April Retail Says +0.5%—But the 90% Confidence Interval Includes Zero; Gasoline Did the Heavy Lifting

In the official press release dated May 14, 2026, April U.S. retail and food services sales rose 0.5% m/m (±0.4%) and 4.9% y/y (±0.5%) to $757.1 billion. The same page carries a starred note that “the 90 percent confidence interval includes zero,” implying April’s change may not be statistically different from flat. That directly muddies the takeaway from the headline ±0.4% band.

Only one revision is explicitly disclosed: March’s monthly gain was trimmed to +1.6% (±0.2%) from +1.7% (±0.4%)—a downward tweak that slightly lowers the carry into April.

Crucially, sales are “not for price changes.” The release highlights year-over-year strength—nonstore retailers +11.1% y/y, total retail trade +5.2% y/y—but omits clear category-level declines visible in the tables and the fact that gasoline stations’ surge likely reflects price effects, not volume.

Here’s what the data reveals:
- The 0.5% m/m headline comes with a starred caveat: the 90% CI includes zero, undermining claims of a definitive April rise.
- March was revised down to +1.6%, a small but directionally important step lower.
- Gasoline stations up sharply y/y—April 2026 at $61,512 million vs $50,887 million in April 2025—suggesting price-driven “strength” in nominal totals.
- Nonstore retailers +11.1% y/y stands out, but autos (-1.2% y/y), department stores (-1.2%), furniture (-3.6%), and pharmacies (-1.3%) are down.
- Food services and drinking places +2.7% y/y (±1.8%)—muted growth with a wide uncertainty band, at odds with a “broad strength” narrative.

Statistical Fog: A 0.5% That Might Be 0.0%

If April is truly +0.5% m/m with a ±0.4% sampling error, the implied 90% interval should be above zero. The starred note that the 90% CI includes zero sends the opposite message. At minimum, communication is inconsistent; at worst, the move may not be statistically distinguishable from flat. In a market leaning heavily on month-to-month momentum, that nuance matters.

The only disclosed revision—March nudged down to +1.6%—slightly lowers the base for April. On the margins, that direction matters: investors typically fade a sequence of softer revisions more than they cheer equally small upward ones.

Nominal Mirage: Gasoline’s Price Tailwind Is Doing Real Work

April’s tables make clear that gasoline stations rose to $61,512 million, up from $59,852 million in March and far above $50,887 million last April. That’s roughly +2.8% m/m and +20.9% y/y. Since figures are nominal and not price-adjusted, a meaningful piece of April’s “strength” likely came from prices at the pump rather than additional gallons sold.

This is exactly the kind of distortion that can inflate headline retail without signaling stronger real consumption. It also means that if fuel prices ease into late Q2, some of this nominal tailwind could reverse, even with steady real demand. Conversely, persistent energy firmness would keep flattering the topline while pressuring real spending elsewhere.

Winners and Quiet Losers Behind the Nonstore Boom

The narrative spotlights nonstore retailers +11.1% y/y, and it should: April printed $137,558 million, up from $136,011 million in March and well above $123,854 million last year. But the same tables show softness elsewhere that went unmentioned:

• Motor vehicle & parts dealers: $139,229 million in April, below $139,814 million in March and $140,863 million in April 2025 (-1.2% y/y). Autos remain heavy.

• Department stores: $3,187 million in April vs $3,226 million last year (-1.2% y/y).

• Pharmacies & drug stores: $33,216 million vs $33,665 million in April 2025 (-1.3% y/y).

• Furniture & home furnishings: $11,082 million vs $11,495 million last year (-3.6% y/y).

Meanwhile, the ex-motor-vehicle-and-gasoline measure ticked up to $556,344 million from $553,704 million—a directional positive that helps temper the gasoline distortion, but not enough to erase the mixed category picture.

Food services and drinking places climbed to $100,970 million from $100,330 million in March, but the release itself notes only +2.7% y/y (±1.8%)—a small real gain once menu-price inflation is accounted for. That’s not the soundtrack of a demand boom.

Selected Categories: Direction and Scale

All figures are nominal and not adjusted for price changes.

Methodology Matters: More Revision Risk, Less Clarity

The April release is upfront about data quality—and investors should take it seriously.

• The advance estimate is built from a ~4,800-firm subsample using a link-relative estimator, with no imputation for most nonrespondents. “Influential” nonrespondents are filled using historical behavior, which can miss turning points. Translation: revisions are the rule, not the exception.

• A structural break landed in April 2025: the survey now covers only businesses with paid employees. Prior vintage series included nonemployers. Cross-benchmark comparisons are imperfect and require caution.

• The ARTS-to-AIES transition delays the Annual Revision Report for MRTS. With benchmarking lagged, seasonal factors and level estimates are less reliable for longer—raising the odds that today’s story looks different after fuller MRTS data and annual integration.

• Quality flags (S, NA, and “() advance estimates not available”) and a stray formatting artifact in the nonstore line are a reminder: granular transparency is uneven*.

• The release shows ex-motor-vehicle-and-gasoline totals but does not provide explicit revision detail for that GDP-relevant proxy. With GDP hinging on “control group” retail (a related but narrower concept), the absence of a clean, revised control figure leaves a gap for macro modelers.

In other words, the May 14 update keeps the headline neat and the caveats dense. For positioning, the caveats matter more.

What This Means for Markets

• Equities:

• Rates and FX:

• Commodities and energy equities:

• Macro modeling and GDP nowcasts:

The Investor Takeaway

• Fade the headline gloss. A reported +0.5% m/m that comes with a note saying the 90% CI includes zero is not the stuff of “reacceleration.”

• Separate nominal from real. Gasoline’s +20.9% y/y is doing disproportionate work; real goods demand looks flatter once fuel is stripped out.

• Position for dispersion, not a broad retail lift:

• Keep powder dry for revisions. With a ~4,800-firm subsample, minimal imputation, a structural coverage break, and delayed annual benchmarks, the revision tape is far from over.

April’s retail print tells two stories: a headline that leans positive and a footnote that questions it. The market will cheer the former on first read. The better trade is to prepare for the latter—and own the parts of retail that don’t rely on gasoline or wishful thinking to look good.