“Virtually Unchanged” Isn’t Stable: December Retail Sales Flat at $735.0B, But the Math Says “Indeterminate”

On February 10, 2026, the Census Bureau’s Advance Monthly Sales report for December 2025 led with “virtually unchanged”—and a 0.0% month-over-month reading on $735.0 billion in sales. That phrasing sounds calm. The fine print is not. The 90% confidence interval includes zero, meaning December could be slightly positive or negative—statistically indistinguishable from flat.

Here’s what the release makes clear:

• December 2025 total sales: $735.0B, 0.0% MoM, +2.4% YoY (nominal, “not for price changes”).

• Retail trade sales: +2.1% YoY; strength centered in nonstore retailers (+5.3% YoY, ±1.4%) and food services & drinking places (+4.7% YoY, ±1.8%).

• November 2025 was labeled “(revised) $735.1B”, yet the October–November percent change remained “unrevised” at +0.6% (±0.3%)—a level shift with an unchanged growth rate.

• The report again stresses nominal figures and flags insufficient statistical evidence that December’s change differs from zero.

• Methodology: “Advance” estimates rely on a subsample of roughly 4,800 firms using a link-relative estimator with limited imputation—improving timeliness but increasing sensitivity to sample mix and nonresponse.

• Timeliness took a hit: “Next release: To be determined.” The report is delayed due to a lapse in federal funding.

Stable? Statistically Indeterminate Is the Real Story

“Virtually unchanged” reads like equilibrium. It isn’t. When the agency emphasizes that the 90% CI includes zero, it’s admitting the point estimate is soft. December’s 0.0% MoM is a placeholder, not a pronouncement. For investors, that matters. A “flat” headline masks a distribution of outcomes in which December could be marginally contracting or expanding—both consistent with the data.

That ambiguity also infects the growth narrative. The report spotlights +2.4% YoY (total) and +2.1% YoY (retail trade), then reminds readers these are not adjusted for price changes. In other words, without deflators, we can’t tell how much of that is volumes versus prices. With momentum cooling into year-end, nominal-only framing invites overconfidence.

Revisions Without Regret: When Levels Move but Rates Don’t

November’s dollar level was revised to $735.1B, yet the October-to-November change stayed “unrevised” at +0.6% (±0.3%). This is how concurrent seasonal adjustment and benchmarking complicate the game: levels can move while growth rates appear stable. The headline suggests continuity; the ledger shows motion.

• Why it matters: GDP tracking, retail control aggregates, and quarterly run-rate assessments are driven by levels, not just rates. An “unrevised” rate amid a revised base can quietly alter the path of consumption, muddying quarter-end estimates.

• The release even leans into rate stability language while acknowledging revised levels—a communications choice that risks understating data drift.

Cherry-Picked Bright Spots vs. Broad Retail Drag

The narrative highlights nonstore (+5.3% YoY, ±1.4%) and restaurants (+4.7% YoY, ±1.8%), which are indeed stronger pockets. But the broader retail trade line sits at +2.1% YoY, and total sales at +2.4% YoY, pointing to softer core retail than the showcased categories imply. Composition did work:

• Services propped up the aggregate: food services growth outpaced retail trade, helping keep total sales near +2.4% YoY.

• E-commerce held its line, but it’s a relative win. In an environment drifting toward nominal softness, outperformance can still mean slower absolute growth.

Crucially, the press text offers no headline summary for the control group (retail sales excluding autos, gasoline, building materials, and food services)—the series most relevant for near-term GDP contribution. For now, we know where the strength was highlighted; we don’t get a clean read on the piece that matters most for macro tracking.

From August Momentum to Year-End Drift

The year finished with a softer gait than it started. Compare the August 2025 narrative to December’s:

• The October–December period rose +3.0% YoY, undershooting the +3.7% full-year pace—a clear downshift into year-end.

• December’s 0.0% MoM (statistically indeterminate) contrasts with August’s +0.6%. The story moved from acceleration to ambiguity.

Layer on the release delay (“Next release: To be determined”) and the “advance” read loses timeliness—hardly ideal for markets that trade on marginal information.

Methodology Matters—and Blurs the Signal

The “advance” program is a feature, not a bug—but investors need to understand its seams:

• Nominal-only: With +2.4% YoY for total and +2.1% YoY for retail trade unadjusted for prices, real demand is opaque. Without deflators, volume vs. price is guesswork.

• Sampling: A subsample of ~4,800 firms using a link-relative estimator with limited imputation raises sensitivity to sample composition and nonresponse. It’s why month-to-month can wobble and levels can be revised even when rate headlines stick.

• Concurrent adjustment: Seasonal factors update in real-time; that’s efficient, but it can reallocate growth across months, keeping rate narratives “unrevised” while reshaping levels.

• Missing control-group narrative: The text doesn’t summarize the GDP-relevant control group. We can dig it out from tables, but it’s conspicuously absent from the headline narrative.

What This Means for Markets

• Rates and duration: A statistically flat December and Q4 growth running below full-year support a modestly bid tone for duration, especially if upcoming deflators show limited real momentum. Curve strategy: favor intermediate duration; keep optionality for front-end repricing if the control group firms up.

• Equities—consumer complex:

• Credit: Retail HY spreads should reflect softening topline and revision risk. Skew toward shorter duration and secured paper where inventory collateral is verifiable.

• Macro tracking: Without real adjustments, PCE goods contributions remain uncertain. Watch the full retail report’s control group, corporate earnings commentary on January/February traffic, and early-quarter inventory builds for confirmation or contradiction.

What to Watch Next

• The delayed “advance” schedule: Any further lag erodes signal value; markets will lean more on corporate disclosures and card-spend trackers.

• Deflators: If goods prices are flat-to-down, real volumes could look better than nominal headlines—if not, the softness is real.

• Revisions: Another pass that shifts levels while rates stay “unrevised” is a tell. It means the path, not just the endpoint, is changing.

The Investor Takeaway

Don’t trade the adjective; trade the interval. December’s 0.0% is a midpoint in a confidence band that includes zero, not evidence of stability. The quarter ran below the full-year pace (+3.0% YoY vs +3.7%), November’s level shifted even as the rate didn’t, and the narrative highlights pockets of strength over a softening core. Position for a consumption path that’s cooler, noisier, and more revision-prone than the headline suggests:

• Tilt toward selective e-commerce and services with pricing power.

• Be cautious on broadline retail lacking traffic momentum.

• In rates, keep a measured duration bias with room to add if real demand underwhelms.

• Demand higher spreads in retail credit unless collateral and cash conversion are watertight.

In short: the data didn’t say “steady.” It said “uncertain.” Price that difference.