Which US Sectors Could Benefit as the AI Trade Cools?

The next stock-market winner might make something considerably less exciting than an AI chip. Think medicines, toothpaste or insurance. 

After a rally dominated by a handful of giants, investors are starting to ask a different question: which businesses can keep earning money when expectations come down and borrowing stays expensive? That could give several overlooked corners of the US market a chance. But there is a catch: an AI sell-off does not automatically make every other stock a winner.

Let's break down where money could move as the AI trade cools, which companies illustrate those opportunities, and how to distinguish a lasting change in market leadership from a brief search for shelter.

How Much of the US Stock Market Rally Came From Big Tech?

Imagine a cricket team that keeps posting impressive totals because a few star batters score most of the runs. The scoreboard looks healthy. It tells you much less about the rest of the batting line-up.

That is the problem with reading the US stock market through the S&P 500 alone.

The index weights companies by their float-adjusted market value. A move in a giant therefore matters much more than the same percentage move in a smaller member. Strong performance at the top can conceal a very different experience elsewhere.

S&P Dow Jones Indices’ figures show how much the largest names mattered:

Sources: S&P Dow Jones Indices, December 2024 and December 2025 market reviews. Return contribution is different from a group’s own investment return.

The Magnificent Seven are Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta and Tesla. They are a useful measure of market concentration, although their businesses extend well beyond AI.

This also corrects an overly simple version of the story. Other stocks did participate. Hundreds advanced. The issue was that a small group contributed disproportionately to the headline returns.

Nor can we attribute every dollar of Big Tech’s gains to AI. Advertising, cloud computing, consumer spending, earnings and interest-rate expectations also influence these companies.

The opportunity now is therefore more specific: investors may become willing to pay for a wider range of dependable earnings, after spending years paying particular attention to the biggest growth stories.

Why Is the AI Trade Facing Pressure Now?

A great technology and a great investment at any price are two different things.

An AI business might attract customers, grow revenue and still disappoint shareholders if its valuation assumed faster growth or if building the infrastructure costs more than expected.

The week ending October 16 offered a glimpse of that tension.

Source: Associated Press, October 16 market summary. These are price changes, not total returns.

The divergence extended to sectors. Barron’s reported that consumer staples and energy each gained roughly 5% over the week, while information technology fell 6.4%. That is evidence of changing leadership over those sessions; it does not establish a permanent rotation.

Inflation is adding another complication.

Source: US Bureau of Labor Statistics, September CPI release.

Gasoline accounted for more than half of September’s monthly headline increase. That matters because an energy-led inflation shock can benefit oil producers while raising costs for households and many other businesses.

Higher yields create two pressures on stocks. They can increase companies’ borrowing costs, and they give investors a more competitive alternative to uncertain future profits.

However, the timing needs care. Yields had risen substantially, but they fell during the latest week of AI weakness. The Treasury’s 10-year par yield declined from 5.24% on October 9 to 5.09% on October 16, while remaining well above its 4.75% reading at July’s end.

So the story is not simply “yields went up, therefore tech went down.” Expensive financing was already part of the backdrop. As investors worried about AI spending and growth, government bonds also attracted demand. MarketWatch reported a move towards Treasuries and defensive stocks during the sell-off.

Does Money Leaving AI Stocks Have to Enter Other Sectors?

No and this is where many rotation stories become misleading.

Sector rotation means investors change their relative exposure to different industries. It is not a rule requiring every dollar of lost technology market value to reappear in healthcare or energy.

Share prices reflect what investors will pay at the margin. A fall in market capitalisation is not a matching pile of cash waiting to be redistributed. Investors can also reduce equity exposure, favour government securities or repay borrowing.

More importantly, the reason AI cools determines who might benefit.

If valuations fall while the economy stays healthy, other profitable businesses have room to attract attention. If AI investment itself contracts sharply, suppliers, construction companies and power projects could lose business too. If inflation squeezes household spending, some companies outside technology will face their own earnings problems.

The strongest first destination is likely to be dependable earnings at sensible prices.

That leads to six areas worth examining.

Which Sectors Could Attract Money as AI Stocks Cool?

