Eco3min US Inflation Regime Score (2004–current)

The eco3min Inflation Regime Score (US) is a monthly composite indicator combining headline CPI, core CPI, and 5Y5Y forward inflation expectations into a single weighted reading, mapped to five inflation regimes calibrated on the eco3min framework. Built from three public FRED series and updated monthly after each BLS CPI release, this dataset provides a one-number diagnostic of where the US inflation system stands at any given moment — Deflation, Disinflation, Target, Elevated, or Very High.

Dataset: eco3min Inflation Regime Score (US) · Updated 2026-06-01

Source: FRED series CPIAUCSL, CPILFESL (BLS), T5YIFR (St. Louis Fed) · Eco3min composite calculation · Score = 0.50 × CPI YoY + 0.30 × Core CPI YoY + 0.20 × 5Y5Y BE

Source: FRED series CPIAUCSL, CPILFESL, T5YIFR · Eco3min composite calculation · Score = 0.50 × CPI YoY + 0.30 × Core CPI YoY + 0.20 × 5Y5Y BE

Macro Takeaway

Score Diagnosis — May 2026

The eco3min Score reads 3.38 in May 2026 — its second consecutive month in the Elevated zone after crossing the 3.0 threshold in April (3.15). These are the first Elevated readings since May 2024, ending a run of 22 consecutive months in the Target zone.

The decomposition shows a headline-driven move rather than a broad inflation re-acceleration:

Contributions rounded to two decimals; the published score is computed on unrounded components (0.50 × 4.17 + 0.30 × 2.82 + 0.20 × 2.27 = 3.385 → 3.38).

Headline CPI has climbed from 2.4% in February to 4.2% in May, as the energy shock linked to Iran and the Strait of Hormuz passed through to consumer prices. Core CPI has risen far less over the same window — 2.6% to 2.8% — and remains below 3%. The 5Y5Y breakeven has moved from 2.12% (March) to 2.27% (May): a drift, not a de-anchoring. With core at 2.82% and expectations at 2.27%, the score holds above 3.0 only while headline CPI stays near or above 4% — as of the May data, this is an energy story, not a diffusion story.

This gap between headline and underlying measures is precisely what separates the score from the eco3min macro regime classification: the score weights headline CPI at 50% and reads Elevated; the regime classification’s inflation axis uses the Dallas Fed Trimmed Mean PCE — which strips out energy by construction — and still reads 2.41% (May 2026), inside its target band. The two indicators answer different questions: what households pay now, versus what persists.

Dataset Overview

Dataset Variables

The CSV and Excel files contain the following columns. Each row represents one month.

Column names match the CSV headers exactly.

Download the Complete Dataset

The full dataset spans 269 monthly readings — every regime transition since January 2004.

Direct CSV Access

The complete dataset is available at a stable URL for direct integration into Python, R, or BI tools:

https://eco3min.fr/dataset/score-eco3min.csv

No signup or API key required. Updated automatically the day after each BLS CPI release (typically the 12th–14th of the month).

Using the Dataset in Python

import pandas as pd

url = "https://eco3min.fr/dataset/score-eco3min.csv"
df = pd.read_csv(url, parse_dates=["date"], index_col="date")

# Latest reading
latest = df.iloc[-1]
print(f"Score: {latest['score']:.2f}  ({latest['regime_id']})")
print(f"Date: {latest.name.strftime('%B %Y')}")

# Months spent in Elevated regime since 2004
elevated = df[df["regime_id"] == "elevated"]
print(f"Months in Elevated: {len(elevated)}")

Using the Dataset in R

library(readr)

url <- "https://eco3min.fr/dataset/score-eco3min.csv"
df <- read_csv(url)

# Latest reading
tail(df, 1)

# Distribution across regimes
table(df$regime_id)

Both examples load the dataset directly from the URL — no download required.

Methodology

Formula

The eco3min Score is a weighted average of three public US inflation indicators:

Score = 0.50 × Headline CPI YoY
      + 0.30 × Core CPI YoY
      + 0.20 × 5Y5Y Forward Breakeven

Why these three indicators

No single inflation indicator captures the full picture. The composite combines three distinct angles:

• Headline CPI measures what households actually pay, including food and energy. It is inflation as experienced.
• Core CPI strips out food and energy to reveal whether inflation has diffused into the broader economy (services, shelter, durable goods).
• 5Y5Y Forward Breakeven is the market-implied average inflation expected over a five-year window starting five years ahead. It is the forward-looking signal indicating whether expectations remain anchored.

The gap between the three components is itself informative. A score of 4 with CPI at 8% and Core at 2% tells one story (transitory energy shock). The same 4 with CPI at 4% and Core at 4% tells another (durably diffused inflation).

Why these weights (50 / 30 / 20)

The weighting prioritizes observed inflation over expectations, and raw measurement over filtered measurement:

• 50% Headline CPI — the reference measure: what appears in headlines, anchors wage demands, and determines real purchasing power erosion.
• 30% Core CPI — filters energy noise without entirely replacing CPI. A persistent gap between Headline and Core is itself a signal.
• 20% 5Y5Y Breakeven — minority weight because market expectations can be noisy (flight-to-quality effects, technical TIPS arbitrages), but their persistent drift remains one of the best predictors of future inflationary regimes.

