Real PCE Goods Computed from Retail Sales Implies a 68 Basis Point Boost to Q4 GDP

Seasonally adjusted retail sales increased in December after statistically insignificant revisions to retail sales for October and November. The Advance Retail Sales Report for December(pdf) from the Census Bureau estimated that our seasonally adjusted retail and food services sales totaled $709.9 billion during the month, which was 0.6 percent (±0.5%) higher than November's revised sales of $706.0 billion, and was 5.6 percent (±0.7 percent) above the adjusted sales in December of last year.   November's seasonally adjusted sales were revised less than 0.1% higher, from $705.7. billion to $706.0 billion, while October's sales were revised less than 0.1% lower, from $703.7 billion to $703.5 billion; however, the November sales change was unrevised at up 0.3 percent (±0.3 percent)* from October.  For the entire year, sales were up 3.2 percent (±0.4 percent) from the 12 months of 2022. Estimated unadjusted sales, extrapolated from surveys of a small sampling of retailers, indicated December’s actual sales rose 8.3%, from $712,614 million in November to $771,433 million in December, while they were up 4.0% from the $741,521 million of sales in December a year ago, so we can see how the large December seasonal adjustment knocked the big holiday sales increase down to a modest print...

Since it's the end of the quarter and the end of the year for retail sales, we'll include the entire table from this report showing retail sales by business type, including the quarter over quarter data.. Again, to explain what this table shows, the first double column below shows us the seasonally adjusted percentage change in sales for each kind of business from the November revised figure to this month's December "advance" figure in the first sub-column, and then the year over year percentage sales change since last December in the 2nd column; the second double column pair below gives us the revision of the November advance estimates (now called "preliminary") as of this report, with the new October to November percentage change under "Oct 2023 r" (revised) and the November 2022 to November 2023 percentage change as revised in the 2nd column of that middle pair (for your reference, the table from the advance estimate of November sales, before this month's revisions, is here).  Then, the third pair of columns shows the percentage change of the most recent 3 months of this year's sales (October, November and December) from the preceding three months of the 3rd quarter (July, August and September) and then from the same three months (October, November and December) of a year earlier.  That first column of the last pair thus gives us a snapshot comparison of 3rd quarter sales to fourth quarter sales, which could be useful in estimating the impact of retail sales on 4th quarter GDP, after those sales are adjusted for price changes.

To compute December's real personal consumption of goods data for national accounts from this December retail sales report, the BEA will initially use the corresponding price changes from the December consumer price index, which we reviewed last week.. To estimate what they will find, we'll first separate out the volatile sales of gasoline from the other totals. From the third line on the above table, we can see that December retail sales excluding the 1.3% price-related decrease in sales at gas stations were up by 0.7%.. Then, subtracting the figures representing the 0.2% increase in grocery & beverage sales and the statistically insignificant change in food services sales from that total, we find that core retail sales were up by more than 0.9% for the month.  Since the CPI report showed that the composite price index for all goods less food and energy goods was unchanged in December, we can thus estimate that real retail sales excluding food and energy sales will also show an increase of 0.9%.  However, the actual adjustment for each of the types of sales shown above will vary by the change in the related price index. For instance, while nominal sales at motor vehicles and parts dealers were up 1.1%, the price index for transportation commodities other than fuel increased by 0.3%, which would suggest that real sales at motor vehicles and parts dealers rose by about 0.8%.  Similarly, while nominal sales at clothing stores were 1.5% higher in December, the apparel price index was 0.1% higher, which would mean that real sales of clothing rose by around 1.4%.  On the other hand, while nominal sales at furniture stores were down 1.0%, the price index for household furnishings and supplies was 0.4% lower, which would suggest that real sales at furniture stores fell by about 0.6%..

In addition to figuring those core retail sales, to make a complete estimate of real December PCE, we'll need to adjust food and energy retail sales for their price changes separately, just as the BEA will do.  The CPI report showed that the food price index was 0.2% higher in December, with the index for food purchased for use at home 0.1% higher and prices for food bought to eat at restaurants 0.3% higher... hence, with nominal sales at food and beverage stores 0.2% higher, half of that was due to higher prices, so real sales of food and beverages were just 0.1% higher. Likewise, unchanged nominal sales at bars and restaurants, once adjusted for 0.3% higher prices, suggests that real sales at bars and restaurants fell about 0.3%. Meanwhile, while sales at gas stations were down 1.3%, there was a 0.2% increase in the retail price of gasoline, which would suggest real sales of gasoline were down around 1.5%, with the caveat that gasoline stations do sell more than gasoline, and we haven’t accounted for those other sales.  By weighing and averaging the real sales changes that we have thus estimated back together, and excluding food services, we can estimate that the income and outlays report for December will show that real personal consumption of goods rose by nearly 0.6% for the month, after rising by an unrevised 0.5% in November, but after falling 0.2% in October, rising by 0.6% in September, and falling by 0.1% in August.  At the same time, the 0.3% decrease in real sales at bars and restaurants would have a modest negative impact on December’s real personal consumption of services...

With those estimates for the relative change in real PCE goods between the months of the 3rd and the 4th quarter, we should be able to also estimate the change in PCE goods between those two quarters. By setting July's real PCE goods as an index equal to 100, we can then say that August's real PCE goods would be equal to 99.9, and from that, we'd get index values of 100.5 for September, 100.3 for October, 100.8 for November, and 101.4 for December. We then compute the quarter over quarter change in those index values at an annual rate to determine the probable change that would be applied to 4th quarter GDP... (((100.3+ 100.8 + 101.4 )/3) / ((100 + 99.9 + 100.5 )/3)) ^4 = 1.02826, which means that real PCE goods are rising at an 2.8% annual rate over the fourth quarter, based on our estimates for the percentage change between real PCE goods of the 3rd and the 4th quarter months.  Since real PCE goods has been running at 24.1% of GDP, that suggests that 4th quarter PCE goods would add roughly 0.68 percentage points to the growth rate of 4th quarter GDP…


Note: the above was first published as part of my weekly economic synopsis at MarketWatch 666.

Subject Meta: 

Forum Categories: 

good calculation

Thanks for this. I keep wondering where the crash is. We need to talk about the very obvious problem of using residential housing as poker chips in the great institional & foreign investor casino.