April Retail Sales “Up” 0.5% to $757.1B—But the Census Footnote Says It Might Be Zero

In the official press release dated May 14, 2026, the Census Bureau led with an April headline gain of +0.5% month over month to $757.1 billion. Then, in an asterisked note, it quietly conceded that “the 90% confidence interval includes zero.” Translation: the reported increase isn’t statistically distinguishable from no change at all. Meanwhile, March’s initially reported +1.7% was revised to +1.6%, gently sanding down the prior month’s momentum. The release also spotlights +4.9% YoY for total sales and +5.2% YoY for retail trade—but repeats that these figures are “not for price changes,” meaning they’re nominal and potentially inflated by inflation itself. Compositionally, nonstore retailers surged +11.1% YoY while food services and drinking places rose a modest +2.7% YoY, hinting at a consumer pivot toward online channels as restaurants lag.

Here’s what the data reveals:
- April’s +0.5% MoM gain is statistically indistinguishable from zero (per the May 14 footnote).
- March’s growth was revised down to +1.6% from +1.7%, a softening the headline downplays.
- Year-over-year gains—+4.9% total, +5.2% retail trade—are nominal and unadjusted for prices.
- Composition matters: nonstore +11.1% YoY vs food services +2.7% YoY signals uneven demand.
- A Special Notice flags delayed annual benchmarking due to the ARTS → AIES transition, raising revision risk.

The 0.5% That Might Be 0%: Significance Matters

The headline says April sales rose +0.5%. The fine print—on May 14, 2026—admits the 90% confidence interval includes zero. That’s not a trivial caveat; it’s a statistical veto on any confident “up” narrative. Treat April as flat until revisions or additional months confirm direction.

Compounding the caution, March’s momentum was trimmed from +1.7% to +1.6%. Not a cliff, but a pattern: the Census frequently revises initial monthly prints—often softening prior exuberance. If you’re trading the flash, budget for the fade.

Nominal Glow, Real Shadow

The release splashes +4.9% YoY for total sales and +5.2% YoY for retail trade. Then it reminds you—multiple times—that estimates are “not for price changes.” Without deflators, we can’t separate volume from price. If goods inflation ran hot in pockets (think gasoline, vehicles, or parts), nominal growth could be doing more price work than real-demand work.

From February to April, sales rose +4.4% YoY (nominal). That’s not a reacceleration story so much as a steady nominal pace, broadly consistent with the August 2025 release’s +5.0% YoY. The through-line: stable nominal, murky real.

Category Spin: E‑Commerce Carries, Restaurants Lag

The narrative highlights nonstore retailers up +11.1% YoY, while food services and drinking places are up +2.7% YoY. That spread matters:
- Strong online points to resilient discretionary spend—or aggressive pricing/promo dynamics—without telling us about unit volumes.
- Restaurants at +2.7% YoY is soft against nominal baselines, suggesting real food-away-from-home demand is tightening once you account for price increases in the sector.
- Volatile heavyweights—motor vehicles and gasoline stations—are under-discussed in the text, despite their ability to swing monthly totals via pricing more than volume. With data “not for price changes,” gasoline especially can distort the topline.

The net effect: the headline stability is being held up by online channels, while on-the-ground services tied to dining look cooler. For equities, that’s a tilt toward efficient, scale e-commerce models and away from traffic-sensitive restaurant operators.

Data Quality: More Noise, More Revisions, More Caveats

The May 14 Special Notice matters. The Annual Retail Trade Survey has transitioned to the Annual Integrated Economic Survey (AIES), and the Monthly Retail Trade Survey Annual Revision Report is delayed. Translation: benchmarking uncertainty just went up, and backward revisions may be larger or arrive later than usual.

Methodology in brief (from the Explanatory Notes):
- The advance estimate uses a subsample of roughly 4,800 employer firms with a link relative estimator.
- For most nonrespondents, imputation is not performed; for a few influential nonresponders, values may be estimated based on historical performance. That design can amplify month-to-month noise and revision risk.
- The universe changed with the benchmark released in April 2025: estimates now include only businesses with paid employees. Previously, nonemployers were included. That structural break complicates time-series comparisons that span the change.
- Some sub-industries are flagged with “(S)” or “(*)” for suppressed or unavailable advance detail—another signal of sampling variability.
- The release does not provide the GDP-relevant “control group” that strips out autos, gasoline, building materials, and food services. Without it, mapping April retail to real-time PCE-goods is guesswork.

In short: this is a useful early read, but it’s a noisy one—especially now.

Historical Echoes, Not Acceleration

Compare the current +4.9% YoY (nominal) to the August 2025 release’s +5.0% YoY. You’re looking at near-identical nominal growth. The narrative drift leans optimistic month to month, but the long arc is stability, not breakout. Both periods also acknowledge revisions—another reminder to discount first prints and watch the second and third.

What This Means for Markets

• Rates and duration: A statistically flimsy +0.5% MoM with modest downward revision argues against a consumer re-acceleration narrative. For Treasuries, that leans slightly dovish at the margin—especially if upcoming deflators show softer real goods demand.

• Equities—retail bifurcation: Nonstore +11.1% YoY supports scale e-commerce and logistics networks. Restaurants +2.7% YoY implies tighter traffic and potential margin pressure as wage and input costs bite. Favor asset-light online models over full-service dining until traffic turns.

• Cyclical read-through: Without the control group, we can’t cleanly infer PCE-goods. But the nominal steadiness plus restaurant softness tilts the consumption mix toward goods/services that can discount or scale digitally. Autos and gasoline remain wildcards—price-led swings could whipsaw monthly prints.

• Positioning and hedges:

What to Watch Next
- The delayed MRTS Annual Revision Report (AIES transition): scope and timing of rebenchmarks that could rewrite recent momentum.
- CPI/PCE goods deflators: how much of the +4.9% YoY is price versus volume.
- The retail “control group” when available elsewhere: crucial for GDP tracking.
- Subsequent revisions to March and April: whether the “up 0.5%” resolves into flat or better.

The headline says “up.” The footnote says “maybe not.” With benchmarking delays, nominal framing, and category skew doing the heavy lifting, the smart trade is restraint: lean into e-commerce strength, stay cautious on restaurants, and let the revisions tell you whether April was a pause—or just statistical static.