PPI vs CPI for energy: does the Producer Price Index lead consumer prices? We tested it

The claim we set out to test is that the Producer Price Index (PPI) shows where the Consumer Price Index (CPI) is heading: in its strongest form, that producer prices lead consumer prices by 3 to 6 months. For fuel buyers that would make each PPI release a preview of pump and heating bills next quarter. This post tests the claim on energy, using index levels we pulled from the BLS public data API on Nov 13, 2026.

PPI vs CPI: what each measures for energy

The PPI measures what US producers receive. In BLS's words, “PPIs measure price change from the perspective of the seller.” For energy that means prices such as a refinery's gasoline, diesel and heating oil, a producer's natural gas, and a utility's electricity. The CPI measures what urban households pay: gasoline at the pump, fuel oil delivered to a home, and electricity and gas bills.

BLS lists three differences that matter for fuel:

• Taxes. The PPI records the producer's revenue, so sales and excise taxes are left out. The CPI records what the buyer pays, taxes included (our paraphrase of the BLS PPI FAQ).
• Imports. BLS: “the CPI includes imports, while imports are excluded from PPI.”
• Coverage. The PPI covers what producers sell to businesses, government and export markets as well as to households. There is a PPI for No. 2 diesel fuel. We did not find a diesel-only CPI series, so for diesel we use EIA's retail price as the consumer side.

The two also differ in how settled they are. Every PPI figure is preliminary when first published and can be revised for up to four months (BLS PPI FAQ). The Jul to Oct 2026 PPI values used here carry BLS's preliminary flag.

The October 2026 numbers

The PPI for final demand rose 0.3% in October, seasonally adjusted, and 6.5% over the 12 months to Oct 2026, unadjusted. Seasonally adjusted, final demand energy fell 0.1% in the month (BLS PPI release, Nov 13, 2026). The table below is our own arithmetic on the unadjusted index levels: for example, PPI final demand energy was 180.653 in Oct 2026 against 131.831 in Oct 2025, a 12-month change of +37.0% (180.653 ÷ 131.831 − 1).

The n/a cells are a gap in the source. The CPI release says BLS “did not collect survey data for October 2025 because of a lapse in appropriations,” and it publishes no 12-month change for Oct 2026 for the affected indexes (BLS CPI release, Nov 10, 2026; BLS note on missing data). CPI gasoline is an exception: it has an Oct 2025 index. We leave the gap empty and compare September where October is missing.

Why PPI and CPI energy diverge

Producer fuel prices are up far more, in percent, than consumer fuel prices:

• Gasoline, Oct 2026: PPI +75.9% against CPI +37.4%, a gap of 38.5 percentage points.
• Diesel, Oct 2026: PPI No. 2 diesel fuel +115.2% against EIA retail diesel +69.9%, a gap of 45.3 points.
• Heating oil, Sep 2026: PPI home heating oil and distillates +103.9% against CPI fuel oil +74.2%, a gap of 29.7 points. In Oct 2026 the PPI series was +118.5%; the CPI side has no 12-month figure.
• Energy overall, Sep 2026: PPI final demand energy +33.8% against CPI energy +20.4%.

Series choice changes the heating-oil gap. The like-for-like consumer series is CPI “Fuel oil” (CUUR0000SEHE01). The broader CPI “Fuel oil and other fuels” (CUUR0000SEHE) was +43.8% in Sep 2026, which would make the gap look like 60.1 points. The extra 30.4 points (60.1 − 29.7) come from the other fuels in the broader index, not from heating oil.

A percent gap of this kind does not mean retail prices have yet to respond. The producer price is one part of what a household pays. EIA lists four parts of the retail diesel price: crude oil, refining costs and profits, distribution, marketing and retail costs and profits, and taxes (EIA, Factors affecting diesel prices). When the refinery price doubles and the other parts move less, the retail price rises by a smaller percentage. That is our reading of the arithmetic, not a BLS or EIA finding.

The lag test

Method: for each pair we take the 12-month percent change of the consumer series in month t and of the producer series in month t − k, for lags k from 0 to 12 months, and compute the Pearson correlation r over CPI months Jan 2012 to Oct 2026. R² is r squared. A month with no value is dropped, never filled. That leaves 176 months for pairs that use a CPI series with the Oct 2025 gap, and 178 for gasoline and for diesel.

Every fuel pair is most correlated in the same month, and the relationship weakens steadily as the lag grows. That holds in both halves of the sample (2012 to 2019 and 2020 to 2026) for energy, gasoline, heating oil and diesel. The one energy pair that peaks later is PPI natural gas, a producer-level price, against CPI utility gas: its highest correlation is at one month.

