Treasury’s 2016 lending paper left ‘true lender’ open and put bank–platform partnerships on bank diligence

Treasury’s May 10, 2016 white paper recorded calls for clarity on which party is the true lender, and asked banks to vet platforms and plan backup servicing. An OCC rule later answered the first question; Congress voided it in 2021.

Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.

Two-minute summaryThe dek and the first sentence of each section, quoted from the article

Treasury’s May 10, 2016 white paper recorded calls for clarity on which party is the true lender, and asked banks to vet platforms and plan backup servicing. An OCC rule later answered the first question; Congress voided it in 2021.

  1. The week The paper described two business models.
  2. The structure: a charter, a platform and a short hold In the platform model the bank contributed its charter.
  3. The open questions: what Treasury recorded and left The paper listed true lender designation among areas of debate.
  4. The partnership burden: diligence, continuity and compliance The paper’s fourth recommendation dealt with partnerships.
  5. The data agenda: transparency, government data and untested models For a bank selling loans to a platform, those standards matter in both directions.
  6. How the paper reached different organizations A community bank considering a platform partnership was a reader directly addressed by the diligence and back-up servicing recommendations. [2] For it, the paper’s message was that a short hold on the balance sheet does not shorten the bank’s responsibility.
  7. Objections to the argument, in order Treasury had no power to settle it by white paper.
  8. Checks a bank or credit union could run These checks apply the paper’s allocation to any arrangement in which a third party originates, services or distributes credit in a bank’s name.
  9. With hindsight The true lender question was answered by rule and then unanswered by Congress.

The week

On Tuesday, May 10, 2016, the Treasury Department issued a white paper titled Opportunities and Challenges in Online Marketplace Lending. [1] [2] Treasury said the paper gave an overview of what it heard in response to its Request for Information, with research on and recommendations for the industry. [1]

Treasury said it received approximately 100 responses to the Request for Information. [1] They came from online marketplace lenders, trade associations, consumer and small business advocates, academics, investors and financial institutions. [1] Treasury said it consulted staff from the CFPB, FDIC, Federal Reserve Board, FTC, OCC, SBA and SEC, while the paper represented Treasury’s views. [1]

The paper described two business models. Direct lenders originate credit to hold in their own portfolios and generally need licenses in each state where they lend. [2] Platform lenders partner with an issuing depository institution, which originates the credit; the platform then buys it for sale to investors whole or through securities. [2]

The paper explained the partnership mechanics. It said platform lenders use the bank’s charter to make loans nationally without individual state licenses. [2] The bank originates credit for applicants on the platform and holds it for one to two days. The platform lender or an investor then buys them. [2]

Treasury listed themes from the responses. It said commenters agreed that using data for credit underwriting is a core element of the industry, holding both promise and risk, including disparate impact and fair lending risk. [1] It noted that applicants do not have the opportunity to check and correct data that may be used in underwriting. [1]

Other themes were that the industry could expand access to credit, that new credit models were untested through a full credit cycle, and that small business borrowers needed stronger safeguards. [1] Treasury also reported support for greater transparency, an undeveloped secondary market, and requests for regulatory clarity about the roles of different participants. [1]

The paper made six recommendations. As Treasury listed them, they were to support stronger small business borrower protections and effective oversight, ensure sound borrower experience and back-end operations, and promote a transparent marketplace for borrowers and investors. [1] The others were to expand access to credit through partnerships that ensure safe and affordable credit, expand access to government-held data, and create a standing interagency working group. [1]

Treasury also named trends to watch. They included the evolution of credit scoring, changing interest rates, liquidity risk, growth in mortgage and auto lending by these firms, cybersecurity threats and anti-money laundering compliance. [1] It said the business models and algorithms had developed largely in favorable credit conditions. [1]

The paper described the market as small but fast-growing, driven by prime and near-prime unsecured consumer credit, followed by small business and student credit. [2] Treasury’s summary said the majority of consumer credit was for debt consolidation, while business credit went to working capital and expansion. [1]

