Opinion

There's Nowhere to Hide in This Cost Cycle

The tell in this cost cycle is not that one input spiked. It is that 12 of 16 of them are climbing at once. When pressure gets this broad, the old trick of substituting toward whatever is still cheap quietly stops working, because nothing is.

Opinion | By Nora Whitfield, Materials & Markets. The argument here is the columnist's own; every figure links to the live series behind it, and opinion is not measurement.

Cost cycles have a personality, and this one's is breadth. It is tempting to fixate on whichever input made the headline this month, but the number that should worry a plant manager is not any single price. It is the count. Of the 16 manufacturing input series I track, from steel and resins to power, wages, and freight-adjacent materials, 12 are on a rising trend right now, with the basket up about 11.4% on average over the past year. That is not a commodity story. That is a diffusion story, and it changes what a buyer can do about it.

Why breadth is worse than a spike

A single input spiking is a manageable problem. You substitute, you redesign around it, you lean harder on the inputs that did not move. The whole discipline of value engineering assumes that when one material gets expensive, another is still cheap. Broad-based inflation breaks that assumption. When 12 of 16 inputs are climbing together, there is no cheap corner to retreat into, and the substitution playbook that has bailed out procurement teams for decades stops paying off. That is the quiet menace of a high-diffusion cycle: it is not louder than a spike, it is just harder to escape.

The lull that taught a generation the wrong lesson

To understand why so many buyers are caught flat-footed, you have to look at what the middle of this archive did to their instincts. Between 2022 and 2024, 8 of the 16 inputs I can track across those years actually fell. Steel closed 2021 at 450 and ground down to 267 by the end of 2024. Procurement teams who lived through that stretch learned a lesson that felt like wisdom: wait, and it comes back down. It did come back down, once. Then between 2024 and 2026, 12 of 16 reversed and climbed. The people still running the 2023 playbook are fighting the last war, and the tape stopped cooperating two years ago.

Six years of steel, as a warning

Read that column of numbers slowly, because it contains the whole argument. Steel did not go up in a line. It exploded, collapsed over three years, and has been climbing again for two, and it sits today 20% below where it closed 2021. Anyone who anchored on the 2021 top thinks steel is cheap. Anyone who anchored on the 2024 bottom thinks it is expensive. Both are anchoring on a single year of a six-year record, which is the most expensive habit in procurement.

Breadth of direction, not breadth of level

Pull the lens back six years and a subtler point emerges, one the last-fourteen-months view hides. These inputs are rising together, but they are rising from wildly different starting points. Of the 16 I can measure against a multi-year archive, 5 sit at or near the top of their whole range, copper, refined aluminum, and factory wages among them, while 1 are still in the lower third of theirs, natural gas and lumber the clearest cases, cheap by the standards of 2021 and 2022 even after their recent bounce. So this is not a story about everything being at a record. It is a story about everything pointing the same way at once, which is the more dangerous condition: a buyer cannot rotate toward the cheap inputs when the cheap inputs are the ones climbing fastest off the floor. Breadth of direction, not breadth of level, is the tell.

The one lever breadth leaves you

Here is the part procurement does not want to hear: when every purchased input is rising, the only cost lever fully in your control is the one inside your own four walls. Yield, scrap, energy intensity per unit, and labor productivity do not depend on a supplier's price list. In a narrow cost cycle those internal gains are nice to have. In a broad one they are the whole game, because they are the only savings a diffusion cycle cannot take back. The plants that come through this with their margins intact will be the ones that treated internal efficiency as a pricing strategy, not a housekeeping project.

A spike you can dodge. Breadth you have to outrun, and the only track you own runs through your own yield and scrap numbers.

What would change my mind

The honest counter to a breadth thesis is that diffusion cuts both ways: if inputs are broadly rising now, a turn could see them broadly falling, handing back the squeeze as fast as it arrived. That is real, and it is why the count is worth watching monthly rather than treating today's reading as permanent. If the number of rising series starts dropping meaningfully, the cycle is loosening and the substitution playbook comes back to life. Until then, breadth is the condition, and internal efficiency is the response. Watch the count, not the loudest line.

Track the full basket of input-cost series, and the composite pressure index built from them, on the live signals page. See the whole cost picture

Published 2026-08-06.