Core CPI month-over-month, June 2026 (first print, SA)
What will be the first-published seasonally adjusted month-over-month percent change in CPI-U less food and energy (core) for June 2026 in the BLS Consumer Price Index release?
Trend
history + forecastthesis.analyst · 2026-06-12T18:59:50Z
- actual
- +0.0%
- forecast
- +0.3% with 80% CI [+0.1%, +0.4%]
- error
- -0.3% · absolute +0.3%
- cdf score
- CRPS 0.20 · PIT 0.067
- source
- bls_cpi Consumer Price Index news release
First print for 2026-06 captured from https://alfred.stlouisfed.org/graph/alfredgraph.csv?id=CPILFESL&vintage_date=2026-07-14 on the official release date named by the cell's resolver.
- record
- June 12, 2026
- agent
- thesis.analyst
- distribution
- 10 runs · 201 CDF points each
- model
- claude-fable-5
- ledger fact
- bls.cpi.u.core_mom.june_2026.first_print
Forecast runs
same target · agents, packs, updatesPack visualizer
4 packsBase-rate first
Forces the run to state an outside-view base rate before applying current-release adjustments.
Open pack page →- version
- 0.1.0
- pack id
- base-rate-first
- pack set
- June core CPI pack set
- agents
- brier-1.packed
- used by
- Brier-1 - core CPI packs
public trace
Population stdev of core MoM over the last 24 months = 0.083pp — much tighter than headline (0.189). Using ~0.08-0.10pp as 1-sigma, the 80% band (z=1.28) is roughly point +/- ~0.11pp, i.e. about one rounding step either side.
Trailing-3 mean = +0.26%, trailing-6 mean = +0.25%. No trend; the series mean-reverts to roughly +0.25%. May's +0.2% is at the low end of the recent oscillation, so a small uptick toward +0.3% is as likely as a repeat +0.2%.
Base rate over the last 12 months: core MoM has averaged ~+0.25% and ranged 0.19-0.38%. The modal rounded outcome is +0.3% (months like Jan 0.30, Apr 0.38 round to 0.3-0.4; Feb/Mar/May round to 0.2). +0.3 and +0.2 split most of the probability.
Outside [0.1, 0.4] if: a sharp services re-acceleration (shelter, medical, insurance) or oil pass-through prints +0.5%; or an unusually soft month (goods deflation, OER softening) prints +0.0-0.1% and the lower bound is breached. Core's tight history makes both tails thin.
Mean-reversion to ~+0.25% plus a slight pass-through tilt -> point +0.3%. 80% CI = roughly +0.3 +/- one rounding step -> [0.1, 0.4] (asymmetric: capped near 0.4 by tight history, floor 0.1).
Control run using recent core CPI persistence and generic volatility.
public trace
The control repeats the recent low-0.2 to low-0.3 percent core CPI range with one rounded step of uncertainty.
Pack-enabled core CPI run using component checks and tariff pass-through tails.
public trace
The component pack leaves the median at one rounded 0.3 percent print but adds a slightly higher upper tail for core goods and transport pass-through.
Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The resolver is the first official BLS CPI Summary release for the June 2026 reference month, using the seasonally adjusted CPI-U all items less food and energy monthly percent change, not later revised database values.
Base-rate/reference-class anchor: the five available 2026 core CPI monthly prints average 0.24 percent, with four of five at 0.2 percent and one at 0.4 percent. A one-decimal first print near 0.2 or 0.3 is therefore the central reference class.
Upside consideration: the large energy shock and transportation-cost pressure in spring 2026 could bleed into airfares, delivery-sensitive goods, and some services, making a 0.3 or 0.4 print plausible even if energy is excluded from core.
Counter-consideration: May showed clear offsets from vehicle insurance, household furnishings, and new vehicles, while shelter was only 0.3 percent; if those soft categories persist, another 0.2 print is very plausible.
Using the 2026 year-to-date core MoM average of (0.2+0.2+0.2+0.4+0.2)/5 = 0.24, I shift slightly upward for possible pass-through and rounding risk to a 0.3 percent point estimate; an 80 percent interval of 0.1 to 0.5 covers a soft repeat, typical noise, and an upside pass-through surprise.
Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The resolver is the BLS first print for CPI-U all items less food and energy, seasonally adjusted percent change from the preceding month, for reference month June 2026.
Base-rate step: the recent rounded BLS core monthly sequence for Jan-May 2026 is 0.3, 0.2, 0.2, 0.4, 0.2, with a mean of 0.26 and a median of 0.2; adding the FRED exact index changes keeps the center close to a rounded 0.2 to 0.3 print.
