Nvidia boosts share buyback by record $150B as AI boom fuels growth

Nvidia added a record $150 billion to its share-repurchase authorization, lifting remaining buyback capacity to $235 billion through fiscal 2028; shares rose more than 2% following the announcement. The authorization exceeds Apple’s $110 billion 2024 approval and is supported by AI-driven cash generation, while Nvidia continues to project roughly 70% revenue growth for fiscal 2028. The buyback comes with Nvidia trading at 16.5x forward earnings—its lowest multiple since January 2015 versus a 15-year average of 30x—reflecting investor concerns about slowing profit growth and durability of AI spending.

Analysis

The authorization is principally a valuation floor, not incremental operating demand. At NVDA's current earnings multiple, retiring shares is economically attractive only if AI revenue durability remains intact; the more important signal is management's willingness to allocate future free cash flow away from external investments at the margin. That can support EPS through fiscal 2028, but investors should not capitalize the full authorization: execution pace, cash conversion after working capital, and any debt-funded component remain undisclosed.

The second-order issue is capital allocation conflict. NVDA's investments in AI customers can preserve the ecosystem's capacity to purchase accelerators, but also raise the risk that reported demand increasingly depends on vendor-enabled financing rather than independently financed cloud ROI. Over the next 1-3 months, hyperscaler capex commentary and NVDA's receivables, inventory and customer-concentration disclosures matter more than the repurchase headline; deterioration in those metrics would turn buybacks into a signal of fewer productive reinvestment opportunities.

Consensus may treat the low multiple as a free call option on sustained AI growth. A more nuanced view is that the multiple already embeds a sharp deceleration from prior growth rates, so the key upside is not multiple expansion alone but evidence that inference demand broadens beyond a concentrated cloud buyer base. Conversely, if customer capex merely shifts from training to lower-dollar inference configurations, unit growth could coexist with lower system-level revenue and gross-margin pressure over the next 6-18 months.

AAPL is not a clean sympathy beneficiary. Large-cap buyback comparisons may direct passive attention toward capital-return beneficiaries, but Apple's return profile is more dependent on mature cash-flow stability while NVDA's depends on a cyclical infrastructure spend cycle. The cleaner relative expression is NVDA versus an AI hardware basket where weaker component suppliers retain higher inventory and pricing risk.

Key Decisions for Investors

• Initiate a modest 3-6 month long NVDA / short SOXX pair only on post-announcement strength that holds above the announcement-day low. The thesis is that buyback support and superior cash generation separate NVDA from the broader semiconductor cycle; exit if NVDA reports sequential gross-margin compression or receivables materially outgrowing revenue.
• For long-only exposure, accumulate NVDA in tranches before the next earnings print rather than chase the initial reaction. Target a 10-15% upside over 3-6 months from EPS support plus renewed confidence in fiscal-2028 demand; risk is 15-20% downside if hyperscaler capex guidance weakens or management does not specify a meaningful near-term repurchase cadence.
• Buy 6-9 month NVDA call spreads rather than outright calls if implied volatility remains elevated: use an at-the-money long call financed by selling a 15-20% out-of-the-money call. This captures a rerating if AI demand visibility improves while limiting premium exposure to a post-results de-rating.
• Set a due-diligence alert, not a trade, for quarterly repurchase dollars, net cash/debt movement, receivable days, inventory growth, and disclosed AI-investment commitments. A large authorization with modest actual execution, rising financing exposure, or weaker cash conversion would falsify the capital-return support thesis.