Datasets:

CyberMax-tools
/

curvewire-us-macro-panel

license: cc0-1.0
pretty_name: 'Curvewire: US Treasury yield curve + BLS macro panel (official sources)'
language:
  - en
tags:
  - finance
  - economics
  - macroeconomics
  - interest-rates
  - yield-curve
  - treasury
  - inflation
  - cpi
  - jobs
  - unemployment
  - bls
  - time-series
  - tabular
  - cybermax
task_categories:
  - time-series-forecasting
  - tabular-regression
size_categories:
  - 10K<n<100K
configs:
  - config_name: treasury_yields
    default: true
    data_files:
      - split: train
        path: data/treasury_yields.parquet
  - config_name: treasury_real_yields
    data_files:
      - split: train
        path: data/treasury_real_yields.parquet
  - config_name: bls_monthly
    data_files:
      - split: train
        path: data/bls_monthly.parquet

Curvewire: US Treasury yield curve + BLS macro panel

Every Monday: the rates and macro brief built from this data (PDF + CSVs), $19/month, cancel any time. Buy now → · Delivered instantly after checkout · no subscription · 14-day refund if it doesn't work as described · help: cybermax.tools@gmail.com

Need it fresh, filtered or via API? This free file is a snapshot (US macro series up to the last refresh), last updated 2026-09-24.

US Rates & Macro Brief: Weekly Edition (yearly) ($190/year): the same Monday rates & macro brief for a year, two months free.

All Three Monday Briefs: Rates, Recalls + Exploited Vulns ($25/month): all three weekly briefs (Treasury/BLS rates & macro, NHTSA recalls, CISA KEV/EPSS) for $25/month instead of $37.

Using it at work? Commercial license + support (from $49/year): invoice, PDF licence certificate, named support, freshness and availability commitments and an SLA; the data stays free and open for everyone.

Get an email when this dataset updates: free, double opt-in, unsubscribe any time.

CyberMax Store: every CyberMax data product, API plan and weekly brief in one place.

Information only, not investment advice.

The US numbers macro models, trading bots and LLM agents ask for most, taken straight from the agencies that publish them (US Treasury and the Bureau of Labor Statistics), not scraped from redistributors, and cleaned into three tidy tables:

config	rows	what
`treasury_yields` (default)	9,189	Daily par yield curve, 1 month to 30 years, 1990-01-02 → 2026-09-23, with 10y–2y and 10y–3m spreads
`treasury_real_yields`	5,936	Daily real (TIPS) par yield curve, 5 to 30 years, 2003-01-02 → 2026-09-23
`bls_monthly`	7,514	15 headline monthly series in long format, with 1-month and 12-month changes

Latest readings in this build (see meta.json): 10-year yield 5.11%, 2-year 4.85% (10y–2y spread +0.26 pt), 3-month 4.19%; August 2026 CPI +3.35% year on year (core +2.45%), unemployment rate 4.1%, payrolls +162k on the month; JOLTS job openings 7.27 million (July).

BLS series included

series_id	series	units
CUSR0000SA0 / CUUR0000SA0	CPI-U, all items (seasonally adjusted / not)	index 1982-84=100
CUSR0000SA0L1E	Core CPI (less food and energy)	index
CUSR0000SAF1, CUSR0000SA0E, CUSR0000SAH1	CPI food, energy, shelter	index
CES0000000001	Total nonfarm payrolls	thousands
CES0500000003, CES0500000002	Average hourly earnings, average weekly hours (private)	USD, hours
LNS14000000, LNS11300000, LNS12300000	Unemployment rate, participation rate, employment-population ratio	percent
JTS000000000000000JOL, JTS000000000000000QUR	JOLTS job openings, quits rate	thousands, percent
WPSFD4	PPI final demand	index

bls_monthly columns: series_id, series_name, units, seasonally_adjusted, date (first day of the month), value, footnote_codes (BLS codes, e.g. P = preliminary), change_1m, pct_change_1m, pct_change_12m (null for series already in percent), change_12m.

treasury_yields columns: date, y_1m, y_6w (1.5-month bill, from 2025), y_2m, y_3m, y_4m, y_6m, y_1y, y_2y, y_3y, y_5y, y_7y, y_10y, y_20y, y_30y (percent), spread_10y_2y, spread_10y_3m (percentage points). Maturities Treasury didn't publish on a date (e.g. 20-year before 1993, 30-year in 2002–2006) are null.

Use it

from datasets import load_dataset
curve = load_dataset("CyberMax-tools/curvewire-us-macro-panel", "treasury_yields", split="train").to_pandas()
print(curve[curve.spread_10y_2y < 0].date.agg(["min", "max", "count"]))   # inversion days

bls = load_dataset("CyberMax-tools/curvewire-us-macro-panel", "bls_monthly", split="train").to_pandas()
cpi = bls[bls.series_id == "CUSR0000SA0L1E"].set_index("date")["pct_change_12m"]
print(cpi.tail(12))   # core CPI inflation, last 12 months

Refresh and revisions

Rebuilt from the agencies' own files (Treasury's daily CSVs, BLS bulk time-series files); meta.json has the build time and each series' latest month. BLS revises recent months (payrolls for two months, CPI seasonal factors each February), and every refresh picks the revisions up, so always use the latest build rather than stitching old ones together.

Use it in production

This snapshot is free. Prefer the weekly picture to raw series? The US Rates & Macro Brief arrives every Monday as a PDF plus CSVs: every tenor's move, curve spreads and TIPS breakevens.

See the US Rates & Macro Brief (weekly)

Licence: Free for any use, including commercial (CC0; Treasury and BLS data are public domain). Using it at work? The Commercial License + Support adds an invoice, a licence certificate, named support and an SLA.

Information only, not investment advice.

Licence

US Treasury and BLS data are works of the US federal government, in the public domain; this compilation is released under CC0 1.0. Please cite the agencies as the source. Compiled by CyberMax. Not affiliated with or endorsed by the US Treasury or BLS. Not investment advice. Series from redistributors with third-party copyright (e.g. some FRED series) are deliberately not included.

Also from CyberMax

Free datasets: Eventwren: SEC 8-K events by Item · Insidewell: SEC Form 4 insider transactions
Tools for AI agents and data teams (pay per use, on Apify), e.g. Insidewell: SEC insider trading tracker and Swellmeter: Google Trends API: https://apify.com/cybermax

More free from CyberMax

Quickstart notebook

Load this dataset with pandas and see real outputs in the quickstart notebook.