Census's 6.87 Backlog Ratio Uses a Narrower Shipment Base

Census's 6.87 Backlog Ratio Uses a Narrower Shipment Base

The Census Bureau's 2 October manufacturing report put August's durable-goods unfilled-orders-to-shipments ratio at 6.87, up from revised July's 6.81. Dividing its $1,609.602 billion backlog by all factory shipments produces 2.44 instead. The difference comes from the shipment base: Census uses the industries with unfilled orders, with stated exclusions.

That scope belongs beside any claim about how much future business manufacturers have secured. The ratio compares an order backlog with one month's eligible shipments. It supplies neither a delivery timetable nor a company's future profit.

Census released the full August report on Friday, 2 October 2026, at 10:00 a.m. Eastern Daylight Time, or midnight at the start of 3 October in Brisbane. Friday was the latest completed relevant U.S. options session when this article was prepared on Saturday morning. No share-price, option-price or market-reaction evidence is used here.

Reproduce the shipment base

Table 3 gives August's seasonally adjusted unfilled orders as $1,609,602 million. Table 1 offers several shipment totals. Using the same backlog numerator with each denominator shows why the row matters:

Industries with unfilled orders. Table 1's August shipments are $234.407 billion. Dividing the $1,609.602 billion backlog by that shipment denominator gives a calculated ratio of 6.87.

All durable-goods industries. Table 1's August shipments are $333.589 billion. The same backlog divided by this broader shipment denominator gives a calculated ratio of 4.83.

All manufacturing industries. Table 1's August shipments are $658.622 billion. The same backlog divided by this broadest shipment denominator gives a calculated ratio of 2.44.

These three ratios are author arithmetic from the published million-dollar cells, rounded to two decimals. Only the first reproduces Census's published unfilled-orders ratio. The broader-denominator calculations answer different questions and should not replace it.

Table 7 places 6.87 in its durable-goods row. Its footnotes exclude industries with no unfilled orders, including wood products and automobiles, and exclude semiconductor manufacturing. The table marks the all-manufacturing and nondurable unfilled-orders ratios as not applicable. Those marks are not zero readings.

A fictional manufacturer makes the coverage issue easier to see. Its made-to-order line has a $600 backlog and ships $100 during a month. A separate shelf-stock line has no backlog and ships $200. The backlog divided by made-to-order shipments is 600 / 100 = 6. Including both lines' shipments gives 600 / 300 = 2. Neither calculation changed the backlog; the second included activity outside its coverage. These invented amounts explain the denominator choice, not the Census result.

A ratio does not date the deliveries

Census defines unfilled orders as an end-of-period backlog, generally reconciled from opening unfilled orders, new orders net of cancellations, and net sales. Shipments measure the dollar value of products sold during the period. Orders can be for immediate or future delivery.

The ratio therefore expresses backlog relative to a monthly shipment flow. Calling 6.87 a guaranteed 6.87 months of sales would add assumptions about delivery timing, cancellations, production pace and prices that this aggregate does not establish. It also says nothing about the costs needed to fulfil those orders or the amount belonging to a particular listed company.

This full-report vintage is separate from the 25 September advance durable-goods comparison. August remains preliminary, while July is revised. The figures are seasonally adjusted, not adjusted for price changes. Census warns that its survey panel is not a probability sample, so sampling error and confidence intervals cannot be calculated. The absence of an interval does not establish precision or statistical significance.

Keep the company and contract evidence separate

An options research premise based on backlog needs evidence about the underlying company's own orders, delivery terms and financial results. A national industry ratio cannot fill those gaps. Preserve the 2 October release, August reference period and eligible-shipment denominator before comparing it with an issuer's disclosures.

Then check the contract's underlying, strike, expiration, exercise and settlement terms, quote timestamp, bid and ask, and available size. Our options-chain guide and liquidity and bid-ask guide explain those records. The Census ratio provides no executable price, expected option payoff or profitable direction.

Options Matrix Pro publishes this article and offers commercial options-analysis software. This is general education, not personal financial, legal or tax advice or a recommendation to trade. An option buyer can lose the premium; sellers can face substantial losses and assignment obligations. Liquidity, concentration, transaction costs, exercise funding, settlement and taxes depend on the position and account. If a research note calls 6.87 an all-factory ratio or a promised delivery horizon, correct that premise before using it to compare contracts.

Sources and scope

• Census Bureau, August 2026 Full M3 Report, CB 26-150 M3-2 (26)-08, released 2 October 2026 at 10:00 a.m. EDT. Tables 1, 3 and 7 supply the figures, row coverage, revision labels and limitations. This current-release PDF rotates; identify the release and reference month before reuse.
• Census Bureau, M3 definitions, accessed 3 October 2026 Australia/Brisbane. Source for shipments, orders and backlog definitions.
• Census Bureau, M3 release schedule and economic-indicator calendar, accessed 3 October 2026 Australia/Brisbane. Both identify 2 October's full August release. The M3 schedule lists the next full report, for September, on 3 November; this article uses no September estimate.
• Cboe, U.S. options hours and holidays, accessed 3 October 2026 Australia/Brisbane. C1's Friday Curb ended at 5:00 p.m. EDT, or 7:00 a.m. Saturday in Brisbane. This venue-specific clock does not establish every contract's closing time or a market reaction.
• FINRA, Options and OCC's options-disclosure page, including its currently linked June 2024 disclosure document, accessed 3 October 2026 Australia/Brisbane. General options, exercise, assignment, settlement, cost and loss-risk context only.