It noted that, as of publication in the second quarter of 2016, it was not yet clear whether an uptick in delinquencies would prove temporary. [2] It also said the health of investor-driven lending models relied on platforms continually attracting funding from investors. [2]

Among the areas of debate it recorded was the true lender question. The paper said many participants noted that clarity on whether the issuing bank or the platform is the true lender would reduce uncertainty. [2] Treasury did not propose an answer. [2]

Treasury’s 2016 white paper is easy to read as a market survey. It is more useful as a map of where risk would land in bank and platform partnerships, and of which questions Treasury chose not to answer. [2]

Our argument is that the paper left the legal question of who lends unresolved, and placed the practical burden of partnerships on bank due diligence and contracts. Its recommendations asked banks to assess platforms, plan for servicing continuity and set duties before signing. [2] In our reading, that allocation outlasted the attempts to settle the legal question by rule. [3] [4] [5]

The same allocation applies to any arrangement in which a nonbank reaches customers through a bank’s charter, whether the product is a loan, a deposit account or a card. The 2016 paper is a compact statement of the questions such arrangements raise, written before the arrangements multiplied. [2] [6]

We take the paper in four parts:

  1. The structure: how platform lending used a bank’s charter, and which party held what, for how long.
  2. The open questions: true lender, assignment of loans and risk retention, as the paper recorded them and left them.
  3. The partnership burden: diligence, servicing continuity and compliance duties that Treasury placed on the parties, especially the bank.
  4. The data agenda: transparency for investors and borrowers, government data for underwriting, and untested models.

The parts are separate. The first describes the arrangement, the second covers law Treasury could not change, the third covers the controls the parties could choose, and the fourth covers the information the market ran on. After them we look at how the paper reached different organizations, take the objections in order, and end with checks a bank or credit union could run.

The structure: a charter, a platform and a short hold

Two models

Treasury distinguished direct lenders from platform lenders. [2] Direct lenders that originate under state licenses, the paper said, are not subject to a federal banking regulator’s supervision, except to the extent the CFPB may supervise them. [2] Platform lenders work through an issuing depository institution, which originates the loans and holds them for one to two days. [2]

The paper said direct lenders increasingly rely on credit facilities, whole loan sales and securitizations to fund originations. [2] Platform lenders fund purchases by passing payment streams to investors through notes whose payments depend on the loans’ performance. [2] The paper called these member payment dependent notes, and said platform lenders using them do not retain credit risk on the loans. [2]

What the bank contributed

In the platform model the bank contributed its charter. The paper said platform lenders use it to lend nationally without obtaining individual state licenses. [2] The bank is the originator of record. The platform acquires the customer, runs the application and, in the paper’s account, buys the loan within one to two days. [2]

In our reading, that short hold is the economic center of the arrangement. The bank’s balance sheet exposure is brief. Its legal and compliance exposure, as originator, is not. The paper recorded commenters’ view that anti-money laundering compliance in these arrangements was set by the platform’s third-party agreement with the issuing bank. [2]

The investor’s side

The paper named Prosper and Lending Club as the only online marketplace lenders then registering their offerings of member payment dependent notes with the SEC. [2] It said very few companies provided information in public securities filings. [2] For investors, that meant limited public data on loan performance, which is why Treasury’s transparency recommendations were aimed at them. [2]

The investor base was varied. The paper named venture capital, institutional investors, hedge funds and financial institutions as sources of funding. [2] In our reading, that mix mattered to a partner bank. A platform that depends on continuous investor funding can stop buying loans quickly if that funding dries up, which leaves the originating bank holding more than a day or two of exposure.

Who supervised whom

The paper said most online marketplace lenders are not directly regulated by prudential regulators, so compliance was not uniform. [2] The bank was supervised. The paper recorded commenters’ view that anti-money laundering compliance in these arrangements was set by the platform’s agreement with the bank. [2] In our reading, the platform reached supervision mainly through the bank’s oversight of that relationship.

