Lender Performance in the Personal Loan Market from the Perspective of Consumer Complaints

Short-term consumer installment loans (commonly referred to as personal loans) are provided to consumers by banks and other regulated depository institutions as well as nonbank financial institutions. Since at least 2015, there has been a major shift in the market share of personal loans to so-called fintech companies. Generally speaking, fintechs are nonbank consumer finance companies that are distinguished from banks and other, more traditional nonbank providers of financial services by their reliance on technological innovations or data applications.[1]

Fintechs have drawn market share mostly from other nonbank lenders, which include payday loan companies, title lenders, consumer finance companies offering personal installment loans and lending affiliates of retail stores. Fintechs’ gains versus these other, more traditional nonbanks may in part reflect innovations in data analysis and risk modeling that have provided them with cost advantages in some contexts. Banks have also lost some market share to fintechs since 2015, although they regained a portion after 2020. Banks’ increasing regulatory costs over this period have likely been a factor contributing to their loss of market share to less heavily regulated fintech companies.

This note compares the lending performance across these three categories of lenders in the personal loan market—banks, fintech companies and other nonbanks—regarding complaints submitted to the Consumer Financial Protection Bureau. The analysis demonstrates that banks have offered superior customer experience by this measure. This assessment suggests that banks’ loss of market share may have left consumers worse off, and the shift in market share from traditional nonbanks to fintechs does not appear to have markedly benefited consumers.

The analysis relies on the CFPB’s complaint database covering the period 2015 through 2022, supplemented with a specially and meticulously constructed lender-type classification. Various performance metrics, including trends in complaint volume, share of complaints by type and frequency of remuneration or remediation provided, are compared across three lender categories: banks, traditional finance companies and fintechs.

Key findings include:

  • Both fintechs and other nonbanks have a comparatively high concentration of complaints in the arguably most concerning complaint category, alleged deceptive practices.
  • Fintechs are over-represented in the critical category of disputed information in the credit report, accounting for close to half of the complaints in this category, with an increasing trend over the analysis period.
  • More frequently than banks, fintechs and other nonbanks fail to provide a timely response to complaints filed with the CFPB.
  • For nearly every complaint category, banks have provided both monetary and non-monetary restitution more frequently than either fintechs or other, more traditional nonbank lenders.
  • Banks’ share of complaints has declined significantly relative to their share of originations since 2015, suggesting that the quality of banks’ lending conduct has improved.

In addition, the note tabulates CFPB enforcement actions related to personal loans conducted since 2015. In none of these actions have the defendants been depository institutions. Fintechs were defendants in eight of the cases, and other nonbank companies or associated individuals were defendants in the other cases. This provides further confirmation of the relatively high quality of banks’ customer relationships in the personal loan market and their compliance with consumer protection laws.

Background on the CFPB’s Complaint System and Database

The data used in this analysis are drawn from the Consumer Complaint Database maintained by the CFPB. The duties of the CFPB, as stipulated under the Dodd-Frank Wall Street Reform and Consumer Protection Act, include directly hearing from and responding to consumer feedback and concerns about financial products and services.

The CFPB addresses complaints about institutions subject to the agency’s supervisory authority, among which are banks, credit unions and thrift institutions with assets over $10 billion along with their affiliates. The CFPB also has supervisory authority over nonbank mortgage originators and servicers, payday lenders, private student lenders and various other participants in consumer financial markets as defined by CFPB rules.[2] Any complaints received by the CFPB about companies that are not under the agency’s supervision are routed to the appropriate regulatory authority.

On the same day that the agency went live, July 21, 2011, it established the Office of Consumer Response and a system for addressing consumer complaints that included a toll-free number, a form on the CFPB website and a process for forwarding complaints to companies and receiving responses. The complaint system was initially limited to credit cards and subsequently expanded in phases.

The milestones most relevant to our analysis are those pertaining to the incorporation of complaints about personal loans. The CFPB began accepting complaints about consumer installment loans, including personal installment loans (as well as auto loans) in March 2012. The system was further expanded to incorporate complaints from borrowers encountering problems with payday loans in November 2013. In July 2014, the CFPB began accepting complaints about pawn and title loans.[3] In March 2016, the CFPB began accepting complaints from consumers encountering problems with online marketplace lenders.[4] 

In July 2013, the system was expanded to incorporate complaints from consumers experiencing debt collection problems related to any type of consumer debt, against any company collecting such a debt from them. Moreover, consumers were allowed the option of filing a second, separate complaint against the company with which they had the original account.

