Is the "Magnificent Seven" Era Over? Here Are the Stocks That Replace Them

Broadcom forecasted more than $100 billion in AI chip revenue by 2027, while Nebius projects annual recurring revenue rising from $1.25B last year to $7B–$9B this year and has rallied ~20% YTD; TSMC is highlighted as the primary manufacturer of leading AI chips (e.g., Nvidia). The piece frames these three — TSMC, Broadcom, and Nebius — as potential successors to the 'Magnificent Seven' amid recent AI-revenue and macro/geopolitical concerns. Recommendation-like coverage is bullish but speculative, relying on continued AI demand and execution by chip makers and cloud/AI infrastructure providers.

Analysis

The immediate structural winners are capital-intensive suppliers to the AI server stack rather than a single GPU maker: foundry capacity, bespoke ASIC vendors, and interconnect/switching have asymmetric pricing power because lead times and qualified supply are measured in quarters-to-years. Expect downstream OEM procurement to remain lumpy — a 6–12 month cadence of large orders followed by inventory digestion — which amplifies earnings volatility but preserves long-run margins for constrained suppliers.

Key near-term risks are macro-driven server capex pauses, rapid efficiency gains in model architecture (which can lower chip demand per model), and geopolitical disruption to Taiwan supply chains; catalysts that will move prices materially are quarterly capex updates from hyperscalers (next 3–6 months), export-control announcements, and TSM/AVGO quarterly guides. On the flip side, multi-year consequences of long fab lead times and customer lock‑ins mean a positive impulse from AI could persist for 24–36 months even if growth slows in any single quarter.

The most overlooked dynamics: custom ASIC adoption widens the market for semiconductor fabs and networking specialists, so valuation multiples should rerate away from single-product concentration toward diversified infrastructure franchises. Conversely, small “neocloud” providers face a two-sided bet — they monetize capacity at higher ASPs today but are vulnerable to vertical integration by hyperscalers; their public moves are therefore high-beta trade opportunities, not core holdings.

Key Decisions for Investors

• Long TSM (structural foundry exposure): allocate 3% NAV via a 12–18 month call spread to cap premium spend (buy 12–18m ATM calls, sell higher-strike calls). Rationale: capture sustained AI server capex and pricing power; target ~35–60% upside over a year if demand holds, with premium loss as downside limited to ~3% NAV if sentiment reverses. Monitor Taiwan geopolitical headlines as a binary tail risk.
• Long AVGO with a protective collar (core holding): buy AVGO stock sized to 4% NAV, buy 6–12 month ~10–12% OTM puts and sell 6–12 month ~25% OTM calls to finance protection. Rationale: Broadcom’s bespoke ASIC + interconnect mix has free-cash-flow defensibility; expected 12–18 month upside 30–40% vs downside limited to ~10–12% in adverse cycles.
• Tactical long NBIS / short AMZN pair (isolate neocloud vs hyperscaler risk): equal-dollar long NBIS spot (size 1–2% NAV) and short AMZN futures or buy-write to offset market beta, horizon 3–6 months. Rationale: capture upside from immediate capacity tightness and AR growth while hedging cloud demand shocks; set a hard stop-loss at -25% on NBIS position due to execution risk.
• Underweight/hedge select non-AI mega-caps (AAPL, AMZN, TSLA): implement a 3–12 month defensive put-spread on a basket (buy 3–6 month 10–15% OTM puts, sell further OTM puts) sized to offset 30–40% of growth-book exposure. Rationale: rotation into infrastructure could punish incumbents’ relative performance; this structure limits cost while offering 2:1+ payoff if rotation accelerates.