Retail sales in January were actually up by 4.8% year-over-year.
By Wolf Richter for WOLF STREET.
You read the headlines today: Retail sales, seasonally adjusted, fell by 0.9% in January from December. Winter storms, fires, the consumer collapsed, whatever. But what we’re actually looking at here are… massive seasonal adjustments.
Not seasonally adjusted, retail sales were up 4.8% year-over-year, but month-to-month plunged by 16.5%, or by $132 billion, to $668 billion in January, from December’s record and upwardly revised $800 billion. Confused yet about whether retail sales in January were good or bad? Yup, it’s the time of the year again to do a romp through seasonal adjustments.
Not seasonally adjusted, retail sales always plunge from December, the best month of the year, to January and February, the worst months of the year – this time by 16.5% in January from December, according to the Census Bureau today. Over the past 20 Januarys, retail sales plunged by 15% to 22% from December. This January was less bad than most: There were only two Januarys with smaller plunges: in 2023 (-14.8%) and in 2021 (-15.4%). The rest of the Januarys experienced bigger plunges.
Retail sales are hyper-seasonal. In December, retail sales explode. In January, retail sales implode, and retailers deal with the hangover. Gift-buying and holiday-spending is over, and here come the returns (negative sales).
For department stores, sales in January collapse by 40% to 50% from December. This January, they collapsed by 44.4%, but that was less than last year’s 46.0% collapse and less than the 48.4% collapse two years ago. Other retailers see less of a drop.
Huge seasonal adjustment factors are used to reduce December sales and to increase January and February sales, with the purpose of leveling them out (NSA = not seasonally adjusted; SA = seasonally adjusted).
December:
- NSA sales: $800 billion
- SA sales: $730 billion
- Seasonal adjustment: -$70 billion.
January:
- NSA sales: $668 billion
- SA sales: $724 billion
- Seasonal adjustment: +$56 billion.
The seasonal adjustment factor attempts to account for seasonal variations and differences in “trading days” and how holidays fall. For example, Christmas Day, when most stores are closed, fell on a Wednesday. In years when it falls on a Sunday, there’s an extra trading day in December that needs to be adjusted out. The Census Bureau’s X-13 ARIMA-SEATS software program calculates these seasonal adjustment factors, based on numerous historical data points.
- NSA: -16.5% month-to-month, to $668 billion, +4.8% year-over-year
- SA: -0.9% month-to-month, to $723 billion, +4.2% year-over-year:
Seasonal adjustments at the Big Three.
Sales at auto & parts dealers, at nonstore retailers (mostly ecommerce), and at food services and drinking places (restaurants, bars, cafes, cafeterias, etc.) combined account for over half of total retail sales. And they have different seasonal patterns. But Januarys stink at all of them.
For auto dealers, the best month of the year is in the spring or summer. One of the months from March through August is typically the peak of the year. Decembers are middle-of-the-road. And Januarys are the worst, they just suck. And February is not much better. But in March, the floodgates open. That’s the rhythm with auto sales, they did that in my time, and they still do.
This January:
- NSA sales: -9.9% month-to-month, +6.8% year-over-year.
- SA sales: -2.8% month-to-month, +6.4% year-over-year.
For nonstore retailers (mostly ecommerce), the best month by far is December, while January or February is the worst month of the year.
This January:
- NSA sales: -23.3% month-to-month, +3.8% year-over-year.
- SA: -1.9% month-to-month, +4.7% year-over-year.
For food services & drinking places, late spring and summer are the best months, January and February are the worst.
This January:
- NSA: -6.9% month-to-month, but +6.9% year-over-year.
- SA: +0.9% month-to-month, and +5.4% year-over-year.
So what does this mean?
For most retailers, January always sucks. That’s kind of what this means. But this January, actual retail sales (NSA) sucked less than January a year ago, they were up 4.8% from a year ago, despite winter storms and whatnot. January and February retail sales data are never conclusive about the direction of consumer spending because the huge and erratic seasonality in those months makes it impossible to draw broader conclusions about the consumer. What I can say is that the retail sales data today doesn’t tell me to start worrying about the consumer, suddenly, after the huge record December.
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
Mortgage News Daily: Best Levels of The Week After Downbeat Data
Lots of Hopeium in the real estate space right now.
Good point that YoY change is probably the best gauge. Seasonal adjustments are often helpful, but looks poor here.
Year over year, yes, but to know how a typical person is doing, one must adjust for both inflation and population growth (“real per person retail sales”).
Real per person retail sales are down year over year. The reported growth is all due to inflation and population change, not to improvements in standards of living.
The long-term trend is also interesting. The St. Louis Fed’s database goes back 33 years, and the story goes like this:
Real per person retail sales rose 25% from 1992-2000, then went nearly flat for 8 years from 2000-2007, before plunging 15% in 2008 in the Great Recession. Growth resumed after the Great Recession, but adjusted for inflation and population, retail sales did not return to the 2005-2007 peak for 12 years, until the COVID stims in 2020-2021. Those drove retail sales wild since much of the service economy was frozen. That COVID high has been wearing off ever since.
Real per person retail sales, as of this latest report, are back to just slightly above the 2005-2007 peak. Average “progress” from 2005 to 2025 was just 0.1 to 0.2% per year (2-4% TOTAL over 20 years). This 20-year stagnation in living standards is part of why many are unhappy about the economy.
