Grocery Market Navigator — Executive Summary

Supply Chain Outlook

Sales
The next six months remain well below the previous twelve. Sales TTM YOY% was 2.0%, but the six-month forecast is only 0.7%, down 20 basis points from last month's 0.9%. The decline is not a demand story, it is an inflation story, since lower expected food inflation mechanically pulls dollar growth down. Against the 3% long-run trend this is a weak number, and it sits in the bottom decile of everything we have seen since 2017, at roughly the 5th percentile. The distribution puts the median at 0.4% with a range of -0.3% to 1.2%, so even the optimistic end of this forecast is a soft market.

Units
Units are expected to average -1.9% over the next six months, a 10 basis point improvement over last month's -2.0% and the only place the lower inflation forecast shows up as good news. That improvement is marginal against a TTM of -0.5%, so the trend is still clearly deteriorating. The path bottoms at -2.5% in August before easing to -1.2% by December, and the distribution is tight, with a median of -2.1% and a range of -2.6% to -1.6%. The model is confident about the direction here, and the direction is down.

CPI FAH
Food inflation is expected to average 2.6% over the next six months, down from the 3.0% June forecast and close to the 2.5% TTM. The path peaks near 3.1% in August and then eases to about 2.2% by December as the lower oil price outlook works through. This is the single most important change in this month's forecast, because CPI FAH is what connects the energy story to both the unit forecast and the sales forecast, and it moved in opposite directions on those two lines.

Farm and Food Manufacturing Prices
Farm products are the most energy-exposed link in the chain and they responded accordingly, with the six-month forecast dropping from 6.2% last month to 3.9%. The caveat is that the farm distribution remains extremely wide, with a median of 5.4% and a range of 0.4% to 11.9%, so this is a lower forecast rather than a confident one. Food manufacturing barely moved, going from 3.8% to 3.9%, because manufacturing wages are the dominant input in that model and wages do not fall when oil does. Its distribution is tight by comparison, a median of 4.0% in a 2.8% to 5.1% range. Both forecasts are also flattered by the deflation of late 2025 and early 2026, which means these year-over-year readings are measured against a depressed base.

Next Six Months
The stabilization in energy prices is real and it is showing up in the upstream forecasts, but it arrives with the supply chain already repriced and the consumer already stretched. Actuals for the last three months were 2.3%, 3.0% and 2.2% sales growth with units at -0.7%, 0.3% and -0.5%, and the model expects both to give that ground back. Soft real disposable income, decelerating home prices, and continued SNAP declines running near -13% are all weighing on volume, and there is no offsetting driver of any size in the model. The forecast of 0.7% sales comes from 2.6% food inflation against -1.9% units, which is the same story as last month with slightly less inflation carrying it. Holding 1.5% sales growth for 2026 now looks optimistic.

Retailer Margins and a Strained Consumer
The cost pressure did not disappear, it moved. Retailers are absorbing it, with the first six grocery retailers reporting CY 2026 Q2 at 1.6% EBIT versus a long-run trend near 4.0%, and that is not a position anyone can hold indefinitely. On the other side, the June and July economic data point the wrong way almost everywhere: the savings rate fell from 3.0% to 2.7%, retail sales dropped 0.6% in July for the steepest decline since May 2025, consumer sentiment fell about 8% to its lowest reading since regular tracking began in the 1970s, and credit card balances are up alongside rising delinquencies in cards, autos and mortgages. Most of that deterioration is concentrated in middle and lower income households, which is exactly where grocery is a fixed line in the budget. If retailers stop eating margin before the consumer recovers, the inflation relief in this forecast reaches the shelf later and smaller than the model assumes.

Causal Chain (TTM | 6-Mo Fcst)

↑

↑

↑

PPI Food Manufacturing

↑

OUTLOOK

PPI Farm Products – 6 Month YOY Outlook

+2.6% | +3.9%

ttm | 6-mo fcst

Before the conflict, farm product prices trended sharply lower peaking at 15.4% YOY in Feb-2025 and bottoming at -6.9% YOY in Jan-2026. As expected farm product prices have started to rebound, hitting 5.7% YOY growth in May-2026 before easing to 1.9% in Jun-2026. The pre-conflict deflation was driven by declining oil prices and falling food import prices but both of are now reversing.

