Step Aside, ExxonMobil: 2 Foundational AI Stocks Now Pay Larger Dividends

Nvidia's annual dividend payout has risen to $24.15 billion after a 25-fold increase in its quarterly dividend to $0.25 per share, supported by strong AI data-center GPU demand and pricing power. Microsoft has overtaken ExxonMobil in annual payouts at $27.05 billion, supported by its $3.64 per-share dividend and 21-year dividend-growth streak. Microsoft reported Azure revenue growth of 43% in fiscal Q4 and ended fiscal 2026 with $76.8 billion of cash and investments and nearly $183 billion of annual operating cash flow, reinforcing capacity to fund AI investment and shareholder returns.

Analysis

The dividend framing is not itself a catalyst: both MSFT and NVDA trade on the durability of AI revenue and, more importantly, the return on an unusually large infrastructure investment cycle. For NVDA, a larger absolute cash distribution marginally broadens institutional eligibility but does not offset the risk that hyperscaler GPU capex normalizes before inference revenue scales; a 10-15% reduction in 2027 hyperscaler capex expectations would matter far more to the equity than incremental dividend growth. The relevant read-through is that management sees sufficient free-cash-flow visibility to return capital while maintaining product cadence, but this is not independently proof that demand remains supply constrained.

MSFT has the more defensible capital-return setup because its recurring software cash flows can subsidize AI capacity buildout through a weaker cloud-demand period. The key 1-3 month catalyst is Azure growth versus AI infrastructure depreciation and lease commitments: sustained cloud growth above the low-30% range with stable operating-margin guidance would support multiple resilience. Conversely, further capex acceleration without a commensurate rise in commercial remaining performance obligations would expose MSFT to a narrative shift from AI monetizer to AI financier.

Second-order beneficiaries remain TSM and AI-networking/power suppliers, but their upside is increasingly contingent on utilization rather than headline GPU demand. The contrarian view is that dividends may signal AI leaders are approaching a capital-allocation trade-off: cash returned to shareholders is cash not deployed into proprietary capacity, acquisitions, or customer financing. That is modestly constructive for mature MSFT, but could become a constraint for NVDA if competition from custom silicon and AMD forces more aggressive ecosystem spending over the next 6-18 months.

Key Decisions for Investors

• No standalone trade on the dividend news; treat it as low-impact confirmation rather than a fundamental earnings revision catalyst.
• Maintain MSFT overweight versus NVDA over the next 3-6 months: MSFT offers AI upside with recurring cash-flow support and lower sensitivity to a single hardware procurement cycle. Reassess if Azure growth falls below 30% or management raises capex faster than cloud revenue for two consecutive quarters.
• Use a pair trade long MSFT / short NVDA only after a data point showing hyperscaler capex moderation or weaker NVDA forward supply commentary; target 10-15% relative outperformance, with a stop if NVDA raises forward revenue expectations by more than 10% on Blackwell/Rubin demand.
• Keep TSM on watch as the cleaner AI-volume confirmation vehicle. Initiate only if CoWoS capacity expansion translates into upward 2027 revenue guidance or utilization evidence; the risk is that packaging supply expands ahead of end-customer demand, compressing mix and margins.