IT Spending Will Exceed $6 Trillion for the First Time in 2026 Thanks to Artificial Intelligence (AI). Here's How to Invest.

Gartner forecasts global IT spending of $6.15 trillion in 2026, up 10.8% YoY, while major tech firms are sharply increasing capex (Amazon plans $200B, Alphabet expects capex to double). Nvidia reported record fiscal 2026 revenue of $215.9B (+65% YoY), Q4 data-center revenue $62.3B (+75% YoY), a 55.6% net margin and a 0.07 debt/equity ratio, underscoring continued dominance in AI GPUs (92% market share). Equinix operates 280 data centers serving ~10,500 companies, yields ~2% dividend (11 years of increases), with AFFO +12% and revenue +5% in 2025, and plans a ~10% dividend raise in 2026.

Analysis

The dominant supplier of AI accelerators has converted a technology lead into structural pricing power; that creates durable platform economics but also concentrates execution and policy risk into a handful of vendors and their upstream suppliers (foundries, memory, power delivery). Expect outsized margin capture to flow to firms that control both hardware and the software stack, while commodity component makers will see cyclical revenue tied to server build windows rather than secular demand.

Data-center real estate and interconnect platforms sit one step removed from chip dynamics: they benefit from broader capex but their cadence is stickier and visible through lease roll and colo bookings. This makes them a lower-volatility play on the AI cycle, but implies sensitivity to interest rates, capital-cycle churn (new hyperscale builds), and a lag between hyperscaler capex announcements and measurable cashflow.

Key risks on a 0–24 month view are policy shocks (export controls or tariffs), sudden capacity additions from alternative architectures or foundries that compress realization, and macro-driven capex pullbacks. Over 2–5 years the biggest reversal would be a successful domestic alternative in large markets that forces price competition and accelerates depreciation on existing hardware. Monitor order-book duration, wafer starts, and colo tenancy mix as leading indicators of inflection.

For portfolios: tilt into the asymmetric convexity of the chip leader with disciplined hedges; own selective infra exposure to capture secular hybrid-cloud growth but size for interest-rate sensitivity; and keep a tactical hedge against policy/geopolitical events that can halve near-term multiple expansion in a concentrated leader.

Key Decisions for Investors

• Long NVDA (core): Establish a 3–6% net portfolio weight via long-dated (9–18 month) call exposure or outright shares to capture secular AI hardware adoption; hedge tail risk with 3–6 month put protection sized to 25–40% of the long notional. Risk/Reward: asymmetric upside if adoption continues; downside protected by puts against policy or demand shocks.
• Equinix (income + structural): Buy shares or 2–4% portfolio weight for a multi-year hold to capture sticky recurring revenue and private interconnect pricing power; sell 6–12 month covered calls to boost yield during periods of sideways tape. Risk/Reward: modest yield plus capital upside tied to occupancy/mix; downside is rate-sensitivity—set stop at a 12% drawdown or if AFFO growth decelerates materially.
• Pair trade to express dispersion: Long NVDA / Short INTEL (or AMD) in equal notional terms for 6–12 months to exploit asymmetric execution and software-ecosystem moats. Size as a market-neutral pair (each 2–3% portfolio); unwind if the lagging supplier reports a clear market-share recovery. Risk/Reward: captures widening margin/valuation gap; risk is a sudden competitive technology or capacity-led price war.
• Event hedge (policy/geopolitics): Buy 3–6 month puts on the chip leader equivalent to ~15–20% of the NVDA position ahead of major export-control or earnings events; alternatively, buy protection on correlated ETFs that track AI infrastructure. Risk/Reward: cost is insurance premium; payoff is large in event-driven downside that would also pressure correlated infra names.