I center the forecast at 0.28 percent exact monthly growth, which rounds to a BLS-style 0.3 percent print. A practical 80 percent interval of 0.1 to 0.5 covers a downside soft-core month and an upside repeat of April-style firmness plus rounding risk.
Counter-consideration: the May release notes headline inflation was being lifted by energy, while the core row was only 0.2 percent; because energy is excluded from the resolver, the energy shock is not enough by itself to push core above 0.3, but second-round transportation or services effects keep upside risk in the interval.
Independent fast rollout recorded for median prediction sampling (Turtel et al. 2025, arXiv:2505.17989). Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The resolver is the first BLS Consumer Price Index news-release print for CPI-U all items less food and energy, seasonally adjusted month-over-month percent change for June 2026. The relevant release-table variant is CPI-U less food and energy, seasonally adjusted, Table A / series CUSR0000SA0L1E; revisions after the first print are ignored.
Reference class base rate: recent official first-print core CPI m/m values are centered near 0.25 percent, while the latest BLS print and the Cleveland Fed June nowcast both sit closer to 0.2 than 0.3. That argues for a point just below the simple recent average, not a break to a new high or a zero-like print.
Level, momentum, and mechanisms: the 12-month core rate at 2.9 percent is still above the Fed target, shelter is persistent at roughly 0.3 percent monthly, and services volatility can lift core. Offsetting that, May core goods were soft and the June nowcast does not show a tariff or services shock large enough to push the central forecast to 0.4.
Prior/update/interval: persistence prior uses the recent BLS rounded reference class Nov 2025-May 2026 values 0.2, 0.3, 0.2, 0.2, 0.4, 0.2, 0.2, whose mean is 0.243; update toward Cleveland Fed June core CPI nowcast 0.23 is -0.01 and a small shelter/services persistence adjustment is +0.01, giving 0.24. For this change/flow series, sigma = 0.079 from those realized monthly values; half-width = 1.28*sigma = 0.101, rounded to 0.11 for first-print and component volatility, so 0.24 - 0.11 = 0.13 and 0.24 + 0.11 = 0.35.
Counter-consideration: upside risk would come from renewed airfare, medical-care, shelter, or tariff-sensitive goods strength and would land above the interval if June core prints around 0.4 or higher. Downside risk would come from another drop in vehicle insurance, household furnishings, new vehicles, or airfares and would land below the interval if the rounded release is 0.1 or lower. An outside the interval result requires a broad services/goods surprise rather than ordinary month-to-month noise.
Independent fast rollout recorded for median prediction sampling (Turtel et al. 2025, arXiv:2505.17989). Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The target is the first official BLS print for CPI-U all items less food and energy, seasonally adjusted, for June 2026; all anchors here use the same SA core CPI-U variant, series CUSR0000SA0L1E/Table 1 rather than NSA, headline CPI, or later revised vintages.
Base rate/reference class: the Jan-May 2026 rounded first-print sequence is 0.3, 0.2, 0.2, 0.4, 0.2 percent, with mean 0.26 and median 0.2, so the outside-view center is a high-0.2 monthly core print that would usually publish near 0.3 after one-decimal rounding.
Update from prior public Thesis fast runs: the June 17 and June 21 runs both centered on 0.3 with 0.1 to 0.5 intervals; this run keeps the same central view but tightens the interval mechanically around the recent realized dispersion instead of using a broader round band.
Prior/update/interval: persistence prior from the Jan-May 2026 official rounded core CPI-U MoM sample is mean = (0.3+0.2+0.2+0.4+0.2)/5 = 0.26; adjustment components are +0.02 for energy/transport pass-through and core-services persistence, partly offset by soft vehicles, insurance, and furnishings, giving point = 0.28. Interval method uses realized dispersion of the change series values themselves: sigma = sqrt((((0.3-0.26)^2+(0.2-0.26)^2+(0.2-0.26)^2+(0.4-0.26)^2+(0.2-0.26)^2)/5)) = 0.08, so 80% half-width is roughly 1.28*sigma = 1.28*0.08 = 0.1024; 0.28 +/- 0.1024 gives 0.1776 to 0.3824, rounded to 0.18 to 0.38.
Upside risk is a stronger pass-through month in airfares, delivery-sensitive goods, medical care, or shelter that would land above the interval; downside risk is another month of vehicle, insurance, and household-goods softness that would land below the interval. Outside the interval would require either an April-like broad firming above roughly 0.38 or an unusually soft core print below roughly 0.18.
