Market Data

The Manufacturing Cycle Right Now: Six Leading Series, One Honest Verdict

No single indicator calls the manufacturing cycle. Six of them, read together and weighted by how early each turns, get much closer. Here is what the leading data says this month, and where it disagrees with itself.

The question every manufacturer wants answered, is the sector heading up or down, has no single-number answer, because every indicator that claims to give one is either lagging, noisy, or narrow. The honest approach is to read a panel of leading series together, weight each by how early in the cycle it turns, and accept that the verdict is a balance of evidence rather than a verdict. Assembled that way, today's panel leans expansionary. This piece walks all six, shows where they agree, and is explicit about where they contradict each other, because the disagreements are as informative as the consensus.

The panel, ordered by how early each turns

Leading indicators lead by different distances, so a useful panel is ordered from earliest to latest. New orders come first: they are the demand that has not yet been produced, currently $657B (Jun 2026), up about 7.4% from a year ago. The inventories-to-sales ratio is next, at 1.47 ratio (May 2026), down about 6.4% from a year ago, and it is the one series that reads backwards, a rising ratio is bearish, because goods piling up faster than they sell force production cuts. Truck tonnage, at 114.30 index (2015=100) (May 2026), up about 0.8% from a year ago, measures goods physically moving. Weekly hours, 41.7 hours/week (Jul 2026), up about 1.5% from a year ago, show employers stretching or shrinking the workweek before they hire or fire. The regional Fed composite, 19.43 index (Jul 1, 2026), is survey sentiment that leads the hard data. And capacity utilization, 75.56% of capacity (Jun 2026), little changed from a year ago, tells you how much slack the whole system has.

Scoring the balance of evidence

Counting the panel by direction, 4 of the five straight-reading leaders are on a rising trend and 0 on a falling one, with the inventories-to-sales ratio not rising, which is the benign reading. That tally is why the panel leans expansionary rather than delivering a clean call. The value of the count is not precision, these are diffusion-style reads with real noise, but discipline: it forces every series into the verdict instead of letting the loudest headline dominate. Industrial production, the hard output number at 98.70 index (2017=100) (Jun 2026), up about 1.1% from a year ago, is the confirming series that these leaders are trying to anticipate, and it lags them by design.

The cycle is never called by one number. It is called by a panel, weighted by how early each series turns, and honest about the votes that go the other way.

Where the panel disagrees, and why that matters

The most useful information in a mixed panel is the specific contradiction. New orders diverging from truck tonnage can mean orders are being booked but not yet shipped, an early-expansion pattern, or that shipments are running down a backlog while new demand fades, an early-contraction one. Weekly hours falling while capacity utilization holds suggests employers trimming labor ahead of a slowdown they see coming. The panel's job is not to erase these tensions but to surface them, so a manager can watch the specific series most likely to resolve the ambiguity next month rather than waiting for the lagging output number to make it obvious.

What to do with a mixed read

A balanced-to-mixed panel argues for optionality rather than a big directional bet: keep capacity flexible, avoid committing to large inventory builds or deep cuts, and watch new orders and the inventories ratio as the earliest tie-breakers. A decisively one-sided panel, most leaders pointing the same way with the inventories ratio confirming, is the signal to act with more conviction on hiring, capacity, and stock. The honest posture today, given a panel that leans expansionary, is to treat the direction as a lean rather than a certainty, and to let next month's earliest-turning series either confirm it or break the tie.

What six years of orders actually show

Over the 34-year record new orders have moved decisively rather than oscillated: from $244B at the close of 1992 to $657B today, up 169%, and now at the top of its 34-year range. A change of that size across a span this long is a level shift, not a cycle, and planning that assumes a return to the 1992 figure is planning against the whole record.

Use the capacity planning calculator to test how much headroom your operation has if the panel's lean plays out. Plan against the cycle

Published 2026-08-06.