Over the six-month forecast window, PPI Farm Products is projected to average +3.9% year-over-year, a meaningful step up from the trailing twelve-month average of +2.6%, with oil prices acting as the single dominant driver. Crude oil, averaging roughly +22.9% above year-ago levels across the period, contributes +2.82 percentage points to that forecast — the clearest transmission of energy costs into farm input expenses, from diesel-powered equipment and irrigation to fertilizer feedstocks. Food import prices, running +2.1% higher on average, add a secondary but still meaningful +1.52 percentage points, reflecting broader global commodity pressures flowing into domestic farm-level pricing. Natural gas, averaging -2.2% below year-ago levels, provides a modest partial offset of -0.17 percentage points, though its dampening effect is too small to materially alter the directional picture. Critically, the +3.9% forecast average understates the underlying pricing pressure: PPI Farm Products peaked at +8.5% year-over-year in October 2025 before collapsing to -6.9% in January 2026, and that prior deflationary trough is now acting as a low comparison base that cushions the year-over-year readings — absent that deflation episode, the same level of farm prices would be generating meaningfully higher reported inflation numbers than executives will see in the headline figures.

Forecast Distribution Simulation

This distribution is built on 1,000 simulations using holdout errors from the rolling validation period. Small differences between the point forecast and the simulated median may occur as a result.

The median forecast is +5.4% with a forecast distribution of +/- 5.8%. We would expect Ppi Farm Products to fall between +0.4% and +11.9% over the next six months.

INPUTS

PPI Farm Products – Inputs

Forecast Decomposition — 6-Month Average Contribution

Variable	Avg Input (YoY%)	Contribution (ppts)	% of Total
Oil Prices	+22.9%	▲ 2.82 ppts	+63%
Food Import Prices	+2.1%	▲ 1.52 ppts	+34%
Natural Gas Prices	-2.2%	▼ 0.17 ppts	-4%
Structural Offset (model constant)	+1.0%	▼ 0.29 ppts
Total Forecast		▲ 3.88 ppts	100%

Oil Prices (Lead 1)

Oil prices are forecast to average +22.9% year-over-year over the next six months, a level that continues to climb and that sits at the root of rising farm input costs across the supply chain. That single driver is projected to contribute +2.82 percentage points to PPI Farm Products inflation, accounting for 63% of total input cost movement in the forecast period.

Natural Gas Prices (Lead 2)

Natural gas prices are forecast to average -2.2% year-over-year over the next six months, a trajectory that continues to pull in a deflationary direction. This decline subtracts an estimated 0.17 percentage points from farm input cost pressure, accounting for roughly 4% of total input movement — a modest but measurable drag operating beneath oil's far more dominant influence on the overall cost picture.

Food Import Prices

Food import prices are expected to average +2.1% year-over-year over the next six months, with the trend continuing to move higher. This contributes an estimated +1.52 percentage points to the overall PPI Farm Products forecast, accounting for 34% of total input movement in the period.

MODEL PERFORMANCE

PPI Farm Products – Model Performance

24-Month Holdout

Over the 24-month holdout period, the model produced a mean absolute error of 6.5 percentage points, meaning that on any given month the forecast was off by that amount on average relative to the actual PPI Farm Products reading. For a series as volatile as farm-level prices, that margin is meaningful — it reflects genuine difficulty in pinning down the exact timing and magnitude of price swings, not a model misspecification.

Holdout Errors Over Time

Rolling errors ranged from 2.9 to 11.1 percentage points across the holdout window, with the widest misses clustering around January 2025. That peak in error is consistent with what happens when this model's key inputs — oil prices and natural gas prices — move sharply or reverse direction in a short window, since fast-moving inputs compress the lead-lag relationships the model is built around and make precise near-term forecasting harder.