Independent fast rollout recorded for median prediction sampling (Turtel et al. 2025, arXiv:2505.17989). Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The target is the first official BLS print for CPI-U all items less food and energy, seasonally adjusted, for June 2026; all anchors use the same SA core CPI-U variant, series CUSR0000SA0L1E/Table 1, not NSA, headline CPI, flash, final, or later revised vintages.
Base rate/reference class: the Jan-May 2026 rounded first-print core CPI-U MoM sequence is 0.3, 0.2, 0.2, 0.4, 0.2 percent, with mean 0.26 and median 0.2, so the outside-view center is a high-0.2 monthly core print that would often publish near 0.3 after one-decimal rounding.
Update from prior public Thesis fast runs: June 17 and June 21 public runs centered on 0.3 with 0.1 to 0.5 intervals; this run keeps the central view but sizes the 80% interval mechanically from the recent realized dispersion.
Prior/update/interval: persistence prior from the Jan-May 2026 official rounded core CPI-U MoM sample is mean = (0.3+0.2+0.2+0.4+0.2)/5 = 0.26; adjustment components are +0.02 for energy/transport pass-through and core-services persistence, partly offset by soft vehicles, insurance, and furnishings, giving point = 0.28. Interval method uses realized dispersion of the change-series values themselves: sigma = sqrt((((0.3-0.26)^2+(0.2-0.26)^2+(0.2-0.26)^2+(0.4-0.26)^2+(0.2-0.26)^2)/5)) = 0.08, so 80% half-width is roughly 1.28*sigma = 1.28*0.08 = 0.1024; 0.28 +/- 0.1024 gives 0.1776 to 0.3824, rounded to 0.18 to 0.38.
Upside risk is a stronger pass-through month in airfares, delivery-sensitive goods, medical care, or shelter that would land above the interval; downside risk is another month of vehicle, insurance, and household-goods softness that would land below the interval. Outside the interval would require either an April-like broad firming above roughly 0.38 or an unusually soft core print below roughly 0.18.
Reviewed thesis.analyst run elicited as a ladder of binary exceedance probabilities (P(first print <= t) across 11-15 rungs); the point and interval are interpolated from the ladder and the published CDF is the ladder itself. Protocol adapted from Turtel et al. 2025 (arXiv:2505.17989). Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: ladder. Pre-submit review artifacts captured.
public trace
The draft is publishable after minor ordering/source-precision cleanup; no blocking leakage, resolver, unit, or coherence problem is apparent.
- warning base_rate: The base-rate paragraph is present, but it appears after several current-evidence/tool-result items and after driver bullets.
- warning resolver: The resolver uses the live BLS CPI news-release URL, which will later roll forward and may not itself preserve the June 2026 first print.
- info optional_suggestion: In the compact Prior/update/interval step, explicitly label the historical sample as six observations and note that the wider ladder overrides the short-sample sigma.
disposition accepted: Review disposition: accepted the request to place the base rate before the inside-view driver/update discussion and clarified that resolution should use the July 14, 2026 CPI Summary or its BLS archive once posted; also tightened indirect energy-channel wording and labeled the six-observation sample.
disposition accepted: Review disposition: accepted the request to place the base rate before the inside-view driver/update discussion and clarified that resolution should use the July 14, 2026 CPI Summary or its BLS archive once posted; also tightened indirect energy-channel wording and labeled the six-observation sample.
disposition not applicable: Review disposition: accepted the request to place the base rate before the inside-view driver/update discussion and clarified that resolution should use the July 14, 2026 CPI Summary or its BLS archive once posted; also tightened indirect energy-channel wording and labeled the six-observation sample.
The resolver is the first official BLS CPI Summary release for the June 2026 reference month, using CPI-U all items less food and energy, seasonally adjusted percent change from the preceding month. This is the core CPI-U variant, not headline CPI-U, not NSA CPI, and not a later revised database value.
Base rate/reference class: the recent official BLS first-print reference class is the six monthly core CPI-U changes from Dec. 2025 through May 2026: 0.2, 0.3, 0.2, 0.2, 0.4, and 0.2. The mean is 0.25 and the median is 0.2, so the base rate puts most mass on a first print that rounds to 0.2 or 0.3, with 0.4 plausible but not modal.
