BHPH Economics and Portfolio Benchmarks 2026

Quick answers

Short answers to the questions dealers, lenders and reporters ask most often about buy-here-pay-here lending. Each links to the table and source behind it.

How big is the buy-here-pay-here market in the United States?

The Federal Reserve's BHPH proxy held about $32 billion in outstanding balances at Q3 2025, roughly 2% of all US auto loan balances and about 5% of subprime auto balances. The figure is conservative because it can miss loans booked through dealers' related finance companies. By a different measure, Experian put BHPH at 9.97% of US auto financing in Q2 2026. Table 1A [1] [4]

Is buy-here-pay-here lending growing?

Yes. Balances in the Federal Reserve's BHPH proxy grew 214% from Q1 2018 to Q3 2025, compared with 34% for traditional auto loans over the same period. Both figures are cumulative nominal growth, not annual rates. Table 1A [1]

What interest rate do buy-here-pay-here loans carry?

In the Federal Reserve's sample of loans observed from Q1 2018 to Q3 2025, subprime BHPH loans carried a balance-weighted derived interest rate of 25.39%, compared with 14.60% for traditional subprime auto loans. The rate is imputed from principal, payment and term; it is not an observed contract APR or a 2026 origination average. Table 1B [1]

What are the average BHPH loan amount, payment and term?

For subprime BHPH loans in the Federal Reserve's Q1 2018 to Q3 2025 sample, the average origination principal was $15,402, the average monthly payment was $405 and the average term was 55 months. About 14% of those balances were on weekly or biweekly payment schedules. Table 1B [1]

What is the delinquency rate on buy-here-pay-here loans?

At Q3 2025, 10.0% of balances in the Federal Reserve's BHPH proxy were delinquent (more than 30 through 120 days past due), compared with 3.8% for traditional auto loans. Table 1C [1]

What is the default or charge-off rate for BHPH dealers?

There is no current industry-wide rate. The available reference points are: America's Car-Mart reported net charge-offs of 27.6% of average finance receivables for fiscal 2026; a historical Subprime Analytics composite averaged a 36.04% default (write-off) rate across 2018 to 2021; and the Federal Reserve found BHPH balances in default or charge-off status at 1.88 times the traditional rate at Q3 2025. These measure different things and should not be averaged. Table 1D Table 8A [1] [2] [8]

How much do BHPH customers put down?

America's Car-Mart reported an average down payment of 5.1% for fiscal 2026, which works out to roughly $1,023 on its $20,064 average retail price (a derived figure). The older NIADA 2021 composite reported an average cash down payment of $1,192. Neither is a current industry average. Table 5A Table 8A [2] [8]

How much gross profit does a BHPH dealer make per vehicle?

America's Car-Mart reported gross profit of $7,442 per retail unit and a 35.4% gross margin for fiscal 2026; the NIADA 2021 composite reported $6,981 per vehicle. Gross profit is measured before credit losses: Car-Mart's credit-loss provision equalled 40.8% of sales in the same year. Table 1D Table 8A [2] [8]

How common is repossession in buy-here-pay-here lending?

At Q3 2025, about 5% of balances in the Federal Reserve's BHPH proxy were in active repossession status, compared with less than 0.5% for traditional auto loans. That is a point-in-time share of balances, not the share of loans that are eventually repossessed. No reliable average time-to-repossession is established. Table 1C Section 7 [1]

Is the FTC CARS Rule still in effect?

No. The FTC removed the CARS Rule from the Code of Federal Regulations in a Federal Register notice dated 12 February 2026, after a federal court vacated it. Other federal and state consumer-protection requirements continue to apply. Section 7 [14]

What is the TradeBasis USA BHPH Macro Conditions Index?

It is an experimental monthly index, published by TradeBasis, that scores the national economic pressure around buy-here-pay-here lending from 0 to 100 using employment, household-budget and funding data. Higher scores mean more pressure. It read 40.4 in August 2026, down 14.2 points from May 2026. It describes the economic setting and is not a measure of BHPH loan performance. Table 2A Index page [17]

How to read this report

This report brings together public measurements of buy-here-pay-here (BHPH) lending, disclosed operator results and related auto-finance indicators, and adds an original TradeBasis index calculated from public economic data. It is built for practical reference and accurate quotation. Each benchmark belongs to a stated population and period; publication in 2026 does not make older observations 2026 data.

