What Lunch Money Tells Us About the 2027 Market

Themes Heading Into 2027 · Part 1

When we think about the markets, “lunch money” is a good place to start, because it shows what rising prices do to a family budget.

Last month the Bureau of Labor Statistics put out a chart showing the rising cost of school lunches over the past 20 years.

Lunch has climbed faster than prices in general and has roughly kept up with paychecks. Somewhere a kid is paying double for rectangle pizza, and the fruit cup still looks smaller than it used to. I would be intrigued to know the prices of choco-mints, a cafeteria staple in the 80s.

The closest comparison is protein bars. Shoppers have reported Clif Bar boxes dropping from 12 bars to 10 and Atkins packs going from five to four, with no sign taped to the shelf. The price holds and the product shrinks, which is the same squeeze on a smaller tray.

The consumer is the engine of the U.S. economy, roughly two-thirds of it, and every theme in this outlook runs through the household budget. Energy is the first pressure on that budget: gasoline, heating and the cost of moving everything to the shelf. The AI build-out is the second link. The hyperscalers, the giant companies like Amazon, Microsoft, Google and Meta that run the world’s biggest data centers and rent out the computing power, are projected to spend about $1.1 trillion in 2027, and that spending has to be paid back by customers, advertisers and businesses that are all watching their own costs. Data centers also run on electricity, so the same energy shock that lands on the lunch tray lands on the build-out. When the consumer is stretched, both stories get harder, and it shows up in the stock market.

Here are the three consumer themes we’re watching as we head into 2027, and what we think each one means for the market. Let’s dive in.

Here’s the chain reaction behind a possible stock market downturn: oil spikes, prices rise, and the Fed raises rates. The first link is still visible at the pump. Regular gasoline averages about $4.29 a gallon, against $3.08 a year earlier, and crude is doing the heavy lifting. Brent, the global benchmark, is trading in the $110s a barrel, up roughly 70% from $67.09 a year ago.

What matters is the cause. This is a supply shock, not a demand boom. The Strait of Hormuz, the narrow channel a big share of the world’s oil ships through, is carrying about 13.1 million barrels a day, roughly 77% of pre-war levels, and it is looking less certain that flows will snap back the moment the conflict is resolved. In past disruptions, restarting shut-in production and normalizing shipping has taken longer than the headlines suggested.

That distinction is what makes this hard for the Fed, and the Fed has already acted. It raised rates a quarter point in September, to a range of 3.75% to 4.00%, and held there on Oct. 28. The vote was 10 to 2, and both dissenters wanted another hike. Markets now see a December increase as more likely than not. Higher interest rates can cool demand, but they can’t produce a single barrel of oil. If prices keep rising for supply reasons, the central bank has to choose between tolerating inflation and slowing an economy that is already stretched. Markets dislike that choice for a simple reason: when investors expect higher rates, they pay less today for earnings that arrive years from now, and the stocks priced on those distant earnings, mostly growth and technology, feel it first. The Nasdaq fell about 2% last week and about 5% in October, while the 10-year Treasury yield sits near 4.9%.

Families are feeling it. Credit card balances reached $1.26 trillion in the second quarter, up $21 billion from the first, according to the New York Fed’s latest household debt report. That sounds alarming, but dollars alone mislead. Fed data on revolving credit, which is mostly credit cards, puts balances at about 5.6% of annual disposable income today. The same ratio was about 6.7% at the end of 2019 and above 9% at the end of 2007, just before the financial crisis. Households owe more than ever, but measured against what they earn, they owe less than they did before the last two recessions.

The strain shows up in who is falling behind. The share of card balances sliding into serious delinquency was 6.97%, essentially unchanged from 6.93% a year earlier, but the New York Fed’s Joelle Scally noted that “new delinquencies for auto loans and credit cards remain at elevated levels.” Credit limits keep rising too, which suggests lenders are still comfortable lending. This looks like a slow squeeze, not a break. Rising rates are the part that can change that, because they raise the cost of every balance that carries interest.

Then there’s buy now, pay later, the short-term installment loans offered at checkout. Bank of America’s October Consumer Checkpoint shows usage rising, with nearly one in five users owing a repayment over $500. Is that a lot? Pay is up about 33% since the end of 2019, but prices are up about 30%, so $500 today buys roughly what $385 bought then. It’s also about 38% of the average weekly private-sector paycheck of roughly $1,300. That’s meaningful for a household living paycheck to paycheck, and it’s not a crisis.

