Investors are doing just as well this year by shunning AI. How to invest if the tech trade keeps fizzling

Goldman's S&P 500 ex-AI index gained 12.4% through Oct. 22, narrowly ahead of the traditional index's 12%, as investors rotated away from AI-linked stocks; the two indexes were neck-and-neck after tech rallied Monday. The S&P 500 tech sector was down about 7% this month, while defensive-sector ETFs rose 7%–10%; energy gained 10%, which the article attributed mainly to the war in Iran. Analysts highlighted non-AI diversification, while Berkshire Hathaway shares rose 8% this month and Eli Lilly and Viking Holdings were cited as potential non-tech opportunities.

Analysis

The useful signal is not that “non-AI” has won; it is that investors are repricing the duration and funding risk embedded in the AI buildout. That can pressure both AI spenders and suppliers: if cloud and platform companies face a higher cost of capital or weaker investor tolerance for capex, orders may be delayed before end-demand changes. That makes the recent weakness in GE Vernova, Bloom Energy, Quanta Services and Vistra potentially an early-cycle capex-risk proxy—not proof that the long-run power constraint has disappeared.

The Goldman ex-AI index is a relative-performance lens, not a clean factor portfolio: excluding AI enablers also changes sector exposures. Defensive outperformance may therefore reflect both de-risking and sector mix. A sharp rebound in tech, or evidence that AI spending is generating monetizable returns, could reverse the trade quickly. Over 1–3 months, watch hyperscaler/platform capex guidance, financing terms, and earnings revisions for infrastructure suppliers; over 6–18 months, the key test is conversion of AI investment into durable revenue and returns.

Contrarian angle: the move may be underpricing the resilience of AI demand while overpricing near-term capex cuts. Avoid a blanket short of AI or its supply chain. Prefer a relative-value expression: defensive compounders against funding-sensitive AI exposure, with limited risk given headline and positioning reversals. Falsification: renewed upward capex guidance alongside improving monetization, or defensive earnings revisions turning negative.

Key Decisions for Investors

• Consider a measured relative-value position: long Republic Services (RSG) and/or Verisk Analytics (VRSK) against a smaller short in CoreWeave (CRWV) or Oracle (ORCL). The thesis is lower exposure to AI funding-duration risk, not a forecast of outright AI demand collapse. Reassess on earnings revisions and financing conditions; stop adding if AI capex and monetization evidence reaccelerate.
• Keep AI infrastructure names GE Vernova (GEV), Quanta Services (PWR), Bloom Energy (BE) and Vistra (VST) on a watchlist rather than treating the selloff as a confirmed structural break. Stabilizing orders/backlogs alongside unchanged customer capex would argue the pullback is a positioning reset; order deferrals or weaker guidance would validate further caution.
• For the next 1–3 months, favor selective non-tech exposure over a broad “ex-AI” basket: commercial services and insurance names cited by analysts include Cintas (CTAS), Arthur J. Gallagher (AJG), Broadridge (BR) and MetLife (MET). Verify earnings revisions and valuation before adding; the index comparison alone is not a buy signal.
• Treat Berkshire Hathaway as a diversified alternative, not a pure anti-tech hedge: its portfolio still includes large technology exposure. Avoid using it as a direct short-AI proxy; relative performance will also reflect its operating businesses and financial holdings.