This Software Company Is Betting $25 Billion That Its Stock Is Severely Undervalued Right Now

Salesforce authorized a $50B buyback (issued $25B debt and executed ~80% of it) while positioning AI as a growth lever: Agentforce sales rose 169% YoY to $800M and AI-related ARR reached $2.9B. Management raised long-term targets to $63B revenue by fiscal 2030 (from $41.5B in FY2026), implying ~11% annualized revenue growth and an adjusted operating margin target of ~40% (vs 34.1%), driving ~15.5% average operating income growth over four years. Shares trade at ~14x forward EPS and ~10x next‑year EBITDA, supporting management’s aggressive buyback strategy.

Analysis

The AI upgrade cycle creates a two-speed market: vendors that own workflow-level data and can embed agentic copilots have leverage to both raise wallet share and monetize incremental usage, while point solutions or infrastructure-lite vendors face displacement. That bifurcation favors incumbent CRM platforms plus adjacent data-integration vendors (they capture sticky upstream spend) and GPU/data-center suppliers who see step changes in throughput demand; conversely, modular point-tool vendors and any seller reliant on seat-based churn are most exposed to substitution risk.

A non-obvious supply-side dynamic: accelerating token consumption makes compute economics the dominant driver of gross margins going forward. As model efficiency and chip supply evolve, unit economics can flip from margin headwind to tailwind — meaning short-term revenue growth can be monetized far more profitably if the company keeps pricing power on credits or unlimited plans. Conversely, hyperscalers or commoditized model providers could force a price war on token access, compressing realized ASPs even as volume grows.

Balance-sheet choices are the operational wild card. Debt-funded capital returns mechanically shrink liquidity and optionality for M&A or accelerated sales/infrastructure spending; that amplifies downside if adoption stalls. Near-term catalysts to watch are enterprise order cadence, token-price realizations per customer cohort, and GPU availability; any negative surprise in those three can unwind the multiple quickly, while sustained customer ROI from agentic features could re-rate the stock materially over 12–36 months.

Key Decisions for Investors

• Long CRM equity + protective puts: Buy CRM shares sized to conviction (e.g., 2–4% portfolio) and purchase 1-year puts ~15–20% OTM sized at 10–15% of the notional long to cap tail risk. R/R: if AI upsell and margin expansion materialize, target 30–50% upside in 12–24 months; downside limited to put cost plus capped draw to strike.
• Relative-value pair: Long INFA (data-integration exposure) vs short a small-cap pure-play seat-based SaaS name (replaceable by copilots). Timeframe 6–12 months; target asymmetric capture of data-pipeline spend reallocation. R/R: expect 2:1 upside if enterprises prioritize data plumbing; stop-loss at 20% adverse move.
• Directional AI infra play via options: Buy NVDA 6–12 month call spread to capture incremental GPU demand while controlling premium outlay (buy nearer-term ITM call / sell higher strike). R/R: targeted 3:1 payoff if enterprise AI rollouts accelerate; risk limited to spread cost, but be mindful of supply-cycle volatility.
• Hedge / short trade on legacy silicon: Buy INTC 9–12 month put spread to express downside vs NVDA on AI execution risk. R/R: hedge asymmetric downside from competitive erosion in data-center AI; limit loss to premium paid for the spread.