The CFPB has made available to the public some of the consumer complaint data it collects through its Consumer Complaint Database. The first version of the public database—launched in June 2012—was limited to credit card complaints and contained individual-level field data collected by the CFPB, including the date of submission, the state and ZIP Code location of the complainee, a unique complaint id number, an issue categorization and the company that the complaint concerns. It also included information about the actions taken on a complaint, such as whether the company’s response was timely and whether the consumer disputed the company’s response.

In March 2013, the CFPB launched a more comprehensive complaint database, expanding the database from about 19,000 credit card complaints to more than 90,000 complaints on mortgages, student loans, bank accounts and services, other consumer loans and credit cards.[5] The database has been continually updated since then.

Since June 2015 (after the CFPB finalized a policy concerning the publication of complaint narratives), consumers submitting a complaint have had the option of sharing a description of what happened, scrubbed of personal information, in the CFPB’s public-facing database.[6] Since then, more than half of consumers submitting complaints regarding personal loans have opted to share their anonymized stories, with these narratives included in the database.

Data Preparation

The focus of our analysis is complaints about personal loans. Because complaints about payday and title loans were not included in the CFPB’s complaint reporting infrastructure until late 2013 and mid-2014, respectively, we limit the sample for our analysis to complaints submitted during or after January 2015.[7] Note that as of 2015 the data still excluded marketplace lenders, which were not incorporated until March 2016. The latest month spanned by our sample is December 2022.

When submitting a complaint, the consumer chooses the product category and subcategory to which it best belongs and identifies the company that, in the consumer’s view, is at fault. We include in our dataset all complaints tagged in the consumer’s submission as belonging to the payday, title or personal loan product or subproduct categories.[8]

Classifying companies. The consumer also identifies the company that is the target of the complaint. However, the complaints database does not provide a classification of companies. Whether the target of a debt-related complaint is a depository institution (bank or credit union), fintech or more traditional nonbank financial services provider or a non-lender such as a collection agency, can only be determined from the company name.

Making this determination can be challenging, in part because the CFPB populates the company name field in the complaints database with a legal entity name. Often, the legal entity name differs from the name by which the company commonly does business and can, at times, be rather obscure.

The initial step taken was to filter out complaints targeting collection agencies, consumer credit bureaus, payment services providers and other non-lenders. We first dropped all companies with “Collection”, “Collections”, “Recovery”, “Recoveries”, “Resolution” or “Resolutions” in the company name, as these terms are indicative of collection agencies. Next, we applied a mechanical procedure, described in Appendix 1, that identifies larger collection agencies and other non-lenders with a high degree of precision.

We then individually checked each of these tagged companies to confirm that they indeed merited exclusion. Most were confirmed to be collection agencies or other non-lenders. Some were determined to be either banks or nonbank lenders and therefore were reincluded. Classification of companies as either banks, fintechs or other nonbanks likewise was accomplished manually, as also described in Appendix 1.

Confirming the loan type designation. In processing a complaint, the CFPB confirms the consumer’s relationship to the lender that is the subject of the complaint. However, the CFPB does not confirm or revise the product and subproduct designations given by the consumer; these are taken as is.

Consequently, some complaints tagged in the CFPB database as personal loans have been misclassified. They may be complaints about auto loans, student loans credit cards, mortgages, checking accounts, etc.

As described in Appendix 1, a best effort was made to filter out misclassified complaints using the complaint narrative field. Among complaints containing a narrative, about 19 percent of those associated with banking institutions and a negligible number of those associated with fintechs and other nonbanks appear to have been misclassified.[9] Since depository institutions are diversified financial service providers whereas nonbanks and fintechs tend to specialize, it is not surprising that nearly all misclassified complaints are associated with banks or other depository institutions.

