In August 2020, two nonbank fintechs—Avant and Marlette Funding—and their partner banks— WebBank and Cross River Bank—reached a “landmark settlement” with Colorado regulators (the “Settlement”) to determine who is the “true lender” of loans originated under their partnerships. Under the Settlement, the nonbank-bank partnerships are permissible under Colorado law, provided that they comply with the terms of a Safe Harbor that is laid out in detail in the Settlement (the “Safe Harbor”). The string of judicial decisions applying the “true lender” theory has become a significant source of risk for nonbank fintechs relying on bank partnerships by creating uncertainty over the legal framework that applies to loans made under these arrangements and threatening the viability of their business model. The Settlement offers a blueprint for the signed parties on how to structure their partnerships in Colorado.
However, the Safe Harbor’s ability to serve as a template for bank-fintech partnerships in other states is limited. The Safe Harbor might discourage nonbank fintechs from operating in Colorado by diminishing the benefits of a partnership. Additionally, the bargain enshrined in the Settlement makes it unlikely that other state regulators adverse to “rent-a-bank” schemes will allow it to serve as a de facto constraint without a formal and probably renegotiated, agreement. This post provides a summary of the terms of the Safe Harbor to examine the potential impacts of the Settlement and its ability to serve as a template or constraint for nonbank-bank partnerships across the country. It argues that it is unlikely that the Safe Harbor will become a standard for structuring partnerships outside of Colorado and that doing so would discourage nonbank fintechs and banks from using these arrangements.
Source of Risk to Nonbank-Bank Partnerships: True Lender
After the 2008 Financial Crisis, traditional financial institutions retreated from small-dollar credit markets, creating a vacuum that fintech lenders helped fill by using innovative technology to make credit determinations and the borrowing process easier. Some of these fintechs, who otherwise would be subject to different licensing requirements and interest rate laws in every state they operate in, have adopted a bank partnership model that follows an originate-to-distribute framework to circumvent these state regulations and reduce their overall costs and regulatory burden.
Under the bank partnership model, fintechs partner with a chartered bank that originates loans, and then the fintech purchases those loans from the partner bank and sells them to investors in the secondary market. The fintech uses the sale proceeds to fund additional loans and earns revenue through origination and servicing fees. Under the exportation doctrine, the chartered bank can originate all loans under the interest rate laws of the state where it is located, allowing the nonbank purchaser to market a uniform product to consumers nationwide without obtaining lending licenses in every state. As the number of nonbank fintechs entering the marketplace lending industry continues to grow, the competitive edge achieved by bypassing state lending and interest rate laws has pushed many nonbanks to adopt this model.
A significant source of uncertainty and risk for fintechs using these bank partnership models arises from the line of recent judicial decisions applying the “true lender” theory to these arrangements. Under the true lender doctrine, courts look at the economic reality of the relationship between the bank and the nonbank to determine who is the true lender of the loan. By recharacterizing the nonbank fintech as the lender of the loan, the nonbank becomes subject to claims under state law that it is operating without a license or making usurious loans. Courts using the true lender theory to examine the substance of these arrangements have inconsistently applied a subjective and fact-intensive “predominant economic interest” test. Under the test, courts assess the “totality of the circumstances” and substance of the arrangement, rather than just the form of the transaction, to determine which entity has the predominant economic interest in the loan.
On July 20, 2020, the OCC issued a notice of proposed rulemaking to determine when a national bank is the true lender in the context of a nonbank-bank partnership and on October 27, 2020, published a final rule. The rule aims to resolve the uncertainty caused by the “true lender” decisions by specifying that the national bank is the true lender of the loan, if, at the time of origination, (1) it is named as the lender in the loan agreement, or (2) it funds the loan. In January 2021, seven states and D.C. filed suit challenging the OCC’s true lender rule in federal court, claiming the rule violates the OCC’s statutory authority, is an unreasonable interpretation of federal law, fails to comply with provisions of the Dodd-Frank Act, and reverses the agency’s longstanding opposition to rent-a-bank schemes without a reasoned explanation.
