August's retail gain leans on pump prices and a Prime Day rebound

Gasoline accounts for 1.1 percentage points of the 6.0% annual gain in retail sales, or about 18% of it ^[1][2]. Wolf Richter is explicit that the gas station dollars rose because pump prices did, not because anyone bought more gallons ^[3]. Month to month the gap between the two headline rates is a tenth of a point ^[3].

The monthly figure also has a calendar hole in it. Richter attributes part of the August rise to recapturing July's drop, which followed Amazon Prime Day shifting into June this year from July last year, worsened by seasonal adjustments ^[4]. "It's not all based on Americans suddenly throwing their entire inheritance at retailers," he wrote ^[5]. His own smoothing shows the size of the distortion: ecommerce rose 2.6% in the single month, while the three-month average rose 0.6%, under a quarter of that ^[6][7][8].

The mix survives the smoothing. Ecommerce is up 10% year over year on both the monthly and the three-month reads ^[6][7], roughly twice the 4.9% ex-gasoline rate for retail as a whole ^[4]. At an 18% share over 12 months it is now the largest retailer category, ahead of motor vehicle dealers at 17% ^[8][10]. Apply that share to a $773 billion month and ecommerce is running near $139 billion against the $130 billion booked at dealers ^[5]. Dealers grew 1.7% over the year, with the three-month average at 3.2% ^[9][10].

Groceries are where the dollar figures need reading carefully. Food and beverage stores took $86 billion, up 0.5% from a year earlier, the three-month average up 0.7% ^[12]. Restaurants and bars took $105 billion, up 5.8% ^[11]. That is 1.22 times the grocery total ^[6], a crossover Richter dates to 2019 ^[14]. "Consumers are splurging on experiences," he wrote ^[15]. All of these are nominal dollars, so a 0.5% year at grocery stores is consistent with volumes flat or falling.

Richter gives August's ex-gasoline annual rate of 4.9% and no earlier month to set it against, so the account cannot say whether spending is speeding up ^[2]. It does support continued nominal growth. I would expect the ex-gasoline rate to hold near 5% for another print or two, because the two categories carrying it sit on smoothed trends: ecommerce at 10% year over year on the three-month average, restaurants at 5.3% ^[7][11].

The counter-thesis is concentration. Motor vehicle dealers, restaurants and bars, food and beverage stores and general merchandise stores add to $401 billion, or 51.9% of the month ^{[9][11][12][13][7]}. Two of those four grew 1.7% and 0.5% over the year ^[9][12]. General merchandise, at $80 billion, grew 4.5% ^[13]. If ecommerce's 10% decelerates toward the middle single digits, the rest of the table cannot hold a 4.9% ex-gasoline rate on its own. A restaurant three-month year-over-year under about 3% would take the discretionary leg out too.