Weights are fixed by construction — no re-optimization, no back-fitting. This is the same design philosophy as classic fixed-weight composites such as the Altman Z-score for corporate distress: transparent coefficients, a bounded output, and a mapping to named zones — reproducibility over statistical optimality.

Data sources

The T5YIFR series begins in January 2003 at the St. Louis Fed; the published Eco3min series starts in January 2004. The static chart above extends the trajectory back to 2000 for illustrative purposes using alternative expectation components; the downloadable CSV starts in January 2004.

Mapping to the 5 regimes

Thresholds are identical to those in the eco3min inflation framework for direct cross-reference.

Update schedule

The pipeline pulls the three FRED series via the FRED API every business day at 08:00 UTC (Mon–Sat). The score value changes once per month, the day after the BLS CPI release (typically the 12th–14th). Between releases, the score stays at its last official value — the 5Y5Y component moves daily but its 20% weight and monthly averaging stabilize the indicator.

How to Read the Score

Historical Reading: Two Decades Across Five Regime Bands

Over 269 monthly observations since January 2004, the score has occupied three of the five regime bands. Deflation (score < −0.5) and Very High (≥ 15) have never been reached: the series floor is −0.08 (July 2009), its ceiling 6.73 (June 2022).

2004–2007 — Late Great Moderation. Oscillation between 1.7 and 3.4 around a mean of 2.7, mostly inside the Target band, with 10 months in Elevated — including a five-month run in 2006. The Fed’s 2% objective was orbited, never durably undershot.

November 2007 – September 2008 — Oil and food spike. Eleven consecutive months in Elevated, peaking at 3.97 in July 2008 as crude approached $145/barrel amid a global food crisis.

December 2008 – October 2009 — Deflation scare. The score collapses from its July 2008 peak to 0.66 by December 2008 and bottoms at −0.08 in July 2009 — the all-time floor. Eleven months in Disinflation, yet the −0.5 Deflation bound is never crossed: at the trough, headline CPI was falling at −2.0% year over year, but core CPI (+1.5%) and the 5Y5Y breakeven (2.2%) kept the composite out of the Deflation band — the composite doing exactly what it was built for. A second Disinflation episode follows in June–December 2010.

2010–2019 — Calm decade. Mean of 1.89, range 0.78–2.98. Of 120 months, 96 sit in Target and 24 in Disinflation — the second half of 2010, a single month in late 2013, and the oil-collapse episode from December 2014 to mid-2016.

April 2021 – May 2024 — The post-pandemic surge. 38 consecutive months in Elevated — the longest streak in the series. The score crosses 4.0 in May 2021 (4.06), six months before the Fed retired the word “transitory” (November 2021), and peaks at 6.73 in June 2022.

June 2024 — Return to Target. 2.93 — the first sub-3.0 reading since April 2021. The score then oscillates inside the Target band for 22 consecutive months (range 2.39–2.94), touching its local low of 2.39 in January–February 2026.

March–May 2026 — Iran/Hormuz energy shock. The score rebounds to 2.85 in March — upper boundary of Target — as headline CPI jumps from 2.4% to 3.3% on a gasoline price spike (+21% year over year) linked to Iran-related supply disruptions. It crosses into Elevated in April (3.15) and reaches 3.38 in May: the first Elevated readings since May 2024. Core CPI rises far less (2.60% → 2.82%), marking this as an energy-driven episode rather than broad re-acceleration.

Key insight: across 269 months, the score has spent 169 in Target (63%), 59 in Elevated, and 41 in Disinflation — zero in Deflation or Very High. The failure modes are asymmetric by era: the 2010s missed the target from below, persistently; the 2020s missed it from above, in one 38-month streak. The band is the same; the direction of the miss defines the decade.

Methodological Limitations

• US coverage only. The score does not apply to the Eurozone, UK, or emerging markets. A transposition would require equivalent components (HICP, ECB 5Y5Y) with recalibrated weights.
• No sectoral diffusion measure. Two situations with score = 3.0 may mask different realities: inflation concentrated in energy versus inflation diffused across 70% of the basket. The Cleveland Fed Median CPI or sectoral dispersion would complement the reading.
• 5Y5Y Forward potentially noisy short-term. Market expectations can be perturbed by technical TIPS arbitrages. The 20% weight limits the impact, but in extreme months (March 2020, March 2023) the score can over-react.

The score is deliberately simple. Adding components (sticky CPI, trimmed mean, M2 growth) would marginally improve precision at the cost of readability and reproducibility.

Related Macroeconomic Datasets

Related Research

Macroeconomic Dataset Hub

This dataset is part of the Eco3min macro-financial data repository.

Explore the Eco3min Dataset Hub

Sources

• U.S. Bureau of Labor Statistics — Consumer Price Index for All Urban Consumers (CPIAUCSL, CPILFESL)
• Federal Reserve Bank of St. Louis — 5-Year, 5-Year Forward Inflation Expectation Rate (T5YIFR)
• Eco3min Research — Composite scoring methodology and regime classification

Dataset Reference

Eco3min Research. (2026). eco3min Inflation Regime Score (US) — Monthly composite indicator (2004–2026).<br> https://eco3min.fr/en/eco3min-inflation-regime-score-us-dataset/