BofA Credit Card Data: October Spending Cools, Official Retail Figures May Be Overstated
nashnova research
Bank of America's real-time card data shows U.S. consumers pulled back sharply in October, but seasonal adjustment factors may make the official retail print look better than it really is — that gap is the signal to watch.
What will the official numbers say?
BofA economists forecast October retail sales at +0.2% m/m, well below September's +0.8%.
Ex-auto sales are expected at +0.4% m/m; the "control group" — the slice that feeds directly into GDP — also at +0.4%, matching consensus.
This means → the headline won't look terrible, but how much of that "okay" print is a seasonal-factor gift rather than real demand is the real question.
Why can't Wall Street agree?
Goldman Sachs has the highest call: headline retail at +0.3% m/m. JPMorgan and Deutsche Bank sit at the bottom: flat (0%).
The bigger split is in the control group: Goldman sees just +0.2%; Deutsche and BofA both say +0.4% — a two-to-one gap.
In plain terms = even top-tier banks disagree by a factor of two on whether the consumer still has legs. Consensus on real demand is far from settled.
What does BofA's own card data say?
October total card spending per household rose +5.8% y/y, down from +6.2% in September. Stripping out gas, growth slowed further to +4.6% (vs. 5.1%).
Seasonally adjusted, monthly card spending grew just +0.1%; the ex-auto retail slice was flat.
This means → the actual dollars swiped tell a clear story: consumers tightened their wallets in October, with growth decelerating across the board.
Which categories cracked first?
Furniture led the decline at -2.7% m/m, followed by department stores at -1.9% (the fourth drop in six months), lodging -0.9%, airlines -0.8%, and apparel -0.5%.
Apparel's y/y growth plunged to just 1.0%. Retail's contribution to overall card spending growth narrowed from 1.6 pp in September to 1.2 pp.
The lone bright spot: online retail rebounded +0.7% m/m after a flat September. Put simply = consumers didn't stop spending entirely — they shifted more of it online.
Why trust BofA's card data?
In the past six months, four major beats or misses occurred. BofA's card data called the direction correctly every time: May's blowout, July's plunge, August's rebound, September's upside surprise.
The 12-month ex-auto correlation with Census data is 86%, slightly below consensus's 89%. But the mean squared error is 0.14, beating consensus's 0.19.
This means → directional tracking is roughly on par with the Street, but at critical turning points the error is smaller — and turning points are where the information is most valuable.
What does this mean for markets?
The 10-year Treasury yield has already climbed to 5.14%, driven by a hot CPI print and hawkish Fed commentary.
If the official retail number overstates reality thanks to seasonal factors, the market may briefly overlook the actual cooling in demand.
This reflects a dangerous mismatch: the statistical release says "okay," but the card swipes say "cooling." Any soft signal in the consumption data becomes a key validation point for the direction of yields.
市场有风险,内容仅供研究参考,不构成投资建议。