OLS Regression Results
R-squared: 0.668			Adj. R-sq: 0.658
F-statistic: 65.70			Prob(F): 2.28e-23
No. Observations: 102			Df Residuals: 98
Variable	Coef	Std Err	t	P>\|t\|	[0.025, 0.975]
const	-0.2775	0.990	-0.280	0.780	[-2.242, 1.687]
oil_prices_lead1_yoy	0.0962	0.025	3.841	0.000	[0.046, 0.146]
nat_gas_prices_lead2_yoy	0.1115	0.021	5.435	0.000	[0.071, 0.152]
import_index_yoy	0.6075	0.255	2.380	0.019	[0.101, 1.114]

Omnibus: 0.263		Prob(Omnibus): 0.877		Durbin-Watson: 0.426
Skew: 0.037		Kurtosis: 3.078

Model Performance

The model's R-squared of 0.668 means it explains roughly two-thirds of the variation in PPI Farm Products over the estimation sample, a reasonable result for a noisy, commodity-driven series. All three inputs are statistically significant: oil prices carry a t-statistic of 3.841, natural gas prices 5.435, and the import price index 2.380 — each comfortably above the 2.0 threshold that signals a reliable, non-random relationship. The Omnibus probability of 0.877 sits well above 0.05, confirming that the model's residuals are normally distributed and that no systematic pattern is being left on the table. The Durbin-Watson statistic of 0.426, however, is well below the ideal value near 2.0, which flags positive autocorrelation in the residuals — meaning consecutive errors tend to run in the same direction, a limitation worth monitoring when interpreting short-run forecasts.

OUTLOOK

PPI Food Manufacturing – 6 Month YOY Outlook

+2.6% | +3.9%

ttm | 6-mo fcst

Prior to the conflict, food manufacturing prices followed farm products lower. PPI Food Manufacturing peaked at 4.9% YOY in Sep-2025 and dropped to 0.8% by Feb-2026, driven by falling farm input costs, declining oil prices, softer manufacturing wages, and weak food import prices. But food manufacturing prices are reversing course as expected but not as significantly as farm products.

Over the next six months, PPI Food Manufacturing is forecast to average +3.9% year-over-year, a meaningful step up from the +2.6% recorded over the trailing twelve months. The single largest driver is manufacturing wages, which are running at an average of +4.6% and contributing 4.37 percentage points to the forecast — reflecting sustained labor cost pressure that food processors have been unable to fully absorb. Oil prices, averaging a sharp +36.7% increase over the period, add another 1.05 percentage points as elevated crude costs flow through to packaging, transportation, and energy-intensive processing operations, while rising PPI Farm Products contribute an additional 0.75 percentage points. A structural offset of -2.33 percentage points embedded in the model partially dampens these pressures, preventing the headline forecast from climbing further. Critically, even this +3.9% average understates the true pricing momentum in the system: PPI Food Manufacturing peaked at +4.9% in September 2025 before deflating to a trough of just +0.8% in February 2026, and the current rebound is now being measured against that depressed base — meaning the low comparison period is acting as a statistical cushion that makes a genuine and accelerating price recovery look far more modest than it actually is in level terms.

Forecast Distribution Simulation

This distribution is built on 1,000 simulations using holdout errors from the rolling validation period. Small differences between the point forecast and the simulated median may occur as a result.

The median forecast is +4.0% with a forecast distribution of +/- 1.1%. We would expect Ppi Food Mfg to fall between +2.8% and +5.1% over the next six months.

INPUTS

PPI Food Manufacturing – Inputs

Forecast Decomposition — 6-Month Average Contribution

Variable	Avg Input (YoY%)	Contribution (ppts)	% of Total
Manufacturing Wages	+4.6%	▲ 4.37 ppts	+70%
Oil Prices	+36.7%	▲ 1.05 ppts	+17%
PPI Farm Products	+4.0%	▲ 0.75 ppts	+12%
Food Import Prices	+0.1%	▲ 0.03 ppts	+0%
Structural Offset (model constant)	+1.0%	▼ 2.33 ppts
Total Forecast		▲ 3.87 ppts	100%

PPI Farm Products (Lead 1)

PPI Farm Products is forecast to average +4.0% year-over-year over the next six months, with the trajectory continuing to climb. That gain contributes roughly +0.75 percentage points to PPI Food Manufacturing — representing 12% of total input movement — though executives should note that this figure is held down by the prior deflation trough, meaning the true pricing pressure building through the farm supply chain is more intense than the year-over-year reading suggests.

Food Import Prices (Lead 3)

Food import prices are averaging +0.1% year-over-year across the forecast period, a figure that is drifting modestly higher but one that, given the V-shape in PPI Food Manufacturing — which peaked at 4.9% in September 2025 before bottoming at 0.8% in February 2026 — understates the true upward pressure, since that prior deflation trough is acting as a low comparison base that makes the current rebound appear far more contained than it actually is in level terms. Food import prices contribute +0.03 percentage points to the PPI Food Manufacturing forecast, representing 0% of total input movement and confirming that while the directional signal is additive, this channel is not a

Manufacturing Wages (Lead 1)

Manufacturing wages are forecast to average +4.6% year-over-year over the next six months, sustaining an upward trajectory that keeps persistent pressure on food production costs. This single input is expected to contribute +4.37 percentage points to the PPI Food Manufacturing forecast — representing 70% of total input movement — though that figure almost certainly understates the true pricing pressure, since it is measured against the depressed 0.8% trough of February 2026, a low base created by the prior deflation that is acting as a cushion on the year-over-year reading and would otherwise produce a materially higher headline number.

Oil Prices (Lead 4)

West Texas Intermediate crude is forecast to average +36.7% year-over-year over the next six months, a trajectory that continues to climb and feed directly into food manufacturing input costs. That oil price surge is projected to contribute +1.05 percentage points to PPI Food Manufacturing inflation, accounting for 17% of total input cost movement in the forecast period.

MODEL PERFORMANCE

PPI Food Manufacturing – Model Performance

24-Month Holdout

Over the 24-month holdout period, the model produced a mean absolute error of 1.5 percentage points, meaning the average forecast landed within roughly one and a half points of the actual PPI Food Manufacturing reading. For executives tracking margin exposure, that translates to a forecast that is directionally reliable and useful for planning, though individual months can deviate more in either direction during periods of rapid input price movement.

Holdout Errors Over Time

Rolling errors ranged from a low of 0.5 percentage points to a peak of 2.7 percentage points, with the largest misses clustering around July 2024. That timing is consistent with what this model's inputs would predict — oil prices, farm product prices, import costs, and manufacturing wages were all moving sharply and at different speeds during that stretch, and when multiple fast-moving inputs shift simultaneously, even a well-specified model will struggle to track the combined effect with precision.

OLS Regression Results
R-squared: 0.897			Adj. R-sq: 0.892
F-statistic: 167.98			Prob(F): 7.51e-46
No. Observations: 102			Df Residuals: 96
Variable	Coef	Std Err	t	P>\|t\|	[0.025, 0.975]
const	-2.6738	0.598	-4.474	0.000	[-3.860, -1.488]
ppi_farm_products_lead1_yoy	0.1803	0.016	11.533	0.000	[0.149, 0.211]
import_index_lead3_yoy	0.2620	0.061	4.267	0.000	[0.140, 0.384]
mfg_wages_lead1_yoy	1.0296	0.168	6.112	0.000	[0.695, 1.364]
oil_prices_lead4_yoy	0.0270	0.005	5.575	0.000	[0.017, 0.037]
covid	2.3996	0.623	3.852	0.000	[1.163, 3.636]

Omnibus: 5.118		Prob(Omnibus): 0.077		Durbin-Watson: 1.368
Skew: 0.505		Kurtosis: 3.244

Model Performance

The model's R-squared of 0.897 means it explains approximately 90 percent of the historical variation in PPI Food Manufacturing, which is a strong fit for a macroeconomic forecasting model of this type. Every input — farm product prices, import prices, manufacturing wages, and oil prices — carries a t-statistic well above 2.0, confirming that each variable is individually significant and not a statistical accident. The Omnibus probability of 0.077 sits above the 0.05 threshold, indicating that the model's residuals are approximately normally distributed and no systematic bias is distorting the forecasts. The Durbin-Watson statistic of 1.368 is somewhat below the ideal value of 2.0, suggesting a mild degree of positive autocorrelation in the residuals — meaning consecutive errors tend to lean in the same direction, which is worth monitoring but does not undermine the model's overall reliability.

