Nondepository Credit Intermediation (U.S.) — Industry-Group Primer

NAICS 2022 code 5222 — a rollup primer covering three child industries. NAICS is the North American Industry Classification System, the standard code set the U.S. and its trading partners use to group businesses by their primary activity. Figures are reported federal facts with citations; statements about the future are labeled as judgments, not forecasts. Relevant to both public-market and private investors.

1. Overview

This is the part of American lending that runs without a deposit base. The Census Bureau groups the nation's credit businesses under subsector 522, Credit Intermediation and Related Activities, and splits it three ways: 5221 Depository Credit Intermediation (banks and credit unions, which fund loans with customer deposits), 5223 Activities Related to Credit Intermediation (the plumbing — payment networks, brokers, loan servicers), and, in the middle, 5222 Nondepository Credit Intermediation — lenders that make loans but take no deposits [1].

Because they cannot fund themselves with checking and savings balances, the firms in 5222 borrow wholesale instead — warehouse credit lines, asset-backed securities (ABS, bonds backed by pools of loans), mortgage-backed securities (MBS), repurchase agreements, and ordinary capital-markets debt — and a large share of them originate to distribute, selling or securitizing the loans they write rather than holding them to maturity. That single structural fact defines the whole level. It makes 5222 a spread-and-fee business exposed to two external variables at once: the cost and availability of wholesale funding, and the credit-loss rate — both of which turn against owners in the same downturn. It is not the steady, deposit-cushioned utility that banking can be; it is the leveraged, market-funded edge of consumer and commercial credit that produced both the 2008 mortgage rescue and the recurring distress in subprime lending.

5222 contains three very different children [1]:

• 52221 Credit Card Issuing — revolving unsecured credit to cardholders (card banks and monoline issuers).
• 52222 Sales Financing — lending tied to a specific purchase (autos, trucks, equipment, furniture, insurance premiums), usually arranged at the point of sale and secured by the thing bought; the signature players are manufacturer-owned captive finance arms.
• 52229 Other Nondepository Credit Intermediation — everything else: nonbank mortgage lending, cash consumer and payday loans, the mortgage secondary market (Fannie Mae and Freddie Mac), plus trade finance, factoring, farm credit, and pawn.

Ways in (previewed here, detailed in sections 4 and 10). Public markets: listed card issuers, near-pure-play auto/consumer lenders and the manufacturer parents that own captives, nonbank mortgage originator-servicers, mortgage real estate investment trusts, pawn and factoring names, and over-the-counter shares of Fannie Mae and Freddie Mac. Private markets: the dominant instruments here are bonds, not equity — card, auto, equipment, and consumer-loan ABS; agency MBS — alongside private-equity ownership of branch, captive, and mortgage platforms.

2. What's inside — the three children and how they differ

The distinctive fact about 5222 is that its three children compete in different markets, price differently, and are owned differently. A household shopping for a credit card does not substitute a tractor loan; a bond investor buying agency MBS is not buying subprime auto risk. So the level is best understood as three return engines under one structural roof, not one industry. The table contrasts them; every share is reconciled against this level's federal totals in Section 3.

ABS = asset-backed security; MBS = mortgage-backed security; GSE = government-sponsored enterprise; REIT = real estate investment trust; PE = private equity; CR4 = combined revenue share of the four largest firms; HHI = Herfindahl-Hirschman Index, a 0–10,000 concentration gauge where U.S. antitrust agencies have historically treated below 1,500 as "unconcentrated." Receipts shares use the 2022 Economic Census; jobs and branch shares use 2023 County Business Patterns [2][3].

Four contrasts to carry into the rest of this primer:

