Thanksgiving Weekend Retail Down 14%. Chances Of Hitting NRF’s Optimistic Holiday Forecast Fades.

Right before Thanksgiving, the National Retail Federation (NRF) gave hope to retailers left struggling from the pandemic. NRF predicted November-December holiday sales would increase between 3.6% and 5.2% over 2019, excluding automobile dealers, gasoline stations and restaurants.

Its forecast was so optimistic that the low-end of the projection exceeded the five-year average growth of 3.5%.  

Then right afterward, it crushed that hope. Consumers spent 14% less compared to last year over the five-day Thanksgiving weekend through Cyber-Monday, $312 on average vs. $362 last year.

Rather, this year’s spending squared with that in 2018, according to the survey conducted by NRF and Prosper Insights and Analytics among 6,500+ American consumers.

And since these five days are the “official-unofficial” start of the holiday shopping season, that weekend’s results may be a better gauge of how the retail economy will fare for the rest of the season.

If that is the case, we are looking at a final result in the $700 billion range as in 2018, rather than the $755.3-to-$766.7 billion that NRF projects. In other words, a 4% decline from 2019’s final tally of $729.1 billion.

Admittedly, economic forecasting is a risky business and more risky this year with so many unknowns, like the rapidly rising Covid infection rates. And to its credit, NRF often gets its forecasts within or close to its range, as it can typically depend on American shoppers’ enthusiasm for the holiday season.

But not all years, most recently in 2018 when it predicted growth from 4.3% to 4.8%, but the actuals came in at 1.1%. Then there was 2008, when NRF got it spectacularly wrong, expecting a 2.2% increase, only to see a drop of 4.8%.

Given all the unknowns, you’d think the NRF would error on the side of caution and give a more modest forecast. But then the headlines wouldn’t look so good or its members feel so confident.

Battle of the forecasts

IBM’s
IBM
Karl Haller, who heads up its Consumer Center of Competency that advises retailers and consumer goods companies, offers a more realistic and subdued forecast for November-December retail sales.

IBM expects only a 1.8% increase heavily weighted toward the retail categories that prospered during the earlier pandemic shutdowns. These essential retailers – home improvement, grocery, mass discounters, and e-commerce – have become consumers’ “habitual first-choice shopping destinations,” he says.

Further, Haller expects a lot of that spending will not be holiday related because consumers are cutting the number of people they buy gifts for and reducing their participation in holiday-related activities, like sending cards or attending parties.

The result is that holiday-related spending will decline 19% from last year, $539 as compared with $670 last year.

Acknowledging that forecasting this year is particularly tricky – “Our first forecast in August was 1%, then a month later 2.4% and now it’s 1.8%, so it basically doubled then dropped by 33%. That’s a lot of movement.” – Haller critiqued the NRF forecast for presenting such a broad spread, especially considering that it came so late in the season.

And he took more substantive aim at these three factors missing in NRF’s predictions:

Too narrow time period

To accurately capture holiday spend, Haller believes this year’s and any year’s forecast must capture the full fourth quarter and extend into January, when most gift cards are redeemed. “Gift cards are a $100+ billion industry so forecasts should definitely extend into January,” he says.  

Further, consumers are shopping for gifts earlier, especially this year.

“Rather than Thanksgiving, Amazon
AMZN
Prime Day (October 13-14) was when the holiday really started,” he says, and adds, “It wasn’t just an Amazon thing either, with people ordering earlier to allow for shipping and worrying about availability because supply chains have been broken” he shares.

The IBM survey found that about 35% of consumers had already started holiday shopping in October. “Retailers have been trying to pull forward spending and it worked this year,” he notes.

Incomplete data

NRF’s forecasts ignore three important categories in retail which IBM includes in its calculations – food service, gas stations and automobile.

Food service last year was virtually equal to food and beverage stores, $766 billion and $765 billion respectively. But this year, food services are way down, -19% through the first ten months, while food and beverage stores are way up, 12%.