1. Healthcare: Demand That Does Not Depend on an AI Budget

Someone receiving treatment does not stop needing medicine because a cloud company trims its spending plans.

That gives established pharmaceutical and medical businesses a useful starting point: demand comes from patient needs, treatment adoption and healthcare budgets rather than enthusiasm for the next data centre.

Two examples are Johnson & Johnson (JNJ) and AbbVie (ABBV).

Sources: Johnson & Johnson Q3 results and AbbVie Q2 results. Growth rates are reported year-on-year figures; AbbVie’s number covers its immunology portfolio, not the whole company.

The attraction is independent demand. But the defensive label needs limits.

Drug companies face patent expiries, competition and pricing pressure. Medical-device demand can depend on procedure volumes. Health insurers have a different problem altogether: medical costs can rise faster than premiums. A loss-making biotechnology company awaiting a trial result is another kind of investment again.

J&J’s latest release itself illustrates the distinction: reported earnings included a special charge connected to a proposed talc litigation resolution. Dependable product demand does not remove legal liabilities.

Our preference within this theme is for established revenue, several sources of earnings and a credible pipeline. Investors should ask how profits would behave if the largest product stopped growing. Healthcare becomes much less defensive when one medicine carries the entire story.

2. Consumer Staples: Repeat Purchases, With a Pricing Test

Toothpaste, detergent and drinks do not attract the same attention as advanced chips. They do have something investors may appreciate when confidence weakens: customers return frequently.

Procter & Gamble (PG) and Coca-Cola (KO) illustrate this recurring demand. Yet their recent results also show why the entire sector should not be treated as one investment.

Sources: P&G fiscal 2026 results and Coca-Cola Q2 results. Operating cash flow is before capital expenditure. Organic growth adjusts for specified effects such as currency and acquisitions and is a company-defined non-GAAP measure.

P&G demonstrates substantial cash generation, but flat quarterly organic sales also warn against assuming that essential products guarantee strong growth. For staples, the critical question is pricing power: can a company recover higher costs without losing too many customers?

Think of a neighbourhood shop raising the price of a familiar product. Revenue might initially rise. But if shoppers start choosing the cheaper alternative, that price increase has not solved the problem for long.

This is why sales volume deserves as much attention as revenue. Coca-Cola’s reported volume growth provides evidence of demand beyond higher prices, although one quarter does not settle the long-term case.

Staples could attract investors seeking repeat purchases and cash generation. The better candidates must also protect volume and margins. A famous brand with falling demand or an excessive share valuation offers less shelter than its reputation suggests.

3. Insurance: A More Selective Way to Think About Financials

“Higher rates help financial stocks” sounds neat. It leaves out quite a lot.

Banks may earn more on some loans, but also pay more for deposits and face weaker borrowers. Property-and-casualty insurers have a different business model, making Progressive (PGR) and Chubb (CB) useful examples.

They collect premiums before paying many of the related claims. The money held in the meantime can earn investment income. Higher available bond yields can improve income as new premiums are invested and existing investments mature.

But first, the insurance itself must be priced properly.

The combined ratio compares claims and underwriting expenses with premiums earned. Below 100% generally indicates an underwriting profit, before investment income.

Sources: Progressive September and quarterly results and Chubb Q2 results. Periods and business mixes differ; these are not directly comparable rankings.

Progressive’s quarterly ratio implies roughly $88.70 of claims and underwriting expenses per $100 of premiums earned, leaving $11.30 before investment income, taxes and other items.

That gives investors two things to examine: underwriting discipline and investment earnings.

Neither is guaranteed. Inflation raises vehicle-repair and rebuilding costs. Catastrophes can overwhelm a good quarter. Rising yields can reduce the market value of existing bond holdings, even while improving future reinvestment opportunities.

Disciplined insurers deserve a closer look than a blanket bet on financials. The test is whether premiums keep pace with claims and reserves remain adequate, not simply whether interest rates are high.

4. Essential Services and Defence: Look Inside Industrials

“Industrials” includes businesses with very different customers. A machinery supplier dependent on new factories has different risks from a company collecting household waste every week.

Waste Management (WM), offers an example of a recurring service that households and businesses continue to need.