1. One child is most of the level — but is itself three-in-one. 52229 books nearly six in ten revenue dollars, about seven in ten jobs, and nine in ten locations. It is the bulk of 5222 by every count. Yet it is the least coherent child: a labor-heavy mortgage machine, a small high-yield consumer-loan business, and a government-anchored securitization near-monopoly, bundled together. Read its own primer before treating "52229" as a single thing.
2. Dollars and people are inverted. Credit card issuing earns the highest revenue per worker in the level — roughly $2.1 million, versus ~$1.6 million in sales financing and ~$1.0 million in the "other" bucket — because card issuing is a balance-sheet business run by a small, highly paid staff (monolines and card banks), while nonbank mortgage origination inside 52229 is hands-on and headcount-heavy. Card issuing occupies just ~1% of the level's locations but ~23% of its revenue.
3. The measured concentration is a blend, not a market. The level's CR4 of 37.8% and HHI of 515.7 look "unconcentrated," but that is an average across three unlike structures: card issuing is genuinely concentrated (CR4 63.1%), sales financing and the mortgage/consumer parts are fragmented, and the secondary-market piece is a chartered duopoly. No single firm competes across all three, so the level-wide figure describes an accounting group, not a competitive market.
4. Ownership could hardly differ more. Card issuing is a public-pure-play-plus-megabank story; sales financing is a captive-subsidiary story (the giants are wholly owned by Ford, GM, Toyota, Deere); the "other" bucket is anchored by government-chartered enterprises under federal control. There is no single "owner profile" for 5222 — which is exactly why you invest in the child, not the code.

3. How big it is

Federal business statistics for NAICS 5222 as a whole. County Business Patterns (CBP) reports establishments, employment, and payroll for 2023; the 2022 Economic Census concentration tables report firms, receipts, and concentration ratios [2][3]. The children reconcile almost exactly to these totals, which is why the shares in Section 2 can be trusted.

How the children add up. Receipts, establishments, and employment sum to the level almost to the dollar: the three children's receipts ($161.6B + $133.6B + $421.7B) equal $716.9B; their establishments (615 + 3,353 + 38,403) equal 42,371 exactly; their employment (75,504 + 82,691 + 411,884) equals 570,079 exactly. The firm counts sum to 13,691 against a level total of 13,616 — a tiny gap because a company operating in more than one of the three industries is counted once at the level, confirming that almost no firm spans the segments. Annual payroll works out to roughly $111,000 per worker [2], a high-skill, high-pay lending workforce rather than a headcount-heavy one.

How to read the concentration. Do not read the level's CR4 of 37.8% or HHI of 515.7 as one competitive market. They are a weighted blend: card issuing (CR4 63.1%, HHI 1,246) is concentrated, sales financing (CR4 29.3%, HHI 379) and the mortgage/consumer world are fragmented, and the secondary-market piece is a Fannie/Freddie duopoly (its subsegment CR4 is 80.2%) [4][5][6]. The level number is essentially the average of a competitive origination world and a duopoly-dominated secondary market — informative about the group, misleading about any real market inside it.

The undercount caveat — read this before quoting $716.9 billion. These are operating-business measures — revenue and headcount — not loan balances or total credit outstanding, and several forces make them understate the activity:

• Receipts are interest and fee income, not the size of the market. The loans these firms touch dwarf their own revenue line, and the figures must never be added to household- or mortgage-debt totals. For scale: U.S. mortgage debt was roughly $21 trillion in 2025, auto-loan balances about $1.67 trillion, and credit-card balances about $1.28 trillion [11] — each an order of magnitude above the relevant child's receipts.
• The biggest card and auto books live outside this subsector entirely. The largest credit-card issuers (JPMorgan Chase, Bank of America, Citigroup) and the banks that lead auto lending run those portfolios inside enterprises classified as Commercial Banking (NAICS 522110) — a depository code (5221), not 5222. What 5222 mainly captures is the monolines, credit-card banks, captives, and independents [4][5].
• Captive giants look tiny as firms. A single captive such as Ford Credit or GM Financial carries well over $100 billion of finance receivables on its balance sheet [5] — more than the whole sales-financing code's ~$134 billion of annual receipts — because balance-sheet assets and income are different things. The 2,180-firm count for that child makes it look small and fragmented; it is not.
• Government "plumbing" is excluded. The Economic Census counts private and shareholder-owned businesses. It leaves out the government corporations that do this exact work — Ginnie Mae (over $2.8 trillion of guaranteed MBS), the Export-Import Bank, and the cooperatively owned Farm Credit System (~$438 billion in loans). Fannie and Freddie are counted (they are shareholder-owned), which is why 52229's receipts are so large, but much of the surrounding federal credit apparatus is not [6][9].
• Small and individual owners are undercounted. CBP covers only employer businesses with paid staff; it excludes the self-employed and no-employee firms. That undercount is heaviest exactly where small, individually owned operators dominate — pawnshops, single-branch payday and small-loan storefronts, sole-proprietor mortgage shops, and small factoring firms. The true count of the long tail is higher than 42,371.
• Mixed vintages. Receipts, firms, and concentration are 2022; establishments, employment, and payroll are 2023. Treat the two federal snapshots as adjacent, not simultaneous.

