partnership failure is Synapse Financial Technologies, the BaaS middleware provider that filed for bankruptcy in April 2024.
Synapse sat between fintech companies and their partner banks, managing the ledger, compliance workflows, and API integrations. At its peak, Synapse facilitated over $400 million in end-user deposits across multiple fintech-bank partnerships.
When Synapse filed for bankruptcy, chaos followed. The bankruptcy trustee reported a $65-85 million shortfall between what end users thought they had in their accounts and what was actually held at the partner banks. Customer funds were frozen for months. The FDIC, which insures deposits at the partner bank level, had difficulty determining which deposits belonged to which customers because Synapse's records were disorganized.
The Synapse collapse revealed the fragility of the middleware model. When the intermediary between the fintech and the bank fails, no one has a clear, real-time view of which customer owns which dollars. The FBO (For Benefit Of) account structure — where the bank holds a single omnibus account containing all fintech customer deposits, and the middleware maintains the sub-ledger — works only as long as the sub-ledger operator remains solvent and competent.
In response to the Synapse failure, the FDIC proposed new rules in January 2025 requiring banks to maintain detailed, reconciled records of individual beneficial owners within FBO accounts. The proposed rules would also require banks to be able to calculate deposit insurance coverage for each end user within 24 hours if the bank or the fintech fails.
What Makes Partnerships Work vs. Fail
After covering dozens of fintech-bank partnerships, patterns emerge in what separates the successful ones from the ones that blow up.
Successful partnerships have direct bank-fintech relationships. The partnerships that survive regulatory scrutiny are those where the bank has a direct contractual and operational relationship with the fintech — not mediated entirely through a middleware layer. The bank's compliance team reviews the fintech's onboarding flows, monitors transaction patterns, and has escalation paths when issues arise.
Successful partnerships have aligned economics. When the bank earns meaningful revenue from the partnership — not just a token fee — it invests in the compliance infrastructure needed to support it. When the bank is earning $50,000 a year from a fintech partnership that creates $5 million in compliance cost, the math does not work.
Failed partnerships have mismatched growth expectations. A fintech growing at 300% per year partnered with a bank whose compliance team grows at 10% per year is a collision waiting to happen. The fintech wants to onboard customers faster; the bank cannot review them fast enough. Corners get cut. Regulators notice.
Failed partnerships have unclear responsibility for compliance. When neither the fintech nor the bank takes clear ownership of BSA/AML monitoring, KYC verification, or complaint handling, gaps emerge. The fintech assumes the bank is handling it. The bank assumes the fintech is handling it. Nobody is handling it.
Failed partnerships have no contingency plan. If the fintech fails, what happens to customer deposits? If the bank receives a consent order and must offboard fintech partners, how quickly can the fintech migrate to a new bank? The Synapse collapse demonstrated that most participants had not thought through these scenarios.
Where the Market Is Heading
The BaaS model is not dying — it is maturing. The partnerships that survive the current regulatory tightening will be the ones that invest in compliance infrastructure proportional to their scale.
Several trends are shaping the next phase:
Fewer, larger partner banks. The era of tiny community banks serving as BaaS platforms is ending. Regulators are effectively pushing BaaS toward banks with sufficient compliance resources — meaning larger banks with dedicated fintech partnership teams. Column, Lead Bank, and Coastal Community Bank have invested heavily in this infrastructure.
Direct charter applications. Some fintechs, tired of partner bank dependency, are pursuing their own bank charters. SoFi acquired Golden Pacific Bancorp in 2022 to obtain a national bank charter. LendingClub acquired Radius Bank in 2021. Varo Bank received an OCC national bank charter in 2020. Chartering is expensive and slow, but it eliminates the partner bank dependency entirely.
Enhanced middleware with compliance built in. The next generation of BaaS platforms — Unit, Treasury Prime, and others — is building deeper compliance infrastructure, including real-time transaction monitoring, automated suspicious activity reporting, and individual-level ledger reconciliation designed to prevent another Synapse scenario.
The fundamental value proposition of fintech-bank partnerships remains intact: banks have licenses, fintechs have technology and distribution. The question is no longer whether these partnerships work, but how to structure them so that compliance, economics, and customer protection are built in from the start rather than bolted on after a crisis.
Key Takeaways
- The four fintech-bank partnership models — BaaS, white-label, referral, and strategic investment — each have fundamentally different economics, control dynamics, and risk profiles.
- BaaS is the model that built modern fintech, but regulatory pressure on partner banks is forcing the model to mature, with consent orders at Blue Ridge Bank, Cross River Bank, and others driving enhanced compliance requirements.
- The Synapse bankruptcy in 2024 exposed the fragility of middleware-dependent BaaS, with $65-85 million in end-user funds misaccounted and customer deposits frozen for months.
- Successful partnerships require direct bank-fintech relationships, aligned economics, and clear compliance ownership — not just API integrations.
- The market is shifting toward fewer, larger partner banks, direct charter applications by major fintechs, and enhanced middleware with built-in compliance infrastructure.