Market context

July Retail Sales Fell 0.6% in a Dollar Series With No Price Adjustment

Census reported July 2026 retail and food-services sales of $763.6 billion on 14 August. The advance estimate is seasonally adjusted but not adjusted for price changes, so an options reader needs a nominal, scope and revision check before drawing an event conclusion.

By Options Matrix Pro Editorial TeamPublished 6 min read
Share

July Retail Sales Fell 0.6% in a Dollar Series With No Price Adjustment

Friday, 14 August 2026 delivered two figures that can be placed beside each other and should not be blended together. At 8:30 a.m. Eastern Daylight Time, the Census Bureau released its July estimate for retail and food-services sales. At 4:40 p.m. Central Time, Cboe’s public market-statistics page displayed 66,955,059 matched U.S. options contracts for the completed session.

The Census headline was $763.6 billion, down 0.6% from June and up 5.0% from July 2025. The agency gives the label that controls the interpretation: the estimate is adjusted for seasonal variation and holiday and trading-day differences, yet it is not adjusted for price changes. That makes the report a timely record of dollar receipts, with a defined scope and revision path. It is not a direct reading of the number of goods and services purchased.

The July number has four useful versions

The report publishes four broad monthly changes. They remove different categories from the same retail-and-food-services universe, so they answer different descriptive questions.

Census seriesJuly 2026 adjusted salesChange from JuneChange from July 2025
Retail and food services, total$763.602 billion-0.6%5.0%
Total excluding motor vehicle and parts dealers$622.176 billion-0.3%5.8%
Total excluding gasoline stations$703.723 billion-0.6%4.2%
Total excluding motor vehicle and parts dealers and gasoline stations$562.297 billion-0.2%4.8%

The four numbers are not competing verdicts on the consumer. They are category filters. Removing motor vehicles changes the July monthly decline from 0.6% to 0.3%. Removing gasoline leaves the reported 0.6% decline in place. Removing both leaves a 0.2% decline. A headline should carry its filter before it is compared with an index, an individual company or an option chain.

Census also reported that the May-to-June total-sales change was unrevised at +0.2%. The July total’s stated margin of sampling error is plus or minus 0.4 percentage points. The report explains that a 90% confidence interval containing zero lacks sufficient evidence of a nonzero change; Census did not attach that marker to the total July change.

Sales receipts are different from the shopping cart

A shop receipt offers a simple way to picture the price boundary. A family can leave a store with the same basket and pay more because prices rose. The register prints a higher dollar total even though the quantity did not change. The Census release measures the register total after its published seasonal, holiday and trading-day adjustments.

Model output from an invented retailer

The following is a teaching model. It does not estimate July inflation, consumer volume, company revenue or any option price.

AssumptionBase monthNext month
Comparable basket quantity100 unitsCalculated from receipts
Price per unit$1.00$1.04
Sales receipts$100.00$105.00

Under those assumptions, the implied next-month quantity is $105.00 / $1.04 = 100.96 units. The model output is a 5.0% increase in dollar receipts alongside a 0.96% increase in basket quantity. At the original 100-unit quantity, the higher price would have produced $104.00 of receipts. The remaining $1.00 comes from the assumed quantity change.

The calculation is deliberately narrow. Retail categories contain different goods, prices, promotions and reporting patterns, so a single price change cannot turn the published total into a real-sales measure. It does show why the phrase “not adjusted for price changes” belongs beside a nominal headline.

Advance estimates have sampling and revision limits

Census describes the advance report as an early estimate based on a subsample of its full retail and food-services sample. The advance survey selects about 4,800 firms, then weights and benchmarks their sales to represent a universe of more than three million retail and food-services firms. The agency says the figures can contain sampling and nonsampling error, and it describes planned revisions as more complete information becomes available.

That procedure is valuable because it supplies a prompt monthly read. It also makes “advance” an operating label rather than a decorative one. The July table marks the current month as an advance estimate and June as preliminary. Census states that revised not-adjusted and corresponding adjusted estimates, reflecting historical corrections and the 2023 and 2024 Annual Integrated Economic Survey, are tentatively scheduled for 28 September 2026.

What Friday’s options count can and cannot add

Cboe’s volume summary lists 66,955,059 matched contracts at 16:40 Central Time on Friday, 14 August. That is useful activity context for the completed options session. The page identifies the aggregate market total and its venue groups; it does not identify trader intent, opening or closing status, account ownership, strike selection, expiry, execution quality or a causal response to retail sales.

The timing matters as well. A monthly release can arrive before the regular equity-options session, while a later full-session contract count pools the day’s many inputs. The two series use different units, scopes and clocks. A daily contract tally cannot convert a monthly nominal-sales estimate into a particular implied-volatility reading or an expected option outcome.

For the separate price concepts in current economic releases, read OMP’s guides to July PPI’s final-demand components and CPI’s four labels. The implied-volatility guide and options-chain guide cover the contract-specific inputs that remain necessary after any macroeconomic release.

Limits and the decision rule

Retail-sales data can be revised, category exclusions change the measure, and nominal dollars can move with prices as well as quantities. Options add liquidity, bid-ask spread, time-to-expiration, exercise, assignment, transaction-cost, tax and loss considerations. The release provides general market information, not personal financial, legal or tax advice or a recommendation to buy, sell or hold an option or security.

Before attaching an options-market conclusion to Friday’s report, record five fields: the July reference month, the 14 August release date, the chosen category filter, the nominal and seasonal-adjustment labels, and the estimate’s advance or revised status. Then inspect the actual contract’s expiry, quote, liquidity and risk separately.

Sources

Sources

Verified August 15, 2026

  1. 1U.S. Census Bureau, July 2026 Advance Monthly Sales for Retail and Food Services
  2. 2U.S. Census Bureau, Monthly Retail Trade - Sales Report
  3. 3Cboe U.S. Options Market Statistics
  4. 4OCC, Characteristics and Risks of Standardized Options

Put the framework to work

Test the framework against real options setups

Use the OMP Matrix, scanners and visualizer to compare yield, risk, liquidity and capital before making your own decision.

Analytics cookies help improve our product. Partner attribution may run separately. Privacy