Our ground-truth file gives no level-wide figure for loan balances, originations, average annual percentage rate (APR), net interest margin, delinquencies, or charge-offs, so none is estimated here. Treat 5222's federal numbers as the operating core of deposit-free lending — not the whole flow of credit it intermediates.

4. The investable universe — where value concentrates across the children

There is no clean single-ETF or single-screen way to own the code, because value concentrates in different vehicles in each child and many of the biggest players are private, quasi-public, or classified in adjacent codes. Group the opportunity by engine. (Tickers and multiples belong here, not in the earlier sections.)

• Credit-card issuing (52221) — a concentrated set of public pure-plays. American Express (AXP), Capital One (COF) — the largest U.S. issuer after its ~$35 billion purchase of Discover closed in May 2025 — Synchrony (SYF, the largest private-label/store-card issuer), and Bread Financial (BFH). Card operations inside diversified banks (JPMorgan JPM, Citigroup C, Bank of America BAC, U.S. Bancorp USB) are a large segment but a bet on the whole bank. Member-owned credit-union issuers (Navy Federal, PenFed) and private card banks sit outside the code. The private route is mainly credit-card ABS [4][7].
• Sales financing (52222) — mostly through the parent. Few pure-play equities exist because the captives are wholly owned subsidiaries; equity exposure comes through the manufacturer parent — Ford (F), General Motors (GM), Toyota (TM), Deere (DE), Caterpillar (CAT), PACCAR (PCAR), CNH (CNH). The closest listed proxies are subprime auto lender Credit Acceptance (CACC), bank-owned prime lender Ally Financial (ALLY), and point-of-sale financier Synchrony (SYF), plus used-car retailers with in-house lending (CarMax KMX, America's Car-Mart CRMT, Carvana CVNA). The deepest institutional exposure is auto and equipment ABS and captive senior notes [5].
• Other nondepository credit (52229) — the deepest public roster, plus bonds and special situations. Nonbank mortgage originator-servicers: Rocket Companies (RKT, now servicing roughly one in six U.S. mortgages after buying Mr. Cooper), UWM Holdings (UWMC), PennyMac (PFSI), Rithm Capital (RITM), loanDepot (LDI). Consumer lenders: OneMain (OMF), Enova (ENVA), OppFi (OPFI), Oportun (OPRT), Regional Management (RM), World Acceptance (WRLD). Mortgage REITs: Starwood Property Trust (STWD), Blackstone Mortgage (BXMT), Annaly (NLY), AGNC (AGNC); ETFs REM and MORT bundle the group. Specialty operating names: pawn lenders FirstCash (FCFS) and EZCORP (EZPW), factoring lender Triumph Financial (TFIN), and agricultural GSE Farmer Mac (AGM). The GSE commons — Fannie Mae (FNMA) and Freddie Mac (FMCC), traded over-the-counter under federal conservatorship — are speculative options on the terms of an eventual government exit, not ordinary dividend stocks. And the mainstream way most investors already own this segment is indirectly, through agency MBS inside nearly every bond fund [6][10].

The synthesis for an allocator. 5222 offers roughly four distinct return profiles across the same cycle: high-return-but-cyclical card credit (52221 equities and card ABS); purchase-tied consumer/commercial credit whose share rotates with the rate cycle (52222 parents and auto/equipment ABS); origination-and-servicing cyclicality plus high-yield consumer risk (52229 equities and consumer-loan ABS); and low-risk, government-backed carry (agency MBS), with the GSE commons as a separate binary policy bet. In private markets, the level's exposure is dominated by debt, not equity — card, auto, equipment, and consumer ABS, agency and Ginnie Mae MBS, whole loans, and mortgage servicing rights — plus PE ownership of branch, captive, and mortgage platforms.

5. How the money works

Every business in 5222 runs the same core formula — interest and fees, plus gains on selling loans, minus funding cost, minus credit losses, minus operating and compliance cost — but the emphasis differs sharply by child, and that is where the analysis lives.

