print-icon
print-icon
Add ZeroHedge as a preferred source on Google

Retail Sales Preview: BofA's "Omniscient" Card Data Sees A Sleepy October... And A Seasonal Gift From The Census Bureau

Tyler Durden's Photo
by Tyler Durden
Authored...

American shoppers took October off. The Census Bureau may not have noticed.

That, in two sentences, is what Bank of America's real-time debit and credit card data says about today's October retail sales report (due at 8:30am ET), which lands in the worst possible spot on the calendar: one week after a hotter than expected CPI and Fed Governor Waller's 50bps trial balloon sent the 10Y yield up 28bps to 5.14% (its biggest weekly jump since April 2025), and a few hours before Nvidia reports its first quarter without Jensen Huang.

Regular readers know that we have leaned on BofA's card data to front-run this report for well over a year (see "July Consumer Spending Unexpectedly Surges, Signals Big Retail Sales Beat: BofA Card Data" from August 2025), and that when the government shut down last October and the data went dark, it was the only retail sales report anyone had. This time the bank's economists Aditya Bhave and Shruti Mishra are delivering a split verdict in their latest retail sales preview (available to pro subs): the card swipes were soft, the official print will probably look fine, and the reason for the gap is a statistical quirk that has little to do with the consumer.

First, the numbers. BofA expects headline retail sales to rise 0.2% in October, a sharp slowdown from September's 0.8% surge, with sales ex-autos up 0.4% and the all-important control group (the one that feeds directly into GDP) also up 0.4%. That is right on top of a consensus which has drifted up to meet it: the Street was at 0.1% on the headline when BofA published last Thursday, and is at 0.2% now, with 0.4% for the control group.

The rest of Wall Street is less united than the consensus number suggests. Goldman is the high bid on the headline at 0.3% and the low bid on the control group at just 0.2%; Deutsche Bank and JPMorgan both see a flat headline, although DB matches BofA's 0.4% on the control group. Put differently, the two banks with the most to say about the consumer this week disagree by a factor of two on the only line that matters for GDP.

Why Listen To The Card Data? Because It Called Every Big Surprise Of 2026

Before getting into the details, a quick reminder of why we give BofA's forecast more weight than the usual sellside guess.

The bank sees card swipes across millions of households in real time, days before the Census Bureau publishes anything, and in the past six months that has translated into a track record no survey of economists has come close to matching. In June the card data pointed to a blowout beat for May and got one (0.9% vs 0.6% expected). In August it flagged a big miss for July, and retail sales duly plunged 0.6% when consensus was looking for a 0.1% gain. In September it saw the rebound coming (1.2% vs 0.8%). And last month, with economists expecting 0.3%, BofA said beat and the number came in at 0.8%.

Tally it up and the card data got the direction of all four big surprises right. The two months where it leaned the wrong way (April and June) were months when the print landed exactly on consensus, so nobody lost money on it.

BofA's own math is a little more modest, and worth repeating for the sake of honesty. Over the past 12 months the bank's ex-autos forecast had an 86% correlation with the Census data, slightly below the 89% for consensus, but its mean squared error was 0.14 versus 0.19 for the Street. Translation: economists as a herd track the wiggles about as well, the card data just misses by less when it counts. And yes, "omniscient" has always come with an asterisk: last December, when the delayed October 2025 report finally came out, BofA was wrong, very wrong. One bad miss in the middle of a shutdown-driven data blackout is probably forgivable.

Card Swipes Stall, But The Seasonals Ride To The Rescue

So what did the cards say about October? Well, not much good.

Total card spending per household rose 5.8% Y/Y, down from 6.2% in September, and ex-gasoline growth cooled to 4.6% from 5.1%. On a seasonally adjusted basis spending rose just 0.1% M/M, the second soft month in a row, and retail ex-autos was flat.

Under the hood it was a sea of red. Furniture tumbled 2.7%, department stores fell 1.9% (their fourth drop in six months), lodging slid 0.9%, airlines 0.8% and clothing 0.5%. Groceries, general merchandise and restaurants all declined too. Gas spending, which had been doing the heavy lifting in the nominal data for months, was basically unchanged at +0.1%. The one bright spot was online (card not present) retail, which rebounded 0.7% after a flat September.

On a Y/Y basis, the Bank of America Institute's companion Consumer Checkpoint report (also available to pro subs) shows the slowdown was broad: clothing growth collapsed to just 1.0%, leisure and lodging cooled to 2.1% each, and the contribution of retail to overall card growth shrank to 1.2pp from 1.6pp in September and 2pp at the summer peak. Only non-store retailers (+13.5%) and miscellaneous stores (+8.4%) accelerated.

With swipes that weak, why is BofA at consensus and not below it?

The answer is seasonal adjustment. In BofA's own card data the October seasonal factor for the control group is much less generous than it was a year ago, which is what pushes the bank's adjusted numbers down. The Census Bureau's projected October factor, on the other hand, is about the same as last year's. As BofA's economists explain:

In the BAC card data, the Oct control group SF is much less favorable than last year. But the Census’ projected control group SF for Oct is similar to last year (Exhibit 5). We think this discrepancy is likely to raise Census Bureau retail sales in Oct by 0.2-0.3pp, relative to the signal from the BAC card data.

