October PPI, called before it prints.
Producer prices for final demand, rebuilt from six pieces: food, energy, core goods, trade margins, transportation and other services. Energy follows weekly pump, diesel, jet fuel and natural gas prices; the rest follows its own momentum, with fuel costs squeezing trade margins and stock prices feeding portfolio fees.
What September 2026 told us
The last release against what was expected, what drove it, and what it changes.
The print
| First print | Now | Consensus | Model | Key Data call | Surprise | |
|---|---|---|---|---|---|---|
| Final demand | +1.3% | +1.34% | +0.6% | +0.91% | +0.60% | +0.7 pt |
| Core (ex food & energy) | +0.7% | +0.68% | +0.3% | +0.53% | +0.30% | +0.4 pt |
What it showed
- September 2026 final-demand PPI rose +1.34% (core +0.68%) against a consensus of +0.6% (core +0.3%).
- Largest contributions: energy +61.9 bp, trade services (margins) +26.9 bp, transportation & warehousing +14.3 bp.
- Producer prices are up 6.3% over the year; trade margins, the noisiest piece, widened 1.4%.
What drove it
Contribution to the headline change, basis points.
Analyst notes
A fuel shock passing through the pipeline. Final-demand PPI jumped 1.3%, more than double the 0.6% consensus. Energy rose almost 11% and contributed about 62 basis points on its own; transportation and warehousing, which buys that fuel, added another 14.
Margins widened, which is unusual in a fuel spike. Trade services, the margin retailers and wholesalers earn, rose 1.4% and added 27 basis points. Normally rising input costs squeeze margins; this time sellers passed costs on, which is worth watching for consumer prices in the months ahead.
Core was hot too. PPI excluding food and energy rose 0.7% against 0.3% expected, and producer prices are up 6.3% from a year ago.
October. With diesel and jet fuel easing from September's peak, the model expects final demand around +0.3% and core around +0.25%, with energy still positive but far smaller.
Where October 2026 should land
Odds of each outcome, sized by the model's own recent misses with fat tails for surprise months.
Final demand model +0.34%
Core (ex food & energy) model +0.25%
What the model expects, piece by piece
Expected contribution to the headline, basis points.
Track record
Every month the model is refit using only data available at the time, then scored against the number first published, the same yardstick used for the consensus.
| Period | Releases | Model miss | Key Data call miss | Consensus miss | Exact rounded print: model / call / consensus | Model closer / further | |
|---|---|---|---|---|---|---|---|
| Since 2023 | 44 | 0.274 | 0.277 | 0.277 | 14% / 9% / 9% | 25 / 19 | Model ahead |
| Since 2019 | 88 | 0.270 | 0.245 | 0.245 | 11% / 16% / 16% | 39 / 49 | Even |
| Since 2014 | 130 | 0.246 | 0.225 | 0.225 | 11% / 17% / 17% | 54 / 76 | Even |
| Period | Releases | Model miss | Key Data call miss | Consensus miss | Exact rounded print: model / call / consensus | Model closer / further | |
|---|---|---|---|---|---|---|---|
| Since 2023 | 44 | 0.227 | 0.237 | 0.232 | 20% / 18% / 18% | 24 / 20 | Even |
| Since 2019 | 88 | 0.235 | 0.234 | 0.231 | 16% / 14% / 15% | 46 / 42 | Even |
| Since 2014 | 130 | 0.220 | 0.217 | 0.215 | 18% / 14% / 15% | 64 / 66 | Even |
How it works
Each of the six final-demand components is forecast separately and re-fitted every month using only data that existed before that month. Energy is forecast on the unadjusted index, where pump and wholesale fuel prices map cleanly, then converted with the seasonal factor; that halves the error of modelling the adjusted index directly.
Components are combined with weights re-estimated each month from the preceding seven years, so the aggregation tracks BLS's own changing weights.
PPI is the noisiest of the inflation reports: trade margins swing with retailer pricing and the first print is often revised. The track record below scores everyone against the number BLS first published.
The Key Data call: once Wall Street's consensus is published, the model's view is combined with it. The combination starts from the consensus and moves toward the model by an amount learned from past releases, also weighing last month's surprise and revision. It is used only while it has beaten the consensus over the previous three years of releases, scored out of sample; otherwise the call is the consensus.