• 52221 (spread plus a payments kicker). Card issuers borrow cheaply and lend at high card APRs, earning a net interest margin (NIM); add interchange (a ~2.3% slice of each purchase) and cardholder fees; and pay it back in funding cost and — the swing factor — credit losses tracked by the net charge-off rate. Federal Reserve research attributes roughly 80% of issuer profitability to lending, with fees around 15% [12]. Losses show up on a lag, so aggressive growth today can mean pain in two years.
• 52222 (contract yield minus funding minus losses). Owners earn the spread between what they collect on installment loans and leases and what they pay to fund them, minus credit and operating costs [5]. Two captive-specific twists matter: subvention (the parent subsidizes a below-market "0% APR" as a marketing cost to move its own product) and residual-value risk on leases (the lender owns the vehicle and bets on its resale value). Funding is deposits for bank-owned lenders (Ally, Synchrony) versus wholesale bonds and ABS for the captives.
• 52229 (gain-on-sale, servicing, and thin-fee-at-scale). Nonbank mortgage lenders earn a gain-on-sale margin times volume, then keep the mortgage servicing right (MSR) — an annuity-like cash flow whose value rises when rates rise, because borrowers stop refinancing. The secondary-market core earns a guarantee fee of a few hundredths of a percent on a multi-trillion-dollar book — a leverage-and-scale model where tiny losses are survivable and a loss spike is not. Consumer lenders earn a risk-adjusted yield (portfolio APR minus charge-offs); pawn and factoring earn discount spreads and collateral recoveries [6].

The through-line. Because none of these lenders holds deposits, profitability is unusually sensitive to two external variables — the cost and availability of wholesale funding and the credit-loss rate — and both turn against owners at the same time in a downturn. The number that matters everywhere is the spread after credit losses, not the headline yield.

6. Demand drivers

The children draw on different demand pools, but interest rates touch all of them and are the master variable:

• Interest rates, cutting differently by child. For mortgages, a one-point move can trigger or kill a refinancing wave. For sales financing, rates hit both funding cost and buyer affordability, because financing is a monthly-payment decision. For cards and consumer loans, higher rates raise both the APRs charged and the issuer's own funding cost, and push near-prime borrowers down-market as banks tighten. For the secondary market, rates govern the origination flow that feeds securitization.
• Consumer spending and employment. Card volume and consumer-loan demand track how much people buy and whether they have jobs; the propensity to carry a balance drives most card profit. Record card balances keep debt-consolidation demand for personal loans firm even as credit softens.
• Vehicle, equipment, and home sales — and asset prices. No purchase, no purchase-tied loan; higher asset prices raise loan sizes even when unit volume is flat. Commercial demand comes from acquisitions and a wall of maturing loans that must refinance.
• Structural shifts. The long move of spending from cash to cards and e-commerce; the steady migration of lending from deposit-funded banks to nonbanks (which now originate most U.S. home loans and a rising share of consumer credit); and fintech's reshaping of origination.
• Government guarantees. Because so much of the level rests on federal backing (GSE, Ginnie Mae, FHA, SBA, EXIM, USDA), demand is unusually sensitive to guarantee-fee settings, program budgets, and reauthorization votes.

7. Regulation

There is no single regulator for 5222 — oversight follows the activity and the charter, and that fragmentation is itself a standing cost and a barrier to entry.