In other words, the government's seasonal factors are about to gift the control group roughly a quarter of a point that consumers never actually spent. Last year's October data was collected in the fog of a government shutdown, so maybe nobody should be surprised that the seasonals around it are a mess.

There is one more wrinkle. If the control group does print 0.4% and goods inflation runs at September's pace, BofA calculates that real control group sales would accelerate on a three-month annualized basis. Good news for Q4 GDP trackers. For a Fed that is busy hiking, it's the opposite.

Goldman Fades The Control Group, Deutsche And JPMorgan Fade The Headline

Echoing the card data on the part that matters, Deutsche Bank's Jim Reid writes in his Early Morning Reid this morning (available to pro subs) that the bank's economists expect an unchanged headline, held back by a step-down in auto sales, and a 0.3% rise ex-autos, but a much firmer control group:

Our economists expect headline sales to be unchanged in October, with the step-down in auto sales weighing on the total. Ex-autos, our economists expect a +0.3% MoM increase after +0.8% previously, with lower gasoline prices acting as a drag. But the underlying consumer picture should look firmer than the headline: our economists expect the retail control measure to rise +0.4% MoM after just +0.1% in September. That will be an important cross-check on consumer momentum after recent data have continued to paint a picture of resilient US growth.

DB's US economists, who after last week's CPI now expect three more 25bps hikes in December, January and March taking fed funds to 4.625%, add that the control group should point to still healthy consumer demand.

Goldman begs to differ. In its US Week Ahead note (available to pro subs), the bank's economists pencil in 0.3% on the headline (consensus 0.2%), 0.4% ex-autos, and just 0.2% for both the ex-autos and gas measure and core retail sales, half of the 0.4% the Street expects:

We estimate nominal core retail sales increased 0.2% in October (ex-autos, gasoline, and building materials; month-over-month SA) based on the signal from alternative measures of consumer spending. We estimate nominal headline retail sales increased 0.3%, reflecting higher seasonally-adjusted gasoline prices.

Attentive readers will notice something odd. Goldman says gasoline lifts its headline number. Deutsche says gasoline is a drag on ex-autos. BofA's cards show spending at the pump up all of 0.1% on the month. Three banks, three opinions on which way a number posted on every street corner in America went last month. Diversification at its finest.

JPMorgan is the most downbeat of the lot. The bank's chief US economist Michael Feroli sees a flat headline and 0.3% for both ex-autos and ex-autos and gas, and thinks the consumer is finally starting to tire as the boost from this year's tax stimulus fades:

Overall, as the earlier boost to spending from the tax stimulus continues to fade and consumers remain pressured by high gas prices, overall consumption could begin to tick slightly lower after the strong prints in the recent quarter.

That said, the bank's Market Intel desk adds some much needed perspective on levels versus changes:

Putting numbers to it, in 2014 Retail Sales averaged $426bn/month increasing to $505bn/month in 2019. YTD, Retail Sales is averaging $723bn/month implying the growth rate of consumption has roughly doubled since COVID.

The desk notes that consumer cash (checking, savings and money market funds) is up $7.9 trillion, or 54%, since the end of 2019, and household net worth is up $78 trillion. It also has its own card data: according to JPM's Matt Boss, discretionary spending on Chase credit cards grew 5.9% in the August-October quarter, an acceleration from 5.3% in Q2, although October slowed to 5.4% from 6.2% in August/September, with the second half of the month (5.1%) weaker than the first (5.6%). So two of the largest card issuers in the country agree: October was a step down, from a very high step.

One housekeeping item, also courtesy of JPMorgan: don't look for the usual Y/Y comparison. Because of the October 2025 shutdown, the bank notes that annual changes won't be published.

The K Has Flipped: Lower-Income Households Now Lead... Because They Have To

Back in August, we pointed out that the BofA data showed K-shaped dynamics were waning. Three months later the K has turned upside down.

In the week ending Oct 31, total card spending rose 7.3% Y/Y for lower-income households, 5.9% for the middle and just 4.9% for higher-income households. Gen Z spending is up 10.7%, Baby Boomers 4.8%.

According to the Institute, lower-income households (+5.9%) even outspent the top 5% (+5.7%) in October, and the gap between the two groups in discretionary spending is the smallest since May 2025. The reason is wages: after-tax pay is growing 4.7% for the lower-income cohort, just under 4% for the middle and 3.4% at the top.

Before anyone declares victory for Main Street, two caveats.

  • A big chunk of that "leadership" is at the pump. Gas spending is still running 25% above last year, and BofA shows lower-income households have seen a disproportionate rise in spending on necessities (gas, groceries, utilities) since the Iran war started. As a result, the share of discretionary spending in the wallet was lower in October than a year ago for every income cohort. This was the one finding that gave BofA's economists pause: "This tempers our optimism somewhat."
  • Everyone is spending faster than they earn. Card spending has outgrown after-tax wages for five straight months in all income groups, and the gap is widest for the top 5%, whose spending runs on stock prices more than paychecks.

