These AI stocks have seen the biggest jump in popularity among the pros

Palantir Technologies led Bank of America's monthly analysis of actively managed U.S. large-cap mutual fund holdings, with ownership rising to 32% from 7% a year ago. Broadcom and GE Vernova each saw ~+16 percentage-point increases in fund ownership; Amphenol and AppLovin climbed ~+12 percentage points, while Arista Networks, Western Digital, Meta Platforms and Netflix also registered notable ownership gains. The data indicate a broad rotation of professional managers into AI- and data-center-related names, per BofA's aggregated bottom-up composite of fund positions.

Analysis

The recent breadth rotation into AI is creating a classic capital-goods cascade: sustained fund flows are not just inflating software multiples but are re-shaping supplier economics across optics, interconnect and power. That elevates pricing power for vendors with tight supply (high-margin connectors, specialized ASIC interposers, and data‑center power systems) and compresses payback periods on hyperscaler capex, increasing revenue visibility for incumbents with long lead times.

Crowding risk is material on two horizons. In the near term (days–weeks) a risk-off delta‑shock or a single large negative print from an AI darling can trigger cross‑strategy de‑risking via redemption-driven selling; over months, it will be fundamentals — slower enterprise implementation of custom AI stacks, normalization of GPU supply, or a pause in hyperscaler capex — that can reverse flows and expose stretched multiples.

Competitive differentiation will matter more than “AI exposure” headlines: companies that control firmware, proprietary interconnect stacks or recurring software monetization (versus one‑time hardware sales) will capture most upside and sustain margins. Secondary beneficiaries include industrial power players and specialty connectors, while legacy low‑margin storage suppliers face asymmetric downside if demand skews toward compute-heavy architectures with different storage patterns.

Given flow-driven momentum, priority is to express convexity via option structures and pairs rather than outright long beta. Size exposure to execution risk (sales cadence, contract delivery) and prefer instruments that limit downside from a rapid sentiment unwind while keeping upside if capex continues to accelerate.

Key Decisions for Investors

• Long AVGO (6–12 months): Buy stock or a 6–9 month call spread to capture continued hyperscaler interconnect demand. Target +25–35% if data‑center capex stays strong; max downside ~15% if semicapex softens. Size 1–2% NAV.
• Conditional directional on PLTR (12–24 months): Buy LEAP call spread (decent delta long, financed by nearer‑dated call sales) to capture enterprise/defense AI deal acceleration while capping premium. Reward asymmetry favorable if execution proves repeatable; downside limited to paid premium.
• Pair trade (6 months): Long ANET / Short WDC equal‑dollar exposure to play rotation into networking optics and away from legacy capacity storage. Target gross pair return +20% with a hard stop at 10% on the pair; this isolates infra vs archival demand risk.
• Infrastructure exposure to power & connectors (12–36 months): Establish a small core position in GEV and APH (1–2% NAV combined) to play secular electrification and fiber demand. Hedge sector cyclicality by keeping position via long dated calls or buying stock with tight stop‑loss; expect multi‑quarter realization of value.