Step 1 · Where it stands
US inflation is 3.7%; a year ago it was 3.0%
Data through Sep 2026. Against its own history it sits at position 64 (warm). This page does not say what comes next; it shows which factors have historically moved first in this site’s data, and which way each is pushing now.
Step 2 · What moves first
7 factors confirmed by the data, in 4 categories
This site ran predictive tests on every pair of its 59 indicators (method on the Network page). The most robust tier of factors that move ahead of US inflation — passing all three tests, genuinely improving forecasts, and holding up under a second method — has 7 members:
- Credit and Financial Stress: Credit spread and OFR stress — faster-than-usual increases (costlier borrowing, tighter markets) tend to be followed by falling US inflation
- Energy: Oil — faster-than-usual price rises tend to be followed by rising US inflation
- Commodities and Agriculture: Commodities, Food prices and Copper — faster-than-usual price rises tend to be followed by rising US inflation
- Real Estate: Real house prices (8 economies) — faster-than-usual price rises tend to be followed by rising US inflation
“Moves ahead of” is not “causes”: the tests show that these changes come first and help forecasting; they cannot rule out a third force driving both. Oil and food are themselves components of inflation (energy and food), so part of their lead is arithmetic.
Step 3 · Which way each is pushing now
Commodities and Agriculture; Energy push up; Real Estate; Credit and Financial Stress pull down
Push = how far a factor’s recent change differs from its usual change, carried through the historical relationship to tell how many percentage points higher (+) or lower (−) US inflation would be afterwards. Future changes are all set to usual; nothing is guessed.
- CommoditiesCommodities and Agriculture · same direction6 months +0.4812 months +0.69
- OilEnergy · same direction6 months +0.3812 months +0.55
- Food pricesCommodities and Agriculture · same direction6 months +0.2612 months +0.37
- CopperCommodities and Agriculture · same direction6 months +0.1112 months +0.15
- Credit spreadCredit and Financial Stress · opposite direction6 months −0.0112 months −0.09
- OFR stressCredit and Financial Stress · opposite direction6 months −0.0512 months −0.16
- Real house prices (8 economies)Real Estate · same direction6 months −0.2612 months −0.70
| Factor | Relationship with US inflation | Past 12 months | Push strength versus its own past | How often the relationship holds | Borne out when the push was strong |
|---|---|---|---|---|---|
| Commodities | faster-than-usual price rises tend to be followed by rising US inflation | +47.8% data through Sep 2026 | percentile 92 of its history | 67% of ten-year windows | 69% (259 months) |
| Oil | faster-than-usual price rises tend to be followed by rising US inflation | +71.8% data through Sep 2026 | percentile 90 of its history | 58% of ten-year windows | 69% (259 months) |
| Food prices | faster-than-usual price rises tend to be followed by rising US inflation | +15.8% data through Sep 2026 | percentile 64 of its history | 42% of ten-year windows | 63% (259 months) |
| Copper | faster-than-usual price rises tend to be followed by rising US inflation | +45.0% data through Sep 2026 | percentile 32 of its history | 42% of ten-year windows | 64% (259 months) |
| Credit spread | faster-than-usual increases (costlier borrowing, tighter markets) tend to be followed by falling US inflation | +0.03 percentage points data through Sep 2026 | percentile 34 of its history | 56% of ten-year windows | 57% (207 months) |
| OFR stress | faster-than-usual increases (costlier borrowing, tighter markets) tend to be followed by falling US inflation | +0.08 data through Oct 2026 | percentile 51 of its history | 50% of ten-year windows | 71% (99 months) |
| Real house prices (8 economies) | faster-than-usual price rises tend to be followed by rising US inflation | −1.5% (median of 8 economies) data through Q2 2026 | percentile 45 of its history | 40% of ten-year windows | 79% (73 quarters) |
The 7 pushes come from 7 separate models and overlap (for example, oil makes up about 57% of the commodities index, and food is in it too), so the pushes cannot be added up, and this site does not combine them into a score. What counts is direction: 4 pushing up, 3 pulling down.
Step 4 · How reliable has this been?
When most factors pushed the same way, US inflation moved that way over the next year 62% of the time
From Feb 2001 to Sep 2025, there were 245 months in which most of the 6 monthly factors (Real house prices (8 economies) is quarterly and not counted here) pushed the same way; over the following 12 months US inflation moved in that direction 62% of the time. Put the other way, 38% of the time it did not.
When a single factor’s push was strong (the strongest third of its own history), the later move went the same way 57%–79% of the time (last column of the table); recomputed using only data available at the time, 56%–77%. Nor does each relationship hold in every era: only in 40%–67% of ten-year windows.
Step 5 · What to watch next
3 things to watch
- Whether the rally in Commodities and Oil continues. This issue’s upward push comes mainly from last year’s sharp rises (Commodities +47.8% and Oil +71.8%). If the rally stops, the push fades over time.
- The next quarter of Real house prices (8 economies). It is the factor pulling down hardest; data run through Q2 2026, 3 months behind the other factors.
- Credit spread and Copper. Their pushes are the smallest right now (Credit spread −0.09 and Copper +0.15). Small does not mean unimportant: a large move would make their push grow.
Think it through: if Commodities stopped changing from next month, which cells in the table would shrink first? In the tug-of-war between Commodities and Agriculture and Real Estate, whose push lasts longer?
What this page does not say
Before you read
- This is not a US inflation forecast: a push is the historical average effect of one factor with everything else unchanged.
- The 7 pushes come from 7 separate models and cannot be added up.
- No factor in this site’s most robust tier moves ahead of US core inflation (excluding food and energy); this page is about headline inflation.
- No buy or sell advice.
Method and revisions
How it is calculated, and what was changed
- Factors: every relationship pointing to US inflation in the most robust tier on the Network page, all listed, none picked; recomputed each October with the data analysis.
- Push: the same predictive-test model (US inflation’s change on its own past 12 months plus the factor’s past 12 months; 4 quarters for quarterly data), carrying the gap between the factor’s last 12 months and its usual into the next 6 and 12 months; recomputed every issue. A separate program recomputed it by a different method, with identical results.
- Revision 1 (made after seeing results, recorded as is): the first version counted whether the 7 factors had risen faster or slower than usual over the past 12 months and looked 12 months ahead. It did not hold: after most pushed up, inflation rose 61% of the time; after most pulled down, it also rose 63% of the time. Commodities act fast and fade fast — a rise from a year ago drops out of the year-on-year rate a year later. So the push is now computed with the predictive-test model itself.
- Revision 2 (recorded as is): the plan was to split the factors into “next few months” and “about a year” by the month of their largest response; all 7 peaked in months 11–15 (a year-on-year rate needs a full year to absorb a change), so they could not be separated. Instead the same factors are shown at two horizons: the next 6 months and the next 12 months.