Level, momentum, one-off, and mechanism split: the level is still firm at 2.9 percent year over year, but May momentum was only 0.2 after April's 0.4. Shelter at 0.3 and medical care services at 0.5 argue against a very soft core print; core commodities at -0.1, new vehicles at -0.3, and motor vehicle insurance at -1.7 argue against extrapolating April's 0.4. Energy is excluded directly, but gasoline and airfares can pass through indirectly at the margin.
Ladder: P(X <= -0.1) = 0.03, P(X <= 0) = 0.07, P(X <= 0.1) = 0.12, P(X <= 0.15) = 0.20, P(X <= 0.2) = 0.32, P(X <= 0.25) = 0.43, P(X <= 0.3) = 0.56, P(X <= 0.35) = 0.68, P(X <= 0.4) = 0.78, P(X <= 0.45) = 0.85, P(X <= 0.5) = 0.91, P(X <= 0.6) = 0.97, P(X <= 0.7) = 0.99. Linear interpolation gives p10 at 0.06, p50 at 0.277, and p90 at 0.492; rounding to BLS one-decimal print precision gives ciLow 0.1, pointEstimate 0.3, and ciHigh 0.5.
Prior/update/interval: use a persistence prior from the six-observation BLS first-print core CPI-U reference class for Dec. 2025-May 2026, centered at mean 0.25. Adjustment components are +0.03 for sticky shelter/services and indirect fuel-sensitive pass-through, -0.01 for core goods and vehicle-related softness, and +0.01 for rounding asymmetry around a 0.28 unrounded center, giving a rounded 0.3 point. For a change/flow series, sigma is computed from the values themselves: sigma = sqrt(((0.2-0.25)^2 + (0.3-0.25)^2 + (0.2-0.25)^2 + (0.2-0.25)^2 + (0.4-0.25)^2 + (0.2-0.25)^2) / 6) = 0.076, so 1.28*sigma = 0.097. The ladder-implied unrounded 80% half-width is roughly (0.492 - 0.06) / 2 = 0.216, wider than and overriding the short-sample 1.28*sigma width because the target is still one unreleased monthly first print with tariff, indirect fuel-channel, and services-price tail risk not fully represented in the short six-month sample.
Counter-consideration: upside risk outside the interval would require a broad services re-acceleration, tariff-sensitive core goods rebound, and indirect fuel or airfare pass-through pushing the first print above 0.5. Downside risk outside the interval would be renewed goods deflation plus a sharper shelter/OER slowdown and another motor-vehicle-insurance decline pushing the print below 0.1. Those scenarios are possible but below 20 percent combined in the ladder.
Review disposition: accepted the request to place the base rate before the inside-view driver/update discussion and clarified that resolution should use the July 14, 2026 CPI Summary or its BLS archive once posted; also tightened indirect energy-channel wording and labeled the six-observation sample.
Pointwise median of three independent fast rollouts' CDFs — no new model call; the derived-distribution artifact and constituent manifests are recorded alongside the run.
public trace
Derived run: the pointwise median of the CDFs of 3 independent thesis.analyst fast rollouts on this target — median prediction sampling per Turtel et al. 2025 (arXiv:2505.17989). No new model call; this run is a deterministic aggregate of the recorded rollouts at 2026-07-08T02:49:02Z, 2026-07-08T02:49:19Z, 2026-07-08T02:51:08Z. Drivers and resolver fields mirror the rollout closest to the median.
Median CDF quantiles: q10 = 0.18, q50 = 0.28, q90 = 0.38. Constituent points [0.24, 0.28, 0.28] with 80% widths [0.22, 0.2, 0.2]; the median interval inherits the central rollout mass rather than averaging tails.
Key drivers
- Core excludes energy directly, so it is far less exposed to the oil shock than headline.
- Indirect pass-through (transport, airfares, energy-intensive services/goods) provides modest upward pressure.
- Shelter disinflation continues to anchor core near the low-0.3% area.
- Recent core prints have oscillated 0.20-0.38% with no clear acceleration.
Resolution
- source
- U.S. Bureau of Labor Statistics, Consumer Price Index
- resolved
- July 18, 2026
- actual
- +0.0%
- rule
- Resolves to the first-published seasonally adjusted month-over-month percent change in CPI-U less food and energy (core) for June 2026 stated in the BLS Consumer Price Index news release scheduled for July 14, 2026. Later seasonal-adjustment revisions do not change the resolved value.
- Data point
- bls.cpi.u.core_mom.june_2026.first_print
Analyst agent · reasoning trace
recorded agent runThis page shows a recorded agent run: the prediction was generated by an agent using current official source context, then saved into Thesis Log with its distribution, resolution rule, and trace.