Back to News
Market Impact: 0.45

Nvidia Shares Are Actually Cheaper Than They Were Before ChatGPT. Here's Why.

NVDAINTCNFLXNDAQ
Artificial IntelligenceTechnology & InnovationCorporate EarningsCapital Returns (Dividends / Buybacks)Company FundamentalsAnalyst InsightsInvestor Sentiment & PositioningMarket Technicals & Flows
Nvidia Shares Are Actually Cheaper Than They Were Before ChatGPT. Here's Why.

Since ChatGPT's Nov. 30, 2022 debut, Nvidia shares have climbed from an adjusted $1.69 to $191.10 (≈11,208%), while quarterly net income rose from $680 million to $39.1 billion (a 4,591% increase). Last quarter EPS grew 65.3% year-over-year, the stock trades at roughly 46x earnings (about 34% cheaper on a P/E basis vs. Nov. 30, 2022), and aggressive share-repurchase programs ($60 billion plus an additional $50 billion announced) have materially boosted EPS amid four consecutive quarters of upside analyst surprises, supporting the case for further upside.

Analysis

Market structure: NVDA is the primary beneficiary — outsized AI inference/datasheet demand gives it durable pricing power in datacenter GPUs while cloud providers (AMZN/MSFT) and HBM memory suppliers also gain; legacy CPU vendors (INTC) and commodity GPU suppliers are at risk of margin erosion. Massive buybacks ($60B + $50B) have mechanically boosted EPS and tightened free float by a material amount, amplifying price moves on earnings beats or misses. The supply-demand picture remains demand-driven (AI model training/inference) with upstream capacity (TSMC/HBM) as the chokepoint — expect pricing power to persist until foundry capacity expands meaningfully over 12–24 months. Risk assessment: Tail risks include US/China export controls or anti‑trust action that could remove key markets, a sharp slowdown in hyperscaler capex, or rapid competitor pricing pressure; any such event could wipe out 30–50% of consensus forward EPS. Near-term (days–weeks) sensitivity is high around earnings/guidance and volatility; medium-term (3–12 months) depends on buyback cadence and fab capacity; long-term (multiple years) depends on AI adoption curves and memory supply growth. Hidden dependencies: NVDA’s economics are tightly coupled to TSMC node availability and HBM pricing; memory cost shocks would compress gross margins quickly. Catalysts: new product launches, cloud contract announcements, and quarterly buyback execution reports. Trade implications: Direct play — establish a 2–3% long NVDA equity position now and scale to 5% on pullback to ~$150 (≈20% off current) with 8–12% stop; size to risk budget. Options — buy a 6–9 month call-debit spread (e.g., NVDA Sep/Dec 2026 200/350 call spread) to target 30–60% upside with defined max loss, financed by selling near-term OTM calls if comfortable capping upside. Pair trade — long NVDA / short INTC equal-dollar 1–2% to isolate AI exposure vs legacy CPU cycle risk. Rotate 3–5% from long-duration growth into AI infrastructure suppliers and selective semicap names ahead of fiscal 2026 capex cycles. Contrarian angles: Consensus underweights buyback leverage risk — EPS growth has been as much buyback-driven as operating income-driven, so earnings per-share could decelerate if buybacks slow, creating a re-rating risk. Market may be underpricing the probability of cyclical memory/foundry oversupply within 12–18 months which would pressure margins; historically semicycle peaks (2017–18) show rapid mean reversion. Unintended consequences include increased regulatory scrutiny as valuation and market dominance grow; plan exits if guidance misses or P/E expands beyond 60 without commensurate forward EPS revisions.