Federal Reserve Finance Companies G.20

The Board of Governors' monthly read on the lenders banks are not. Federal Reserve Finance Companies G.20 tracks receivables outstanding at US finance companies - the non-depository specialty lenders behind auto sales financing, equipment leasing, factoring and mortgage finance - split three ways: consumer (motor vehicle loans, motor vehicle leases, revolving, other), real estate (one-to-four family, other) and business (retail and wholesale motor vehicle loans, motor vehicle leases, equipment loans and leases, other). Each sector reads out in seasonally adjusted and not-seasonally-adjusted form, at level, in flow, and as annual percent change.

Two things lift it above an ordinary credit series. First, the terms-of-credit table: interest rates, maturity in months and amount financed on new and used car loans - pricing, tenor and check size in one place, which is as close as public statistics get to a rate card for specialty lending. Second, the asset dimension separates owned, managed and securitized receivables, so the balance-sheet lending business and the servicing business never blur into one number.

Datadory delivers the whole release as typed rows: 156 series, monthly observations back to January 1943, definitions verified against the live record. Within our catalog of 1,744 datasets across 159 viable industries this record scores 10/10 - one of only a handful that reach the ceiling. Get a sample of this dataset and see the shape before you commit.

Captured during the August 2026 research pass, straight from the seasonally adjusted levels table:

date=Jun 1985   total_millions=270,964.88   consumer=106,488.66
                real_estate=25,876.00       business=138,600.22

date=May 2026   total_millions=1,969,225.72 consumer=924,046.27
                real_estate=316,956.97      business=728,222.47

Read them together and the industry's biography writes itself. Between June 1985 and May 2026 total receivables grew 7.3x - from $271 billion to $1.97 trillion - while the mix rotated underneath: consumer's share climbed from 39.3% to 46.9%, real estate's from 9.5% to 16.1%, and business fell from 51.2% to 37.0%. Specialty lending started the period mostly a commercial trade and ended it mostly a household one.

The latest month carries its own story. From April to May 2026 total receivables rose $11.4 billion, and 95% of that increase came from business lending (+$10.8 billion) rather than consumer (+$0.6 billion) - the kind of rotation a credit-cycle watcher wants weeks before it reaches an earnings call. Every row lands pre-split by sector, so the rotation is arithmetic, not re-aggregation.

Twelve fields carry the full machine-readable bundle, mapped below with types, definitions and examples verified against the live record during the August 2026 research pass. Six of them - ASSET, ITEM, SERIES, DATAREP, SA, UNIT - do the dimensional heavy lifting: every observation resolves to a sector, an asset class, a receivable category and a seasonal treatment through enum columns rather than string parsing, which is why two analysts pulling the same series always land on the same row.

Geography - United States, national aggregates. Domestic finance companies report to the Federal Reserve and the release publishes country-level totals; there are no state or regional splits, because the panel is designed to read the national specialty-lending book, not a county map.

Temporal - the deepest run in the specialized-finance slice: monthly observations beginning January 1943 in the machine-readable bundle, roughly a thousand observations per series across 156 series. Each current release presents the latest five years plus year-to-date months, and displayed historical tables reach to June 1985.

Granularity - monthly national totals split four ways at once: by asset dimension (owned / managed / securitized / all), by sector (consumer / real estate / business), by receivable category within each sector, and by seasonal adjustment. Quarterly depth covers the historic balance sheet and the terms-of-credit tables.

Set against the wider Datadory catalog - 1,744 datasets averaging a 7.81 quality score - the specialized-finance slice holds 4 cataloged datasets and this record scores 10/10: fully verified field dictionary, eighty-plus-year temporal coverage, and a published methodology behind every figure. Where it ranks: best specialized finance datasets.

API, files, or your warehouse. Daily, weekly, or hourly.

You pick the channel and set the cadence to match the decision the receivables feed. Credit-cycle dashboards want the fast lane; portfolio models want warehouse loads sized for batch; BI tools want tabular extracts that land SQL-friendly. Cadence changes are a settings conversation, not a re-integration project.

Every delivery ships with the field dictionary above unchanged and sample rows for validation, flattened to one observation per series per period - so a consumer-versus-business share computation stays arithmetic even when the seasonal-adjustment flags differ. Name the sectors and the cadence when you request the sample; get a sample of this dataset first and let the ongoing arrangement follow it.

