Q3 GDP Update, Estimate Revisions, Retail Sales

By in
Q3 GDP Update, Estimate Revisions, Retail Sales

Three “Data” items today:

#1: Not to start on a down note, but the Altanta Fed GDPNow model just printed its lowest Q3 estimate to-date at just 1.2 percent. We are well within the window where GDPNow is most accurate since it has almost complete data for the most recent quarter. We had hoped this model would have reconsidered its take after Friday’s strong Advanced Retail Sales report, but no …

Here are the major economic releases since GDPNow was looking for +5 percent Q3 growth back in early September and which ones pulled that estimate down to its current 1.2 percent expected growth rate. As you can see, the step function decreases occur when the Census Bureau’s manufacturing data and auto sales data are released (September 2nd, October 4th). Other reports do not reset expectations higher.

Takeaway: GDPNow is predicting a slow Q3 growth rate when the number comes out on October 28th; we can extrapolate that supply chain disruptions combined with labor shortages will be how Wall Street’s commentary frames that disappointment. Markets tend to see GDP as a rearview mirror sort of report, but it will certainly put renewed pressure on the current administration to do what it can to ease these bottlenecks going into 2022. Mid-term elections are in 387 days, after all …

#2: It’s time to check in on what we only half-kiddingly call “the most important chart to US stock investors”. The image below is courtesy of FactSet, and it shows the last years’ worth of weekly S&P 500 earnings per share estimates based on Wall Street analysts’ financial models for each company in the index. The estimates for 2021 are the bottom line and 2022’s are the top line.

Our annotations (in red) show that analysts have waited until quarterly earnings announcements to meaningfully change their 2021 and 2022 estimates. The rate of change during the last 3 earnings seasons has run +3 – +7 percent. Outside of earnings seasons, the pace of their upward revisions has been much slower, just +1 – +2 percent. Going into the current reporting cycle Wall Street S&P 500 earnings estimates have actually declined for 2021 and only held flat for 2022.

That analysts have been consistently surprised by the strength of US corporate earnings and continued to underestimate future earnings leverage is important for 2 reasons. First, since their numbers are a proxy for expectations imbedded in stock prices, this explains why the S&P 500 is up 19 percent YTD (2021 estimates are up 21 pct). Second, we’re beginning to see the same sort of upside earnings surprises in Q3 results that powered estimate revisions in the last 3 financial reporting seasons. Too early to call it a trend, true: only 8 percent of the S&P 500 has reported. But thus far companies are beating estimates by an average of 15 percentage points, so we’re off on the right foot at least.

Takeaway: markets are always smarter than Wall Street analysts, but we doubt another strong (+10 percentage points of annual growth versus Street estimates) earnings season is fully baked into US large cap stock prices. That sort of comp remains our base case since Q3 estimates are still 7 percent below Q2 actuals ($49.24/share versus $52.80/share). Even if Q3 GDP was only 1-2 percent, as GDPNow predicts, that should be enough to see earnings improve further.

#3: Two brief thoughts on Friday’s Advanced Retail Sales report, which came in better than expected.

First, we need to adjust the data for inflation. The report is in nominal dollars and does not isolate inflation as a component of rising or falling retail sales. Using the latest CPI report as a proxy for inflation, we get the following results for the headline results as well as a few areas of interest:

  • Total US Retail Sales comp in September 2021 versus September 2020: +8.5 percent (13.9 pct reported, 5.4 pct CPI)
  • Retail Sales excluding gas stations and car dealers: +10.7 pct (13.5 pct reported, 2.8 pct core CPI inflation less new/used vehicle inflation)
  • Food and beverage stores: +2.5 pct (7.0 pct reported, 4.5 pct “food at home” CPI inflation)
  • Restaurants and bars: +24.8 pct (29.5 pct reported, 4.7 pct “food away from home” CPI inflation)

The upshot here is that, unlike the decade before the pandemic, inflation is now meaningful enough to skew the data from the Retail Sales report. At one level, it doesn’t necessarily matter because consumers are clearly spending on the items covered by the report. We do think it is important to understand inflation’s effect, however, because it makes comparing 2021 to prior periods more challenging. Consumers are, simply put, not buying as much “stuff” for the same dollar of outlays.

Second, even with the inflation adjustment in mind, there were several segments that really stood out in Friday’s report:

  • Furniture and home furnishing stores: +0.2 percent versus August and +13.4 pct vs. September 2020.
  • Clothing and clothing accessory stores: +1.1 pct versus August, +22.4 pct vs. September 2020.
  • Food Services and drinking places: +0.3 pct versus August, +29.5 pct vs. September 2020.

In this data we see the reflection of Jessica’s suburbanization theme, where Americans remain focused on upgrading their homes/wardrobes and spending time at nearby restaurants and bars. They are not travelling much, for leisure or work. They may not even yet be commuting into the office regularly. The money they would be spending on those activities (and did, prepandemic) is going into categories that the Advanced Retail report captures.


Atlanta Fed GDPNow: https://www.atlantafed.org/cqer/research/gdpnow

FactSet Earnings Insight: https://www.factset.com/hubfs/Website/Resources%20Section/Research%20Desk/Earnings%20Insight/EarningsInsight_101521.pdf

Retail Sales Report: https://www.census.gov/retail/marts/www/marts_current.pdf