Because nearly two-thirds of complaints in the database lack a narrative, this filtering is necessarily incomplete. By inference (assuming that the presence of a narrative is unrelated to whether a complaint is misclassified), the sample overcounts complaints regarding banks by about 12 percent. The overcount is a consequence of complaints that have been misclassified but cannot be identified due to the lack of a narrative.

Additional useful data items. Consumers submitting a complaint are also required to select, from a menu of choices, an issue category. Depending on the product type, they also may have to select a sub-issue that best characterizes their complaint.[10]

The menu of issues and sub-issues specified by the CFPB underwent a substantial change in April 2017, and thus, it is not possible to group complaints by issue category in a consistent manner for the full 2015-2022 analysis period.[11] Therefore, when analyzing issues, we restrict the sample to July 2017 forward.

Moreover, the CFPB’s menu of issues and sub-issues is more detailed than necessary or suitable for our analysis. Therefore, we regrouped the issues and sub-issues from the complaints database into a consolidated list, as detailed in Appendix 2.

The data also indicate whether the company’s response to the complaint was considered “timely” and provide information on the type of response. In regard to “timely response”, the agency asserts that it “expects companies to provide complete, accurate and timely responses tailored to the issues described in each consumer’s complaint, generally within 15 calendar days.”[12] Responses are characterized by the agency according to three possible outcomes: the complaint is closed with monetary relief, with an unspecified type of non-monetary relief or with an explanation but no relief.[13]

As described by the CFPB, closing a complaint with non-monetary relief could include such actions as changing account terms or correcting submissions to a credit bureau. Closing a complaint with explanation “means the company providing an explanation to the consumer that substantively meets the consumer’s desired resolution or explains why no further action will be taken.” 

Consumers are notified when the company has provided a response and are then given a 60-day window within which to dispute the company’s response.[14] Originally, the CFPB complaints database included an indicator for whether the consumer chose to dispute the response, but this data item was discontinued in 2017.

Finally, as noted previously, the narrative description a consumer provides when submitting a complaint is shared publicly only if the consumer affirmatively checks a consent box. For consumers who opt-in, the narrative description is included in a narrative field within the complaint database, after the CFPB has taken steps to remove personal information.[15]

Complaint Trends

As shown in Panel A of Figure 1, the aggregate number of complaints related to personal loans has averaged about 3,140 per year over the analysis period. Total annual complaints peaked at about 3,800 in 2016, then steadily declined to a low near 2,600 in 2021, followed by a sharp increase in 2022, at least partly attributable to increased lending. Panel B breaks down the semi-annual number of complaints by issue category for the second half of 2017 through the end of 2022. A substantial drop in complaints about alleged deceptive practices occurred during the pandemic

period. We observe a generally declining trend for complaints about difficulties curing delinquency and increasing trends for complaints about the loan application process and disputed information on credit reports.

Each of the issue categories shows a rise in the volume of complaints during 2022. Apart from reflecting loan growth, the rise in complaints (particularly the spike for payments not properly credited) might be tied to the expiration of COVID-19 debt relief programs for households near the end of 2021. The resumption of scheduled loan repayments may have been accompanied by some confusion about which payments were due when.

Figure 1: Complaint Trends for Personal Loans

Panel A: Overall Market (2015 – 2022)

Figure 1: Complaint Trends for Personal Loans

Panel B: By Issue Category (July to Dec. 2017 – 2022)

Panel B: By Issue Category (July to Dec. 2017 - 2022)

Source: CFPB Consumer Complaints Database

Shares by lender type. Figure 2 shows the annual share of complaints by lender type—banks, fintechs and other (more traditional) nonbanks—from the CFPB data, while Figure 3 shows the annual share of personal loans originated by lender type, as calculated using the TransUnion Consumer Credit Database. Credit unions are excluded here and from the remainder of the note to facilitate visual comparison between the two charts—although credit unions have a substantial share of the market, very few are subject to CFPB supervision due to the size threshold applied to individual depository institutions.[16]

Loan origination volume is one important factor affecting the share of complaints associated with an institution category, so we would expect the former to be somewhat proportionate to the latter. The complaint shares in Figure 2 are broadly consistent with the origination shares in Figure 3, confirming this expectation.

