Market Data

The Six-Month Fuse: Your Margin Two Quarters Out Is Already Being Set

Margin compression is not a surprise. It is a fuse with a known burn time. Our pass-through model measures how much input inflation reaches output prices, and how long it takes, which means today's input surge is a scheduled margin event you can already see coming.

Manufacturers treat margin compression as weather, something that happens to them. The data says it is closer to a fuse: lit now, detonating on a schedule. Our cost pass-through model, which measures how input prices flow into output prices over time, finds the tightest fit at a lag of about 6 months with a pass-through coefficient of 0.207, meaning only a fraction of an input increase ever reaches the customer, and it arrives two quarters late. As of Jun 2026, input prices are running 21.9% year over year while output prices are up just 2.4%, a gap of 2.1 points. That gap is the fuse, and it is already lit.

What a ~0.2 coefficient actually means

A pass-through coefficient near 0.207 is the whole story compressed into one number. It says that for every point of input inflation, only about a fifth reaches output prices, at least within the window the model measures, and the rest is absorbed into margin or offset by productivity. The fit is not loose, either: the correlation at the best lag is 0.712, strong enough to treat the relationship as a genuine signal rather than a coincidence. Low pass-through is a double-edged finding. It means manufacturers are shielding customers from the full input shock, and it means they are eating the difference, quarter after quarter, on a predictable delay.

Turning the lag into a forecast

Here is the practical power of a measured lag: it makes margin partly forecastable. If input prices are up 21.9% now and only about 0.207 of that reaches output prices roughly 6 months later, then the margin pressure of two quarters from now is not a mystery, it is a calculation you can run today. A finance team that knows the coefficient and the lag can put a number on the compression heading toward them, size the escalation clauses needed to blunt it, and stop treating each quarter's margin as a fresh surprise. The fuse has a burn time; the model tells you roughly when it reaches the powder.

Margin compression is not weather. It is a fuse with a 6-month burn time, and the input data lit it two quarters before you feel it.

The honest limits of the model

A single coefficient and lag are a simplification of a messy reality: pass-through is not constant, it rises when demand is strong enough to make increases stick and falls when the market has slack, and the lag shifts with contract structures and inventory. So the model is a well-fit average, not a law, and it should be read as a base case to plan around rather than a precise prediction. But even as an average it beats the alternative, which is pretending margin compression is unforecastable and being surprised by it on schedule. For the current input, output, and gap figures as they update, the cost pass-through signal keeps the live read.

Six years of the fuse burning

The 36-year record shows steel making a full round trip, which is why point-in-time comparisons mislead so badly here. Its high came at the close of 2021 around 449.71, gave way over the following years to 266.51 by the end of 2024, and has climbed since to 361.44. That leaves it 20% below the peak and well off the floor, so whether today looks high or low depends entirely on which year you anchored to.

The cost pass-through signal tracks the input-versus-output gap, the best lag, and the coefficient as they move. See the pass-through signal

Published 2026-08-06.