In our reading, that made the bank’s third-party risk management the main supervisory channel into the platform’s operations. The paper’s recommendation that prudential regulators evaluate partnerships and identify their risks points in the same direction. [2]

Who held the customer

In the paper’s description, the platform handled the borrower-facing work. [2] The paper said it was common for online marketplace lenders to service loans until they became delinquent and then to outsource servicing to collection agencies. [2] It said some engaged back-up servicers to ensure continuity and mitigate counterparty risk. [2]

For the borrower, the bank was the originator, while the platform took the application and commonly serviced the loan. In our reading, that split is why the true lender question mattered to consumers as well as to lawyers. It determined which rules governed the price of the loan, and who answered for the experience.

Incentives in the structure

The paper said direct lenders earn the majority of their revenue from interest income and fees on loans, and may earn fees for servicing loans sold to others. [2] For platform lenders funded by member payment dependent notes, it said they do not retain credit risk if borrowers default. [2]

In our reading, that combination sets the incentive problem the rest of the paper circles. A party paid on volume and free of credit loss has a reason to originate. The investor bears the loss, the bank bears the name, and the borrower bears the product. Treasury’s recommendations on transparency, servicing and risk retention debates are each attempts to put some of the loss back where the decisions are made. [2]

Small business loans

The paper drew attention to small business borrowers. It cited the 2015 Small Business Credit Survey, in which small businesses approved by online lenders reported a 15 percent lender satisfaction score, against 75 percent for community banks. [2] The top frustrations were high interest rates, unfavorable repayment terms and lack of transparency. [2]

Treasury said strong evidence indicated that small business loans under $100,000 share characteristics with consumer loans but lack the same protections. [2] It said it was willing to work with Congress on legislation addressing oversight and borrower protection. [2]

The open questions: what Treasury recorded and left

True lender

The paper listed true lender designation among areas of debate. It said many participants noted that clarity on which entity, the bank or the platform, is the true lender would alleviate uncertainty. [2] Treasury made no recommendation on the point. [2]

The question matters because, in our general reading, the answer decides which interest rate rules apply: the bank’s own authority, or the licensing and rate rules of the state. The paper did not set out the legal analysis, and we do not attempt it here; organizations should take that question to counsel.

Loans after assignment

The paper recorded a second uncertainty. RFI commenters described regulatory uncertainty around a 2015 Second Circuit decision as one of the primary hurdles for growth of a securitization market. [2] The paper’s footnote said the court held that the National Bank Act does not preempt state-law usury claims against third-party debt collectors collecting on a loan assigned to them by a national bank. [2]

The footnote said the Supreme Court was then deciding whether to hear the case. [2] For a platform whose model depends on buying loans from a bank, the question of whether the bank’s rate authority travels with the loan goes to the value of every loan bought.

Risk retention

The paper recorded a third debate. Some advocates argued for applying a form of risk retention to marketplace loans, pointing to the Dodd-Frank requirements for asset-backed securities. [2] Many others, the paper said, stated that risk retention would have different outcomes for different business models. [2]

In our reading, risk retention was the market’s own answer to the alignment problem the paper raised: a platform that keeps no credit risk has weaker reasons to underwrite carefully. Treasury recorded the debate and did not resolve it.

What uncertainty did to contracts

When the law is unsettled, contracts carry the risk. A partnership agreement can allocate it through representations about compliance, indemnities, repurchase obligations and termination rights. The paper did not prescribe such terms, but its call for the parties to determine the distribution of duties and costs before partnering points to them. [2]

In our reading, three questions decided how much legal risk a bank kept. The first was whether the agreement obliged the platform to buy every loan, whatever happened to the law. The second was whether the platform had the capital to honor an indemnity. The third was whether the bank could stop originating quickly if a court or regulator changed the answer. A bank unsure on all three held more risk than its brief hold suggested.