In contrast, the FDIC has yet to propose a similar “true lender rule.” In June of 2020, Chairman Jelena McWilliams stated that the agency was planning to craft a true lender rule soon. However, during a panel on Consumer Financial Services held in December 2020, Leonard Chanin, Deputy to the FDIC Chairman, stated the FDIC does not have “authority under the Federal Deposit Insurance Act to determine when a loan is made by a state bank.” Deputy Chairman Chanin commented that the FDIC can “fill in certain ‘gaps,’” but “state law controls when a loan is made by a state bank and the FDIC cannot preempt state law on this issue.” Without a bright-line rule from the FDIC on true lender standards, many nonbanks partnering with a bank regulated by the FDIC continue to face recharacterization risk. Although the bank partnership model is widely used by a growing number of nonbank fintechs, only a select number of banks serve as the partner in these arrangements.
Colorado Settlement & Safe Harbor Terms
In 2017, the Administrator of the Colorado Uniform Consumer Credit Code (the “Administrator”) brought actions in state court against two nonbanks fintechs—Avant and Marlette Funding—alleging that they were the true lenders of loans and were assessing finance charges in violation of Colorado law. After years of litigation, and adverse decisions in federal and state court, the parties reached an agreement with the Administrator to settle the cases in August 2020. Rather than just charging a fine, the Settlement created a Safe Harbor under which the nonbank fintechs and their bank partners—WebBank and Cross River Bank—could continue to originate loans to Colorado consumers as long as they complied with the Safe Harbor.
Under the Settlement, as long as the partnership lending programs (the “Program(s)”) meet the requirements under the Safe Harbor Terms, the Administrator and Attorney General will not bring any claims alleging that (i) the loans are not subject to federal preemption, (ii) the banks are not the true lenders, or (iii) that the assignment of the loans from the bank to a non-bank assignee affects the assignee’s ability to enforce the loans. To comply with the Safe Harbor, the Program must comply with five specified criteria: (1) Oversight, (2) Disclosure and Funding, (3) Licensing, (4) Consumer Terms, and (5) Structural. Broadly, the Safe Harbor requires: (1) oversight of the loans and the nonbanks, (2) the bank discloses its role and funds the loans according to certain requirements, (3) the nonbank obtains necessary licenses, (4) that “Specified Loans” do not exceed 36% APR and Colorado law applies unless preempted, and (5) the process for selling “Specified Loans” to the nonbank conforms with one of four options.
The terms of the Safe Harbor rely on a distinction between two kinds of loans: “supervised loans” and “Specified Loans.” Supervised loans are defined under Colorado law, and Specified Loans are created and defined by the Settlement. “Supervised loans” are consumer loans that have APRs between 12% and 21%. A “Specified Loan” defined in the Settlement, is a consumer loan exceeding the maximum interest rate allowed under Colorado’s usury laws, subject to certain further limits outlined in the Safe Harbor Terms. The Consumer Terms Criteria set an upper limit of 36% APR that can be charged on Specified Loans. Therefore, Specified Loans under the Settlement have an APR of 21% to 36% and exceed the state’s cap of 21%.
The Oversight Criteria lists oversight responsibilities for the bank’s prudential regulators, the banks, and the nonbank fintechs. The FDIC and the bank’s state banking regulators have oversight authority over the loans offered under the Program, and have “access to examine, review, and audit the Partner Fintech.” The bank must oversee and have “ultimate approval authority” on loan origination, all marketing content related to loans offered and originated by the bank under the Program, and all content related to the Program on the nonbank’s website. The bank also “controls all terms of credit” under the Program and has the “absolute right” to change the credit policy for determining whether to originate a loan and the terms of the loans. The bank has the authority to oversee and approve the nonbank’s risk management program for significant third-party sub-vendors used under the Program. The bank’s oversight program must follow the compliance requirements set by the FDIC in its guidance for third-party lending relationships, found in FDIC FIL 44-2008 and the proposed FDIC FIL 50-2016. Finally, the nonbanks must maintain a compliance management system approved by the bank, reviewed or audited at least annually, that includes a mechanism to report and handle consumer complaints about the nonbank, the Program, or any sub-vendors.
The Disclosure and Funding Criteria require that the loan agreements must identify the bank as the lender of the loan, and this fact must be reflected in all website content, pre-origination consumer disclosures, and marketing materials relating to the Program, and that the nonbank may not provide funds to the bank for the express purpose of funding the origination of loans under the Program. Instead, the bank must fund loans using sources allowed under banking regulations, including its own capital, reserves, retained earnings, deposits, and credit facilities. However, the bank can require the nonbank to keep a deposit account at the bank to secure its obligations to the bank, subject to certain limitations discussed in the Structural Criteria terms of the Safe Harbor.