Consumer credit reporting data are the best available option for measuring the share of originations by type of institution. However, it is important to bear in mind that there are very significant data limitations associated with directly comparing the share of complaints by lender category to the share originations observed using credit reporting data.

On the one hand, smaller depository institutions are not covered by the CFPB complaint database, as they are not subject to the CFPB’s supervisory authority. The limited information available suggests that non-covered institutions account for fewer than 20 percent of personal loans originated by banks and savings and loan institutions.[17] The implied overstatement of banks’ origination share relative to their complaint share would be offset to a large degree by the presence of misclassified complaints regarding banks, discussed previously.

On the other hand, consumer credit reporting data have significant gaps in reporting by nonbank financial institutions. Many payday loans and title loans, which are associated with a major share of complaints about personal loans, do not get reported to the major national credit bureaus, implying substantial understatement of the origination shares of traditional nonbank lenders.[18]

Figure 2: Complaint Shares by Lender Type

Source: CFPB Consumer Complaints Database

Figure 3: Origination Shares by Lender Type

Source: TransUnion Consumer Credit Database

Moreover, it is important to recognize that the frequencies of complaints across lender categories may reflect not only the quality of customer relationships but also the credit quality composition of borrowers and the credit standards applied to loan applications. Borrowers with lower incomes and lower credit scores will have higher delinquency rates, resulting in more complaints about difficulties curing delinquency. Although, in general, they may be less inclined to file complaints when facing difficulties. Lenders applying stricter credit standards are apt to have more complaints about the loan application process.

These caveats aside, two striking differences between the trends shown in Figure 2 versus those in Figure 3 merit consideration. They are not readily attributable to data imperfections or risk segmentation and they seem telling.

First, traditional nonbanks’ share of complaints rose markedly between 2015 and 2018, despite a declining market share of originations. Their share of complaints over this period might be related to an expanding share of online relative to storefront lending activity by these lenders.[19] That is, the growth in online lending within this segment might have led to some deterioration in the quality of customer relationships.

After 2019, traditional nonbanks’ share of complaints declined and by 2022 had returned to its 2015 level, despite the much reduced market share of these institutions. Their more recently declining share of complaints might reflect their continuing decline in market share. It might also be tied to the improved financial position of their predominantly lower-income clientele due to COVID-19 assistance programs. The direct economic assistance provided by the federal government to U.S. households particularly aided lower-income households with debt repayment.

Second, banking institutions’ share of complaints dropped off sharply relative to their share of originations, particularly before 2020. Banking institutions’ share of complaints stepped up between 2019 and 2020—the mirror image of the drop in share of traditional nonbanks—but this increase only partly offset the longer run reduction. The overall trend suggests improving the quality of banks’ lending conduct.

Lender Performance by Issue Category

Figure 4 breaks down the number of complaints about personal loans by issue category during the second half of 2017 through 2022, separately for the three lender classifications. Note that some complaint categories may be better indicators than others as to the quality of customer service and lending relationships. As noted previously, borrower credit quality might play a role.

Figure 4: Complaints by Lender Type and Issue Category (July-Dec 2017 through
2022)

Source: CFPB Consumer Complaints Database

On the one hand, the volume of complaints about delinquency difficulties may be more reflective of the risk profile of an institution’s clientele than of the lenders’ quality of services. A large volume of such complaints may be primarily driven by the number of customers encountering financial difficulties and only secondarily by the lenders’ actions. Thus, nonbanks, including fintechs, can be expected to have a higher volume of these complaints.[20]

Similarly, complaints about the loan application process often are tied to a loan having been denied based on a customer’s qualifications. Lenders that have more restrictive lending standards may be subject to more complaints regarding the application process, but this cannot be equated with customers being mistreated.

On the other hand, complaints about alleged deceptive practices may be particularly concerning. An elevated volume of such complaints might indicate unfair, deceptive or abusive lending practices that are illegal and can lead to significant financial harm for consumers that may be comparatively difficult to remediate.