In Q2 2026, WM reported revenue of $6.684 billion, up 4%, and nearly 12% growth in operating cash flow. Its operating margin, calculated from reported operating income divided by revenue, rose from approximately 17.9% to 18.7%.

The useful observation is that profit grew faster than revenue. Investors should investigate whether pricing, productivity and recurring customer relationships can sustain that pattern.

WM is still exposed to fuel, labour, investment requirements and economically sensitive waste volumes. Essential service does not mean cost-free service.

RTX (RTX) illustrates a separate industrial theme: defence demand and aircraft-related activity. Its Q2 2026 backlog was $289 billion, including $119 billion in defence and $170 billion in commercial business.

Backlog means orders awaiting fulfilment. It offers visibility, but it is not cash already earned. Delivery delays, contracts with fixed prices and rising production costs can still damage margins. RTX’s substantial commercial business also means it is not a pure defence investment.

Recurring essential services and well-funded defence programmes offer reasons for earnings that are independent of AI spending. Investors still need to distinguish dependable orders from profitable delivery and avoid assuming that every industrial company shares these characteristics.

5. Energy: A Possible Inflation Beneficiary, With Cyclical Risk

If expensive oil is part of the inflation problem, oil producers have an obvious place in this discussion.

ExxonMobil (XOM) and Chevron (CVX) combine production with downstream operations such as refining. Their economics differ from those of businesses that primarily consume fuel.

Sources: ExxonMobil Q2 results and Chevron Q2 results. Figures are rounded. Free cash flow uses each company’s definition and should not be treated as a perfectly standardised comparison.

Those figures show the cash-generating potential of favourable commodity conditions. They also create a valuation trap.

A stock can look inexpensive when its profits are temporarily unusually high. If oil prices fall, the earnings beneath that apparently low price-to-earnings ratio can shrink quickly.

Investors should therefore test energy companies using a range of oil prices, normal spending requirements and cash flows without unusually favourable working-capital movements. Chevron specifically identified working capital as one contributor to its quarterly cash-flow improvement.

Energy is a conditional beneficiary of persistent commodity inflation, rather than a general-purpose defensive sector. If AI weakness develops into a broader demand slowdown, oil consumption and prices could weaken too. Production disruptions can also prevent a company from fully benefiting from higher prices.

Energy belongs in a different part of the rotation argument from medicines and toothpaste. Its strongest case depends on what happens to commodity supply and demand.

6. Utilities: Essential Demand, but Financing Still Matters

Electricity is essential. That makes utilities sound like an obvious destination when investors become cautious.

The difficulty is that delivering it requires expensive infrastructure, frequently financed partly with debt.

Duke Energy (DUK) illustrates a large regulated utility, serving 8.7 million electricity customers in its latest quarterly announcement. NextEra Energy (NEE) combines its Florida utility with a substantial energy-development business.

Their investment cases require more than a forecast of electricity demand. Investors need to examine borrowing costs, regulatory permission to recover spending, construction budgets and the time between spending money and earning a return.

A high dividend does not erase those questions. If bonds offer more income, investors may demand a lower utility share price before accepting its risks.

There is another complication: some power-sector growth expectations depend on data centres. NextEra’s earnings discussion explicitly referred to developing a national data-centre hub footprint.

A different sector label does not necessarily give you a different economic exposure. Moving from AI chips into an electricity project whose economics depend on AI expansion may leave the underlying bet largely intact.

Selected utilities become more compelling if yields settle or decline and investment plans remain well funded. They are a less straightforward answer to a scenario in which inflation and financing costs keep climbing.

How Can Investors Tell Whether a Quality Stock Is Too Expensive?

The sectors above contain businesses worth researching. They are not a valuation-ranked list of stocks to purchase at any price.

Quality means evidence of durable earnings, sensible use of capital and an ability to fund obligations through a difficult period. It is a business characteristic, not a guarantee about the share price.

A simple model shows why the price matters.

The price-to-earnings ratio, or P/E, is the share price divided by annual earnings per share. If a company earns $1 per share and trades at $20, investors are paying 20 times earnings.

Ignoring dividends for a moment:

Share-price change = (1 + earnings growth) × (ending P/E ÷ starting P/E) − 1.

