Prediction: This Is Where Nvidia Stock Will Be When September Ends

Nvidia’s fiscal Q2 revenue surged 106% YoY to $96.2B and non-GAAP EPS rose 120% to $2.22, beating estimates of $2.09 EPS on $92.3B revenue. The company guided for fiscal 2028 revenue growth of ~70%, well above the 44% analysts expected, citing supply constraints. Shares jumped nearly 9% the day after results, supported by TSMC capex expansion (TSMC 2026 capex +$8B to $62B; additional $100B U.S. spend) and expectations of higher earnings momentum (analyst FY2027 EPS estimate $9.29, +95% YoY).

Analysis

The real signal is not that one GPU vendor keeps winning; it is that AI demand is still outrunning industrial capacity. In that regime, the first durable beneficiaries are the bottleneck owners — advanced foundry, packaging, lithography, and power/networking names — because scarce wafer starts and substrate supply force customers to prepay for capacity, not just chips. That makes TSM the cleaner expression of the capex wave, while equipment vendors like AMAT, LRCX, and ASML should see duration extension if spending stays elevated.

Near term, the stock can still drift higher as sell-side estimates re-anchor, but the next 1-3 months are about digestion risk, not demand risk. If hyperscaler commentary turns more disciplined, or if custom silicon/TPU adoption keeps climbing, NVDA’s share of AI spend can compress even if the overall pie grows. The key second-order effect is that a supply-constrained outlook pushes incremental economics upstream to the supply chain, meaning NVDA can remain strong while relative outperformance shifts to TSM and the equipment complex.

The contrarian read is that the market may be underpricing how much of NVDA’s future growth is already “paid for” in the multiple. A 29x earnings multiple is not obviously cheap when the thesis depends on scarcity pricing persisting and no meaningful demand substitution. Falsifiers over the next quarter: any cut to 2027/2028 growth expectations, softer cloud capex commentary, or evidence that new supply ramps faster than pricing power. If those show up, the sector rerates lower even if absolute revenue keeps rising.

Key Decisions for Investors

• Buy TSM on 2-3% pullbacks over the next 1-2 weeks; better risk/reward than NVDA because it monetizes the bottleneck directly and has lower single-name execution risk.
• Long NVDA / short AMD for 1-3 months as a relative-value expression if estimate revisions keep moving higher; thesis breaks if AMD’s accelerator share gains accelerate or sector multiples re-expand broadly.
• If already long NVDA, trim into strength after the post-earnings gap extension; the stock likely has upside, but the asymmetry worsens if consensus keeps chasing the guidance rather than fundamentals.
• Accumulate AMAT, LRCX, or ASML on weakness as a 6-18 month infrastructure trade; upside comes from persistent capex rather than product-cycle volatility.
• Set an alert on hyperscaler capex guidance and TSM capex conversion over the next 1-2 quarters; a slowdown there is the earliest falsifier for the AI spend compounding story.