Where a lead does show up, and how far to trust it

Outside energy the picture is different. PPI final demand less foods and energy against CPI less food and energy peaks at a 5-month lag (r = 0.88, against 0.82 in the same month). That is inside the 3 to 6 month range. It is not a stable rule: in 2012 to 2019 the correlation between the two was below 0.25 at every lag from 0 to 12 months. The lead comes from the 2020 to 2026 half of the sample alone.

For the headline indexes, PPI final demand against CPI all items is highest at a one-month lag (r = 0.94, against 0.93 in the same month). That is close to what Bentley Reid wrote on Oct 17, 2024: an “incredibly tight correlation between the two lines,” with “the CPI measure typically lagging the movement in PPI by just 1 month” (Follow The PPI Piper). The difference between lag 0 and lag 1 in our sample is 0.01, too small to call a lead.

Menzie Chinn reviewed the research on Aug 17, 2025 and called the evidence mixed: one study he cites is skeptical that the PPI adds predictive power, another finds producer prices lead consumer prices in G-7 economies, and a third finds a few detailed PPI series, fuels among them, carry most of the signal (our paraphrase of “PPI => CPI?”, Econbrowser). Our result for fuel does not contradict that. Fuel PPI and fuel CPI move together closely; they do so in the same month.

Limits of this test:

• It measures correlation, not cause, and it is not a forecast.
• The data are monthly. A lead of days or weeks inside a month cannot be seen.
• 12-month changes overlap from one month to the next, which makes correlations at nearby lags look alike.
• The four most recent PPI months are preliminary and the Oct 2025 CPI gap removes two months from most pairs.

What it means for fuel buyers and pricing analysts

• Read energy PPI as a same-month measure. In this sample it told you about the month it covers, at the producer level. It did not carry extra information about consumer energy prices three to six months on.
• Do not read the percent gap as a retail increase still to come. EIA's weekly US retail diesel price was $6.127 a gallon on Nov 9, 2026, down from $6.352 on Oct 26, 2026. Regular gasoline was $3.995 a gallon on Nov 9, 2026, down from $4.269 on Oct 26, 2026 (EIA Gasoline and Diesel Fuel Update). Those weekly retail prices fell into November. The Oct 2026 12-month PPI changes were above 75% for gasoline and 115% for diesel.
• Match the series. Compare PPI home heating oil and distillates (WPU057302) with CPI fuel oil (CUUR0000SEHE01), not with fuel oil and other fuels. Compare PPI No. 2 diesel fuel (WPU057303) with a retail diesel price such as EIA's EMD_EPD2D_PTE_NUS_DPG.
• Expect revisions. A contract or model that uses a PPI fuel index should say which vintage it uses, because the first print can change for four months.
• Use daily and weekly prices for timing. The October indexes were published on Nov 10 and Nov 13, 2026, after the month had ended. See current diesel prices, gasoline prices and the heating oil price, the CPI history, and our Nov 2026 posts on why diesel is so expensive and why gasoline prices are going down.

What would change this

The verdict rests on where the correlation peaks. It would change if a later sample showed fuel pairs peaking at a lag of three months or more, or if revisions moved the recent PPI months enough to shift the peak. BLS schedules the November 2026 CPI for Dec 10, 2026 and the November 2026 PPI for Dec 15, 2026. The figures on this page are as of the October 2026 releases.

Sources

• BLS, Producer Price Indexes, October 2026, released Nov 13, 2026. Series WPUFD4, WPUFD412, WPUFD49104, WPU0571, WPU057303, WPU057302, WPU0531, WPU054, WPU0541 and WPU0551 from the BLS public data API, retrieved Nov 13, 2026.
• BLS, Consumer Price Index, October 2026, released Nov 10, 2026. Series CUUR0000SA0, CUUR0000SA0E, CUUR0000SA0L1E, CUUR0000SETB01, CUUR0000SEHE, CUUR0000SEHE01, CUUR0000SEHF01 and CUUR0000SEHF02, retrieved Nov 13, 2026.
• BLS, PPI frequently asked questions (PPI and CPI differences, revisions).
• EIA, Gasoline and Diesel Fuel Update: monthly series EMD_EPD2D_PTE_NUS_DPG through Oct 2026 and weekly retail prices through Nov 9, 2026, from the EIA API, retrieved Nov 13, 2026.
• Bentley Reid, “Follow The PPI Piper”, Oct 17, 2024.
• Menzie Chinn, “PPI => CPI?”, Econbrowser, Aug 17, 2025.

BLS.gov cannot vouch for the data or analyses derived from these data after the data have been retrieved from BLS.gov. Nothing here is investment advice.