BHPH dealers sell vehicles and provide financing directly. Related finance companies, dealer subsidiaries and indirect lenders may be classified differently across datasets, so this edition keeps sector measurements, operator reference points and adjacent-market indicators separate.

All dollar amounts are US dollars. Percentages keep the source's precision and definition. Figures are descriptive reference points, not operating targets or recommendations. Bracketed numbers link to the source list at the end of the report. This report is not a census of BHPH dealers; observation dates vary by metric.

1Benchmark tables

3Market coverage and borrower mix

What the public datasets cover

The Federal Reserve identifies independent-auto-dealer-coded loans in the NY Fed/Equifax Consumer Credit Panel as a proxy for BHPH lending. It can miss loans coded to finance companies, including subsidiaries of BHPH dealers. The reported $32 billion therefore describes the identified sample rather than a complete industry inventory. [1]

Experian's financing-source classification and NIADA's combined BHPH/other category answer different questions. NIADA's 32.6% share does not show that one-third of independent dealers offer BHPH financing, and "other" financing cannot be silently recategorized as BHPH. [4] [5]

Borrower mix

In the Fed's sample, the deep-subprime share fell from about 70% in 2018 to slightly above 50% in 2025. The historical Table 1 puts 77.52% of BHPH balances in its broader subprime grouping. These use different credit-group definitions and are not competing measurements. [1]

Changes in observed credit mix do not explain why customers entered BHPH financing or show that loan performance improved. This report supplies no sector borrower-income benchmark: dividing an average payment from one dataset by a payment-to-income ratio from another does not give a reliable estimate of average borrower income. The macro index's wage and spending inputs are also national proxies, not estimates of BHPH borrower income; see Table 2A. [17]

Independent dealer market context

NIADA reports about 9.8 million independent-dealer used-vehicle sales in 2025, within an estimated 38.6 million total used transactions. These are broader independent-retail figures, not BHPH sales volumes. [5]

4Credit performance and measurement

Fitch index revision

Fitch revised the inclusion method for its prime and subprime auto ABS indices from the July 2026 index and applied the change retrospectively. The values below come from one account of the revised series; earlier originally published values are not mixed in. [6] [16]

Credit Acceptance collection forecasts

Stocks, flows and losses

A delinquent balance share is a stock at a date. New delinquencies are a flow over a period. Annual net charge-offs use average receivables, while cumulative losses use an origination-pool denominator. Allowances are accounting estimates of expected losses, and provisions are period expenses. Comparisons must keep these distinctions.

The TradeBasis macro score is a separate contextual measure. Comparing its movement with credit data requires matching observation periods and allowing for loan seasoning; the series does not establish a causal link. [17]

5Deal and portfolio economics

Down payments and cash collected

Car-Mart reported a 5.1% average down-payment percentage for FY2026 and 6.1% for its fourth quarter. Multiplying these by the matching average retail prices gives about $1,023 and $1,228. These are scale illustrations derived from two averages, not directly observed mean cash down payments. [2]

Gross profit and credit expenses

Car-Mart reported FY2026 gross profit per retail unit of $7,442, a 35.4% gross margin and a credit-loss provision equal to 40.8% of sales. These measures have different timing and denominators. Gross profit is not a prepaid cash loss reserve, and the provision-to-sales ratio is not a lifetime loss percentage. [2]

Comparing this operator's gross with franchise-dealer front-end gross requires reconciling ancillary revenue, accounting policy, credit expenses, overhead and financing. This report makes no claim that net profitability is higher or lower than conventional used-car retail.

Contract terms and collateral

The Fed's $15,402 average subprime BHPH principal and Car-Mart's $20,064 average retail price come from different populations and periods. Dividing one by the other does not give a typical financed share, loan-to-value ratio, down payment or markup. [1] [2]

Loan-to-value comparisons need the same collateral valuation basis: retail price, wholesale value and book value are not interchangeable. Payment-to-income comparisons need matching payment schedules and the same gross or net income definition.

In August 2026 the macro index's essential-cost and earnings indexes were 107.8 and 108.3, each based on its 2024 average of 100, giving a ratio of 99.5. This describes selected costs relative to average pay, not the share of a borrower's income available for a vehicle payment. [17]

Two numbers in every BHPH deal depend directly on vehicle value: the acquisition and reconditioning cost at the start, and the recovery value at the end if the vehicle is repossessed. A dealer's own loss and return figures are only as reliable as the valuation behind them. Valuing each unit against current comparable listings at acquisition and again at repossession gives a consistent basis for both. That is the work TradeBasis is built for.