Designing around the question

Some risk could be reduced without a legal answer. A partnership could price loans within the limits of the borrower’s state, so that the rate would be lawful whichever party was the lender. That choice constrains the product, and the paper did not discuss it. We raise it as a design option the uncertainty made relevant, not as a recommendation Treasury made.

Why Treasury left them open

Treasury’s chosen instrument was coordination. Its sixth recommendation was a standing interagency working group, including Treasury, the CFPB, FDIC, Federal Reserve Board, FTC, OCC, SBA, SEC and a state bank supervisor. [2] The paper said the group could identify where additional regulatory clarity could protect borrowers and investors and expand access to credit. [2]

A white paper is not a rule, and Treasury could not settle questions of bank powers or state law by recommendation. In our reading, the paper’s choice was realistic. It also meant that, for the life of the arrangements then being signed, the legal questions remained a risk the parties had to price and allocate in their contracts.

The partnership burden: diligence, continuity and compliance

Diligence before signing

The paper’s fourth recommendation dealt with partnerships. It said financial institutions and online marketplace lenders should determine the distribution of duties, direct and indirect costs, and compliance requirements of any partnership. [2] It said thorough due diligence and risk assessments should take place before any strategic partnership, to ensure new products fit the institution’s customer needs and risk tolerance. [2]

The recommendation was framed around community development financial institutions, which Treasury saw as partners for reaching borrowers who are creditworthy but not scoreable by traditional models. [2] Its diligence language, though, applies to any bank considering a platform. [2]

CDFI partnerships in detail

Treasury suggested that CDFIs and platforms entering referral, co-branded or white label partnerships create financial literacy materials on credit management. [2] It said the CDFI Fund’s Capacity Building Initiative could develop a module to help CDFIs build business advisory services, small business education and debt management guidelines. [2]

It also said CDFIs, platforms and prudential regulators should work together to ensure the safety and soundness of such collaborations. [2] In our reading, the CDFI case is the hardest version of the partnership problem. The borrowers are the least scoreable by traditional models, so the platform’s model is least tested on them, and the CDFI’s mission makes a failure especially costly.

Servicing continuity

The second recommendation dealt with servicing. Treasury said servicing challenges were then limited by the favorable credit climate and the use of ACH collection, and that it was unclear whether servicing infrastructure would cope with rising delinquencies. [2]

Treasury said it was vital for all platforms to develop back-up servicing plans so loans continue to be managed if the firm ceases operations. [2] It recommended that depository institutions and CDFIs in partnerships with platforms work with them to follow industry standards and identify back-up servicing options. [2]

In our reading, this is where the paper turns operational. It asked the bank, which held the loans for a day or two, to care about servicing that would run for years. The reason is that the bank’s name stays on the loan agreement, and borrower harm from a failed servicer returns to the originator’s reputation and supervision.

Borrowers in distress

Treasury said lenders and contracted servicers should exercise prudence with borrowers in financial distress, and give accurate and actionable information. [2] It named Fair Debt Collection Practices Act guidelines, dispute resolution options and credit counseling as useful information. [2] It recommended regulators continue to monitor servicing for consumer, small business and student borrowers and hold servicers accountable. [2]

Who carries cost

Treasury asked the parties to determine direct and indirect costs before partnering. [2] The direct costs are visible: fees, technology and staff. The indirect ones are harder: the bank’s compliance monitoring, its examination time spent on the partnership, and the reputational cost if borrowers are harmed. In our reading, a pricing model that covers only the direct costs leaves the bank subsidizing the arrangement.