Nonbanks, both traditional lenders and fintechs, have a higher number and concentration of complaints about alleged deceptive practices (about one-third of their complaints) compared to banks (about one-fifth). This differential persists if we exclude (in consideration of the factors noted above) the delinquency difficulties and loan application process categories.[21]

Banks incur relatively few complaints about unauthorized charges or debits, accounting for a mere three percent of these complaints. Fintechs are over-represented in the category of disputed information in the credit report, with close to half of the complaints of this type. Inaccurate credit reporting is a significant concern, as it may harm consumers’ future access to affordably priced credit.

Banks’ largest concentrations of complaints are in the categories of customer service and payments not properly credited. These areas might merit a closer look for exploring potential steps toward improvement.

Examining the time series of complaints within each issue category by lender type yields a couple of additional notable insights, specifically for the categories of customer service and disputed information on credit reports. The trends by lender type for these categories are shown in Figures 5 and 6, respectively.

The volume of complaints regarding customer service spiked for banks during 2020, while dropping for fintechs and traditional nonbanks. The jump in complaints for banks might have been tied to the pandemic-related, temporary branch closings during 2020, which took away an important delivery channel for customer support. Altogether, 2020 accounted for two-fifths of banks’ total complaint volume over the five-and-a-half-year period.

Figure 5: Complaint Trends by Lender Type: Customer Service

Source: CFPB Consumer Complaints Database

Figure 6: Complaint Trends by Lender Type: Disputed Information on Credit Reports

Source: CFPB Consumer Complaints Database

Regarding disputed information on credit reports, an upward trend is observed for all three lender types. This trend is relatively mild for banks. It is particularly steep for fintechs, which is concerning.

Complaint Remediation

The CFPB complaints data include three additional potential indicators of the quality of a lender’s conduct, which assess the extent to which a lender is responsive to complaints and seeks to remediate perceived wrongdoings. These are: (1) whether the lender met the CFPB’s required response timeline (“timely” versus “non-timely” response); (2) whether the complainant was granted monetary compensation (“monetary relief”); and (3) whether some other form of remediation was provided (“non-monetary relief”). We find that according to each of these metrics, banks exhibit clearly superior performance compared to fintechs and traditional nonbank lenders.[22]

Non-timely response generally means that the lender failed to provide a substantive response within the requisite 15-day timeline and thus can be considered an objective indicator of substandard conduct. Figure 7 shows the frequency of non-timely response by lender type from 2015 through 2022, both overall and for the two subperiods 2015-2017 and 2018-2022. Although a non-timely response is uncommon—well under 10 percent of complaints have a non-timely response—there are notable differences by lender type and across subperiods.

Over the full period, the non-timely response rate for banks is 2.3 percent, less than half that of fintechs and about one-fourth that of traditional nonbanks. Banks’ non-timely response rate fell from close to 3.7 percent to 0.8 percent across the two subperiods. Traditional nonbanks also improved some over time, while fintechs did not.

Figure 7: Frequency of Non-Timely Response

Source: CFPB Consumer Complaints Database

Most complaints are closed with the lender providing an explanation without offering monetary or non-monetary remediation. In contrast to response timeliness, whether a lender provides compensation does not necessarily indicate better treatment of the consumer. On one hand, many complaints may be based on misunderstandings, or there might often be valid reasons for the lender’s conduct that the consumer does not fully appreciate. In such cases, there is no need for any sort of compensation. On the other hand, in some cases compensation might have been offered because the conduct that generated the complaint caused undue harm to the consumer.[23]

With these caveats acknowledged, we proceed under the assumption that conditional on the issue category, the harm associated with a complaint would not systematically differ by lender type. That is, conditional on a complaint being filed and on the issue category, we see no reason to expect systematic differences in complaint severity by lender type. Therefore, a higher frequency of remediation in a lender category should indicate superior conduct in the sense of lenders making a greater effort to mitigate potential consumer harm.

Because the applicability and composition (monetary or not) of remediation are apt to vary by the nature of the complaint, it is best to examine relief provision by issue category. Figure 8, panels A through F show the frequency of each type of restitution by lender type and issue category for the period July 2017 through 2022 (excluding the two smallest categories). The aggregate frequencies are reported in Figure 9 for this period, and in Figure 10 for the full analysis period of 2015 through 2022.

