Bank Partnership or Your Own Lending Licences: How US Fintech Lenders Choose

US fintech lenders reach borrowers in one of two ways: they hold state lending licences and lend themselves, or they partner with a bank that makes the loans and uses federal law to charge its home state's interest rate nationwide. The bank route gives one national product faster, but it carries true lender risk, the chance that a state or court treats the fintech as the real lender and applies state licensing and rate limits. That risk is live in 2026: Illinois tests who holds the predominant economic interest in a loan, and Colorado's opt-out from federal rate exportation, in force since 1 July 2024, is before the full Tenth Circuit.

This guide explains both models in plain words, where the law stands at 20 October 2026, and how lenders decide. For the state licences themselves, see our guide to US lending licences by state and the US consumer lending licence page.

How the bank partnership model works

In a bank partnership, a bank is the lender of record. The fintech designs and markets the product, runs the application and servicing technology, and often buys the loans, or an interest in them, shortly after the bank makes them.

Two federal powers make this work. A state-chartered bank insured by the FDIC may charge borrowers in other states the interest rate allowed in its home state, under 12 U.S.C. 1831d. National banks have the same power under 12 U.S.C. 85. This is called rate exportation. In 2020 the OCC and the FDIC each adopted a rule confirming that the interest rate that was lawful when the bank made the loan stays lawful after the loan is sold. These are known as the valid-when-made rules.

The result is one product, one price and one set of documents for borrowers in most states, without the fintech holding a lending licence in each of them.

What your own licences give you

Holding state licences means the fintech is the lender. It sets credit policy within each state's law, keeps the full loan margin, and does not depend on a bank's appetite, pricing or approval. Its regulators are the states and the CFPB.

The cost is that every state's rules apply directly: rate limits, such as Florida's caps of 36% a year on the first US$10,000 of a consumer finance loan; capital and bond minimums; disclosures; annual reports; and examinations. Licences are obtained one state at a time, usually through NMLS.

The true lender question in plain words

The true lender question asks who really made the loan. If the answer is the fintech, not the bank, the loan is judged under the borrower's state law as if the fintech had made it without the bank. Where the fintech has no licence, or the rate is above the state's limit, the loan may be unenforceable in part or in full, and the regulator may seek refunds and penalties.

Courts and regulators look at substance over form. Questions they ask include:

• Who carries the credit risk once the loan is made?
• Who receives most of the economic benefit?
• Who designed the product and controls underwriting?
• How quickly, and on what terms, are loans or receivables sold to the fintech?
• Does the bank exercise real control over the programme?

Federal law once offered a simple answer. In October 2020 the OCC adopted a rule that a national bank was the true lender if it was named as lender in the loan agreement or funded the loan. Congress disapproved that rule under the Congressional Review Act, and it was signed out of force on 30 June 2021 (Public Law 117-24). There is now no federal true lender rule.

Where states have acted

Illinois shows the modern state approach. Its test does not ask who signed the loan agreement. It asks who holds the economic interest, who arranges the loan with a right to buy it, and whether the arrangement as a whole is designed to evade the 36% cap.

Colorado shows the second approach: removing rate exportation itself. Federal law lets a state opt out of the provision that allows state-chartered banks to export rates, and Colorado did so from 1 July 2024. Trade associations sued. The district court granted a preliminary injunction limiting the opt-out to loans made by banks located in Colorado. On 10 November 2025 a divided Tenth Circuit panel reversed, holding that a loan is made in Colorado if either the lender or the borrower is located there. The full court then agreed to rehear the case, which vacated the panel decision and left the injunction in place while the rehearing proceeds. Iowa and Puerto Rico have also opted out.

The federal position at 20 October 2026

The valid-when-made rules protect a loan's rate after sale. They do not decide who the true lender is, which remains a question of state law and the facts.

How banks oversee their fintech partners

A bank partner answers to its own federal regulator for the programme. Under the 2023 interagency guidance, the bank must assess the fintech before signing, set contractual controls, monitor performance and compliance, and keep the ability to exit. In practice this means the bank approves credit policy, marketing, complaints handling, vendor selection and compliance testing, and audits the fintech.

The September 2026 proposal would replace the 2023 guidance with an approach that scales oversight to the risk of each relationship. For a fintech lender, the practical message does not change: the bank must be able to show it controls the lending, which is also what a court will look for in a true lender case.

Choosing between the two models

Three questions usually decide it:

• Price: if the product's APR is within the rate limits of the states that matter, own licences remove true lender risk entirely. If it is above them in key states, a partnership may be the practical national route, and the true lender analysis decides whether it holds.
• Economics: if the fintech will buy most loans soon after origination and carry the credit risk, it looks more like the lender under tests such as Illinois'. Retaining more risk and benefit at the bank reduces that exposure but changes the economics.
• Time: a partnership can be faster to a national launch, but bank onboarding under the third-party guidance takes time, and licences filed in parallel shorten the route to independence.

Reducing true lender exposure in a partnership

No structure removes true lender risk, but the factors courts and states look at point to how a programme can be built so the bank is, in substance, the lender:

• The bank approves and can change the credit policy, pricing, marketing and loan documents, and records that it does so
• The bank keeps a real economic interest, for example by holding loans or a share of receivables for a period rather than selling everything at once
• The fintech does not guarantee or insure the bank against losses on the loans, which Illinois lists as a factor pointing to the fintech as lender
• Complaints, compliance testing and audit rights run to the bank, with documented oversight under the third-party guidance
• Borrowers in states that have acted, such as Illinois above 36% APR, are either excluded, priced within state limits, or served under the fintech's own licence

Each of these changes the economics. The aim is a programme whose paperwork and practice both show the bank in control, because a court will test the practice, not just the contract.

Hybrid models and moving to your own licences

Many lenders run both models. A common pattern is licences in the largest states where the product fits within state rate limits, and a bank partner elsewhere. Another is a partnership at launch while state licences are filed, then a move of new lending onto licences state by state.

Moving requires planning. Loans already made by the bank stay bank loans. New borrowers in each state move to the fintech's licence from a set date, with new disclosures, loan agreements and reporting. Some states also license servicers, collectors or loan buyers, so the fintech may need licences for the existing book even while the bank remains lender of record.

A decision checklist

• List target states and the APR, fees and loan sizes of each product
• Test each product against each state's licensing and rate rules
• Map who carries credit risk, who receives the economics and who controls underwriting in any partnership
• Check the Illinois anti-evasion test and the Colorado case status for borrowers in those states
• Check which states license brokers, servicers or loan buyers
• Prepare for the bank's third-party due diligence and ongoing audits
• File licences in parallel where independence is the long-term plan
• Review BNPL and earned wage access features separately, because several states now regulate them directly; see BNPL and earned wage access in the US

Regulatory Counsel advises fintech lenders on both routes as part of our consumer credit practice, from state licence applications to the documentation a bank partner needs.