Nvidia's Stock Hasn't Been This Cheap Since 2019. Here's Why It's the Best Buy in the Market Now.

Nvidia is trading at ~31x trailing P/E, its lowest level since 2019, and the article argues this AI-chip selloff leaves valuation mispriced versus peers. Wall Street expects Nvidia revenue growth of ~82% in the current fiscal year and ~42% next year, with the author claiming next year’s growth may be underestimated. Relative to Apple (~40x) and Alphabet (~27x), and even faster-growth expectations for Nvidia versus peers, the piece frames the stock as a discounted buy as the AI infrastructure build-out continues.

Analysis

The market is treating NVDA less like a hypergrowth story and more like a cash-flow compounder, which is why a lower multiple matters more now than the raw growth rate. That usually favors a re-rating only if earnings keep beating by enough to pull estimates higher; otherwise the stock can stay cheap for longer even in a strong end-market. Near term, the biggest mechanism is index/portfolio rebalancing: if investors rotate from “AI beta” into “AI picks-and-shovels,” NVDA can lag even while fundamentals stay strong.

Relative-value dynamics look more interesting than outright longs. AMD is the obvious valuation pressure point because it still trades on a far more optimistic future-share-gain narrative; if NVDA remains the category leader on both growth and margin, AMD is vulnerable to multiple compression first. AVGO is a better-quality peer, but the market may increasingly reward diversified AI exposure over pure accelerator exposure, which caps NVDA’s upside multiple relative to software-adjacent infrastructure names.

The contrarian miss is that a low P/E on a cyclical semiconductor leader is not automatically cheap if the denominator is peaking. The stock can underperform for months if hyperscaler capex growth decelerates, export restrictions tighten, or gross margin normalizes faster than expected. What would falsify the bullish setup is a guide-down in next-quarter data-center growth or any evidence that customers are stretching refresh cycles beyond the current AI build-out window.

Over 6-18 months, the best setup is still a continued earnings compounding story, but the easy money is likely in relative trades, not an outright chase. If NVDA keeps beating while the group de-rates, the spread between it and AMD should widen materially; if AI spending broadens beyond a few hyperscalers, AVGO and the broader SOXX basket could catch up faster than NVDA expands from here.

Key Decisions for Investors

• Long NVDA / short AMD as a 1-3 month relative-value pair: NVDA has the stronger earnings momentum and lower valuation risk, while AMD remains more exposed to narrative disappointment and multiple compression.
• Buy NVDA only on weakness into earnings or post-capex headlines; do not chase after a single-session rerating. Use a staggered entry and size for 6-18 month compounding, not a quick trade.
• If already long AVGO, keep it as the higher-quality AI infrastructure exposure versus adding more NVDA here; AVGO has less multiple risk if AI capex broadens but chip leadership becomes less concentrated.
• Set a watch item on next hyperscaler capex commentary and NVDA data-center growth guidance. A clear slowdown there would be the fastest thesis breaker and should trigger de-risking across NVDA and AMD.
• For investors seeking sector exposure without single-name beta, favor SOXX/SMH over chasing NDAQ-style commentary. The trade works only if AI capex stays durable; if it rolls over, the basket will de-rate together.