The S&P 500 rose 2% last week and finished its strongest quarter in six years as investors increasingly price an AI-driven earnings breakout. The key debate is less current valuation and more whether future earnings can sustain “extraordinary” levels tied to AI optimism.
The market is still rewarding the part of the AI stack where revenue visibility is shortest but pricing power is highest: compute, networking, power, and cooling. That argues the winners are not just the obvious megacap platforms, but the suppliers that sit behind them and can convert demand into backlog and margin before the rest of the ecosystem catches up. The second-order effect is that capital gets pulled away from broader IT budgets, which can leave non-AI software and services fighting for the same enterprise wallet.
The biggest loser is likely the long-duration software cohort whose AI story is still mostly “feature parity” rather than new budget creation. If customers are funding AI by shifting spend from application software to infrastructure, the beneficiaries are a narrower group than the index suggests, and dispersion inside tech should widen further. That also makes passive exposures more fragile: QQQ/XLK can keep melting up while breadth deteriorates, but they will likely underperform sharply if one or two AI bellwethers miss.
The contrarian risk is that the market is underpricing capex digestion and depreciation. Earnings can stay exceptional only if hyperscaler spend keeps accelerating and end-user monetization shows up within the next 1-2 quarters; otherwise the narrative shifts from growth to cash conversion, and multiples can compress even without a fundamental demand break. Key falsifiers are a slowdown in AI capex guidance, a sustained rise in real yields, or evidence that enterprise AI usage is not translating into incremental ARR.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request DemoOverall Sentiment
mildly positive
Sentiment Score
0.20