OUTLOOK

CPI Food-at-Home – 6 Month YOY Outlook

+2.5% | +2.6%

ttm | 6-mo fcst

Food inflation was also softening ahead of the conflict peaking at 2.7% YOY in Aug-2025, and hitting 2.0% in Mar-2026. That pre-conflict decline was driven by falling farm and food manufacturing costs working through the supply chain. The other key drivers — PPI Grocery Retail, stable around 3.0% YOY, and retail wages, trending up since Jan-2025 — were providing some upward momentum. The result was a slow, measured decline heading into the conflict. But we are starting to see, the recent higher farm and food manufacturing prices are starting to push food inflation higher.

Over the six-month forecast period, CPI Food-at-Home is projected to average +2.7% year-over-year, a modest step up from the trailing twelve-month average of +2.5%, with the path running from +2.8% in July to a softer +2.2% by December 2026. The single largest contributor is PPI Grocery Retail, averaging +4.3% and adding roughly 1.70 percentage points to the forecast, reflecting the upstream cost pressures that have built through the supply chain even as they have been slow to reach shelf prices. PPI Food Manufacturing, averaging +3.9% and contributing 1.29 percentage points, reinforces that pipeline pressure, while retail wages averaging +3.2% add a further 1.48 percentage points — though it is worth noting that slowing wage growth relative to earlier in the cycle is one reason CPI Food-at-Home has held relatively stable despite meaningful upstream acceleration. Critically, a structural offset of -1.82 percentage points is suppressing the headline YoY reading, meaning that without this drag — which captures margin compression from grocery retailers absorbing costs rather than passing them through — the forecast average would be closer to +4.5%, a level that would register as a more serious inflation episode for executives planning procurement and pricing strategy.

Forecast Distribution Simulation

This distribution is built on 1,000 simulations using holdout errors from the rolling validation period. Small differences between the point forecast and the simulated median may occur as a result.

The median forecast is +2.8% with a forecast distribution of +/- 0.6%. We would expect Cpi Fah to fall between +2.2% and +3.5% over the next six months.

INPUTS

CPI Food-at-Home – Inputs

Forecast Decomposition — 6-Month Average Contribution

Variable	Avg Input (YoY%)	Contribution (ppts)	% of Total
PPI Grocery Retail	+4.3%	▲ 1.70 ppts	+38%
Retail Wages	+3.2%	▲ 1.48 ppts	+33%
PPI Food Manufacturing	+3.9%	▲ 1.29 ppts	+29%
Structural Offset (model constant)	+1.0%	▼ 1.82 ppts
Total Forecast		▲ 2.65 ppts	100%

PPI Food Manufacturing (Lead 1)

PPI Food Manufacturing is forecast to average +3.9% year-over-year over the next six months, with the index trending higher as upstream oil-driven cost pressures continue to work their way through the supply chain. That acceleration contributes an estimated +1.29 percentage points to the CPI Food-at-Home forecast, representing 29% of total input movement — though its full pass-through to shelf prices is being partially absorbed by slowing retail wage growth and margin compression at the grocery level.

PPI Grocery Retail

PPI Grocery Retail is forecast to average +4.3% year-over-year over the next six months, with the index trending higher as upstream oil-driven cost pressure increasingly passes through to the retail shelf. That acceleration contributes an estimated +1.70 percentage points to the CPI Food-at-Home forecast, accounting for 38% of total input movement across the supply chain.

Retail Wages (Lead 3)

Retail wage growth is forecast to average +3.2% year-over-year over the next six months, holding at a pace that continues to add meaningful cost pressure to grocery operations. This wage growth contributes an estimated +1.48 percentage points to the CPI Food-at-Home forecast, accounting for 33% of total input movement and making it the single largest moderating-yet-additive force in the outlook.

MODEL PERFORMANCE

CPI Food-at-Home – Model Performance

24-Month Holdout

Over the 24-month holdout period, the model produced a mean absolute error of 0.8 percentage points, meaning that on average its CPI Food-at-Home forecasts landed within roughly four-fifths of a percentage point of the actual reading. For grocery executives using this output to plan pricing strategy or negotiate supplier contracts, that margin is tight enough to be operationally useful — a miss of that size rarely changes the directional call or the magnitude of the pricing signal.