The Licensing Criteria requires the nonbank fintechs to obtain any necessary licenses under Colorado law and imposes additional reporting requirements in connection with the license. The nonbank must obtain a “supervised lender license” under Colorado law if the Program offers supervised loans, and it must include with its license application a description of the products offered under the Program and how they comply with the Safe Harbor. Thus, if any loans offered under the Program are greater than 12%, then the nonbank must obtain a supervised lender license.
Furthermore, licensed nonbanks must submit an additional Compliance Report—along with the Supervised Lender Annual Report already required under Colorado law—containing information on every Specified Loan originated under the Program and “a reasonably comprehensive explanation of the manner in which the Partner Fintech has complied with the Structural Criteria” to make the Specified Loans. The bank is obligated to stop originating Specified Loans if the nonbank fails to submit the Compliance Report, other than Specified Loans previously approved.
Finally, the Structural Criteria requires that the purchase of loans under the Program by the nonbank from the bank comply with one of four specified structures: (1) Uncommitted Forward Flow Option, (2) Maximum Committed Forward Flow Option, (3) Maximum Overall Transfer Option, or (4) Alternative Structure Option. Broadly, the options under the Structural Criteria set procedures for purchasing Specified Loans, place various limits on the volume of loans that can be purchased under committed and uncommitted obligations under the Program, and set additional limits on collateral and indemnification in uncommitted obligations. The fourth option, the Alternative Structure Option, provides that the Administrator can approve in writing an alternate set of terms under which the Program will fall under the Safe Harbor.
Under the “Uncommitted Forward Flow Option” the nonbank may not enter into a committed obligation to purchase Specified Loans from the bank. However, the nonbank and bank may create a process for the sale of Specified Loans following certain procedures. First, the bank must provide notice to the nonbank of Specified Loans it wants to offer for sale, and the nonbank then must provide notice of the Specified Loans offered that it wants to purchase. If the nonbank does not purchase the offered Specified Loans, the bank can retain and service those Specified Loans, sell them to a third party other than an affiliate of the nonbank, put the Specified Loans into a securitization sponsored by the nonbank or its affiliates so long as the bank participates on similar terms as other investors, or put the Specified Loans into a securitization sponsored by a third party.
The Uncommitted Forward Flow Option creates additional restrictions on what the nonbank can indemnify the bank for and how collateral can be used in relation to Specified Loans. The nonbank can indemnify the bank for losses related to services that the nonbank agreed to perform, fraud on the part of the nonbank or borrowers, and the representations and warranties made by the nonbank. However, the nonbank cannot indemnify the bank for losses related to the nonbank’s failure to purchase Specified Loans it has not agreed to purchase or for the performance of Specified Loans. The same restrictions apply to the nonbank’s ability to seek indemnification from its vendors. In addition, the bank can require the nonbank to maintain cash collateral to secure the obligations of the entire Program, not just Specified Loans, but it can only be used to secure payment for Specified Loans if the nonbank provides proper notice as described above.
The “Maximum Committed Forward Flow Option” provides two options for the nonbank to enter into a committed agreement with the bank to purchase a set amount of Specified Loans, subject to certain limitations. Under the first option, the nonbank can agree to purchase up to 49% of the total origination volume of Specified Loans in the Program during a calendar year. However, the bank cannot transfer any additional Economic Interests in the Specified Loans on an uncommitted basis to the nonbank. Under the second option, the nonbank can agree to purchase up to 25% of the total origination volume of Specified Loans in the Program during a calendar year. Moreover, the bank can transfer additional Economic Interests in Specified Loans to the nonbank on an uncommitted basis by following the same process set forth in the Uncommitted Forward Flow Option. There are no limits on collateral or indemnification under the Maximum Committed Forward Flow Option, except for Specified Loans purchased on an uncommitted basis, which must comply with the terms of the Uncommitted Forward Flow Option. Banks are not restricted from transferring Specified Loans to unaffiliated third parties or from contributing “Specified Loans into a bona fide securitization transaction, including a securitization sponsored by the Partner Fintech.”
