The Census Bureau will publish its advance estimate of October retail sales on Tuesday, Nov. 17, at 8:30 a.m. Eastern, according to CNBC's calendar, and economists surveyed by The Wall Street Journal expect a gain of 0.2 percent after an increase of 0.8 percent in September, MarketScreener reported. The report is short, but reading it well requires knowing what it measures, how early it arrives and how much it tends to change afterward.
What the report covers
The Census Bureau describes retail sales as "one of the primary measures of consumer demand," according to its description of the survey. The figures cover retail and food service companies that sell to final consumers, which means a restaurant meal counts alongside a purchase at a department store.
The advance report is an early estimate, and Census bases it on a sample of about 4,800 firms. The Bureau of Economic Analysis uses the retail estimates as an input when it calculates gross domestic product, so the monthly report also feeds into the broadest measure of the economy.
Seasonal adjustment
The headline numbers are seasonally adjusted, Census says. Spending rises and falls with the calendar every year, and the adjustment is meant to remove those regular swings so that one month can be compared with the next. A reader who sees a monthly change on Tuesday is seeing the adjusted figure.
Two months in one report
Each report contains two estimates. The first is the advance estimate for the month just ended, which on Tuesday will be October. The second is preliminary data for the previous month, so Tuesday's release will also carry a revised figure for September.
That structure means a careful reader looks at two numbers on release day. A September figure that moves up or down from its first reading changes the base against which October is measured.
How much the first estimate moves
The single most useful number for interpreting the advance figure comes from Census itself. The "average absolute revision to the advance estimate is now two-tenths of one percent," according to the bureau's survey description.
Tuesday's consensus offers the comparison that matters. The Journal's survey median for October is 0.2 percent, the same size as the average revision that Census reports for the advance estimate, which means the change economists expect for the month is about as large as the amount by which a first estimate typically moves once more data arrive. The September increase of 0.8 percent was four times that size.
The same survey expects sales excluding autos to rise 0.4 percent in October, after 0.9 percent in September, according to MarketScreener.
Where it fits in the larger picture
Retail sales are one piece of consumer spending, not the whole of it. Real consumer spending grew 3.8 percent in the third quarter, investingLive reported. Federal Reserve Governor Waller said in a Nov. 10 speech that the initial estimate of third-quarter real GDP growth was 3.4 percent.
The October report is the survey's first monthly reading for the fourth quarter. It will arrive on the same day that Home Depot reports earnings, and a day before Target, Lowe's and TJX, according to CNBC's calendar, so the government's estimate and the retailers' own accounts of the month will be published within the same week.
A checklist for Tuesday
On Tuesday morning, a reader can work through the release in order. First, find the October advance estimate and compare it with the 0.2 percent consensus. Second, find the revised September figure and compare it with the earlier 0.8 percent. Third, keep in mind that the October figure itself will be revised next month, by two-tenths of a percent on average, according to Census.