Big Tech stocks suddenly look cheap

Goldman strategist Peter Oppenheimer says tech valuations, relative to expected consensus growth, have fallen below the global aggregate and the valuation premium of US hyperscalers has compressed to near-market levels, creating attractive buying opportunities. The 'Magnificent Seven' have lost a collective $1.1 trillion in market cap as of early April 2026, amid massive hyperscaler capex and investor worries about returns and cash-flow strain. Concrete pain points include Oracle raising debt and laying off 30,000 workers to fund AI infrastructure, and investors rotating into oil and defense due to geopolitical shocks from the Iran conflict and fears of AI-driven disruption to incumbents.

Analysis

The hyperscaler-led capex wave is creating a durable two-tier market: firms that provide the platform stack and those that bear the build cost. That bifurcation produces recurring revenue capture for software, tools, and middleware vendors that sit one layer above raw infrastructure, while raw capex spenders face compressed near-term returns but optionality on future monetization.

Second-order supply effects matter: sustained data-center buildouts favor component suppliers with concentrated capacity (custom ASICs, power gear, high-end switches) and create intermittently attractive pricing power that can last quarters — not days — when supply tightness surfaces. Conversely, incumbent enterprises with high fixed-cost on-prem footprints are at structural risk of margin erosion as they cede workloads to hosted platforms.

Key risks are non-linear and multi-horizon: acute rotation from growth to macro/geo risk can compress long-duration multiples in days, while meaningful monetization of AI (pricing models, inference-as-a-service) plays out over 6–24 months and would re-rate winners. Regulatory friction or rapid model commoditization are multi-year threats that disproportionately hurt incumbents with low differentiation. The path to mean reversion is clear — signs include capex deceleration, improving free-cash-flow conversion, or a string of monetization beats from platform-hosted services — each a 1–4 quarter catalyst to unwind the current discount.

Key Decisions for Investors

• Buy MSFT via 12-month call spread (buy ATM, sell +30% strike) sized at 1–2% NAV; thesis: capture re-rating if enterprise AI monetization accelerates; max loss = premium, target net return ~+3x premium within 6–12 months; add-on on 5–10% pullback.
• Long AMZN stock with a 9–18 month horizon, hedge with 6–9 month puts (protective collar) sized to limit downside to ~15%; reward scenario = 25–40% if AWS/commerce synergy reacceleration occurs; hedge caps cost and preserves upside.
• Relative-value pair: Long ORCL (stock or deep-in-time calls) / Short IBM (buy puts) over 6–12 months — size 1% NAV each leg; rationale: asymmetric recovery optionality in software/managed-services vs legacy hardware/service decay; target 2.5:1 reward:risk, stop-loss if pair fails to diverge after two earnings cycles.
• Tactical short of legacy telecom/hardware exposure (e.g., NOK/BB) via 3–6 month puts (staggered) to limit capital at risk; expected payoff if platform shift accelerates is -20% to -35% per name within 6 months, use strict Vega cuts into major macro headlines.