Jensen Huang Told Investors in June to 'Buy at a Discount' During Nvidia's Sell-Off. Nearly Three Months Later, Nvidia Is Up 4%, and a Broader AI Basket Gained 8%. Did His Call Pay Off?

Nvidia CEO Jensen Huang told investors to buy Nvidia during AI-stock price weakness ~3 months ago; since then NVDA is up ~4%, roughly in line with the S&P 500 but trailing an equal-weight AI basket that gained ~8% (driven by Microsoft +24%, Alphabet -5%, Amazon +8%). The article cites strong fundamentals for Nvidia’s fiscal Q2 2027 results, with revenue +18% sequentially and +106% year over year, supporting the view that AI chip demand remains robust. Valuation is framed as a relative discount with NVDA trading at ~17.5x P/E versus a 5-year average of ~23x and below the S&P 500’s ~25x.

Analysis

The better read-through is not about one chip vendor’s near-term stock path; it is about where AI cash flows are actually accruing. In the current phase, hyperscalers with the cheapest cost of capital and the widest distribution are better positioned to monetize AI through cloud, software attach, and workflow lock-in, while semiconductor returns become more sensitive to any deceleration in capex growth or a normalization in multiples.

MSFT looks like the cleanest beneficiary because AI can be sold as an incremental software feature with limited incremental customer acquisition cost. AMZN also has a favorable setup if AWS utilization keeps rising, but the market will punish it quickly if AI spend compresses cloud margins before revenue catch-up. GOOG is the most fragile of the mega-cap AI names because AI can both defend and cannibalize its core search economics; if monetization lags user adoption, the stock can underperform even in a strong AI tape.

The contrarian point is that NVDA’s short-term underperformance may actually be a valuation problem, not a demand problem. If enterprise and sovereign AI spend broadens beyond the current hyperscaler cohort over the next 6-18 months, the stock can rerate again; if not, the market will increasingly value recurring AI revenue at the software layer more highly than one-time hardware sell-through. The key falsifier is any evidence that capex budgets flatten or that cloud providers start talking about ROI payback periods rather than expansion, which would pressure NVDA first and the broader AI basket shortly after.

Key Decisions for Investors

• Favor a 1-3 month relative long MSFT / short NVDA pair: MSFT has more visible AI monetization and lower earnings fragility, while NVDA is more exposed to multiple compression if capex growth slows; cover if NVDA guidance implies another leg of shipment acceleration.
• For fresh capital, prefer AMZN over NVDA on pullbacks for a 6-12 month horizon: AWS can convert AI demand into recurring revenue, but only if margin commentary stays stable; trim if cloud margins compress without a corresponding revenue inflection.
• Avoid chasing GOOG into strength until AI search monetization is clearer; use it as a hedge short against MSFT in a basket trade if you want to isolate the best AI monetization model.
• Watch next earnings/capex guides as the catalyst: a pause in hyperscaler AI spend would be a tactical short NVDA signal, while another round of raised capex plans would invalidate that view quickly.