Ignorant braindead math-challenged BS
1. You linked a chart that some dufus made and saved on FRED that adjusted retail sales to overall CPI. That is braindead ignorant BS. CPI is dominated by services, core services account for 64% of CPI, and inflation is in services. But retailers don’t sell services, they sell goods. Inflation, even before the pandemic, has been in services, not in goods. And you idiots are adjusting retail sales of goods to inflation in services (rent, healthcare, insurance, subscriptions, etc.). You gottta adjust retail sales to the inflation of goods. But wait…
2. Durable goods prices — the biggest category of goods that retailers sell – have dropped YOY = DEFLATION in durable goods. We have discussed this here endlessly. See link below from two days ago. In January, durable goods CPI was down 1.2% yoy. So that has the opposite effect, retail sales adjusted for deflation in durable goods were HIGHER than not adjusted. Inflation is in services, but retailers don’t sell services. Don’t you ever read anything here???
https://wolfstreet.com/2025/02/12/beneath-the-skin-of-cpi-inflation-worst-month-to-month-acceleration-of-cpi-since-aug-2023-on-spikes-in-used-vehicles-non-housing-services-food-energy/
3. Population growth in 2024 was 0.9%.
So do the math if you can:
— retail sales: +4.8% YOY
— durable goods inflation: -1.2% yoy,
— so “real” retail sales, adjusted for this deflation: ca +5.5% yoy
— population growth: +0.9%
— real per capita growth in retail sales: ca. 4.5%
4. I’m so sick and tired of people abusing my site to spread this stupid ignorant BS day after day after day. Go somewhere else to spread ignorant crap.
Drunken sailors maybe had 1 less day at port and spent the day drinking instead of retail shopping
Those sailors can still shop online while the vessel is underway – you have cell service if you’re not too far from shore.
SA = seasonally mal-adjusted
Wolf – So you covered most of the year for auto sales. Where do September to November fall? I assume they are average months? I had heard previously that October was a good month to buy.
Thanks.
On a scale of 1 to 10, with 10 being best, I would say about:
Sept 5
Oct 4
Nov 3
Dec 5
Jan 1
Feb 2
Mar to Aug: 8-10
But the worst month for retail sales in countries in the Southern Hemisphere is July, so it can’t be due to weather.
Winter in the Southern Hemisphere is June, July and August.
I purchased quite a few car parts mail-order last month. Doing my part to push up those NSA January sales!
“Decembers are middle-of-the-road.”
I always wondered how many people actually purchased cars as X-mas gifts. All the holiday car commercials seem to imply it’s a very common thing to give for a gift.
ShortTLT – Oversized bow sales track December auto sales.
It’s not as gifts normally, though there is some of that. But it’s the end of the year, and a lot of loose ends get tide up, some people have tax reasons, some people spend their bonuses, etc. There is also a lot of advertisement and specials to pull out the year, etc.
It feels like markets are always looking for reasons to send yields lower (and stock prices higher), consistent with the Federal Reserve’s still-dovish bias for the longer term.
They overreact especially strongly to downside misses on retail sales, manufacturing, housing starts, and other notoriously noisy data sets that are secondary to the Federal Reserve’s 2 mandates.
The hot CPI report yield spike lasted for all of a day. 10Y+ yields have declined for the past 5 weeks, with the Jan 10th top aligned perfectly with the S&P 500’s year-to-date bottom.
It just goes to show that some indicators might be better left at year to year (like retail sales), and some like cpi are probably okay month-to-month, realizing of course these are all estimates. Seasonal adjustment is always risky, because it is based on models made using historical data. There is this thing called structural change, that occasionally can make a mockery of seasonal adjustment models.
It might be informative if a data expert like Wolf (I am not being facetious, I think Wolf is excellent) would rank the most reliable indicators, based on how they actually reflect reality, in his opinion. And then also rank changes in such indicators based on how they actually reflect changes in reality. Perhaps a useful measure would be the size of revisions over time, in percent if possible. One might have to look at monthly, three month moving average, six month moving average, and annual time frames.
Of course, now we are talking about reality, which can vary from person to person, but I do know if I drop a 100 pound barbell directly onto my toe, there is a near 100 percent probability that my toe will hurt and it will be the same probability for almost everyone, except perhaps for those strung out on some drugs. That is an example of reality.
So Wolf,
Do you think retail sales would explode again just like in 2021 and 2023 (from looking at your graphs). We have a first ‘blip’ down from the usual convexity up. High probability is just a Fake up and we go higher in retail sales in the next few months? Based on your analysis so far, consumers are doing pretty well. Inflation has been going up consistently for the last 3 months.
Your thoughts? Thanks.
I’d say let’s wait until March maybe April to declare a decline in consumer spending.
Especially with the up revisions this was a nothing burger.
One possible driver for the January sales is the very cold weather we have been having in the South and the East. In Tennessee, we had a lot of days barely above freezing and several snow/ice events. Less people going out shopping plus everyone is broke from Christmas. Supposedly it was the coldest January in 11 years.
I think February sales could also be lower due to the impact of much higher utility bills from the January freeze. I just got my electric bill for January, and it was 40% above the previous 3 year January average.
I blame it on that damn groundhog.