In nearly every issue category, banks provided both monetary and non-monetary restitution much more frequently than either fintechs or traditional nonbank lenders.[24] The largest divergences are in the categories of alleged deceptive practices, payments not properly credited and disputed information on the credit report.

In general, traditional nonbank lenders have provided relief with somewhat greater frequency compared to fintechs. Fintechs were much less likely to provide relief for difficulties with curing delinquency compared to both banks and traditional nonbank lenders.

For alleged deceptive practices, banks provided monetary or non-monetary relief more than three times as frequently as fintechs and more than four times as frequently as traditional nonbank lenders. Banks provided relief for more than a quarter of the complaints in this category.

For payments not properly credited, as well as for disputed information on the credit report, banks provided monetary or non-monetary restitution more than twice as frequently as either fintechs or traditional nonbank lenders. Banks provided relief for more than 20 percent of complaints about payments not properly credited.

Figure 8: Frequencies of Monetary and Other Relief, by Issue Category (Jul-Dec 2017 through 2022)

Panel A: Alleged Deceptive Practices

Figure 8: Frequencies of Monetary and Other Relief, by Issue Category (Jul-Dec 2017 through 2022)
Panel A: Alleged Deceptive Practices

Panel B: Customer Service

Panel B: Customer Service

Panel C: Difficulties with Curing Delinquency

Panel C: Difficulties with Curing Delinquency

Panel D: Disputed Information on the Credit Report

Panel D: Disputed Information on the Credit Report

Panel E: Loan Application Process

Panel E: Loan Application Process

Panel F: Payments Not Properly Credited

Source: CFPB Consumer Complaints Database

Figure 9: Aggregate Frequencies of Monetary and Other Relief (Jul to Dec 2017 – 2022)

Figure 9: Aggregate Frequencies of Monetary and Other Relief (Jul to Dec 2017 - 2022)

Source: CFPB Consumer Complaints Database

Figure 10: Aggregate Frequencies of Monetary and Other Relief (2015 – 2022)

Figure 10: Aggregate Frequencies of Monetary and Other Relief (2015 - 2022)

Source: CFPB Consumer Complaints Database

As shown in Figure 10, from 2015 through 2022, banks provided monetary relief for about 13 percent of their total complaints, more than three times as frequently as fintechs and traditional nonbank lenders. In addition, banks provided non-monetary relief for about 8 percent of their total complaints, more than three times as frequently as fintechs and more than one-and-a-half times as frequently as traditional nonbank lenders.

CFPB Enforcement Actions

Public enforcement actions undertaken by the CFPB offer another vantage point from which to assess the quality of lending conduct in the personal loan market. As stated on the CFPB website:[25]

Congress provided the Consumer Financial Protection Bureau with four important tools to carry out the mission of protecting consumers: rulemaking, supervision, enforcement and education. When a depository institution, company, individual or other entity subject to our enforcement authority breaks the law, we may take enforcement action against them. In many cases, we will partner with other federal regulators or state agencies to investigate the wrongdoing and coordinate the enforcement action.

Thus, enforcement relates directly to lenders’ compliance with consumer financial protection regulations.

As described by the CFPB, several sources of information in addition to consumer complaints are used to identify potential issues that may warrant an investigation that might culminate in an enforcement action, including whistleblower reporting, referrals from federal regulators and other local, state and federal agencies and the results of supervisory exams.

In assessing whether to open an investigation, the CFPB will weigh, among other factors, whether “there is a plausible set of facts that, if proven, would amount to a violation of one or more federal consumer financial laws” and “whether there is evidence of a magnitude of harm that justifies investment of resources.” When warranted by an investigation, the CFPB may take a public enforcement action, which may involve either bringing the matter before a state or federal court or instituting an administrative adjudication proceeding.[26] Alternatively, the agency may close the investigation without taking public action or may choose to resolve the matter through the supervisory process.

According to the CFPB, a primary objective of the agency’s enforcement actions is to obtain redress for the consumer. As described by the CFPB, “We also have obtained a wide range of injunctive relief designed to stop unlawful conduct and prevent future violations. In some instances that relief has included banning individuals and companies from future participation in the marketplace.”

