Guide

Factory orders explained

Harbor Manufacturing's demand desk once sized raw-material buys from durable goods orders alone — and missed a Q2 restocking wave in petroleum-based lubricants and industrial chemicals that showed up only in total factory orders. The nondurable slice of the Census Bureau's M3 Manufacturers' Shipments, Inventories, and Orders report rose 1.8% month-over-month while headline durables fell on civilian aircraft; Harbor's steel PO model under-ordered by 12% until the team added the full M3 total-manufacturing book to their release-day checklist. Factory orders (the colloquial name for total manufacturing new orders in M3) capture every dollar purchase order booked at U.S. factories — durables and nondurables — not just the durable-goods advance that dominates financial headlines. This guide covers the four M3 pipeline stages (new orders, shipments, unfilled orders, inventories), how durables and nondurables split the total, the relationship to PMI and industrial production, the GDP manufacturing bridge, a Harbor Manufacturing total-book refactor, a technique decision table versus durable-goods-only models, pitfalls, and an analyst checklist.

What factory orders measure

The Census Bureau publishes M3 monthly, typically on the first business day after the durable goods advance (often the same 8:30 a.m. ET slot, one week after the reference month ends). The survey covers manufacturers in NAICS sectors 31–33 plus select non-manufacturing establishments that ship factory products.

Factory orders in market parlance means total manufacturing new orders — the sum of:

Durable goods orders — products with expected useful life of three years or more (machinery, autos, aircraft, appliances).
Nondurable goods orders — shorter-lived factory output (food, petroleum refining, chemicals, paper, textiles, apparel).

All series are reported in nominal dollars, seasonally adjusted. Month-over-month and year-over-year percent changes are standard; analysts often smooth three- to six-month annualized growth to filter aircraft and energy noise.

Factory orders vs durable goods orders

Media often says “factory orders” when they mean the durable goods advance released one week earlier. The advance is a fast subset; the full M3 release adds nondurables and revises durable detail. For total manufacturing demand, read the full report line: Manufacturing: New Orders (total), not only durables.

Release	Scope	Typical timing
Durable goods advance	Durables only; core capex ex aircraft highlighted	~3–4 weeks after month-end
Full M3 / factory orders	Durables + nondurables; total manufacturing	Same day or week after advance
Annual benchmark revision	Revised levels back several years	Typically May

The four M3 components: orders, shipments, backlog, inventories

M3 is a pipeline accounting system for the factory sector. Each month factories report four dollar flows:

New orders — customer purchase commitments received. Leading indicator: today's PO is tomorrow's production schedule.
Shipments — goods leaving the plant. Coincident with production and the manufacturing slice of GDP.
Unfilled orders (backlog) — cumulative orders not yet shipped. Rising backlog with stable shipments implies future production ramp.
Inventories — finished goods and work-in-process on factory floors. Rising inventories with falling orders signals margin pressure.

The identity linking them (simplified): unfilled orders at month-end equals prior unfilled plus new orders minus shipments (plus/minus revisions). Analysts watch orders minus shipments as a proxy for backlog build; sustained positive gaps foreshadow higher industrial production one to three months later.

Durables vs nondurables in the total

Durables typically account for roughly 55–60% of total manufacturing orders; nondurables the remainder. Nondurables matter because:

Petroleum and coal products — refinery throughput tracks energy prices and export demand; volatile month-to-month.
Chemicals — plastics, fertilizers, and industrial inputs feed downstream durable production (Harbor's missed lubricant wave lived here).
Food manufacturing — less cyclical but large in dollar weight; smooths total orders during recessions.
Paper and textiles — smaller but sensitive to housing (paper) and retail apparel cycles.

A durable-goods-only model can misread total factory momentum when energy and chemical orders swing on commodity prices rather than capex cycles.

How factory orders link to GDP and industrial output

Shipments (not orders) enter GDP as part of personal consumption, business investment, government purchases, and net exports depending on product category. New orders lead shipments; unfilled orders stock future output.

The bridge to quarterly GDP manufacturing:

Orders accelerate, shipments lag — typical early-cycle pattern; IP and earnings follow in 1–3 months.
Shipments rise, orders flat — backlog drawdown; production may soften even if current output looks strong.
Both orders and shipments fall — contraction signal; confirm with PMI and capacity utilization.

GDP nowcasts incorporate M3 shipments when available; orders inform the next quarter's revision path. Pair nominal orders with PPI for manufacturing to separate price from volume effects.

Factory orders vs PMI and other leading indicators

The ISM Manufacturing PMI surveys purchasing managers on whether conditions improved or worsened (diffusion index). Factory orders record actual dollar POs. They complement rather than duplicate:

PMI New Orders below 50, factory orders rising — managers feel pessimistic while customers still place POs; common in late-cycle inventory normalization.
PMI above 50, factory orders falling — sentiment holds while real bookings soften; watch for PMI catch-down.
Both falling three months — high-confidence industrial slowdown signal.

