2 Companies That Will Join Nvidia, Apple, and Alphabet in the $3 Trillion Club by 2028

Broadcom targets $100 billion in AI chip revenue by end-2027 and reported $8.4B from its AI semiconductor business in Q1 FY2026; the author projects Broadcom could reach a $3T market cap by end-2027 (currently $1.59T, needs ~89%+; analyst EPS est $17.54 for 2027 and a 40x multiple implies $701/sh, ~109% upside). TSMC (current value $1.75T, needs ~71% gain) is expected to grow ~25% CAGR 2024–2029 and could reach $3T by 2028 if growth and multiple hold (trading at ~32x trailing earnings).

Analysis

The market is pricing two different monetization models into the same AI narrative: asset-light monetizers of design/IP vs capital-intensive foundry providers. Design/IP-centric firms can expand gross margin and FCF conversion by converting hourly/volume exposure into multi-year, step-up contracts; that creates asymmetric upside but concentrated counterparty risk if a handful of hyperscalers drive most bookings. Foundries, by contrast, act as the plumbing — volume diversification but exposed to capex cycle, node transitions and wafer-utilization mismatch that can compress margins quickly when demand normalizes.

Second-order beneficiaries and pain points matter more than headline chip demand. EDA and IP licensors, advanced packaging and substrate suppliers, and test/assembly partners will see pull-through with shorter lead times and stickier revenue profiles than pure memory OEMs; conversely, GPU incumbents could see erratic RNG-like share movements as workloads standardize and buyers oscillate between general-purpose and bespoke fabrics. Geopolitical and logistics risk is asymmetric: a short disruption in cross-strait shipping or power availability can crystallize a multi-week revenue outage for foundries while leaving remote design partners relatively intact.

For portfolios, think of positions as timing-capex and contract-consolidation trades rather than pure AI exposure. Near-term catalysts are cadence of disclosed hyperscaler design wins, node-shifts and ASPs, and quarterly capex guidance — each can re-rate the two business models in 3–12 months. The consensus underestimates customer concentration and the potential for hyperscalers to internalize more of the silicon stack; that makes option-structured, hedged exposure superior to naked long volatility on valuation re-rating expectations.

Key Decisions for Investors

• Long AVGO equity with protection: size 2–3% NAV long AVGO (buy shares) and buy 12–18 month 10–15% OTM puts to cap tail risk. Rationale: asymmetric upside if contract economics scale; protection limits drawdown to ~10–15% at modest cost.
• AVGO call spread (cheaper upside): buy 18-month AVGO 25% OTM calls and sell 18-month 60% OTM calls to finance premium. Target: 2–3x upside if design wins compound; max loss = premium paid, suitable for 1–2% NAV allocation.
• Pair trade — long AVGO / short TSM (12–24 months): size ~0.6 AVGO exposure per 1.0 TSM to neutralize macro beta and express monetization vs capex risk. Hedge tail: buy 9–12 month TSM puts to protect against geo disruption; expected payoff if monetization premium outperforms foundry/multiple contraction.
• Buy suppliers of advanced packaging/EDA on dips (select names with 6–12 month catalysts) as a lower-beta way to play AI build-out. Size 1–2% NAV positions, take profits on 20–30% moves and monitor customer concentration disclosures closely.