Compliance set by contract

The paper recorded that anti-money laundering compliance in platform arrangements was determined by the platform’s third-party agreement with the issuing bank. [2] Its list of debated areas said platforms and financial institutions faced threats of both cybersecurity and fraud, and that regulators should continue to study safeguards. [2]

Treasury encouraged financial sector firms to adopt baseline cyber protections, join the Financial Services Information Sharing and Analysis Center, and develop response and recovery arrangements. [2] For a partnership, those recommendations become contract terms: who monitors, who reports incidents to whom, and how fast.

Monitoring after signing

The paper focused on diligence before a partnership. Ongoing monitoring is the other half. The bank needs reports on applications, approvals, pricing, complaints, delinquencies and servicing performance, at a frequency that lets it act. [2] Treasury’s recommendation that regulators monitor servicing and hold servicers accountable implies that the bank, as originator, should see the same data first. [2]

Concentration

A bank that originates for a single platform depends on that platform’s funding and operations. The paper flagged liquidity risk: investor-driven models relied on continually attracting funding. [2] In our reading, a bank should know what share of its origination volume and fee income comes from each partner, and what it would do if the largest one stopped.

Concentration also runs the other way. A platform with one bank partner depends on that bank’s appetite and its examiners’ view of the program. If the bank exits, the platform needs another charter quickly. In our reading, that mutual dependence is a reason for both parties to plan the transition in the contract, not after a dispute.

Incidents and fraud

The paper named cybersecurity and fraud as threats to both platforms and financial institutions. [2] An incident at the platform can expose applicants’ data collected under the bank’s name. The contract has to say who detects, who notifies, who pays and who speaks to borrowers. Those terms are cheap to write before an incident and expensive to negotiate during one.

What the bank owns in the end

Read together, the recommendations allocate the bank three duties. It must understand the platform before signing, make sure servicing survives the platform, and set compliance duties in the contract. [2] None was new to bank supervision. The paper’s contribution was to name them for a model in which the bank’s exposure looked brief.

The data agenda: transparency, government data and untested models

Investor transparency

The third recommendation asked the industry to adopt standardized representations, warranties and enforcement mechanisms. [2] It also asked for consistent reporting standards for loan origination and portfolio performance, securitization performance transparency, and consistent pricing methodologies. [2]

Treasury recommended a private sector registry, available to the public, to track data on transactions, including note issuances, securitizations and loan-level performance. [2] It encouraged the private sector to work with consumer advocates to meet strict privacy and data security standards. [2]

For a bank selling loans to a platform, those standards matter in both directions. Representations and warranties define when the bank must take a loan back. Loan-level reporting is what lets an investor, or a supervisor, see whether the platform’s underwriting is holding up.

Government data for underwriting

The fifth recommendation dealt with government-held data. Treasury described smart disclosure, meaning the release of information in standard machine readable formats, and recommended that the CFPB and FTC include it in their guidance on consumer disclosures. [2]

Treasury also recommended automating the IRS Income Verification Express Service with an application programming interface. [2] It said the IRS should study feasibility and resources and launch a pilot with dummy data. [2] It said lenders lacked comprehensive real-time sources to verify borrowers’ financial capacity. [2]

In our reading, this recommendation treated income verification as shared infrastructure. A lender’s ability-to-repay process depends on documents the borrower supplies or data the lender buys. A government interface with the borrower’s consent would change the cost and reliability of that step for every lender.

Comparison shopping

Treasury said smart disclosure would let third parties and nonprofits build comparison sites for loans, similar to travel comparison sites. [2] It said that in complex markets it can be too time consuming for consumers to read and process every disclosure themselves. [2]

For a lender, machine readable terms are a discipline. They require prices and fees to be expressed consistently enough for software to compare them. A product whose cost depends on undisclosed factors does not fit the format. In our reading, that is why smart disclosure would have helped borrowers especially where pricing was opaque.

Verification in a partnership

In a bank and platform arrangement, the platform collects the application and the documents. The bank, as originator, needs to know how income and identity were verified, because the loan is made in its name. [2] A government verification interface of the kind Treasury proposed would have produced a record of the check itself, rather than a copy of a document the borrower uploaded. In our reading, that kind of record is easier for an originating bank to rely on, and easier for an examiner to test.

