Nvidia Is the Cheapest It's Been Since 2019. Why Investors Should Load Up Now.

Nvidia shares slid from a mid-May all-time high to around $202, with roughly $800B of market cap lost and forward P/E down to 22.6 (vs ~18x forward earnings earlier in the decline). The company lifted its quarterly dividend from $0.01 to $0.25 and announced an additional $80B buyback program, signaling confidence in cash returns. Fundamentals remain strong with record revenue of $81.6B (+85% YoY) and data center revenue up 92%, even as investors weigh rising cost/margin pressures and broader semis weakness.

Analysis

This looks more like a positioning reset than a thesis break: when a $5T platform starts leaning harder on dividends and buybacks, it signals the stock is entering a “cash-return plus growth” regime that can attract a different shareholder base and dampen downside volatility. That said, the market will eventually force a harder question than share count reduction—whether AI demand is merely strong or still strong enough to justify a premium multiple after the initial capex wave is absorbed.

Near term, the biggest second-order winner is not necessarily NVDA itself but the surrounding infrastructure stack: networking, power, and cooling vendors should benefit if hyperscalers keep spending while the GPU leader trades more like a mature compounder than a pure growth story. The clearest relative loser is AMD: if NVDA’s valuation de-risks without a meaningful demand slowdown, competitors have less room to argue for share gains, and any AI spending pause will hit the smaller player’s multiple harder.

The contrarian risk is that the current rebound is being read as “cheap” when it may just be the market pricing in digestion risk. Over 1-3 months, any sign of flattening data-center growth or margin compression would likely re-rate the stock lower even with buybacks in place; over 6-18 months, export restrictions, customer concentration, or a pause in hyperscaler capex are the real falsifiers. NFLX is effectively noise here—there is no direct fundamental read-through, so the trade is really about semis and AI spending breadth, not broader growth stocks.

Key Decisions for Investors

• Long NVDA on weakness using a 3-6 month call spread or staged equity entry; thesis is mean reversion plus capital-return support, but only if data-center growth stays above 50% YoY and gross margin holds.
• Pair trade: long NVDA / short AMD for 1-3 months to express relative moat and cash-return quality; best if AI spending stays strong but the market continues to reward the category leader over share-gain stories.
• If already long semis, rotate part of the basket from high-beta names into NVDA and infrastructure proxies (SMH/QQQ exposure, with preference for network/power vendors) over the next 4-8 weeks to reduce earnings sensitivity.
• Set a downside alert on NVDA if forward revenue guidance or gross margin inflects lower on the next print; that would be the first credible signal that the buyback story is masking a deceleration rather than offsetting one.