Will US per-capita real grocery spending decline year-over-year in any 2026 quarterly BLS Consumer Expenditure release?

The macro math remains unchanged: 16% adoption * 5.3% reduction = ~0.85% aggregate drag on food-at-home spending. But the new basket composition data introduces a significant offset. If GLP-1 users shift to healthier, pricier items with baskets 'larger by a double digit percentage,' the per-capita spending effect could be neutral or even positive in dollar terms, even as caloric consumption falls. The BLS Consumer Expenditure Survey measures dollar spending, not caloric intake. This basket composition offset weakens the core mechanism by which GLP-1 would drive a spending decline. The 4-quarter resolution window remains the primary driver of non-trivial probability, as macro headwinds (food deflation, mild recession) could independently cause a quarterly decline.

Management's 'wash' framing applies to Walmart specifically, not to aggregate BLS data. The basket composition offset may hold for Walmart's mix (which skews toward fresh food and higher-income customers) but the broader grocery market includes convenience stores, discount grocers, and households where GLP-1 users may simply buy less without trading up. The aggregate BLS effect is a weighted average across all shopping behaviors. Additionally, the 'double digit percentage' larger baskets could reflect frequency reduction -- fewer trips but larger baskets -- which may still net to lower total spending. The macro uncertainty for 2026 remains: tariff concerns, potential consumer spending deceleration, and food-at-home CPI normalization could compound with GLP-1's marginal drag. Previous assessment of 20-30% base rate for at least one quarterly decline across 4 quarters remains valid.

The basket composition offset is the most material new information since the initial prediction. If GLP-1 users are spending more per shopping trip on healthier/pricier items, the net effect on dollar-denominated per-capita grocery spending could actually be positive, not negative. This fundamentally undermines the core thesis of this market -- that GLP-1 reduces per-capita grocery spending. The remaining probability comes almost entirely from non-GLP-1 macro factors: food deflation, recession risk, or consumer retrenchment. The committee noted that GLP-1 adoption data is now 20 months stale, which adds uncertainty, but the direction of new evidence (basket composition offset) cuts against the YES case. Reducing from 0.20 to 0.16 to reflect the weakened mechanism.

The population-level effect was already estimated at only ~0.85% aggregate drag, and now the basket composition offset suggests even that estimate may be too high. GLP-1 users shifting to healthier, pricier foods could mean per-capita dollar spending remains flat or increases despite reduced food volume. The BLS Consumer Expenditure Survey tracks nominal and real dollar spending, not physical quantities. This means the key mechanism (GLP-1 reduces grocery spending) is weaker than previously assessed. The 4-quarter window gives multiple chances to catch a macro-driven decline, but the structural GLP-1 contribution to that probability is now diminished. Base rate for at least one quarterly real decline across 4 quarters is perhaps 15-25% in normal macro conditions.

The real vs. nominal distinction remains important but now cuts less clearly toward YES. If GLP-1 users are buying pricier items, this increases nominal spending per capita. For real spending to decline, food-at-home CPI inflation would need to outpace nominal spending growth. With food-at-home CPI decelerating from the 2022-2023 surge but still positive, and nominal spending potentially supported by the GLP-1 basket composition offset, the likelihood of real per-capita declines is lower than previously estimated. The market now hinges almost entirely on whether broader macro forces (recession, food deflation, tariff impacts on consumer behavior) create a quarterly decline -- GLP-1 is effectively a neutral factor rather than a headwind for dollar-denominated spending.

Calibrating against the previous aggregate of 0.21, the new information (basket composition offset) should modestly reduce probability. However, several countervailing factors limit the downward revision: (1) GLP-1 adoption has likely continued rising since the stale mid-2024 data, potentially from 16% to 18-20% of households; (2) the committee flagged that management may underweight long-term structural effects -- the 'wash' may not hold at higher adoption rates; (3) the 2026 macro environment includes tariff uncertainty that could dampen consumer spending broadly. A modest 1pp reduction from 0.21 to 0.20 reflects the net of basket offset (downward) vs. rising adoption and macro uncertainty (upward). The key question remains whether any single quarter catches a decline from any cause.

The basket composition offset weakens the GLP-1 mechanism for reducing dollar-denominated grocery spending. CFO's 'wash' characterization, while potentially self-serving, aligns with the logic that healthier/pricier items could maintain spending levels. The 0.85% aggregate drag estimate was already marginal; with the composition offset, the effective GLP-1 contribution to per-capita spending decline may be near zero. Remaining probability reflects base rate for macro-driven quarterly declines across 4 quarters.

Risk-weighted assessment: the downside risk (YES outcome) is now lower because the primary mechanism (GLP-1 reducing grocery dollar spending) has been partially invalidated by basket composition data. The remaining risk scenarios are almost entirely macro-driven: US recession, food deflation from commodity price collapse, or tariff-induced consumer retrenchment. None of these are high-probability in 2026 baseline forecasts. Historical base rate for real per-capita food-at-home spending decline in any single quarter is roughly 4-6%, giving a cumulative ~15-22% over 4 quarters. With GLP-1 no longer a meaningful contributor, anchoring to the lower end of this range.

Evidence-weighted: the strongest evidence (peer-reviewed Cornell/JMR study) showed 5.3% household grocery spending reduction. But the new Walmart CFO commentary provides an important qualifier -- GLP-1 users shift to healthier, pricier items. The study measured total grocery spending, so the 5.3% reduction already accounts for any composition shift in its sample. However, the Walmart-specific observation of 'double digit percentage' larger baskets suggests the composition shift may be more pronounced at retailers with strong fresh/healthy offerings. The BLS aggregate may still see some net reduction, but likely smaller than the 5.3% per-household figure implies. Effective aggregate drag may be 0.4-0.6% rather than 0.85%, which is definitively too small to register in BLS data against normal growth rates.