Nvidia, Palantir AWOL, But Top Funds Go Bonkers For This AI Mobilizer

Top mutual funds materially increased exposure to AI-related names, with money managers scooping up roughly $20.15 billion in new positions, while AI bellwethers Nvidia and Palantir did not appear on the latest list of new buys. Notable beneficiaries included Alphabet and Broadcom, underscoring continued institutional appetite for AI plays and providing potential support to valuations and momentum in the sector ahead of near-term catalysts such as Broadcom's Q4 results.

Analysis

Market structure: Large active managers rotating into GOOGL and AVGO (and away from NVDA/PLTR) reinforces a bifurcation: hyperscaler cloud and semiconductor infrastructure (GOOGL, AVGO, ORCL) are winners with expanding pricing power for datacenter services and custom silicon, while pure-play analytics/AI integrators (PLTR) face tougher fund flows and multiple compression. Expect 6–12 month revenue multipliers favoring infrastructure names by ~0.5–1.0x EV/S relative to small-cap AI names as capex re-accelerates. Cross-asset: risk-on flows should push real yields +10–30bp near-term, lift USD marginally, raise equity implied vols for small AI names and press energy commodity demand from datacenter expansion.

Risk assessment: Key tail risks are (1) US/China export controls on datacenter CPUs/GPUs within 30–90 days, (2) an adverse antitrust action vs GOOGL within 6–18 months, and (3) a liquidity shock if concentrated fund positioning reverses (fast 15–25% downside in crowded names). Immediate (days) risk is flow-driven dispersion; short-term (weeks/months) is earnings and guidance; long-term (quarters/years) is adoption cadence and regulatory regimes. Hidden dependencies include hyperscaler procurement cadence and Broadcom supply-chain constraints that can amplify swings.

Trade implications: Favor overweight infrastructure: establish 2–3% long in GOOGL and 2% in AVGO (6–12 month horizon) before next earnings, with 10–15% stop-losses. Pair trade: long AVGO vs short PLTR (1.5:1 dollar exposure) to capture margin and flow differential over 3–9 months. Use options: buy AVGO 3-month 10–15% OTM call spreads ahead of earnings to limit max loss to premium; sell short-dated covered calls to finance 12–18 month LEAP buys on GOOGL if you want lower cost exposure.

Contrarian angles: The market may be underpricing NVDA pullbacks as temporary de-leveraging rather than secular decline — a disciplined dip-buy (1–2% position) on >12% NVDA pullback could pay off over 6–12 months. Conversely, consensus momentum into GOOGL/AVGO could be overbought; trim longs if either rallies >20% pre-earnings. Historical parallels: 2016–17 cloud hardware rotations show momentum chases can reverse sharply on one earnings miss, so size positions to withstand a 15–25% volatility shock.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in GOOGL with a 6–12 month horizon ahead of next earnings; use a 12% stop-loss and plan to trim 40% of the position if the stock rallies >20% pre-earnings.
• Establish a 2% long position in AVGO and buy a 3-month 10–15% OTM call spread (cost = premium) ahead of earnings to capture upside while capping downside; target 50% return on the spread, max loss = premium paid.
• Initiate a relative-value pair: long AVGO (notional X) and short PLTR (notional 0.67X) for a 3–9 month trade to exploit infrastructure demand vs. pure-play multiple compression; set combined stop at a 15% P&L drawdown.
• If NVDA drops >12% from current levels, allocate a tactical 1–2% buy-the-dip position for a 6–12 month hold, otherwise avoid initiating new NVDA exposure given crowding risk.
• Reduce exposure to small-cap pure-play AI/software names by 50% of current weights and redeploy into cloud infrastructure and semis (GOOGL, AVGO, ORCL) over the next 30 days; monitor export-control headlines and hyperscaler capex reports as triggers to re-adjust within 30–90 days.