Historical cash-on-cash returns

Car-Mart's September 2024 disclosure reports the pool returns below as of 31 July 2024. The FY2025 originations were largely outstanding when forecast. These are historical disclosures, not 2026 estimates. [9]

What comparable dealer economics require

A dealer-level comparison should identify origination month, cash invested, timing of collections, defaults, net recovery proceeds, servicing costs, financing costs and shared overhead. A cumulative cash multiple alone does not describe payback time or peak capital required.

This edition does not establish a current industry-wide cash-on-cash or store-payback benchmark. Franchise aggregator estimates and values read approximately from chart axes are not presented as measured benchmarks.

6Funding and operator events

Bank financing in the Fed matched sample

These are facilities to dealer businesses, not consumer vehicle loans. The matching exercise covers identified large-bank exposures and is not a complete census of industry funding. [1]

The TradeBasis funding pillar measured 75.3 out of 100 in August 2026. Its inputs were a 6.75% monthly average prime rate and 1.8% net tightening for small-firm commercial and industrial loans, from the bank lending survey released 3 August 2026. These broader measures supplement the dated BHPH facility disclosures above; they do not measure dealer warehouse pricing or credit availability directly. [17]

America's Car-Mart funding and operations

Car-Mart's FY2026 release disclosed a need for additional financing and substantial doubt about its ability to continue as a going concern. Its September 2026 release reported limited inventory and origination capacity, with retail units down 81.9% and revenue down 57.3% year over year for Q1 FY2027. These are company-specific disclosures, not industry growth rates. [2] [3]

Period-end finance receivables, net of allowance, were about $909.8 million at 31 July 2026. Gross principal balance and net carrying amount must be labeled separately whenever used. This report does not assume that a previously disclosed covenant-relief date remained the operative deadline after later amendments. [3]

Tricolor as an event reference

Tricolor's bankruptcy and the December 2025 federal charging announcement are relevant to collateral reporting and funding controls. The DOJ alleged fraudulent collateral practices. Allegations and charging status remain attributed; they do not establish how common similar conduct is among BHPH dealers. Trial dates, litigation outcomes and recovery forecasts need separate, current docket checks. [1] [15]

7Collections, repossession and regulation

Assignments, recoveries and proceeds are different measures

A repossession assignment is an instruction to recover a vehicle. A completed recovery is a recovered unit. Auction or resale proceeds are a dollar amount. A balance marked as repossessed is an account-status observation. These describe different stages and cannot substitute for one another.

Recovery value and elapsed time

The RDN unit-recovery ratio differs from Fitch's financial recovery index. Likewise, the Fed's approximately 5% active-repossession balance share is a point-in-time stock and cannot show how many days pass from the first missed payment to recovery. [1] [6] [11]

A widely repeated comparison of 30 days or less for BHPH against about 92 days for the wider market is not used here as a benchmark: its sources do not share a population, start date, endpoint or method. Dealer comparisons should state whether timing ends at assignment, physical recovery, sale or final cash receipt.

Historical research and resale

The CFPB's January 2025 repossession report is a useful primary research reference, with its own lender sample and observation window; comparisons with platform data need matching definitions. [12] Repeated resale examples do not establish an average recovery cycle or its profitability. Down payments, intervening payments, recovery costs and resale proceeds must stay distinct.

Regulatory context

This part records selected developments relevant to collections and vehicle finance. It is not a state-by-state statement of dealer obligations. Event dates, effective dates and the parties covered by an action should accompany any quotation.

Credit Acceptance settlement. On 17 September 2026 the New York Attorney General announced a multistate Credit Acceptance settlement described as about $700 million, including more than $630 million in debt relief, $60 million in restitution and $15.5 million in penalties. The amounts are rounded in the announcement and are not an independently reconciled total. For certain covered at-risk borrowers who default within 12 or 18 months and whose vehicles are repossessed and sold, the announcement describes forgiveness of 95% of the remaining debt, limits on collecting the balance, and restrictions on lawsuits or resale of that debt. These are settlement-specific conditions, not a general rule for BHPH loans. [13]

Federal CARS Rule. The FTC's 12 February 2026 Federal Register notice removes the CARS Rule from the Code of Federal Regulations following the court's vacatur. Its withdrawal does not mean other federal or state consumer-protection requirements have stopped applying. [14]

