2 Artificial Intelligence (AI) Stocks That Won in 2025 Are Losing in 2026. Why This Is a Buying Opportunity

Palantir: revenue growth has accelerated for 10 consecutive quarters, culminating in 70% sales growth last quarter; PLTR rose 135% in 2025 but is down ~18% in Q1 2026, presented as a buying opportunity given its Foundry AI platform and strong government business amid the Iran conflict. Broadcom: AVGO was up 49% in 2025 and is down ~10% in Q1 2026; management projects $100 billion in custom AI chip sales in fiscal 2027 and maintains that AI hardware gross margins will remain unchanged as the business scales. The article attributes part of the Q1 rotation into value to the war with Iran and fears of peak AI infrastructure spending, framing recent pullbacks as tactical entry points for long-term growth exposure.

Analysis

Platform-level data fabrics create asymmetric switching costs that are rarely priced into headline multiples: migrating a mission-critical orchestration layer typically requires 6–12 months of integration work, 20–30% rework of analytics pipelines, and retraining of downstream models — numbers that favor incumbents with installed base economics. That implies durable gross retention and optionality to cross-sell higher‑margin services over multiple years, turning what looks like high revenue growth into pseudo-annuity streams if retention holds.

On the silicon side, multi‑generation co‑design programs front-load engineering and capital and backload margin capture; this produces lumpy near-term revenue but materially higher lifetime customer value once N+1/N+2 variants ship. The supply‑chain consequence is tangible: advanced packaging, substrate allocation and test capacity become choke points that reallocate economic surplus toward firms that control system‑level integration rather than pure-play fabs.

Key tail risks are straightforward and time‑staged: near term (0–6 months) the dominant risk is a capex pause by hyperscalers that compresses multiples and IV; medium term (6–24 months) the risk is successful vertical insourcing by a hyperscaler or a standards push that commoditizes orchestration layers; long term (2–5 years) regulatory/defense procurement shifts or open ecosystems could either cement or erode moats. The market appears to overdiscount multi‑year lock‑in economics while overreacting to quarter‑to‑quarter capex noise — this opens structured entry opportunities where downside is limited and asymmetric upside is tied to multi‑gen program outcomes.

Key Decisions for Investors

• PLTR — Buy a defensive structured long: initiate a 2–3% NAV position by purchasing the equity and hedging with 9–12 month protective puts sized to limit downside to ~25% of position value. Rationale: preserves upside to a multi‑quarter re‑rating if retention/upsell proof points arrive, while capping tail loss from near‑term sentiment shocks.
• PLTR options alternative — If cost-sensitive, buy a 9–12 month diagonal call spread (buy longer-dated LEAP, sell nearer OTM calls around next 2 earnings/contract windows) to capture event-driven IV while funding premium; target 40–80% upside on spread if catalysts prove out, max loss = premium paid.
• AVGO — Core long: initiate a 3–5% NAV long in spot AVGO with a 12–18 month horizon. Hedge cyclical infrastructure risk by shorting an equal notional amount of INTC (12–18 month equity) to express outperformance of custom/system integration vs legacy IDM exposure. Risk/Reward: expect asymmetric payoff if multi‑gen wins crystallize (target +25–40% on AVGO vs -15–25% on INTC); cut if spread tightens >15% in 6 months.
• Tactical volatility — Buy a concentrated calendar on PLTR around the next major defense/earnings date (long-dated call, short nearer-term call) to capture expected IV re‐rating on positive bids/renewals. Keep size small (≤1% NAV) — this trade pays off on catalyst confirmation while limiting carry if the market stays muted.