Holdout Errors Over Time

Rolling errors across the holdout ranged from 0.4 to 1.1 percentage points, with the widest misses clustering around January 2026. That peak in error is consistent with what this model is built to capture: when oil prices move sharply or food manufacturing costs shift quickly, the lagged input structure means the model is partly chasing a fast-moving target, and the one-month lead on PPI food manufacturing and the three-month lead on retail wages can briefly fall out of sync with an accelerating reality.

OLS Regression Results
R-squared: 0.889			Adj. R-sq: 0.884
F-statistic: 194.16			Prob(F): 2.23e-45
No. Observations: 102			Df Residuals: 97
Variable	Coef	Std Err	t	P>\|t\|	[0.025, 0.975]
const	-2.4112	0.376	-6.413	0.000	[-3.157, -1.665]
ppi_food_mfg_lead1_yoy	0.3574	0.031	11.696	0.000	[0.297, 0.418]
ppi_grocery_yoy	0.5152	0.031	16.635	0.000	[0.454, 0.577]
retail_wages_lead3_yoy	0.3912	0.103	3.801	0.000	[0.187, 0.595]
covid	1.2763	0.518	2.464	0.016	[0.248, 2.305]

Omnibus: 0.774		Prob(Omnibus): 0.679		Durbin-Watson: 1.062
Skew: 0.193		Kurtosis: 2.764

Model Performance

The model's R-squared of 0.889 means it explains nearly 89 percent of the historical variation in CPI Food-at-Home, which is a strong fit for a macroeconomic forecasting model operating at this level of complexity. Every input clears the significance threshold comfortably — all four variables carry t-statistics well above 2.0, with PPI grocery wholesale the strongest signal at 16.6 and retail wages the most modest at 3.8, though still clearly meaningful. The Omnibus probability of 0.679 is well above the 0.05 threshold, confirming that the model's residuals are normally distributed and not systematically biased in any direction. The Durbin-Watson statistic of 1.062 is lower than the ideal reading near 2.0, which suggests some positive autocorrelation remains in the residuals — a signal worth monitoring, as it can indicate the model is occasionally slow to fully absorb momentum in the underlying series.

OUTLOOK

US Grocery Units – 6 Month YOY Outlook

-0.5% | -1.9%

ttm | 6-mo fcst

US Grocery Market Units have been trending down YOY since early 2025 and bottomed out at -2.1% just ahead of the conflict.The last 12 months were down -0.5%, but the first half of the year was much stronger than the last half. The higher CPI FAH number in Apr-2026 helped to keep unit growth down about 0.7% YOY.

Over the six months from July through December 2026, US grocery unit volume is forecast to average -1.9% year-over-year, a meaningful deterioration from the -0.5% recorded over the trailing twelve months and a far cry from the roughly 2% growth seen in late 2024 and early 2025. The dominant force driving that decline is food-at-home inflation, which averages 2.6% across the forecast window and subtracts an estimated 2.14 percentage points from unit growth — a direct consequence of elevated oil prices feeding through farm input and food manufacturing costs into shelf prices that squeeze consumer purchasing power in volume terms. Partially offsetting that pressure, pharmacy revenue growth contributes a modest +0.19 points as grocery retailers with integrated pharmacy operations capture some incremental traffic, while home price appreciation adds another +0.14 points by supporting household balance sheets. SNAP cost declines averaging -12.9% subtract a further 0.18 points, reflecting reduced benefit support for price-sensitive shoppers who tend to be among the most unit-elastic buyers in the channel.

Forecast Distribution Simulation

This distribution is built on 1,000 simulations using holdout errors from the rolling validation period. Small differences between the point forecast and the simulated median may occur as a result.

The median forecast is -2.1% with a forecast distribution of +/- 0.5%. We would expect Units Mkt Trend to fall between -2.6% and -1.6% over the next six months.