Supervision and device practices. Changes to supervisory thresholds should be identified as proposals or final rules and are not changes to all substantive legal obligations; withdrawing agency guidance is not repealing the underlying statute. The Fed notes GPS use by BHPH operators, but this edition does not establish a current national penetration rate for GPS or starter-interrupt devices; vendor subscriber counts include other customer types. Device restrictions, repossession procedures, notice requirements and rate limits depend on the applicable law and transaction. This edition does not reproduce state rate-cap tables or penalty schedules from vendor summaries. [1]

8Historical comparisons and outdated figures

Historical benchmarks stay useful when their original definitions and periods are kept. A later publication date, or a different operator's disclosure, does not automatically replace them. The 2021 Year in Review includes separate NIADA and NCM operating composites and loss metrics averaged across 2018 to 2021. [8]

How to replace an outdated claim

For an undated "average APR around 20%," use the Fed's 25.39% derived historical-sample rate only with its Q1 2018 to Q3 2025 period and definition. The two figures do not show a measured 5.4-point increase in a consistent series.

For an undated down-payment amount, give a dated operator reference or a clearly identified historical composite; Car-Mart's derived dollar illustration is not a new industry mean. For "one-third of dealers offer financing," do not substitute NIADA's financing share, which has a different denominator.

Editorial practice: keep the historical value with its date, add a newer comparable measurement when one exists, and say when the newer source measures something different. Values that cannot be traced are omitted rather than relabeled as current.

9Methodology and updates

Evidence and classification

This report combines a secondary compilation by TradeBasis Research (Sections 1 and 3 to 8) with an original TradeBasis index (Section 2). Source entries distinguish primary research, issuer disclosures, accessible secondary reporting and derived indices. Each number is tied to a named source and an observation period. Source authority and population representativeness are assessed separately.

Known gaps

This edition does not establish a current representative BHPH dealer count, borrower-income average, nationwide device-usage rate, time-to-repossession average, industry-wide net return or complete dealer P&L composite. No claim is made that this is the first free report or that all newer data elsewhere is paywalled.

Refresh schedule

TradeBasis Research reviews operator disclosures quarterly, ABS indices when updated, and industry composites when released. Data is replaced only after checking population and method. Forecast dates are preserved, and revisions to old observations are kept separate from newly observed periods. Legal status should be checked separately at the time of use.

The macro index is refreshed monthly using its versioned method. Baseline dates and weights stay fixed within a version, missing observations are kept, and revisions are archived. Cite the index version and release date. [17]

Revision record

3 October 2026, web edition. Published at tradebasis.net with plain-English key findings, per-table citations and the companion index page and data files.

25 September 2026. Added the TradeBasis USA BHPH Macro Conditions Index (Section 2), its method and source [17], with related context in Sections 3 to 6. Sector and operator benchmark values keep their prior observation periods.

22 September 2026. Corrected down-payment periods, cash-flow distinctions, unsupported inferences, mixed-basis Fitch comparisons and historical loss periods.

21 September 2026. First edition.

Sources

Primary and secondary evidence are labeled. Sources [1] to [16] were accessed on 22 September 2026. Source [17] uses inputs retrieved on 26 September 2026 UTC. Observation, retrieval and publication dates are kept separate.

