This Hedge Fund Just Sold Nvidia and Bought AMD Stock. Which Artificial Intelligence (AI) Chipmaker Is the Better Buy?

Tiger Global initiated/added to AMD in Q2 (position ~1.6% of portfolio, ~$400M) while trimming Nvidia by ~7%, with Nvidia still the fund’s third-largest holding at 9.3% (~$2.2B). Despite AMD outperforming in 2026 (+110% vs Nvidia +12%), the article argues Nvidia offers better value on trailing/forward P/E, implying more upside from business performance as AMD’s valuation normalizes. Overall, it’s a bullish tilt toward Nvidia relative to AMD, but based on dated 13F positioning rather than new fundamentals.

Analysis

This looks more like a positioning signal than a fundamentals signal. The key market mechanism is that AMD is trading on a much higher expectations base, so incremental hedge-fund sponsorship can help near term, but the stock has less room for error if revenue ramps or gross margin expansion slip even modestly. NVDA, by contrast, is the cleaner “quality at a price” exposure: a lower multiple plus stronger cash generation gives it a better downside buffer if AI capex growth normalizes.

The second-order effect is relative-value pressure within semis rather than a clean directional call. If more growth funds rotate from NVDA into AMD, the impact is likely temporary because the true marginal buyers of NVDA are still strategic allocators, index flows, and company buybacks, not 13F readers. Over 1-3 months, the real catalyst is earnings guidance and hyperscaler capex commentary; a single strong AMD print can extend momentum, but the setup is fragile if sell-side estimates stop rising.

Contrarianly, the market may be underestimating how much execution risk remains embedded in AMD’s rerating. The bull case requires sustained share gains in AI accelerators plus no margin giveback, which is a high bar versus a leader that already has ecosystem lock-in and scale advantages. On the other side, NVDA could underperform tactically if investors keep extrapolating “too good to be true” growth into a deceleration regime, but that would need evidence in order flow or guide revisions, not a stale quarterly filing.

Key Decisions for Investors

• Initiate a relative-value long NVDA / short AMD pair for 1-3 months ahead of next earnings cycle; thesis is valuation asymmetry with lower earnings risk on NVDA. Falsify if AMD raises AI revenue guide and NVDA signals a capex slowdown or margin compression.
• If already long semis, rotate beta from AMD into NVDA on pullbacks rather than chasing AMD strength; AMD’s premium multiple leaves less room for any execution miss over the next 1-2 quarters.
• Use SOXX or SMH as a hedge around any single-name position: long NVDA vs short SOXX reduces market beta while preserving the idiosyncratic valuation spread.
• Watch AMD’s next guide and hyperscaler capex commentary as the real catalyst; if AMD sustainably outgrows NVDA for two consecutive quarters, the pair trade should be reduced. If not, expect valuation re-rating back toward the leader.
• For higher-conviction risk control, buy NVDA calls funded by AMD call overwrites/covered shorts into earnings; payoff improves if NVDA resumes leadership while AMD’s multiple contracts.