Notifications of enforcement undertaken are posted on the agency’s website, along with court documents and other related materials. They are searchable using product type, date range and other categorical filters, as well as by keywords.[27] Public enforcement actions related to the origination or servicing of personal loans are slotted to the product categories of “Short-Term Small-Dollar Loans” or to “Other Consumer Loans.”

Based on a systematic review of the enforcement actions from 2015 through 2023 within these two categories, we identified 41 actions related to personal loans. These are cataloged in Appendix 3; some involved repeat offenders.

Tellingly, in none of these enforcement actions have depository institutions been the defendants. Fintechs were defendants in eight of the cases and traditional nonbank companies or associated individuals were defendants in the other cases.

Concluding Remarks

Over the past decade, there has been a significant shift of market share in this sector from banks to traditional nonbanks and to a lesser degree from banks to fintech companies. Banks’ loss of market share likely reflects cost disadvantages arising from increased regulatory burdens.

This note demonstrates that banks have offered superior customer experience along at least one important dimension: customer experience as reflected in the volume and nature of complaints submitted to the Consumer Financial Protection Bureau. In this regard, banks’ loss of market share for personal loans has left consumers worse off. By the same measure, the shift in market share from traditional nonbanks to fintechs does not appear to have markedly benefited consumers.

Among other findings, the analysis demonstrates that fintechs and other nonbanks have a relatively high concentration of complaints in the arguably most concerning complaint category—alleged deceptive practices. Also, fintechs are over-represented in the category of disputed information in the credit report, another situation raising concerns about potential material harm to consumers.

Nonbanks, including fintechs, more frequently fail to provide a timely response to complaints filed with the CFPB, compared to banks. In nearly every complaint category, banks have provided both monetary and non-monetary restitution much more frequently than either fintechs or traditional nonbank lenders.

The analysis also demonstrates that banks’ share of complaints has declined significantly relative to their share of originations since 2015. This trend suggests improving the quality of banks’ lending conduct.

In addition, the note reviews CFPB enforcement actions conducted since 2015 that are related to personal loans. We find that in none of these actions have the defendants been depository institutions, providing another indicator of banks’ relative trustworthiness in the personal loan market.

To access the appendices, please click here.


[1] Most fintech companies specialize in certain types of lending, such as personal loans, although some offer more than one type of loan product. Big tech companies including Amazon and (just recently) Apple who now offer buy now pay later financing are also fintechs to the extent that the provide financial services.

[2] See Institutions subject to CFPB supervisory authority. The CFPB expanded its authority over time, through rulemakings, to include new categories of institutions. The CFPB began supervising larger credit reporting companies in July 2012 and began supervising nonbank auto finance companies in 2015.

[3] Also in July 2014, the system was expanded to incorporate complaints from consumers encountering problems with prepaid cards and complaints about debt settlement services and credit repair services.

[4] Other milestones include the following: The system was expanded to incorporate handling complaints on mortgages and other home loans in December 2011. In March 2012, the CFPB began accepting complaints about bank accounts, including checking accounts, savings accounts, CDs and related services, as well complaints about private student loans. In October 2012, the CFPB began accepting complaints about credit reporting.

[5] See CFPB Releases Largest Collection of Federal Consumer Financial Complaint Data | Consumer Financial Protection Bureau (consumerfinance.gov).

[6] See CFPB Publishes Over 7,700 Consumer Complaint Narratives About Financial Companies | Consumer Financial Protection Bureau (consumerfinance.gov).

[7] There might be a lag between when the CFPB began to accept complaints in these categories and when the public became generally aware that this was the case. Moreover, the consumer narrative field, which we rely on to some extent as described below, was absent from the data prior to June 2015.

[8] The CFPB revised its complaint-reporting product and subproduct categories in April 2017. Originally, payday loans had been a separate product category, while title and personal loans had been individual subproducts (alongside vehicle loans, pawn loans and personal lines of credit) within the consumer loan product category. After April 2017, payday, title and personal loans (alongside pawn loans and personal lines of credit) were combined as individual subproducts within a single product category. See the CFPB’s Summary of product and subproduct changes, April 24, 2017.

[9] The percentage of misclassified complaints associated with BHCs and S&LHCs fluctuated some from year to year, mostly ranging between 17 and 21 percent.