• The common federal consumer rulebook. The Consumer Financial Protection Bureau (CFPB) administers the statutes that reach across the level's consumer-facing lending: the Truth in Lending Act (TILA/Regulation Z), the Equal Credit Opportunity Act, the Fair Credit Reporting Act, the Fair Debt Collection Practices Act, and authority over unfair, deceptive, or abusive acts. Card issuing adds the CARD Act of 2009 (rate-change and fee rules); mortgage adds RESPA, TRID, and the Ability-to-Repay standard [14].
• State law binds hardest on consumer credit. Roughly 20 states plus D.C. cap small-loan APRs near 36%, and where voters decide by ballot the caps pass by wide margins; the Military Lending Act caps covered servicemembers at 36% [15]. There is no generally applicable federal APR cap for ordinary consumers.
• Prudential oversight is patchy. Card banks and bank-owned finance arms carry the full banking-regulator overlay (Federal Reserve, OCC, FDIC). But no federal safety-and-soundness regulator oversees nonbank mortgage firms — a structural gap that the Financial Stability Oversight Council (FSOC) and GAO warned in 2024–25 leaves thinly capitalized nonbank servicers as a financial-stability risk; state licensing through the Nationwide Multistate Licensing System is the front line [13].
• The secondary market and specialty tail. Fannie and Freddie are run by the Federal Housing Finance Agency (FHFA) as conservator since 2008, which sets their capital and guarantee-fee framework; Farmer Mac and the Farm Credit System answer to the Farm Credit Administration; the Export-Import Bank operates under a congressional charter that must be periodically reauthorized (currently through end-2026).
• Live political variables. The federal posture swung deregulatory in 2025 (CFPB leadership change, deprioritized enforcement), while the durable constraint stayed at the hard-to-reverse state level [14][15]. For card issuers, the proposed Credit Card Competition Act would force large issuers ($100B+ assets) to enable a second routing network, a threat to interchange revenue [16]. And overshadowing the whole level is the future of GSE conservatorship (Section 10).

8. Consolidation

Consolidation is happening across all three children, but for different reasons, and 2025 was a landmark year:

• Cards are vertically integrating. The defining move is owning the payment network as well as the lending: American Express has always run a closed loop, and Capital One's ~$35 billion purchase of Discover closed in May 2025, buying the same structure and making it the largest U.S. issuer. The child is already top-heavy (CR4 63.1%), so further large moves are possible but hard to execute [4][7].
• Mortgage is consolidating fastest. Rocket acquired Mr. Cooper for $14.2 billion (closed October 2025), the largest independent-mortgage deal ever, joining the biggest originator to the biggest servicer; Rocket also bought Redfin, and Bayview took Guild Mortgage private. Scale, technology, and servicing economics increasingly favor the largest platforms, squeezing thinly capitalized small independents [6][8].
• Sales financing rotates and rolls up in stress. Captives, banks, credit unions, and independents compete for the same borrower and trade share cyclically; consolidation clusters in stressed credit cycles when smaller lenders lose funding access (recent examples: Synchrony buying Ally's point-of-sale business; Santander taking its U.S. consumer arm private) [5].
• The "other" bucket is a duopoly at the core, consolidating at the edges. There is little classic competition in the regulated center (Fannie/Freddie, Ginnie Mae, Farmer Mac, EXIM), but the tail — pawn, factoring, private credit — does roll up. The most consequential story is not a merger at all: the possible partial privatization of the GSEs, which would redraw ownership of the level's single largest revenue pool [6].

Across the level, the durable advantages are the same: cheap and reliable funding, underwriting-data quality, servicing efficiency, compliance infrastructure, and the charters and licenses that are hard to replicate.

9. Risks

The children share a common risk spine — no deposits, wholesale funding, credit losses, and heavy government or regulatory dependence — with different pressure points:

• Credit cycle (dominant). Rising unemployment and falling income lift delinquencies and charge-offs across all three; card and subprime-consumer books and office-exposed commercial loans are hit hardest, and losses generally lag the loans that caused them. The Federal Reserve's 2026 Financial Stability Report noted consumer delinquencies eased late in 2025 but stayed elevated versus the prior decade [17].
• Funding and liquidity. With no deposits and no lender-of-last-resort, these firms depend on ABS, warehouse lines, repo, and MBS markets. When those seize (2008, briefly 2020), lenders cannot fund or offload new loans; mortgage servicers face margin calls and must advance payments on delinquent loans exactly when cash is scarce — the FSOC-flagged vulnerability [13].
• Interest-rate and prepayment risk. Higher rates raise funding cost faster than lenders can reprice and compress margins; lower rates spark prepayment and price competition and cut servicing values. The direction that helps one child hurts another.
• Concentration and systemic importance. Because two firms stand behind trillions in mortgages, a GSE stumble is a national housing and financial-stability event; and the level's card and secondary-market pieces each concentrate risk in a handful of firms [6].
• Policy and regulatory risk. State rate caps and ballot measures, a possible GSE conservatorship exit and EXIM reauthorization, interchange politics (the Credit Card Competition Act), guarantee-fee changes, and fair-lending and servicing enforcement all sit outside management control [14][15][16].
• Model, operational, cyber, and counterparty risk. Credit and fraud models can fail outside their historical data; lenders hold sensitive borrower data; platforms can depend on a few funding partners or loan buyers.
• Private-market opacity. Private lenders, funds, securitization vehicles, and MSR portfolios report infrequently and are hard to value; the Federal Reserve has flagged rising bank interconnection to private-credit funds and BDCs as a growing watch-item [17].

