AI took investors on a date in 2025. In 2026, it may be time to foot the bill
AI fueled a big year for equities in 2025. The S&P 500 finished up roughly 17%, pulled higher by a handful of megacap tech names and the AI trade. That level of concentration created comfort on the way up - and a single point of failure if the story cracks.
"Given the concentration, you need to diversify the portfolio, and you need to understand what you own and what happens if you're wrong," said Max Wasserman of Miramar Capital. That's the playbook for 2026: less hype, more underwriting, and a portfolio that can survive if AI enthusiasm cools.
Concentration risk is the headline risk
Markets leaned heavily on the "Magnificent Seven." If AI stumbles, tech is exposed. Tom Essaye at Sevens Report warned, "If AI goes south on us, tech will go." The unified story of 2025 is giving way to a harsher selection process where cash flows, unit economics, and near-term ROI matter again.
You can see the split already. Memory leaders like Micron ripped higher with AI demand for high-bandwidth components. Former favorites that don't show quick payback are getting screened harder. Expect more dispersion - not fewer headlines, but less forgiveness.
Follow the cash, not the buzz
Concerns over circular financing in parts of the AI stack - where investors fund revenue that flows back to them - should make you scrutinize incentives and revenue quality. Price-to-dream is out. Contracted revenue, backlog, utilization, and cost curves are in.
For chip exposure without riding every rumor, Broadcom stands out, supported by application-specific silicon and software cash flow from VMware. As analyst Lisa Schreiber noted, this mix "diversifies the business from solely a chip company" and can set a higher valuation floor than hardware-only peers.
Still bullish on AI? Be selective
Dan Ives remains positive on core names like Nvidia into 2026 while looking for the next leg across companies such as Nebius, Iren, and Palo Alto Networks. The key shift: position for clear product-market fit and pricing power, not just adjacency to AI.
Play the infrastructure where demand visibility is highest. Track compute capacity, memory supply, power/energy constraints, and the enterprise pipeline for AI workloads. The market is starting to reward firms that convert AI spend into immediate productivity or margin gains.
The 493 could matter again
A softer Fed stance - and the possibility of rate cuts to support a cooling economy - opens the door for the rest of the S&P 500 to participate. Lower short rates lift "boring" cash generators. Retailers like Home Depot and McDonald's can benefit from steady demand plus lower financing costs.
On defense, dividend payers such as Chevron, AbbVie, and Waste Management provide ballast if AI multiples compress. The message: barbell growth exposure with durable cash flow and balance-sheet strength.
Healthcare: growth at a sensible price
Healthcare has a habit of grinding higher regardless of political noise. Essaye argues the market often overstates policy risk, creating entry points backed by aging demographics and recurring demand.
Eli Lilly remains a standout. GLP-1 drugs like Zepbound and Mounjaro have driven strong year-over-year growth and function as a "growth-defensive" hybrid - secular demand with pricing and scale. In a year that could test narratives, that mix is attractive.
Your 2026 portfolio checklist
- Reduce single-point-of-failure risk: Cap position sizes in megacap tech and diversify AI exposure across chips, software, and beneficiaries (energy, utilities, data centers).
- Underwrite the P&L: Prioritize companies with contracted revenue, visibility into AI monetization, and credible unit economics. Defer "promise-only" stories.
- Favor cash generators: Add steady earners that benefit from lower short rates. Screen for free cash flow yield, pricing power, and debt maturity profiles.
- Stress-test AI assumptions: Run scenarios where AI spend flattens or procurement cycles extend. What happens to margins, capex, and payback periods?
- Play dispersion, not direction: Use relative value within semis, software, and services. The winners/losers gap should widen as the market demands proof.
- Mind duration and beta: Blend growth with defensives to stabilize volatility if rates fall slower than hoped or earnings revisions roll over.
- Watch the policy path: Track the Fed's stance and liquidity conditions; they'll set the backdrop for factor leadership and valuation support. Fed policy updates
How this plays out
AI isn't going away. But the easy trade is behind us. 2026 rewards investors who can separate real cash flow from slideware, and who build portfolios that don't rely on one theme to keep working.
Expect more choppiness, more earnings-driven moves, and more stock picking. If AI execution stays strong, leaders will earn it. If it stalls, the market has other places to go - and this time, the 493 have a bid.
For finance teams sharpening their AI edge
If you're evaluating tools, benchmarks, or upskilling pathways specific to finance, this curated list is a useful starting point: AI tools for finance.
One last note: stay process-driven. Update theses as data shifts, keep risk sizing honest, and let cash flow quality be your north star. The story got you here; the numbers will get you through 2026.
Your membership also unlocks:
