Foundry and Forge Shop Products PPI: From Index Changes to Budget Sensitivity

The BLS Foundry and Forge Shop Products Producer Price Index (WPU1015) reads 296.49 index (1982=100) for Aug 2026, up about 3.7% from a year ago. The reference period identifies the observation; a later website refresh does not turn it into a daily price. This is a producer price index, not a delivered quote in dollars per unit.

This series covers foundry and forge shop products. It is broader than iron and steel castings and must not be presented as a casting-only production or cost measure.

Apply the change only to exposed spend

Match alloy, process, weight, tooling, machining allowance and acceptance requirements to the part being quoted. The observed product price includes more than metal input cost; a supplier's conversion cost needs its own evidence.

A useful sensitivity is exposed spend × assumed index change × assumed pass-through. For example, $100,000 of genuinely benchmark-linked spend, a hypothetical 5% move and 60% pass-through produces a $3,000 change. Those are explicit example assumptions, not observed market values. Fixed-price spend, unmatched products and delivery charges should not automatically receive the same increase.

Keep physical efficiency separate

Yield, scrap, consumption and product price affect different parts of a cost calculation. Record them separately so an efficiency improvement is not counted twice or mistaken for an agency price move. Quote comparisons should use cost per accepted unit on the same specification and delivery basis.

A difference between this PPI's growth rate and a steel mill PPI's growth rate does not isolate labor, energy, foundry margins or a fabrication premium. Their product baskets differ, and neither is a supplier's cost ledger.

Source and update notes

The source is BLS series WPU1015, also published by FRED at https://fred.stlouisfed.org/series/WPU1015. The linked data page provides the current observation and available history. A successful refresh changes readings only when the source returns changed data. Missing or failed observations must retain their original periods and be reported as unavailable or delayed.