10. How to invest & outlook

The organizing principle for the whole level: buy the child — the engine — not the code. 5222 is not one industry but three (really four) exposures that behave differently across the same cycle.

Public routes, by engine:

1. Card credit (52221). Pure-play issuers AXP, COF, SYF, BFH give the most direct exposure; diversified banks (JPM, C, BAC, USB) give diluted exposure where you must analyze the card segment separately from the parent. Judge them on NIM, net charge-offs, and reserves, not headline yield [4].
2. Purchase-tied credit (52222). Own the captive as a segment through Ford, GM, Toyota, Deere, Caterpillar, PACCAR, or CNH; or the near-proxies Ally (ALLY), Credit Acceptance (CACC), Synchrony (SYF). Value lenders on price-to-book, return on tangible equity, and charge-off trends [5].
3. Mortgage and consumer cyclicality (52229). Originator-servicers (RKT, UWMC, PFSI, RITM, LDI) are leveraged plays on origination volume and MSR values; consumer names (OMF, ENVA, RM, WRLD) are credit stocks — analyze loan yield, funding mix, charge-offs, and leverage, not revenue growth alone [6].
4. Government-backed carry and the GSE special situation (52229). The mainstream, lowest-risk route is agency MBS inside bond funds — most investors already own this segment there. Farmer Mac (AGM) offers normal secondary-market economics; the GSE commons (FNMA, FMCC) are speculative OTC bets on the terms of an eventual conservatorship exit — a binary policy event, not a yield investment [6][9][10].

Private routes are dominated by debt, not equity — card, auto, equipment, and consumer-loan ABS; agency and Ginnie Mae MBS; whole loans and MSRs — plus private-equity ownership of branch, captive, and mortgage platforms. Diligence is loan-level: performance by vintage and borrower segment; advance rates, covenants, and triggers; servicing oversight; and the legal status of any bank-partnership or securitization structure.

Near-term swing factors (judgment, not a forecast):

1. The rate path — the single biggest variable, reviving mortgage origination and easing consumer funding costs if it falls, while the secondary-market pipeline follows origination flow.
2. Consumer charge-off and delinquency direction — the 2025 softening off the 2024 peak is the master variable for the card and consumer engines [11][17].
3. The GSE privatization/IPO timeline — any concrete step would reprice Fannie and Freddie dramatically and ripple through every mortgage originator's economics [6].
4. The regulatory split — a lender-friendly federal CFPB against the steady, hard-to-reverse spread of state 36% rate caps, plus the interchange fight over the Credit Card Competition Act [15][16].
5. Consolidation — Capital One–Discover and Rocket–Mr. Cooper set scale benchmarks, with continued pressure on smaller lenders across all three children [7][8].

Bottom line. NAICS 5222 is the deposit-free edge of American lending: credit extended without customer deposits, funded in wholesale markets, and mostly sold rather than held. That shared structure makes the whole level a spread-and-fee business whose fortunes turn on funding access and credit losses — and makes it structurally a share-taker from deposit-funded banks over time. But its three children could hardly diverge more in size, concentration, ownership, and cycle: a concentrated card-credit business, a captive-dominated purchase-finance business, and a sprawling "everything else" bucket anchored by a government-controlled duopoly. Size expectations to the specific engine — and to the credit cycle, not to a straight line.