INPUTS

US Grocery Units – Inputs

Forecast Decomposition — 6-Month Average Contribution

Variable	Avg Input (YoY%)	Contribution (ppts)	% of Total
CPI Food-at-Home	+2.6%	▼ 2.14 ppts	-79%
Pharmacy Revenue	+2.6%	▲ 0.19 ppts	+7%
Snap Cost	-12.9%	▼ 0.18 ppts	-7%
Home Prices	+0.4%	▲ 0.14 ppts	+5%
Real Disposable Income	+0.1%	▲ 0.05 ppts	+2%
Total Forecast		▼ 1.94 ppts	100%

CPI Food-at-Home

CPI Food-at-Home is forecast to average +2.6% year-over-year over the next six months, with the trajectory continuing to climb from current levels. That acceleration is the dominant drag on grocery unit volume, subtracting 2.14 percentage points from the unit growth forecast and accounting for 79% of total input movement across all tracked drivers.

Real Disposable Income

Real disposable income is expected to average just +0.1% year-over-year over the next six months, a figure that sits barely above flat and is trending sideways with no meaningful acceleration in sight. That thin growth rate contributes a modest +0.05 percentage points to the grocery unit forecast, accounting for roughly 2% of total input movement — a signal that consumer purchasing power is neither a tailwind nor a meaningful drag in the current cycle.

Home Prices

Grocery unit growth tied to home price appreciation is expected to average +0.4% year over year over the next six months, a modest but positive contribution as rising home equity supports consumer spending confidence. This adds approximately +0.14 percentage points to the unit growth forecast, accounting for 5% of total input movement in the model.

MODEL PERFORMANCE

US Grocery Units – Model Performance

24-Month Holdout

The 24-month holdout mean absolute error came in at 0.01 percentage points, which in practical terms means the model's average forecast deviated from actual US grocery unit growth by essentially a rounding margin over the holdout period. For executives using this output to plan inventory or negotiate supplier contracts, that level of precision is operationally meaningful — the model is not just directionally correct but close in magnitude as well.

Holdout Errors Over Time

Rolling errors across the holdout window ranged from 0.4 to 1.1 percentage points, with the widest gaps appearing around January 2026. That peak in forecast error is consistent with what happens when one or more of this model's key inputs — food-at-home CPI, real disposable income, home prices — move sharply or rapidly in a short window, giving the model less stable signal to work from. These are not alarming deviations, but they are a reminder that forecast confidence narrows when the macro environment is moving fast.

OLS Regression Results
R-squared: 0.890			Adj. R-sq: 0.883
F-statistic: 127.47			Prob(F): 3.45e-43
No. Observations: 102			Df Residuals: 95
Variable	Coef	Std Err	t	P>\|t\|	[0.025, 0.975]
const	22.2927	1.151	19.366	0.000	[20.007, 24.578]
cpi_fah_log	-0.8227	0.070	-11.694	0.000	[-0.962, -0.683]
rdi_log	0.3486	0.127	2.739	0.007	[0.096, 0.601]
home_price_log	0.3461	0.071	4.873	0.000	[0.205, 0.487]
snap_cost_log	0.0137	0.010	1.441	0.153	[-0.005, 0.033]
pharm_drug_log	0.1004	0.039	2.549	0.012	[0.022, 0.179]
covid	0.0335	0.007	4.759	0.000	[0.020, 0.047]

Omnibus: 113.269		Prob(Omnibus): 0.000		Durbin-Watson: 2.073
Skew: 3.459		Kurtosis: 28.115

Model Performance

The model's R-squared of 0.89 means it accounts for 89 percent of the observed variation in US grocery unit growth, a strong fit for a consumer demand series subject to behavioral noise. Most inputs clear the significance threshold comfortably — food-at-home CPI carries a t-statistic of -11.7, real disposable income 2.7, home prices 4.9, and pharmaceutical drug spending 2.5, all well above the 2.0 threshold that signals a reliable relationship. The SNAP cost variable is the exception, with a t-statistic of 1.4, meaning its contribution is not statistically distinguishable from noise at conventional confidence levels. The Omnibus probability of 0.000 indicates the residuals are not normally distributed — a flag worth monitoring — while the Durbin-Watson statistic of 2.073 sits close enough to 2.0 to confirm there is no meaningful autocorrelation in the errors, meaning the model is not systematically leaning on its own past mistakes.