Models without a downturn

Treasury said the new underwriting tools had been developed in a period of very low interest rates, declining unemployment and strong credit conditions. [2] It said higher charge-off and delinquency rates on recent consumer loan vintages might signal concern if conditions deteriorated. [2]

The paper recommended the working group discuss ways to monitor risk through the credit cycle, including forward-looking loss projections from market participants. [2] For a partner bank, that is a model risk question. The bank originates on the platform’s model, and needs some view of how the model behaves when conditions change.

The registry and privacy

A public registry of loan-level performance raises an obvious tension with borrower privacy. Treasury addressed it briefly, encouraging the private sector to work with consumer advocates to meet strict privacy and data security standards. [2] Loan-level data without names can still identify borrowers when combined with other data. Any registry design would have to decide what fields to publish and at what level of aggregation.

Fairness and data quality

Treasury said data-driven algorithms carry the risk of disparate impact and fair lending violations. [1] It noted that more accurate scores are not necessarily better for each borrower, and that some research suggested advanced models may reflect underlying bias. [2] It also recorded advocates’ concern that lenders could use new data sources to identify and target vulnerable consumers. [2]

The paper’s point that applicants cannot check and correct data used in underwriting is the one with the clearest operational answer. [1] A lender that cannot tell an applicant which data drove a decision cannot give a useful adverse action reason. That is a design requirement for any partnership’s decisioning system, whoever builds it.

Fair lending testing

Treasury said objective research was needed on the effects of new credit modeling approaches, and asked the working group to discuss nontraditional data in credit scoring. [2] For a partner bank, the practical step is testing. It needs outcome data by applicant group, and the right to ask the platform to explain and change model features that drive disparities.

Platforms may treat their models as proprietary, and the paper acknowledged that credit scoring models are proprietary and data sources are expensive or unavailable to outside researchers. [2] In our reading, the contract is where a bank secures access to what it needs to test, because the bank answers for outcomes on loans it originates.

How the paper reached different organizations

A community bank partner

A community bank considering a platform partnership was a reader directly addressed by the diligence and back-up servicing recommendations. [2] For it, the paper’s message was that a short hold on the balance sheet does not shorten the bank’s responsibility. The questions are what the platform does in the bank’s name, and what happens if the platform fails.

A larger bank buying loans

A larger bank buying whole loans from platforms was on the other side of the transaction. The paper’s investor recommendations applied to it: standardized representations and warranties, enforcement mechanisms and consistent loan-level reporting. [2] For such a buyer, the representations define what can be sent back, and the reporting is how it learns that it should.

A CDFI

Treasury saw CDFIs as partners for reaching thin-file and no-file borrowers and young small businesses. [2] It suggested CDFIs could use platforms’ underwriting technology and back-end operations, and platforms could use CDFIs’ local knowledge. [2] It offered CDFI Fund support for capacity building and named an innovation challenge winner working on referral tools. [2]

A credit union

Credit unions appeared in the paper as RFI respondents through their trade associations. [2] A credit union buying loans or participations originated through platforms faces the same questions as a bank buyer: representations, servicing continuity and loan-level data. Its regulator is different, but the diligence logic is the same.

A platform

For a platform, the paper set out the market’s expectations: back-up servicing, transparent pricing, standardized data for investors, and fair treatment of borrowers in distress. [2] It also signaled that its bank partners would be asked to show diligence on all of these. [2]

An investor or securitizer

The paper recorded commenters’ agreement that active growth of a securitization market would require transparency and significant repeat issuance. [1] It recorded commenters’ view that legal uncertainty, limited ratings and lack of visibility into collateral were the primary hurdles. [2] For an investor, the open legal questions were a direct valuation risk.