[10] The menu of issues and sub-issues to choose from varies somewhat depending on the product and subproduct.

[11] When the CFPB revised its complaint-reporting product and subproduct categories in April 2017 it also updated its list of issues and sub-issues.

[12] More specifically, the company response is deemed untimely if not provided within 30 days, or if not provided within 60 days after an initial response of “in progress.” See table 2 in the 2015 Audit Report from the Office of the Inspector General of the Board of Governors of the Federal Reserve System.

[13] SeeCFPB Launches Consumer Complaint Database” (June 19. 2012). A few complaints are recorded in the data simply as “closed”, none after 2017.

[14] See the CFPB webpage “Learn how the complaint process works.”

[15] When submitting a complaint, consumers can self-identify as older Americans (62 years of age or older) or as military-related (servicemember or veteran or spouse or dependent of a servicemember or veteran), and this information is recorded in the database as well. Another data item records the channel by which the complaint was submitted. Most complaints are submitted through the CFPB’s online portal. Other channels include phone, referrals from another regulatory agency, postal mail and fax.

[16] Credit unions which are included in the data for Figure 1, comprise an immaterial share of complaints. During 2015 through 2020, only about 1 percent of complaints in the CFPB data were associated with credit unions; their share rose to 1.4 percent in 2021 and then to 2 percent in 2022. As calculated using the TransUnion Consumer Credit Database, the credit union share of personal loan originations was upwards of 25 percent during 2015-2017, after which it declined gradually, reaching 20 percent as of 2022.

[17] According to TransUnion’s Q2 2023 Credit Union Market Perspectives Report (slide 16) Large and Regional Banks, which approximately correspond to banks with more than 10 billion in assets, accounted for more than 5/6ths of personal loan dollars originated in Q4 2022.

[18] See Ben Luthi, “What Is a Payday Loan and How Does It Work?” Experian.com, January 3, 2019. Also see Ben Luthi, “How Do Title Loans Work?” Experian.com, Experian.com, May 4, 2023. In addition to these loan types,

[19] For instance, a 2019 report from the Financial Health Network found that online payday lenders increased their share of the overall payday loan market between 2015 and 2018.

[20] Banks tend to serve the most creditworthy clientele (higher credit scores); traditional nonbank lenders typically serve the higher end of the credit risk spectrum; while fintechs tend to concentrate in the middle range. For example, the December 2023 Credit Industry Snapshot from TransUnion (page 10) reports that the percent of unsecured personal loans 60-plus days past due for banks, fintechs and traditional nonbank finance companies, respectively was 1.2 percent, 3.8 percent and 5.8 percent as of December 2023 and 1.1 percent, 3.4 percent and 6.7 percent as of December 2022.

[21] After excluding these categories, complaints about alleged deceptive practices comprise 49 percent of the complaints targeting traditional nonbank lenders, 40 percent for fintechs and 27 percent for banks.

[22] Until 2017, the data also included an indicator for whether the lender’s explanation was disputed by the consumer. We do not observe any meaningful difference in frequency of disputed explanations by lender type, which is 18 percent for banks, 20 percent for fintechs and 14 percent for traditional nonbanks. The lower frequency of disputed explanations for the traditional nonbanks might be due to their customers being more easily discouraged from pressing the complaint further.

[23] It would be difficult, if not impossible, to determine which complaints may have more or less validity or may have generated undue harm to the consumer. The narrative filed is of limited usefulness for this purpose because it often is not filled in or contains insufficient information to assess the severity of harm to the consumer.

[24] The single exception is for difficulties with curing delinquency (panel C), where traditional nonbanks had a slightly higher frequency of providing non-monetary relief compared to banks.

[25] See Life Cycle of an Enforcement Action | Consumer Financial Protection Bureau (consumerfinance.gov)

[26] As described by the CFPB, “Administrative adjudication proceedings are formal adversarial proceedings conducted by an administrative law judge, who issues a recommended decision to the Bureau’s Director. The Director issues a final decision, either adopting or modifying the administrative law judge’s recommended decision.”

[27] See Enforcement Actions | Consumer Financial Protection Bureau (consumerfinance.gov)