Does it work as a recession indicator? On its own, no. Buy now, pay later only became mainstream in the past several years, so it has never been tested through a downturn, and the number measures how much people have borrowed, not how many payments they’ve missed. Bank of America’s own authors don’t see the rise as overly concerning. The signals with a track record are the rate of new delinquencies, jobless claims and the unemployment rate, and we’d want to see those turn before calling anything a warning. Does it still matter? Yes, because it shows where the pressure is building: lower- and middle-income households that finance everyday purchases, and by extension the companies that sell to them.

What does that mean for stocks? Two things. First, a gloomy mood hasn’t been a bad sign for stocks. Consumer sentiment is a survey of how people feel about the economy. The four other times it fell below 60 (1980, 2008, 2011 and 2022), the S&P 500 gained between 15% and 22% over the following twelve months. That’s only four examples, and this time stocks were already up about 40% from their 2025 low when sentiment bottomed, so we’d treat it as a clue and nothing more. Second, with less savings, spending now leans on jobs and on the wealth of households that own stocks. If energy keeps squeezing budgets, we’d expect it to show up first in the earnings of companies that sell to everyday consumers, not in the AI names.

Put it together and you get a consumer who is still spending but has less cushion underneath. Spending is holding up. The strain shows up in income, savings and credit.

What’s happening. AI spending still drives the earnings, but the stocks are repricing the risk, including financing costs and margins. Investors still like the story. They’re just checking the math now, and last week they checked it out loud.

It appears to be all about the hyperscalers. They’re the ones writing the checks for the chips, the buildings and the power, and their spending is somebody else’s revenue, which is why the chipmakers are booking the profits.

Software is the weak spot. Software gave back its summer rally in September. Software Development fell 5% and UI and Design fell 17%, while Semiconductors rose 9%. The market is picking its AI winners rather than buying all of tech. One month can’t tell us whether that’s a rotation, which is investors moving money from one group of stocks to another, or the start of a new downtrend.

The stress test. Investors are asking harder questions, and the answers got expensive last week. Alphabet fell about 12.6% on Oct. 29 after its earnings, its worst day since it went public in 2004, on softer search revenue and faster spending on AI infrastructure. Alphabet, Microsoft, Amazon and Meta together spent roughly $202 billion on property and equipment in the third quarter, more than the roughly $176 billion in cash their businesses generated. A year ago they had cash left over. The pain spread down the chain: Intel fell about 29% on Oct. 30 after a weak forecast, and chip stocks dropped about 7% that day.

The forecast. The spending is locked in through 2027, so the question is who gets paid for it. Last week was the first real test of whether the market will keep funding it, and it flinched. We still expect the market to favor the companies selling the chips and equipment, and the ones with their own cash, and to be tougher on the biggest spenders until the returns show up. January’s earnings reports will show whether the big spenders keep spending.

What’s happening. A small group of stocks has carried the whole index, and September showed how thin that is. It’s like a group project where one kid does all the work. The grade looks fine until that kid stays home sick. In October, investors started asking what happens to the grade.

The good news. The market has started to broaden, and October made the case. On Oct. 7, big tech had outrun the average stock by more than at any point since 2000. Since then, about three out of four stocks in the S&P 500 have beaten the index. For the month, the average stock gained about 4.9% while the S&P 500 slipped about 0.9%. Wide dispersion with low correlation is what a market looks like when investors are sorting winners from losers one company at a time, instead of buying the whole index. State Street expects 10 of 12 industrial industries to post double-digit earnings growth next year, as AI spending spreads into power, construction and equipment.

The catch. Sorting isn’t the same as broadening, and the sorting can be rough. It’s closer to a rotation than it was a month ago, but not a clean one. Last week’s leaders were health care, materials and utilities, a mix of economy-sensitive and defensive stocks, the things people buy in any economy. Small companies caught a bid, with the Russell 2000 up about 1.3% while the S&P 500 fell, but one week isn’t a trend. And with the 10-year Treasury yield, the interest rate the government pays to borrow for ten years, near 4.9%, borrowing costs are still a headwind.

While rising lunch money prices signal headwinds for the future, profit growth has been tepid and oil remains the wild card. You’ll notice we haven’t mentioned the midterm elections. We note that sentiment about politics moves the market more than the actual policy changes do, and trying to project how that unfolds is like herding cats. But rest assured, it’s on our mind.

We’ll keep you posted on evolving thoughts in our future Meridith on the Markets pieces. Stay tuned for Part 2 of Market Outlook 2027.

— Meridith on the Markets

Source for the school lunch figures: Bureau of Labor Statistics, U.S. Department of Labor, The Economics Daily, “School lunch prices over the last 20 years” (visited October 15, 2026).

Sources for the household figures: Federal Reserve Bank of New York, Household Debt and Credit report, Q2 2026; Bank of