Consider three hypothetical one-year outcomes:

Author’s illustrative calculations. These are not forecasts or current valuations of any named company. Returns exclude dividends, taxes, fees and currency movements.

The first business grows earnings rapidly, yet its shareholders lose money because investors become less willing to pay a high multiple.

The second grows much more slowly but delivers a positive price return.

The third is the warning for anyone rushing into defensives: dependable earnings cannot always compensate for an excessive entry price. Even a cash dividend equal to 2% of the initial investment would only reduce that example’s loss to 13.2%, assuming no reinvestment.

This is the financial case for paying attention to quality now. Investors may place greater value on earnings that are already visible. But if everyone pays too much for that visibility, the supposed shelter can develop its own valuation problem.

A Three-Question Test for Stocks Beyond the AI Trade

Instead of beginning with a sector label, begin with three questions.

1. Who pays the company?

Identify the customer and the budget funding the purchase. A hospital treating patients, a household renewing insurance and a hyperscaler expanding a data centre have different reasons to spend.

Then look through indirect exposure. An equipment supplier might sit outside technology but depend heavily on the same few AI customers. Diversification needs different sources of earnings, not just different stock tickers.

2. Who funds the company?

Ask whether existing operations generate enough cash for investment, debt obligations and dividends. For non-financial businesses, review several years of operating cash flow, capital expenditure and debt maturities rather than one unusually good quarter.

Here is a simple refinancing example. A company replacing $10 billion of debt carrying 3% interest with debt costing 6% would incur an additional $300 million of annual interest expense. Against $2 billion of operating profit, that is equivalent to 15% of operating profit before considering taxes or other changes.

Banks and insurers need different measures, including capital adequacy, asset quality and, for insurers, reserve strength. A generic industrial-company debt ratio is not enough.

3. What price are investors paying?

Compare the valuation with realistic earnings growth, normal cash generation and the business’s own risks. Check what must go right to justify today’s price.

For energy, use earnings across a commodity cycle. For healthcare, allow for product competition. For insurers, do not assume a low-catastrophe quarter repeats forever. For utilities, include the financing needed to deliver growth.

Together, these questions ask whether the business has customers, funding and a valuation that can survive the same adverse scenario. That is a more useful filter than simply searching for whatever has lagged Nvidia.

What Would Confirm a Broader US Stock Market Rally?

A week of defensive strength is encouraging for those businesses. It is insufficient evidence that the whole market is becoming healthier.

There are several checks investors can follow:

The equal-weight S&P 500 gives each company approximately 0.2% at quarterly rebalancing, making it a useful comparison with the conventional index. However, it also changes size and sector exposures; it is not a pure measure of investor sentiment.

The distinction between relative and absolute performance is especially useful. A sector falling 5% while the market falls 15% has outperformed. Its investors have still lost money.

For a durable broad-market advance, more companies need earnings support. Defensive leadership alone can instead signal that investors are worried about the economy.

Which Sectors Fit Each Economic Scenario?

There is no single “after AI” portfolio that fits every outcome. This is how we would organise the possibilities:

Author’s scenario framework, not return forecasts. Outcomes can overlap and change over time.

This is also why smaller companies are not automatically the next winners. A smaller business may face expensive debt and fragile demand despite having little exposure to AI. The Russell 2000’s latest weekly decline illustrates that a technology sell-off can occur without a small-cap revival.

Author’s Take: Follow the Earnings, Then Check the Price

There is a credible case for investors to look beyond the AI trade. The strongest version of that case starts with businesses whose customers continue paying, whose finances can withstand expensive credit and whose share prices do not require flawless execution.

For an AI valuation reset with a still-functioning economy, we see the clearest starting point in established healthcare, selected consumer staples, disciplined insurers and recurring essential services. Energy has a separate case when commodity inflation persists. Utilities need a more careful judgement about financing and how much growth depends on AI-related electricity demand.

These are priorities for research, not a prediction that the sectors will rise together.

The broader market can earn a larger role without AI becoming a failed technology. Equally, technology can remain economically important while some of its stocks deliver disappointing returns.

For investors, the useful question is whether a company can justify its price with earnings it can realistically deliver. Sometimes the answer will still be an AI business. Sometimes it will be the company collecting the rubbish every Tuesday.