A national receivables ledger earns its keep in five specific jobs:

• Investors and quant researchers - the credit-cycle input for nonbank lending models. Total receivables momentum against the consumer/business split turns "credit is tightening" into a measurable series, and the 1943 start date gives a backtest more than eight decades of runway.
• Market researchers and consultants - the citable base for sizing the specialty-lending market. Sector shares, growth rates and terms of credit measured one way by one agency make the chapter-one exhibit defensible without a footnote argument.
• Lender strategy and competitive intelligence teams - the national benchmark for your own book. Comparing your auto lease growth against the national motor vehicle lease line tells you whether you are gaining share or just riding the market.
• Economists, academics and students - the long nonbank credit series textbooks skip. Owned versus managed versus securitized splits support real research questions about off-balance-sheet lending, not just aggregate totals.
• Risk and treasury functions - the terms-of-credit table doubles as a market-rate reference for new and used auto paper: rates, maturity in months, amount financed, published on one schedule with no vendor derivation in between.

Investors & Quant Researchers get the tightest fit: a monthly, eight-decade credit aggregate with the consumer/business split already cut. Market Researchers & Consultants get the definitional source for the specialty-lending market they keep being asked to size. Journalists, Academics & Students get figures that trace to a named Federal Reserve release and survive citation review. Data Scientists & ML Engineers get a fully enumerated enum-based schema that drops into a pipeline without entity resolution. Persona-by-persona detail: market researchers, data scientists.

Inside this slice - four cataloged datasets - the records answer different questions, so pair by question rather than rank. This G.20 record answers how much credit do US finance companies carry, on what terms at 10/10. World Bank Financial Sector Indicators (also 10/10) answers the cross-country version for 265 economies, but stops at the bank-centric indicators and carries no lender-type split. SBA Open Data Portal (9/10) answers which small businesses got government-backed loans, loan by loan - program-level granularity this release deliberately lacks. Hugging Face Finance Topic (7/10) is not a statistics set at all but 1,352 community ML corpora, useful for model training rather than measurement.

The G.20's edge is the nonbank lens plus the depth: no neighbor in the slice reaches 1943, and none isolates finance companies from depository institutions. For the sibling Fed releases, G.19 Consumer Credit reads household borrowing economy-wide while this release reads the finance-company share of it; the direct comparison piece weighs this record against the SBA catalog.

Four caveats, stated up front.

First, geography is national only. The release publishes United States aggregates; there is no state, metro or regional cut anywhere in the 156 series. Regional specialty-lending questions need a different source.

Second, the displayed historical tables begin in June 1985 even though the machine-readable series reach January 1943 - if you want the deep decades, say so when scoping the sample so the feed is cut from the right layer.

Third, preliminary values revise. The most recent month ships flagged preliminary and gets revised in later releases, so any process keyed to the latest observation should expect restatements rather than treat them as errors.

Fourth, scope the sample by dimension, not by table name. Tell us which sectors, which asset classes (owned, managed, securitized), whether terms of credit matter to you, and the cadence you need - daily, weekly, or hourly - and the sample comes back shaped to your book before any recurring arrangement exists.

Provenance note - published by the Board of Governors of the Federal Reserve System, compiled from the benchmark Finance Company Survey (FR 3033s: a stratified random sample of roughly 2,400 firms, about 1,000 responding) and the monthly Domestic Finance Company Report of Consolidated Assets and Liabilities (FR 2248: a panel of about 50 companies). Percent change is computed from unrounded data as flow divided by prior-period level.

Methodology note - units are billions of dollars except as noted, and the machine-readable layer carries UNIT and UNIT_MULT columns so a value never arrives ambiguous about scale. OBS_STATUS codes mark normal observations separately from not-available, no-data and not-calculable cells, which keeps gaps honest in downstream joins.

Completeness note - Datadory scores this record 10/10: the field dictionary maps fully to the published series vocabulary, coverage runs to 1943, and every figure traces to a named release. Nothing in the verified surface is approximated.

Where to go next - the specialized finance data hub holds the rest of the slice, the best specialized finance datasets ranking shows where this record sits among its neighbors, and the glossary entries for finance company receivables and specialty lending define the terms this release measures. For the publishing institution, see the Board of Governors source profile; for household borrowing economy-wide, see G.19 Consumer Credit; and the comparison piece weighs this record directly against the SBA open data catalog.