A small business owner

A borrower especially exposed in the paper’s account was the small business owner taking a loan under $100,000, which Treasury said shared characteristics with consumer loans without the same protections. [2] The paper referred to an industry-led small business borrowers’ bill of rights covering transparent pricing, responsible underwriting, fair treatment by brokers and fair collection practices. [2]

A bank’s board

For a bank’s directors, the paper offered a short list of questions any partnership should answer. They are the duties and costs each party carries, the diligence done before signing, the fate of servicing if the partner fails, and the behavior of the partner’s model as credit conditions change. [2] A board that can get clear answers on those four is overseeing the arrangement. One that cannot is relying on the partner’s assurances.

A state regulator

Treasury proposed that a representative of a state bank supervisor join the working group. [2] State law was outside the paper’s scope, but the true lender and assignment questions were, at bottom, about when state rules apply. [2] In our reading, any durable answer needed the states at the table.

Objections to the argument, in order

1. Treasury should have answered the true lender question

Treasury had no power to settle it by white paper. The question turned on bank powers and state law. The OCC later tried to answer it by rule, and Congress voided the rule. [3] [4] Treasury’s restraint looks, with hindsight, like an accurate reading of how contested the question was.

2. Diligence duties already existed

Banks already had third-party risk management expectations. The paper did not claim novelty; it applied them to a model where the bank’s exposure looked brief. [2] The 2023 interagency guidance later said a banking organization remains responsible for compliance in activities involving third parties. [5]

3. Back-up servicing is the platform’s problem

The platform services the loans, so continuity looks like its concern. But Treasury asked banks and CDFIs in partnerships to identify back-up servicing options. [2] When a servicer fails, borrowers deal with the consequences under loan agreements that name the bank.

4. The paper was too favorable to the industry

The paper recorded the industry’s case for expanded access. It also recorded unfavorable small business borrower experience, disparate impact risk, untested models and weak servicing for a downturn. [2] Its recommendations included small business protections and regulator monitoring of servicing. [2]

5. Market events made the paper moot

Whatever happened to individual business models after 2016, the questions stayed open. Federal banking agencies were still asking in 2024 about roles, risks and compliance in bank-fintech arrangements, including lending. [6]

Even the OCC’s 2020 rule, which would have settled the question for national banks, said it held banks accountable for all loans they make in such partnerships. [3] A clear answer about who lends does not answer whether the lending is done well. Diligence addresses the second question, which no rule on lender status could settle.

7. Treasury lacked supervisory authority, so the paper changed nothing

Treasury did not supervise banks or platforms, and the paper did not bind anyone. [2] Its influence was agenda-setting: it named the questions, and the supervisory guidance that followed addressed the same ones. The 2023 guidance’s attention to planning, contracts and contingency is close to what the paper asked of partnerships. [5]

Checks a bank or credit union could run

These checks apply the paper’s allocation to any arrangement in which a third party originates, services or distributes credit in a bank’s name.

Write down what is done in your name

List every function the partner performs that a borrower would attribute to you: marketing, application, decisioning, disclosures, servicing and collections. For each, name the contract clause that sets the partner’s duty and your right to monitor it. [5]

Test the servicing fallback

Identify the back-up servicer, the data it would need, and the date of the last test transfer. Treasury called back-up servicing plans vital. [2] A plan never tested is a hope.

Map compliance duties

Record who runs customer identification, sanctions screening, fair lending testing and complaints. The paper recorded that anti-money laundering compliance was set by the third-party agreement. [2] Confirm the agreement says what you think it says.