Finally, the “Maximum Overall Transfer Option” allows the bank to transfer to the nonbank up to 85% of overall Economic Interests in all loans originated under the Program on an annual basis, including Specified Loans. However, no more than 35% of the “total originated principal amount” of loans can be Specified Loans per year. Additionally, the nonbank cannot purchase a pool of loans that contains more than 35% Specified Loans or Economic Interests in Specified Loans per year. The same terms apply regarding collateral, indemnification, and the bank’s ability to transfer Specified Loans to third parties or securitizations as the Maximum Committed Forward Flow Option.
Under the Settlement, the banks and nonbank fintechs agreed to revise their Programs to comply with the Safe Harbor for a period of at least five years. However, the banks and nonbanks can terminate compliance with the Safe Harbor if, after two years from the execution of the Settlement, a “Change in Law” has occurred. Change in Law is defined in the settlement as a federal or Colorado statute, decision of the U.S. Supreme Court or Colorado Supreme Court, a final decision of a Colorado appellate court, or a regulation finalized by the FDIC, that sets a test for determining when the bank in a bank partnership is the true lender under Section 27 of the FDIA. A final regulation by the FDIC will not constitute a Change in Law until one year has passed since its publication and it has not been (i) rescinded by joint resolution by Congress, (ii) rescinded, removed or otherwise withdraw by the FDIC, (iii) superseded or invalidated by a federal statute, or (iv) enjoined by a federal court after the end of a one year-period.
Finally, as part of the settlement the banks and fintechs agreed to pay $1,050,000 to the Colorado Attorney General to go towards attorney fees and costs, consumer education, future consumer fraud or antitrust enforcement, or other public welfare purposes. In addition, the banks and nonbank fintechs agreed to pay $500,000 to Colorado’s Money Wi$er program, which offers K-12 financial literacy education.
Potential Impact on Nonbank-Bank Partnerships
The Colorado Settlement was applauded by some observers as offering a potential blueprint for nonbank fintechs that rely on bank partnership models, but its ability to have a lasting impact or bring needed clarity to the industry is limited. By examining two key functions of the Safe Harbor—forcing the bank to retain more risk and better tools for state regulators— and their potential effects on the bank partnership model, it is unlikely that the Safe Harbor will act as a de facto constraint on nonbank-bank partnerships in other states.
To understand the potential effects on nonbank-bank partnerships, it is important to appreciate what the Settlement achieves. One primary function of the Safe Harbor is to force the bank to internalize some risk of the loans originated under the Program to varying degrees. For example, under the Uncommitted Forward Flow Option, the nonbank fintech can purchase every Specified Loan from the bank, leaving the bank with a clean balance sheet regarding the partnership at the end of each year. The catch, under the Safe Harbor, is the lack of enforceability and certainty for the bank. The Uncommitted Forward Flow Option prohibits any private ordering between the bank and the nonbank to buy loans, removing the bank’s ability to hold the nonbank accountable in court if it suddenly stops purchasing Specified Loans. The bank also cannot seek indemnification from the nonbank for losses relating to the nonbank’s failure to purchase the Specified Loans or to the performance of unpurchased Specified Loans. The result is that the bank has to internalize some risk of the Specified Loans because of the possibility that the nonbank will not purchase the loans. This can help reduce moral hazard inherent in originate-to-distribute lending models by forcing the bank to more closely examine the underlying creditworthiness of borrowers and the underwriting standards used by the nonbank.
Under each of the Structural Options of the Safe Harbor, the bank must share some of the risk associated with the loans and therefore has increased incentives to monitor credit determinations more carefully. Although this helps reduce some of the risks associated with nonbank-bank partnerships, it might affect the viability of these arrangements. The more skin in the game the bank has, the greater the cost for the bank to maintain the partnership. The bank must expend additional time and resources examining the creditworthiness of borrowers and underwriting standards used by the bank, as well as monitoring the nonbank more closely. As the benefits of partnering with a nonbank fintech diminish, banks might begin pulling out of these partnerships or avoid states like Colorado that have burdensome requirements. The nonbanks face similar increased costs to comply with the Safe Harbor. Currently, Avant does not offer loans in Colorado. The Settlement’s ability to serve as a model in other states will be limited if the parties to the Settlement seem unwilling to utilize its protections.
