Nvidia's Jensen Huang Thinks the Software Sell-Off Is Wrong: 1 Great Stock to Buy on a Dip

Mgmt expects $1.0B in free cash flow in 2026 and the stock trades at ~19.4x this year’s expected FCF, implying a multiyear-low valuation. PTC is embedding AI into its CAD/PLM solutions to leverage product digital-thread data, while ARR continues to grow at a high single-digit rate despite weak U.S. manufacturing. A cyclical industrial recovery combined with AI-driven product enhancement could materially boost earnings, though near-term sector weakness remains a risk.

Analysis

Embedding generative agents into product lifecycle software creates an asymmetry most investors miss: the data accumulated across design→build→service cycles converts one-off license relationships into ongoing model-training economies. That raises lifetime value non-linearly — each incremental deployment improves inference quality for co-tenants in the same industrial vertical, which can lift gross retention and justify premium feature tiers rather than simple seat-based pricing.

The compute footprint of that strategy is a second lever that will determine margin capture. If the vendor pushes inference to the edge (OEM gateways, on-prem appliances) the value accrues to GPU/accelerator suppliers and to the software vendor via appliance bundles; if heavy cloud inference prevails, cloud providers and variable OpEx win while software margins compress. That bifurcation creates an actionable leaderboard among chip and cloud suppliers where timing of wins matters (near-term edge pilots vs multi-year cloud contracts).

Operational risks are concentrated and measurable: slow integration at large OEM accounts, open-source model substitution, or a manufacturing capex pause would each delay monetization by 6–24 months. Conversely, a string of enterprise proofs-of-value or a non-exclusive hardware partnership would be a catalyst for multiple expansion; monitor large contract announcements and margin trajectories as 3–12 month leading indicators of a re-rating.

Key Decisions for Investors

• PTC — Buy equity (size 1–2% net exposure), 12–24 month horizon. Reward: capture ARR/LTV revaluation if industrial AI features drive stickiness (target +35–50%). Risk: cyclical manufacturing slowdown or failure to monetize AI features; set protective stop at -15–20% and scale into weakness on 2–3 consecutive quarters of stable retention metrics.
• PTC — Buy 12–18 month call LEAPS (small exposure 0.5–1% portfolio) to express asymmetric upside while limiting capital at risk. Entry when implied vol compresses below its 6‑month median; take profit if calls double or if stock up >40%, and cut loss if premium declines 50% from entry.
• Long NVDA / Short INTC pair, 6–12 month horizon, equal notional sizing (1–2% net long thematic). Rationale: capture GPU-led edge/cloud inference upside vs CPU/legacy wafer-cycle risk. Risk/reward: expect NVDA to outperform INTC by ~30–50% in an AI-driven industrial upgrade cycle; unwind if NVDA underperforms by >20% or if Intel announces material accelerator wins.
• Event hedge — Buy PTC into near-term earnings or partner-announcement windows and finance with short-dated call sales (covered or cash-secured) to lower basis. Keep net upside uncapped with modest premium income; exit prior to close when the catalyzing announcement is public to avoid post-event volatility decay.