Ask for model evidence

Request the partner’s model documentation, performance by vintage and any stress results. Treasury warned the models were untested through a credit cycle. [2]

Check adverse action reasons

Sample declined applications and confirm the reasons given match the data used. The paper noted applicants could not check data used in underwriting. [1]

Read the agreement for the downturn

Reread the partnership agreement as if delinquencies doubled and the partner’s investors withdrew. Find the clauses that govern loan purchases, servicing transfer, collections conduct and termination. Treasury’s warning that the models had not been tested through a credit cycle is a reason to test the contract the same way. [2]

Measure concentration

Record the share of origination volume and fee income from each partner, and the action plan if the largest partner stopped buying loans. Treasury named investor funding as a liquidity risk for these models. [2]

Read complaints in your name

Collect complaints about the partner’s product that mention your institution, from your own channels and the partner’s. Complaints are the earliest signal that what is done in your name differs from what the contract says.

Keep your own copy of the loan data

Confirm that the bank receives, on a schedule, a complete copy of borrower, loan and payment data for every loan it originated. A back-up servicer needs that data to take over, and the bank needs it to answer regulators if the partner stops responding. [2]

Plan the exit

Set out how the bank would transition the activity to another party or bring it in-house, as the 2023 guidance described for planning. [5] Include loan data, borrower communications and the timeline.

With hindsight

The true lender question was answered by rule and then unanswered by Congress. On October 30, 2020, the OCC published a final rule on when a national bank or federal savings association makes a loan. [3] Under it, a bank made a loan if, as of the origination date, it was named as the lender in the loan agreement or funded the loan. [3] The OCC said the rule applied including in partnerships with third parties such as marketplace lenders. [3]

The OCC said the rule held banks accountable for all loans they make, including in marketplace lending partnerships, and rejected arrangements in which a bank rents its charter to a third party. [3] Opposing commenters said it would facilitate such schemes and increase predatory lending. [3] The rule took effect on December 29, 2020. [4]

The OCC later recounted what followed. The Senate passed a joint resolution disapproving the rule under the Congressional Review Act on May 11, 2021, and the House passed it on June 24, 2021. [4] The President signed it as Public Law 117-24 on June 30, 2021. [4] On August 5, 2021, the OCC removed the rule from the Code of Federal Regulations, saying it had no legal force or effect. [4]

The diligence burden the paper described moved into supervisory guidance. In final interagency guidance published in the Federal Register on June 9, 2023, and final as of June 6, the Federal Reserve Board, FDIC and OCC replaced each agency’s earlier general guidance on third-party relationships. [5] It said a banking organization is responsible for conducting its activities in compliance with laws and regulations, including activities involving third parties. [5]

The 2023 guidance described planning that includes contingency plans for transitioning an activity to another third party or bringing it in-house. [5] It also addressed contract terms on default and termination, and responsibility for compliance. [5] In our reading, those are the 2016 paper’s back-up servicing and duty-allocation recommendations in supervisory language.

The agencies added supplements. The 2026 proposal listed a May 3, 2024 guide for community banks on third-party risk management, and a July 25, 2024 joint statement on bank arrangements with third parties to deliver deposit products. [7]

In a notice published July 31, 2024, the OCC, Board and FDIC requested information on bank-fintech arrangements. [6] They said supervisory experience had highlighted a range of risks, and that such arrangements could include lending products such as unsecured consumer or small business loans. [6] They said arrangements may increase operational complexity and compliance challenges. [6]

The guidance is changing again. On September 15, 2026, the OCC, Board, FDIC and NCUA proposed new third-party risk management guidance and said they planned to rescind and replace existing guidance. [7] They said they believed the 2023 guidance had frequently been interpreted in an overly broad manner, with insufficient focus on tailoring. [7] The proposal set a comment deadline of November 16, 2026; as checked on that date, it remained a proposal. [7]

What held up from 2016 is the allocation. Whatever the law said about who lends, supervisors kept treating the bank as answerable for what partners did in its name. [3] [5] What did not hold up was the hope of regulatory clarity on true lender. The OCC’s rule was voided, and the paper’s working group question, where additional clarity could help, was still being asked in 2024. [4] [6]

Spotted an error? Tell us and we will check it and correct the article where needed.

  • marketplace lending
  • true lender
  • third-party risk
  • Treasury
  • bank partnerships