OUTLOOK

US Grocery Sales – 6 Month YOY Outlook

+2.0% | +0.7%

ttm | 6-mo fcst

Sales were solid in 2024 and the first eight months of 2025, but the last six months have seen YOY sales drop precipitously. The long-run trend for grocery is about 3% YOY with 2% growth typically coming from CPI FAH and the other 1% from unit growth. CPI FAH reversed course in Apr-2026 which helped Sales growth tick up to 2.3% YOY in April.

Over the next six months, US grocery sales growth is forecast to average just +0.7% year-over-year, a meaningful step down from the +2.0% pace recorded over the trailing twelve months, with the trajectory bottoming near +0.5% in August and September 2026 before a modest recovery to +0.9% by year-end. The dominant driver holding sales in positive territory is food-at-home inflation, which contributes +2.64 percentage points to the forecast as grocery prices continue to reflect elevated oil-driven input costs working through the supply chain into supermarket shelves. Offsetting nearly all of that pricing support is a deterioration in unit volume, which subtracts 1.91 percentage points from the forecast as consumers — squeezed by cumulative price increases — pull back on the number of items they purchase, a pattern consistent with the broader trading-down and trip-consolidation behavior visible since late 2025. The net result is a sales environment that is technically growing but almost entirely price-driven, with real volume contraction quietly eroding the underlying health of the category in ways that top-line dollar figures alone will not reveal to investors.

Forecast Distribution Simulation

This distribution is built on 1,000 simulations using holdout errors from the rolling validation period. Small differences between the point forecast and the simulated median may occur as a result.

The median forecast is +0.4% with a forecast distribution of +/- 0.8%. We would expect Sales Mkt Trend to fall between -0.3% and +1.2% over the next six months.

INPUTS & PERFORMANCE

US Grocery Sales – Inputs and Performance

Forecast Decomposition — 6-Month Average Contribution

Variable	Avg Input (YoY%)	Contribution (ppts)	% of Total
CPI Food-at-Home	+2.6%	▲ 2.64 ppts	+58%
US Grocery Units	-1.9%	▼ 1.91 ppts	-42%
Total Forecast		▲ 0.73 ppts	100%

US Grocery Units

Grocery unit volumes are forecast to average -1.9% year-over-year over the next six months, a contraction that will act as a persistent drag on sales growth throughout the period. This volume weakness is expected to subtract 1.91 percentage points from the sales forecast, accounting for 42% of total input movement and making it the single largest downward force on revenue growth.

CPI Food-at-Home

CPI Food-at-Home is expected to average +2.6% year-over-year across the six-month forecast period, with the index trending modestly higher as oil-driven input costs continue to work through the food supply chain. That inflation rate contributes +2.64 percentage points to grocery sales growth over the period, accounting for 58% of total input movement in the forecast model.

Forecast Errors

US Grocery Sales carries no standalone error statistic because it is a derived model, calculated by multiplying US Grocery Units by the Average Price index (CPI Food-at-Home) rather than estimated directly from data. The forecast distribution characterizes uncertainty in place of a holdout MAE, implying a spread of approximately 0.8 percentage points across the plausible outcome space. At the 90% confidence level, that distribution spans -0.3% to 1.2% for the six-month average year-over-year change.

The median of the forecast distribution sits at 0.4%, meaning half of simulated outcomes fall below that level — notably softer than the 0.7% point forecast, which reflects the influence of higher-probability upside scenarios pulling the central estimate above the midpoint. Practically, the 90% interval from -0.3% to 1.2% tells executives that a mild nominal contraction in grocery sales remains a live possibility, while an outcome above 1% is roughly as plausible as the point forecast itself. The 0.7% point forecast therefore sits in the upper half of the distribution and should be read as an optimistic central case rather than a conservative one.

Because US Grocery Sales is derived entirely from upstream inputs, its forecast accuracy depends on how well the Units and CPI Food-at-Home models each perform on their own terms. Any error in either component — whether from a mis-forecast in oil-driven input costs, an unexpected shift in consumer volume behavior, or a base-effect distortion in the price index — will propagate directly into the sales figure with no offsetting mechanism. Readers seeking R-squared values, holdout MAE, and other statistical diagnostics should refer to the dedicated Units and CPI Food-at-Home sections of this report.