3.4% GDP, Gasoline Up 34%, Hershey Volume Down 10%: The Consumer Crack Is a Price-Level Problem

The Bureau of Economic Statistics put third-quarter real growth at 3.4% annualized, and the composition explains why the print felt disconnected from the shelf. Gross private domestic investment, including information processing equipment, drove the acceleration from 2.2% in the prior quarter. The growth was bought with capital spending, not with household demand.

The labor market is why this stays invisible in the headline. Initial claims of 194,000 for the week ended Oct 17 kept layoffs near 57-year lows, yet payrolls rose just 29,000 in September with unemployment at 4.2%. Low firing keeps unemployment at 4.2% while the low-income cohort quietly loses hours. That is why the aggregate looks calm: the adjustment runs through hiring and hours rather than layoffs.

There is no monetary cushion coming for the squeezed consumer. The Fed held at 3.75%-4.00% and markets priced roughly an 84.6% chance of a December rate hike after the inflation data. Note also that the Fed's preferred gauge is running hotter than the CPI version of the same idea, with core PCE at 3.1% against core CPI at 2.6%. No monetary cushion is coming for a squeezed consumer.

The composition of September's inflation is the whole story. Food at home rose 1.8% over the year while food away from home rose 3.5%, and gasoline rose 34.2%. The grocery shelf is nearly flat while the pump got 34.2% more expensive. A household that gives up a restaurant meal is trading down; a household paying 34% more for the drive to work has no substitute to trade down to.

The arithmetic is impossible to miss. Hershey priced up 11 points and sold 10 points fewer units, leaving organic sales up just 0.8% even as reported net sales rose 3.0% to $3,275.5 million, helped by 2.1 points from the LesserEvil acquisition. Adjusted earnings per share of $2.64 were up 103.1%, and reported gross margin rose 1,350 basis points to 46.1%.

Here is the part that gets missed, and it starts with cocoa. The International Cocoa Organization's daily price was $4,317.67 a tonne on Oct 27, 2026, after cocoa nearly tripled above $12,000 a tonne in 2024. With input costs collapsing while the household budget was taxed 34.2% at the pump, the manufacturer could hold shelf prices and still earn more on every unit: a consumer price shock was converted into a manufacturer margin windfall.

The earliest damage is not in the pantry, it is in the gas-station basket. Candy, gum and mint sales at US convenience stores fell nearly 5% year over year in September, with the weakness continuing into October, while everyday candy, mints and gum grew 3.7% in the quarter. The damage sits in the gas-station basket, not the pantry.

Convenience purchases are bought on the way to fill a tank that now costs 34.2% more, which makes that channel the cleanest read on the mechanism. Hershey said about 35% of seasonal sales land in the week of Saturday Halloween, and the National Retail Federation expects Halloween spending of about $13.5 billion this year against $13.1 billion last year. About 35% of seasonal sales now land in a single week, and the chief executive said the last few days are critical.

The chain splits into four groups. Input-leverage brands collect the windfall: Hershey lifted adjusted earnings per share 103%, and McCormick held volume and mix at -0.3% while organic sales rose 1.9%. The trade-down destination takes the volume: Dollar General grew traffic 2.0%. The squeezed middle is mass and grocery retail, where Walmart decelerated to a 1.1% ticket and Kroger grew sales 0.1% on a 22.4% gross margin. And brands that lost pricing power did not get the relief at all, with Kraft Heinz posting a $5.46 billion quarterly net loss and General Mills taking a $2.81 billion charge. The transfer rewards pricing power and punishes impulse volume.

The margin recovery is not accumulating on the balance sheet. On a trailing-twelve-month basis through the second quarter, Hershey generated $2.66 billion of operating cash flow against $112 million of capital spending, or $2.54 billion of free cash flow, against $1.12 billion of dividends, $762 million of acquisition spending and $439 million of buybacks. Turns the cost windfall into 2.1 points of acquired growth, because with its own volume down 10 points, Hershey bought LesserEvil, which contributed 2.1 points to third-quarter net sales growth.

Management's counterargument deserves weight. Hershey said brand unit trends are already improving, that pricing moves from high double digits in the third quarter to single digits in the fourth, and that the improvement should continue through 2027. If true, the volume damage was the cost of defending price through a fuel shock rather than a permanent loss of the customer, and 2027 lands in the middle of the 2% to 4% sales growth algorithm management reaffirmed on the call.

The evidence arrives on a schedule over the next weeks. The last week of October is the first hard print, because about 35% of Hershey's seasonal sales land in that week and the National Retail Federation's $13.5 billion Halloween estimate is the sector's first tape of whether participation converts into volume.

Over one to three years the question is whether the cost relief outlasts the consumer squeeze. Hershey said cocoa is coming off a historical surplus of 500,000 to 600,000 metric tonnes and that it does not expect prices to stay at current levels, so if cocoa re-inflates while unit volume is still negative the margin reverses quickly. The margin windfall is finite and mean-reverting, and management's own 2027 framework of 2% to 4% sales growth, with the lower end flagged if consumer pressure persists, is the honest version of this thesis.

The three prints are not in conflict once each is asked the question it can answer. GDP measures what the economy produced, the PCE price index measures what households can still buy after prices, and Hershey measured what the lowest-income households decided to skip. The aggregate economy is fine; the bottom of the income ladder is not, and the earnings that prove it sit with the brands that held price while the volume went to the discounters.