The bank’s increased diligence over lending decisions might also result in changes to the nonbank’s lending models in disadvantageous ways. By using innovative algorithms that result in more accurate underwriting, fintechs have been able to assess complex risk profiles at the lower end of the credit spectrum and extend credit to borrowers that would otherwise not have access to credit. As banks are incentivized under the Safe Harbor to more closely examine the nonbank fintech’s underwriting and have the “absolute right” to change the credit policy, this could result in nonbanks retreating from certain credit markets at the bank’s urging.
A second function of the Settlement is to provide Colorado regulators better oversight and enforcement tools against nonbanks to protect consumers. For example, the bank has to maintain an oversight program conforming to the FDIC’s guidance in its Financial Institution Letters (FILs) on Third Party Risk. The banks were likely already complying with the FILs, so it does not reflect a drastic change. However, agency guidance does not have the force and effect of law, unlike statutes and regulations. The Settlement gives state regulators additional oversight ability and leverage against nonbank-bank partnerships. Rather than relying on guidance from federal regulators, the state bargained for the right to remove the Safe Harbor’s protections if the partnership does not comply with the FDIC’s guidance.
This feature of the Settlement makes it unlikely that the Safe Harbor will become a de facto constraint for nonbank-bank partnerships across the country. Even if nonbank fintechs conform their bank partnerships to the structures of the Safe Harbor, the arrangement will not be protected from litigation in other states without a formal agreement like the Settlement. Banking regulators and attorneys general in other states might be unwilling to grant this protection without receiving the same increased oversight and enforcement abilities that Colorado regulators bargained for. Additional regulatory requirements in each state would increase the costs for fintechs and reduce the advantages of the bank partnership model. Further, states may draw the line differently as to the degree of economic interest that the bank partner must have or the permissible interest rates under a Safe Harbor, resulting in the same patchwork regulatory framework that currently exists. The Settlement anticipates that other states will reach different compromises; it stipulates that if any of the nonbanks or banks bound by the Settlement enter into a similar agreement with the attorney general or regulators in another state that provides for a lower APR limit than the Safe Harbor, then the Settlement will be amended, and the lower APR limit will apply to that party.
For similar reasons, it would not be an improvement on the current situation if the Safe Harbor served as a model for statutory safe harbors adopted at the state level. The wide range of state usury caps is a testament to the differing views that states have regarding the appropriate level of protection and regulation for consumer credit services and products. State statutes creating safe harbors would set diverging standards, increasing the regulatory burden for nonbanks reliant on bank partnership models. If nonbanks are unable to circumvent state regulation through bank partnerships, the advantages of a bank partnership would disappear entirely, forcing nonbanks to adopt new business models.
Ultimately, the greatest chance for certainty in this area comes from a uniform requirement applied across the states through regulation at the federal level. The OCC’s true lender rule, finalized in October 2020, can serve this function for nonbank partnerships with national banks – if it can survive legal challenges. As for state banks and industrial loan companies, including WebBank, Cross River, and other banks active in this arena, comments from the FDIC’s Deputy Chairman in December of 2020 reflect the diminished likelihood that the FDIC will step in anytime soon to provide uniformity. The last hope is intervention from Congress. Although bills creating a national cap of 36% APR on all consumer loans have been proposed, serious attempts to remove the states’ ability to set usury laws will be strongly resisted.
The uncertainty surrounding the validity of the bank partnership model for nonbank fintechs is in part a function of the fact that existing regulations for marketplace lending were primarily implemented before fintechs entered the scene. While regulation at the federal level would provide certainty, it might fail to keep up with innovation and could lose its teeth as fintech disruptors find other means of regulatory arbitrage. There is also concern that regulating fintechs too heavily during these relatively early stages will “stifle innovation and potentially derail the adoption of useful technology.” Proponents of fintech highlight its “potential to help reach much of the unbanked and underserved” population, making the stakes higher if poorly-designed regulations hamper future innovation.
Despite being an innovative compromise, the Safe Harbor increases the compliance costs for banks and nonbank fintechs and removes some of the advantages of these partnerships. Because the Settlement reflects the bargain that the Colorado authorities were willing to make in exchange for increased regulatory tools, the Safe Harbor will probably not serve as a de facto constraint nationwide. Now that the dust has settled, it seems that the uncertainty caused by true lender recharacterization risk will continue to define the regulatory framework for nonbank fintechs using bank partnership models in marketplace lending.
Brooke Reczka is a J.D. Candidate at Duke Law