Salesforce, Inc.

Salesforce, Inc. (CRM) — Investment Memo

🐂 The Bull Case (Warren's voice)

• Moat is real and sticky. Switching costs are enormous. You can’t rip out Salesforce without rewriting your entire sales process. Data, workflows, third-party apps – all locked in. That’s a moat that survives mediocre management.
• Economics are exceptional – on the surface. Subscription gross margin ~80%. Cash collected up front. FCF margin 35% of revenue. The cash machine hums, even if the engine is stock‑comp diluted.
• Growth runway in AI agents. Agentforce (per‑conversation pricing) could expand total addressable market beyond seats. If productivity gains are real, unit economics explode.
• At what price does it become attractive? If you can buy at a discount to intrinsic value that compensates for capital allocation sins. A 25% margin of safety on the DCF ($416/share) makes the downside of empire‑building tolerable. A 50% margin ($278/share) would be a Buffett‑style entry – assuming the moat holds for 20 years.

🐻 The Bear Case (Charlie inverts)

• Scenarios that permanently impair this business:

  1. AI commoditization kills the per‑conversation model. A decade from now, cheap open‑source AI agents that hook into any CRM destroy Salesforce’s unit‑pricing leverage. Or Microsoft/Google bundles a free CRM. Switching costs don’t matter when the product becomes a commodity.
  2. Stock‑based compensation addiction catches up. Net income is the real earnings power – and it’s volatile (2022: $1.4B, 2023: $0.2B). When growth stalls, dilution will be revealed as the only way profits were ever “growing.” Owners paid for free cash flow with their own equity.
  3. Debt‑fueled empire building collapses. Debt went from $2.0B to $14.4B while organic growth decelerated. If a recession hits, covenants tighten, and acquisitions (Slack, Tableau, Informatica) never deliver the promised synergies. The balance sheet is a ticking time bomb disguised as a SaaS story.

• Most likely threat over what timeframe: AI commoditization – 5–10 years. The per‑conversation bet is a product of faith, not evidence. If AI doesn’t boost productivity enough for clients to keep paying, the whole model cracks. Cisco in the dot‑com bust, but slower.

💰 Valuation & Margin of Safety

• DCF estimate: $515.9B total / $555.37 per share (15% FCF growth, 10% discount rate, 3% terminal growth). But note: FCF is inflated by stock‑comp. Real intrinsic value is lower.
• 25% margin of safety entry: $416.53 (conservative – assumes moat holds but capital allocation remains mediocre)
• 50% margin of safety entry: $277.68 (Buffett’s ideal – compensates for empire‑building risk and dilution)
• Current price: Not provided. If it’s above $416, it’s expensive relative to honest earnings power. If below $278, it’s a deep value trap? Only if the moat is still intact – but the bear case says it’s eroding.

Verdict: PASS

PASS. The moat is real but leaking, the management is an empire‑builder who dilutes you, and the DCF relies on inflated FCF. Even at a 50% margin of safety, the structural threat from AI commoditization makes this a bet on faith, not fundamentals.