Regional Fed surveys (Philadelphia, Empire State, Dallas) offer mid-month hints but the national M3 total is the hard-data anchor. Cross-check with advance goods trade for export demand and business inventories for wholesale and retail stock positions outside the factory gate.

Seasonal adjustment, revisions, and release quirks

Advance vs full revision — durable goods numbers can shift when the full M3 incorporates more survey responses.
Aircraft and defense lumpiness — still dominate durable volatility; strip transportation for private capex reads (see durable goods guide).
Energy price pass-through — petroleum nondurable orders spike when crude rallies even if refinery volume is flat.
Government shutdown delays — Census releases may slip; stacked releases increase headline volatility.
Concurrent 8:30 data — factory orders often share the slot with construction spending, trade, or claims; read the bundle.

How markets react

Total factory orders move industrial equities, materials, and Treasury yields when the surprise is broad-based (not only aircraft):

Total orders beat, durables in line — nondurable strength (chemicals, petroleum) can lift materials ETFs even when machinery headlines look dull.
Total miss driven by durables — check nondurables before calling an industrial recession; food and chemicals may hold.
Rising unfilled orders across durables and nondurables — backlog build supports IP and cyclical earnings in Q+1.
Inventory build with weak total orders — margin risk for producers; bearish for industrial margins near-term.

Long-horizon investors weight six-month annualized total orders growth and the durables/nondurables split over single-month aircraft-driven prints.

Worked example: Harbor Manufacturing total-book refactor

Harbor Manufacturing supplies precision components to machinery OEMs and chemical processors. After the Q2 lubricant miss, the ops team extended their release-day checklist to the full M3 total:

Log total manufacturing new orders m/m and 3-month annualized before reading durable subcomponents.
Compute durables share of total — if durables fall but total rises, drill into chemicals and petroleum nondurables (Harbor's process-fluid customers).
Read core capital goods ex aircraft for private capex (unchanged from durable-only workflow).
Compare total orders to total shipments — positive gap for two months triggers extended steel and alloy PO lead times.
Cross-check ISM Manufacturing New Orders diffusion against total orders direction; note divergences in the ops journal.
Update raw-material model weights — Harbor shifted chemical-input sensitivity from 8% to 14% of the total-book signal after the refactor; forecast error on lubricant-related inputs fell 52 bps over two quarters.

Outcome: total-book integration did not replace core capex tracking — it prevented blind spots when nondurable factory demand led the cycle.

Technique decision table

Question you have	Best indicator	Why
Total U.S. factory demand this month?	M3 total manufacturing new orders	Includes durables and nondurables
Private business equipment capex?	Core capital goods orders ex aircraft	Strips transport; durable subset only
What actually hit GDP manufacturing?	Total or durable shipments	Shipments count at factory gate
Energy and refinery cycle?	Nondurable petroleum orders subtable	Price- and volume-sensitive slice
Near-term factory output volume?	Manufacturing IP ex-autos	Real hard output; confirms orders with lag
Manager sentiment before PO data?	ISM PMI New Orders	Survey leads at some turning points
Backlog pipeline depth?	Total unfilled orders	Stock of future production
Using only durable goods advance?	Add full M3 nondurables	Avoid missing chemical/energy swings

Pitfalls

Equating durable goods advance with factory orders — nondurables can flip the total sign.
Ignoring petroleum price effects in nondurables — nominal dollar spikes without volume growth.
Trading one month of total orders — smooth three to six months; aircraft and energy dominate single prints.
Using orders for GDP without shipments — orders lead; GDP counts shipments.
Expecting PMI and M3 to match monthly — judge over quarters during inventory swings.
Overlooking advance-to-full revisions — durable detail can change materially in the full report.
Missing durables/nondurables split — a rising total with falling durables tells a different story than broad acceleration.

Analyst checklist

Track M3 total manufacturing new orders alongside durable goods advance.
Log durables and nondurables contributions to total m/m change.
Plot six-month annualized growth in total orders vs manufacturing IP.
Read core capital goods ex aircraft for private capex (durable subset).
Monitor total unfilled orders and orders-minus-shipments gap.
Cross-check ISM Manufacturing New Orders for direction confirmation.
Deflate nominal trends with PPI manufacturing where volume matters.
Note Census footnotes for delays, revisions, and response rates.
Pair with business inventories and advance trade for full pipeline view.
Record release timing on the economic calendar — typically fourth week of month, 8:30 a.m. ET.

Key takeaways

Factory orders means total manufacturing new orders in Census M3 — durables plus nondurables, not the durable advance alone.
New orders lead shipments and industrial production; unfilled orders stock future output; inventories warn on margin risk.
Nondurables (petroleum, chemicals, food) can move the total when durables look soft on aircraft noise.
Harbor Manufacturing cut raw-material forecast error 52 bps by adding the total M3 book to their durable-only checklist.
Pair total factory orders with PMI, IP, and core capex for a complete industrial cycle read.