On that second point, with Nvidia fresh off its worst week since April 2025 and the S&P sliding again this morning, the Institute concedes that the wealth effect is the thing to watch:

The impact on this cohort’s spending from recent stock market gyrations will warrant close attention going forward. But given these wealth effects reflect the accumulation of asset price increases over a considerable period of time, in our view, spending may still be relatively robust with only small to modest pullbacks in equities.

Small to modest. We'll let readers decide whether that describes the past three weeks in semis.

Holiday Preview: More Trips To The Checkout, Smaller Receipts

Today's report also doubles as the opening bell for the holiday season, when (per the Institute) retailers book at least one in every five dollars of annual revenue, and jewelers nearly one in four. Here the early read is surprisingly strong: BofA's daily data shows total card and retail ex-auto spending running well above both 2024 and 2025 levels, and spending on holiday items is up 7.4% from the same period a year ago.

And just like the broader data, it is the bottom of the income distribution doing the buying. Holiday spending by lower-income households, which trailed every other group in October 2024 and 2025, is up about 9% Y/Y, versus 6% for higher-income households and 4% for the top 5%. Gen Y and Gen Z are up about 10%.

Look closer though, and the picture is of a shopper who is busy but careful. Transactions are growing faster than dollars across holiday items overall and in clothing, general merchandise and online, which means the average ticket is shrinking. In the Institute's breakdown, online holiday transactions are up roughly 15% while the average spend per transaction is down about 4% (by our read of the chart). With prices rising almost everywhere else, a falling ticket probably tells you people are trading down.

Is the early strength just a pull-forward? BofA's holiday survey (2,062 respondents, taken in September) suggests the opposite, with more shoppers planning to buy in early to mid-December than last year. The less cheerful finding is that around 46% of lower-income respondents say they plan to spend less this year, which sits awkwardly next to the cohort's 9% card growth. Either the survey is wrong or the cards are about to cool. JPMorgan's own survey of 1,000+ consumers, for what it's worth, has about 80% planning to spend the same or more.

As for whether households can afford it: mostly yes. A higher share of BofA cardholders is paying off the full balance every month than in 2019 or 2025 across all income groups, and deposits remain above inflation-adjusted 2019 levels (though the tax-refund bump has now fully faded). At the margins, things are less festive: the share of cardholders making only the minimum payment keeps creeping higher, and 401(k) hardship withdrawals have risen for two straight quarters to the highest level on the Institute's chart, which goes back to 2024.

Heads The Fed Hikes 50, Tails The Recession Trade Wins

Which brings us to how to trade it, and here the setup is about as awkward as it gets.

JPMorgan's trading desk, which remains tactically bullish, says another slowdown is expected and that a downside print risks feeding the recession narrative (the bank's recession basket is already up 15.8% this quarter while the S&P is down 1.2%). An in-line or stronger number, it argues, would help push flows back into beta and cyclicals. But a number that is too strong runs straight into Waller. With 17 Fed speakers on the tape this week and 27bps of hikes already priced for December, JPMorgan's trading desk frames it this way in its bear case:

A hotter Retail Sales print plus continued hawkish Fedspeak, perhaps seeing multiple members discussing the potential for 50bp; there are 17x Fedspeakers this week. Should the market believe that 50bp is necessary then it seems unlikely that hiking cycle is as shallow as the current pricing of 4x hikes (Dec inclusive).

As we discussed earlier today in "Equities Are Pricing A Slowdown, Rates Are Pricing Inflation", the two markets can't both be right, and today's report is the first hard data point that gets to referee. Positioning doesn't make it any easier: Goldman's consumer desk notes that hedge funds just bought consumer discretionary at the fastest pace in more than three months while shorting staples, so the fast money is already leaning toward a consumer that holds up. Home Depot reports the same morning, followed by Target, Lowe's and TJX on Wednesday and Walmart on Thursday.

Our take: the card data has earned the benefit of the doubt over Goldman's 0.2%, and the seasonal quirk means the control group is more likely to land at or above 0.4% than below it. That is the hawkish outcome. A flat headline will make for gloomy chyrons at 8:31am, but the Fed doesn't set policy off auto sales, and an acceleration in real goods spending hands Waller exactly the ammunition he asked for. BofA's own caveat on the holiday data is the right note to end on:

This could be due to the shift to online retail or a sign that consumers are trading down. It is still early days, though, and we will be watching these trends closely through Nov-Dec.

A consumer who shops more often, pays less each time, spends faster than he earns and still beats the Street. Then again, he has done it every month since the war started.

We'll have the full breakdown today at 8:30am.

Much more in the full BofA "October retail sales: looking past a tepid month" note and the Bank of America Institute's "Consumer Checkpoint: Spending season", both available to pro subs.

0