How the Nowflation gauge is built — every component, every input, every variant, exposed

The Nowflation Gauge reads 2.89% year over year as of 2026-09-02, against official CPI of 3.4% for Jul 2026 (NSA, quoted as BLS quotes it) — a gap of -0.51pp. It is a daily number, but only 39.1% of its basket weight is market-priced today: 3 of 14 components carry a live source (Shelter: Rent, Motor Fuel, Used Vehicles), and the owners'-equivalent-rent weight rides the all-tenant rent index (new-lease rents, tenure-weighted, used as measured). The remaining components hold the Jul 2026 CPI print, released 2026-08-12, until the next release on 2026-09-11. Engine method v1.3; weights from the 2026 BLS relative-importance table.

The variants today

VariantYoYLevel (Jan 2018 = 100)Live coverageShelter treatmentWeights
Nowflation Gauge (CPI-comparable)2.89%136.8539.1%all-tenant rent index (v1.3): new-lease rents (Zillow ZORI 0.8 / Apartment List 0.2, chain-linked) averaged over the tenure of every lease in force — what the renting population pays this month, measured, not forecast — carrying both the rent weight and the owners’-equivalent weightBLS CPI relative-importance table
Gauge (SA fuel, pace basis) (seasonally adjusted fuel)2.87%136.7639.1%same all-tenant rent model as the Gauge; the only difference is motor fuel, seasonally adjusted with the BLS gasoline factors — the basis for the annualized 1- and 3-month paceBLS CPI relative-importance table
Cost of Living (mortgage-payment shelter)3.68%157.5839.1%all-tenant rent model for renters; a principal-and-interest payment index (80% of the Zillow home value at today’s 30-year rate) for ownersBLS CPI relative-importance table
CPI-Tracker (official shelter dynamics)3.34%135.535.1%official BLS rent and owners’-equivalent subindexes only — no market rents — so this variant re-tracks the print instead of leading itBLS CPI relative-importance table
Supercore (services ex-shelter)2.30%128.560.0%excluded, along with every goods, food-at-home and energy componentBLS weights renormalized over food away, medical care, recreation, education & communication and the core residual
PCE-weighted (out-of-pocket scope)2.91%132.8219.2%same all-tenant rent model as the GaugeBEA PCE relative-importance shares (hand-seeded from Table 2.3.5)

Every component on 2026-09-02

ComponentWeightLevelYoYMoMBLS subindex YoYDiffFeedSource codesData throughAge (days)Contribution (bps)
Shelter: Owned (in the Gauge this weight rides the rent blend, 1.68% YoY)26.5%143.433.23%+0.26%3.23%0OFFICIALCUSR0000SEHC2026-07-016344.5
Everything Else18.5%132.692.47%+0.21%2.47%0OFFICIALCPILFESL2026-07-016345.6
Food at Home8.2%134.432.61%-0.07%2.61%0OFFICIALCUSR0000SAF112026-07-016321.4
Medical Care8.1%124.41.70%+0.36%1.70%0OFFICIALCPIMEDSL2026-07-016313.8
Shelter: Rent7.5%147.511.68%+0.16%2.86%-1.19ppLIVEzori_us + aptlist_rent_us (stale)2026-07-313312.6
Food Away5.7%146.363.40%+0.31%3.40%0OFFICIALCUSR0000SEFV2026-07-016319.4
Education & Comm5.5%108.240.55%+0.58%0.55%0OFFICIALCPIEDUSL2026-07-01633
Recreation5.3%122.42.58%+0.23%2.58%0OFFICIALCPIRECSL2026-07-016313.7
New Vehicles3.6%122.560.54%+0.08%0.54%0OFFICIALCUSR0000SETA012026-07-01632
Motor Fuel3.0%166.3629.45%+0.59%29.36%+0.09ppLIVEaaa_gas_national2026-09-02088.3
Electricity2.8%143.854.20%+0.13%4.20%0OFFICIALCUSR0000SEHF012026-07-016311.8
Apparel2.5%109.133.87%+0.09%3.87%0OFFICIALCPIAPPSL2026-07-01639.7
Used Vehicles2.1%143.151.25%-1.36%-1.87%+3.13ppLIVEmanheim_muvvi2026-07-01632.6
Utility Gas0.7%157.084.32%+0.68%4.32%0OFFICIALCUSR0000SEHF022026-07-01633

