History Says This Will Happen to Nvidia's Stock in 2027 (Hint: You're Going to Want to Buy Now)

The article argues Nvidia could nearly double from about $210 to roughly $390 per share by the end of fiscal 2028 if it earns the $15.57 EPS Wall Street expects and trades at the S&P 500's 25.1x P/E multiple. Nvidia's trailing P/E has compressed to about 28x from above 50x in 2024 as revenue rose 106% year over year last quarter, while management guides for 70% fiscal 2028 revenue growth and analysts project 90% growth in Q3 and 65% in Q4. The bullish case depends on sustained AI-computing demand, continued estimate beats, and a valuation near 30x earnings, potentially supporting a $10 trillion market capitalization.

Analysis

The relevant debate is no longer whether AI infrastructure spending grows, but whether Nvidia retains enough of the incremental workload mix to defend gross margin as inference becomes a larger share of demand. Hyperscaler capex is increasingly bifurcating: frontier-model training favors Nvidia’s full-stack advantage, while steady-state inference is the natural opening for internal ASICs and lower-cost accelerators. That makes revenue durability more credible than a simple valuation comparison, but makes terminal-margin assumptions the key multiple risk.

Near-term upside depends on evidence that next-generation system ramps are supply-constrained rather than demand-constrained. Watch hyperscaler capex commentary, lead times for HBM/co-packaged networking, and Nvidia’s networking attach rate; a strong GPU shipment number without corresponding networking growth would imply a less favorable product mix. Over the next 1-3 months, earnings revisions and backlog conversion can support the shares; over 6-18 months, the material risk is that customers use custom silicon to reduce Nvidia’s share of inference spend, compressing both growth and the premium assigned to its software ecosystem.

Consensus bullishness may understate Nvidia’s exposure to a capex digestion cycle: all major buyers can remain strategically committed to AI while still shift purchase timing by one or two quarters. Conversely, the bear case often treats customer ASIC development as immediately substitutive; it is more likely additive in training-heavy deployments, leaving Nvidia positioned to capture the highest-value workloads. The thesis is falsified by sequential data-center revenue deceleration combined with gross-margin erosion and lower customer concentration visibility, not merely by a quarterly revenue miss.

Key Decisions for Investors

• Maintain a core long NVDA only on post-earnings confirmation of sustained data-center gross margin and robust networking attach; add on a 10-15% event-driven pullback rather than chase sentiment. Upside requires continued estimate revisions; exit/reduce if management guides to sequential revenue deceleration alongside a material gross-margin step-down.
• Pair trade for a 3-6 month horizon: long NVDA / short AAPL in equal dollar beta-adjusted exposure. This isolates AI infrastructure monetization from consumer-device valuation risk; stop out if AAPL demonstrates a credible paid AI-services or hardware upgrade-cycle acceleration that closes the earnings-growth gap.
• Monitor AMD and hyperscaler custom-silicon disclosures as an alert rather than initiate a short immediately. A credible sign of substitution would be Nvidia inference mix weakening, HBM availability improving without shipment acceleration, or customer commentary explicitly reallocating workloads; absent those signals, competing accelerators remain more likely to expand total compute demand than displace Nvidia training deployments.
• For defined-risk exposure through the next results, prefer NVDA call spreads 6-9 months out rather than outright near-dated calls. This captures revision-driven upside while limiting exposure to a capex-timing reset; avoid if implied volatility already prices a move materially above the stock's historical post-results range.