1. [1]Federal Reserve Board. Subprime Auto Lending Trends in Buy Here Pay Here Auto Lending. FEDS Notes, 8 May 2026.
   Primary research; CCP consumer loans and Y-14Q bank facilities. Table 1 covers Q1 2018 to Q3 2025.
   federalreserve.gov/econres/notes/feds-notes/subprime-auto-lending-trends-in-buy-here-pay-here-auto-lending-20260508.html
2. [2]America's Car-Mart. Fourth Quarter and Fiscal Year 2026 Results. 14 July 2026.
   Primary issuer disclosure; fiscal year and quarter ended 30 April 2026.
   ir.car-mart.com, Q4 and FY2026 results
3. [3]America's Car-Mart. First Quarter Fiscal Year 2027 Results. 9 September 2026.
   Primary issuer disclosure; quarter ended 31 July 2026. Fiscal 2027 is not calendar 2027.
   ir.car-mart.com, Q1 FY2027 results
4. [4]Experian. Auto Loan Rates and Financing for 2026. Accessed 22 September 2026.
   Publisher's public summary of Q2 2026 automotive finance data; includes BHPH financing share.
   experian.com/blogs/ask-experian/auto-loan-rates-financing
5. [5]NIADA. UCIR Shows Independent Used Sales Top 9.8 Million in 2025. March 2026.
   Primary public summary of UCIR 2025; preserves the combined BHPH/other category.
   niada.com, UCIR 2025 summary
6. [6]Auto Remarketing, reporting Fitch Ratings. Fitch Watches Tax Refund Boost on Securitizations Fade During Half of the Year. 24 August 2026.
   Secondary account of Fitch's 19 August 2026 monitor. Revised historical index basis; not a BHPH-only series.
   autoremarketing.com, Fitch securitization coverage
7. [7]Credit Acceptance, via GlobeNewswire. Second Quarter 2026 Results. 4 August 2026.
   Primary issuer release; collection forecasts as of 30 June 2026. Contractual repayments include principal and interest.
   globenewswire.com, Credit Acceptance Q2 2026
8. [8]NABD, NIADA, SGC CPAs and contributors. BHPH Benchmarks: Market Perspectives and Trends, 2021 Year in Review. June 2022.
   Historical industry report. PDF page 12 contains separate NIADA and NCM operating composites; page 13 states loss metrics use 2018–2021 averages.
   sgcaccounting.com/Resources/BHPHBenchmarks2021.pdf
9. [9]America's Car-Mart, SEC exhibit. Q1 Fiscal 2025 Results and Cash on Cash Returns. 4 September 2024.
   Primary historical issuer disclosure; pool estimates as of 31 July 2024, not current forecasts.
   sec.gov, Car-Mart exhibit 99.1
10. [10]Federal Reserve Bank of New York. Household Debt Balances Decreased Slightly. 11 August 2026.
    Primary Q2 2026 household-credit release; auto balances and annualized transitions into serious delinquency.
    newyorkfed.org/newsevents/news/research/2026/20260811
11. [11]CURepossession, reporting RDN. RDN Repossession Volume Report Q3 2025. 2025; accessed 22 September 2026.
    Secondary publication of RDN platform data. Distinguishes assignments, completed recoveries and publisher projections.
    curepossession.com/rdn-repossession-volume-report-q3-2025
12. [12]Consumer Financial Protection Bureau. Repossession in Auto Finance. January 2025.
    Primary research; lender sample and study period must accompany any extracted benchmark.
    consumerfinance.gov, Repossession in Auto Finance (PDF)
13. [13]New York Attorney General. Credit Acceptance Multistate Settlement Announcement. 17 September 2026.
    Primary announcement. Relief and restrictions concern covered Credit Acceptance loans; not a general BHPH rule.
    ag.ny.gov, Credit Acceptance settlement
14. [14]Federal Trade Commission, Federal Register. Revision and Withdrawal of Rules to Conform to Federal Court Decisions. 12 February 2026. 91 FR 6507.
    Primary rulemaking notice; includes removal of the CARS Rule after vacatur.
    federalregister.gov, 2026-02866
15. [15]US Department of Justice, SDNY. CEO, CFO, COO Charged in Connection with Collapse of Tricolor Auto. 17 December 2025.
    Primary historical charging announcement. Allegations are not findings of guilt.
    justice.gov, Tricolor charging announcement
16. [16]Fitch Ratings. North American Auto ABS Monitor 1H26. 19 August 2026.
    Underlying rating-agency publication; the figures in this report are sourced through accessible coverage [6].
    fitchratings.com, Auto ABS Monitor 1H26
17. [17]TradeBasis. TradeBasis USA BHPH Macro Conditions Index. Version 1.0-experimental. September 2025 to August 2026. Released 26 September 2026 UTC.
    Original derived index using public BLS, US Department of Labor and Federal Reserve data, including FRED-distributed series. Equal-weight national macro pillars; latest-revised retrospective history. Ten complete monthly scores; October and November 2025 are unavailable where required inputs were not published. All source observations, basket weights, formulas and the source list are published as companion CSV files so the index can be reproduced.
    tradebasis.net/research/bhph-macro-index   Index CSV   JSON   Methodology   Source data   Basket weights   Source list

How to cite this report

This report is free to read, quote and reuse with attribution to TradeBasis under a Creative Commons Attribution 4.0 licence. Figures from third-party sources remain the work of those sources; please credit both.