My Top 2 Generative AI Stocks for Long-Term Investors for May 2026

Broadcom’s AI chip sales surged 65% to $20 billion in fiscal 2025 and management expects $60 billion to $90 billion by fiscal 2027, while analysts project revenue and EBITDA CAGRs of 47% and 46% through fiscal 2028. Alphabet is framed as a strong AI monetization play, with expected 2025-2028 revenue and EPS CAGRs of 19% and 16% and valuation at 27x earnings. The article is constructive on both names, but it is primarily an opinion piece rather than new company-specific news.

Analysis

The market is still underestimating the second-order beneficiaries of AI capex migration: custom silicon and distribution-rich software platforms are taking share from generic compute. AVGO is the cleaner way to own inference at scale because hyperscalers are optimizing for cost-per-token, not just raw FLOPs; that shifts budget away from Nvidia-like general-purpose accelerators over the next 12-24 months as workload mix moves from training to serving. The key implication is that AI spend is becoming more operationally embedded, which should keep demand durable even if headline model enthusiasm cools.

GOOGL is the more asymmetric quality compounder because it monetizes AI at the point of intent, not just at the infrastructure layer. The risk the street is missing is not AI disruption of search, but cannibalization of lower-value traffic that improves margins and click quality while expanding pricing power in higher-conversion queries. If AI answers reduce user friction without collapsing query monetization, Alphabet can harvest productivity gains and improve ad ROI simultaneously, making this a rare case where AI is deflationary for internal costs but inflationary for revenue quality.

The main near-term risk is positioning: both names are consensus AI longs, so multiple expansion is vulnerable if investors demand proof that AI capex is translating into incremental free cash flow rather than just higher depreciation and integration costs. NVDA is the relative loser in this framing, not because of any fundamental collapse, but because every incremental dollar of hyperscaler inference spend diverted to custom ASICs compresses its share of the AI wallet over time. Expect the divergence to play out over quarters, not days, with the inflection point coming from 2026 guidance revisions and customer concentration disclosures rather than one-quarter earnings beats.

Key Decisions for Investors

• Long AVGO on 1-3 month pullbacks; use any post-earnings weakness to build, with a 12-18 month target on continued hyperscaler custom-ASIC adoption. Risk/reward favors mid-teens upside if AI revenue mix reaches the upper end of guidance, with downside protected by diversified non-AI cash flow.
• Long GOOGL vs short NVDA as a 6-12 month pair trade. Thesis: the market is overpaying for pure compute while underpricing AI distribution and monetization; if inference shifts toward custom silicon, GOOGL should compound while NVDA's mix grows more competitive and less scarce.
• Buy GOOGL outright for 12-24 months into any broad tech drawdown. This is the best risk-adjusted AI exposure because it can fund AI investment internally, reducing dilution/funding risk versus smaller AI platforms.
• Sell out-of-the-money AVGO calls against a core long if already positioned. The name is strong but consensus-owned; premium capture makes sense while implied volatility stays elevated around AI guidance updates.
• If looking for a hedge, reduce NVDA beta rather than shorting the whole AI complex. The second-order risk is share shift to custom accelerators, not a collapse in AI demand, so the cleaner expression is relative underweight, not outright bearishness.