What fed the gauge today

How the number is constructed

  1. Collect — every input is stored with a vintage date the day it is pulled: 14 BLS subindexes (via FRED) as the carry-forward history plus the live inputs (Zillow ZORI 0.8 / Apartment List 0.2 new-lease rents for the shelter model; AAA daily pump price for fuel; Manheim wholesale index for used vehicles, shifted 30 days for the wholesale-to-retail lag). Nothing is back-filled from memory.
  2. Date by publication — a monthly print enters the daily history on the day it became knowable, never on its reference month: CPI subindexes step on the BLS release calendar, other monthly sources on their observed publication lag (ZORI about 16 days after month-end, Apartment List about 45 days, Manheim within the first week). The July CPI exists in the history from its August release date, so a year-ago comparison only sees what was public a year ago.
  3. Model shelter (v1.3) — new-lease rents measure the marginal tenant, whoever signs this month; the CPI measures every tenant, most of whom are mid-lease. The rent sources are chain-linked into one new-lease index, then averaged over the trailing 36 months with weights that decay at the monthly lease turnover (6% a month, a mean tenure of about 17 months) — the all-tenant rent, used as measured (v1.3: no forecasting map; the Gauge is a measurement). On the record it leads the official rent index by about 6 months (r 0.976); the calibrated “where official rent lands” map is shown on the Lead Test only. Applied raw, new-lease rents measure the marginal tenant and never converged to the print.
  4. Rebase and splice — every series is indexed to January 2018 = 100. A live overlay is scaled to equal the official subindex on the day it starts (or re-starts), so a splice can never jump the level. Blends are chain-linked: each day’s move is the weighted average of the day-over-day moves of the sources that are fresh on both days, so a source dropping out or returning changes the composition without moving the level.
  5. Stale fallback — a live source stops counting once its newest data is older than its staleness window (14 days for daily feeds, 60 for monthly). The component then holds its last live level and moves with its official subindex from there: a dead feed can neither freeze the component nor snap it back to the official level. Coverage falls, the number keeps publishing.
  6. Official between prints (v1.3) — every component without a better daily source rides its BLS subindex at its last measured level until the next print lands (no extrapolation: the Gauge measures, the CPI-Tracker and the print-day model forecast). Electricity and utility gas are deliberately official (EIA’s residential price series publish roughly 12–13 weeks after the month, later than the CPI subindex they would stand in for, and spot Henry Hub is not a regulated retail tariff); food at home waits for a USDA composite with 13 months of history.
  7. Aggregate — headline = Σ weight × component index (Laspeyres-style) with the BLS relative-importance weights for the year; in every variant but Cost of Living the all-tenant rent index carries both the rent and the owners’-equivalent weight. Coverage = the share of basket weight fed by a live source that day; a day is only published when at least 95% of the weight has a value. Any live-source move over 5% in a day is held one day at the prior value.
  8. Compare — YoY and MoM are same-calendar-day comparisons (September 1 against September 1, clamped to the shorter month), not 365- or 30-day offsets. The official figure quoted next to the Gauge is the BLS YoY on the NSA index at one decimal, exactly as published; the Gauge itself is built on the seasonally adjusted subindexes.
  9. Pace (v1.3) — the annualized 1- and 3-month pace is computed on the cpi_comparable_sa variant: the Gauge with motor fuel seasonally adjusted by the BLS gasoline factors, so a summer pump-price swing does not read as a change in trend. YoY is half last year’s prices; the pace is what today’s prices are doing.

Engine v1.3: daily component indexes (base 2018-01 = 100) built from publication-dated BLS subindexes (held at their last measured print between releases) with live overlays spliced on — an all-tenant rent index for shelter, daily pump prices, wholesale used-vehicle values — aggregated with BLS relative-importance weights into the variants above; YoY on the same calendar day a year earlier, pace annualized on the seasonally adjusted variant. Coverage counts the basket weight fed by a live source that day. Read-only over the engine’s headline_index, component_index and the freshness record written at each run. Full methodology and changelog: methodology. Feed health: system status. Why the gauge and the print differ: why the gap.

Explore: dashboard · methodology · status · heatcheck · my inflation · vs bls · jobs · housing · farm to shelf · data · scoreboard · gap · gdp · oil · oil vs cpi · gauge · pipeline · natgas · reserves · stockpiles · about · privacy · contact · analysis · expectations · activity · labor · conditions · income · trade · surveys · today · forecast · compare · embeds · money · fiscal · stress · rates · affordability · matrix · indicators · mortgage · calculator · cart · states · counters · recession · commodities · energy · countries · macro · calendar · liquidity · ratios · housing market · revisions · century · nowflation difference · nowflation terms · nowflation how to · supercore · credit · growth · real wages · jobs preview · fx · metros · prediction markets · heating season · thanksgiving · fed watch · nowflation vs truflation · lead test · eggs · gas · coffee · ground beef · electricity · milk · bread · bacon · chicken · chocolate · butter · cheese · flour · rice · potatoes · tomatoes · bananas · oranges · cola · i bonds · tax brackets · all 1,629 series