---

Sources

1. U.S. Census Bureau, 2022 NAICS Definitions — Subsector 522 and Industry Group 5222 (Nondepository Credit Intermediation), industries 52221 / 52222 / 52229 (scope, the deposit-free structure, and the 5221/5223 boundaries), 2022. https://www.census.gov/naics/?input=5222&year=2022
2. U.S. Census Bureau, County Business Patterns 2023 — NAICS 5222 and children (establishments, employment, annual and Q1 payroll). https://www.census.gov/programs-surveys/cbp.html
3. U.S. Census Bureau, 2022 Economic Census — Concentration of Largest Firms — NAICS 5222 (firms, receipts, CR4/CR8/CR20/CR50, HHI). https://data.census.gov/table/ECNSIZE2022.EC2200SIZECONCEN
4. Histometrics primer — NAICS 52221, Credit Card Issuing (child figures and company detail; 2022 Economic Census concentration CR4 63.1%, HHI 1,245.6).
5. Histometrics primer — NAICS 52222, Sales Financing (child figures, captive economics; 2022 Economic Census concentration CR4 29.3%, HHI 378.9).
6. Histometrics primer — NAICS 52229, Other Nondepository Credit Intermediation (child figures, GSE and mortgage detail; 2022 Economic Census concentration CR4 53.2%, secondary-market subsegment CR4 80.2%).
7. Capital One Financial Corp., Form 8-K, Completion of Acquisition of Discover (closed May 18, 2025; ~$35B; closed-loop network), 2025. https://www.sec.gov/Archives/edgar/data/927628/000119312525122059/d934475d8k.htm
8. Rocket Companies, Rocket Closes $14.2 Billion Acquisition of Mr. Cooper (largest independent-mortgage deal; also Redfin), 2025. https://www.rocketcompanies.com/press-release/rocket-companies-closes-14-2-billion-acquisition-of-mr-cooper/
9. Federal Housing Finance Agency, Conservatorship of Fannie Mae and Freddie Mac; Ginnie Mae, EXIM, and Farm Credit System program data (government "plumbing" excluded from the Economic Census), 2025–2026. https://www.fhfa.gov/conservatorship
10. Affordable Housing Finance / FHFA, Conservatorship Crossroads: What's Next for Fannie Mae and Freddie Mac? (GSEs backing ~$7.7T; agency-MBS role), 2025. https://www.housingfinance.com/finance/conservatorship-crossroads-whats-next-for-fannie-mae-and-freddie-mac
11. Federal Reserve Bank of New York, Household Debt and Credit (Q4 2025 — mortgage ~$21T, auto ~$1.67T, card ~$1.28T); TransUnion, Q4 2025 Credit Industry Insights (unsecured personal-loan balances ~$276B), 2026. https://www.newyorkfed.org/microeconomics/hhdc
12. Board of Governors of the Federal Reserve System, FEDS Notes, Credit Card Profitability (lending ≈80% of profitability, fees ≈15%), 2022. https://www.federalreserve.gov/econres/notes/feds-notes/credit-card-profitability-20220909.html
13. U.S. Financial Stability Oversight Council, Report on Nonbank Mortgage Servicing (2024); U.S. Government Accountability Office, Nonbank Mortgage Companies (GAO-25-107862, 2025); Conference of State Bank Supervisors, NMLS/SAFE Act state licensing. https://home.treasury.gov/system/files/261/FSOC-2024-Nonbank-Mortgage-Servicing-Report.pdf
14. Consumer Financial Protection Bureau, Regulations and Nonbank Supervision (TILA/Reg Z, ECOA, FCRA, FDCPA, UDAAP; CARD Act; RESPA/TRID/Ability-to-Repay), 2026. https://www.consumerfinance.gov/rules-policy/regulations/
15. Center for Responsible Lending, 36% Rate Caps (state small-loan caps, Military Lending Act, ballot measures), 2024. https://www.responsiblelending.org/research-publication/36-cap-annual-interest-rate-stops-payday-lending-debt-cycle
16. U.S. Senate (Durbin/Marshall), Credit Card Competition Act (issuers with $100B+ assets; reintroduced 2025; not enacted), 2025–26. https://www.durbin.senate.gov/newsroom/press-releases/durbin-marshall-introduce-bipartisan-credit-card-competition-act
17. Board of Governors of the Federal Reserve System, Financial Stability Report, May 2026 (consumer delinquencies eased late 2025 but remained elevated; bank interconnection to private credit/BDCs). https://www.federalreserve.gov/publications/files/financial-stability-report-20260508.pdf
18. U.S. Small Business Administration, Table of Small Business Size Standards (NAICS 5222 industries, $47M average annual receipts), 2023. https://www.sba.gov/document/support-table-size-standards