Nvidia Hits a New All-Time High. Is the AI Stock a Buy?

Nvidia set a new intraday record of $243.37 and reached a $5.8 trillion market capitalization, while the S&P 500 and Nasdaq also hit record highs. The article cites a 30 P/E versus higher multiples for AMD (166), Intel (103 adjusted) and Broadcom (48), and forecasts Nvidia fiscal 2028 revenue of $687 billion, up 67%, with adjusted EPS of $15.80. It calls the shares undervalued and a strong buy, while flagging growth-maturation and competition risks; cooler PCE inflation and weak employment data have reduced the likelihood of future Fed rate hikes.

Analysis

The bullish setup is increasingly a rates-and-positioning trade layered on top of an operating thesis: lower yields support long-duration growth, while fresh highs can attract momentum flows. That can extend the move near term, but also makes NVDA more exposed to an inflation or rates reversal than its headline valuation suggests.

The key medium-term risk is not simply slower AI spending; it is a change in where the economics accrue. Amazon and Alphabet developing in-house accelerators could cap GPU pricing power even while total compute demand grows. That substitution may shift value toward networking and custom silicon suppliers such as Broadcom and Marvell, but does not guarantee they capture equivalent profits. Conversely, Nvidia’s software ecosystem and deployment friction can delay substitution; monitor customer adoption and workload performance, not announcements alone.

Treat peer P/E comparisons and distant-year EPS as weak evidence of cheapness: earnings bases, accounting adjustments, and forecast horizons may differ. Verify the forecast’s fiscal-period and scope before underwriting it. An IPO valuation for an AI customer is not proof of funded compute demand or Nvidia revenue.

Over days, momentum may persist; over 1–3 months, rates, cloud-capex commentary, and Nvidia’s next results matter more. Over 6–18 months, custom-chip penetration and data-center returns determine whether growth converts to durable pricing power. The thesis weakens on cloud-capex reductions, slowing data-center revenue, margin deterioration, or rising yields; a broad AI selloff could compress multiples before estimates fall.

Key Decisions for Investors

• Do not chase strength after a sharp run. For an existing position, retain core exposure but add only on a pullback or after earnings confirm sustained data-center demand and margins; size for high volatility.
• For incremental bullish exposure, consider a defined-risk NVDA call spread rather than outright calls, conditional on acceptable implied volatility and a catalyst within the option tenor. If volatility is expensive, wait rather than force the trade.
• Track Amazon and Alphabet custom-chip deployments, cloud-capex guidance, Nvidia data-center revenue and gross-margin trend, and the source and fiscal basis of forward EPS estimates. These are more informative than peer P/E rankings or IPO headlines.
• Falsify the near-term momentum thesis if inflation or yields reverse higher and NVDA loses its post-breakout support; reassess the structural thesis if cloud providers cut capex, Nvidia’s data-center growth decelerates materially, or margins weaken while custom silicon adoption accelerates.