“There has been a $100 billion shift from restaurant to grocery. You can’t just measure retail growth with grocery without putting restaurants back in,” Haller asserts.

Gasoline stations, a $500 billion category last year, too have seen a 16% reduction in sales through October, which would be a moderating factor in NRF’s forecast.

And then there is automobile, which is the biggest single category in the Census’ retail reporting,$1.2 trillion last year.

“What if people actually acted like the Lexus
PLXS
commercial?” Haller quips. “It doesn’t take that many people buying his-and-hers cars to add up to a lot of spending. NRF wouldn’t get that data.”

Bifurcation in the consumer economy

And this is where Haller sees the biggest flaw in NRF’s forecast. “The top 30% of consumers [based on income] spend about as much as the bottom 70%. It’s basically a 50/50 split,” he says, based on the Bureau of Labor Statistics Consumer Expenditure Survey (CEX).

The NRF’s forecast factors in a strong stock market and rising home values, but those mostly impact the upper-income consumers. The middle-and-lower incomes households, which represent roughly 90 million households, haven’t fared so well, suffering more job losses and not having a cushion to fall back on.

Can the top 30% affluent consumers punch above their current 50% weight and fulfill NRF’s expectations?

Haller believes they have a chance, but it won’t be in gift purchases, since they along with middle-and-lower income consumers are reducing gift spending. Lower and middle-income consumers are cutting back 20-to-30%, while upper-income consumers will spend 7% less.

“What appears to be happening is the affluent consumers, who have this pot of money and are feeling wealthy, have shifted their spend from discretionary services [travel, hospitality, dining] into goods, mostly durable goods, some into non-durable goods,” he says.

So the question is will the affluent indulge in gifts for themselves this holiday season, like that Lexus car, a new living room or kitchen, or whether they will sit tight and wait to spend on discretionary services like travel once the pandemic lifts?

“It’s a conundrum,” he says. “Macy’s
M
or Best Buy
BBY
don’t care if it’s a million customers buying $1,000 each, or if it’s 100 million customers spending $10 each. They just want the sale.”

Garbage In, Garbage Out (GIGO)

Ultimately, retail forecasts are only as good as the data that goes into them, and in that regard, Haller believes every retail forecaster suffers from inadequate reporting on consumer spending. The government provides three sources of consumer economic data and none of them line up adequately.

“We are all missing the boat,” Haller believes. “The BLS, BEA and Census data need to be more aligned, because if you’re in a consumer-facing business, that is what you’re battling against.”

The Bureau of Labor Statistics CEX doesn’t tell you where consumer purchases are made, but does a good job in breaking down expenditures by demographics, like age, income, ethnicity and size of household.

Likewise, the Bureau of Economic Analysis (BEA) personal consumption expenditures data doesn’t break out where purchases are made. And because expenditures by nonprofit institutions serving households are also included in the same table, its numbers are further blurred. There are roughly 1.5 million non-profits representing over $1 trillion in contributions to the economy.

“Consumer spending is roughly 70% of the economy, but retail is only 40%,” Haller says. “Discretionary services spending like travel, entertainment and hospitality isn’t captured in retail.”

And then the Census Department’s monthly retail survey is veering further and further from what is actually happening in retail. The e-commerce data is particularly problematic.

“Walmart
WMT
 and Macy’s measure it in different ways,” he says. “And then Amazon reports its Whole Foods
WFM
data differently. A Whole Foods order made online but picked up in-store is tracked as an e-commerce sale. But for most other retailers, it would count as a store sale. The e-commerce percent of retail is becoming less and less irrelevant.”

And since retailers hang on the Census Department retail data, it encourages tunnel-vision so that the more important shifts in consumer behavior is missed.

“The problem is companies are only focused on the competitor next door and ignoring the fact that spending is shifting out of the category entirely from a good to a service,” Haller concludes. “You go out of business because you are battling to be number one or two in the department store business, let’s say, but you are missing the fact that people are renting clothes or buying by subscription.”