Salesforce Stock Just Soared. Thank Anthropic.

Salesforce shares jumped nearly 23% to about $252 after fiscal Q2’27 adjusted EPS rose 103% to $5.90, but $2.53 of that came from unrealized investment gains tied to its Anthropic stake. Excluding the investment gains, adjusted EPS grew ~16% Y/Y, while revenue increased 11% to $11.3B and non-GAAP operating margin was 34.1%; operating cash flow rose 71% to $1.3B. Salesforce also raised guidance to FY’27 revenue of $46.1B–$46.4B (11%–12% growth) and reported Agentforce + Data 360 ARR of nearly $3.9B (+210% Y/Y), with Agentforce ARR passing $1.5B.

Analysis

The market is rewarding CRM as if its AI optionality is already monetized, but the core issue is that a large share of the quarter’s earnings was mark-to-market rather than recurring software economics. That matters because the rerating can persist only if the company converts AI distribution into durable attach, retention, and margin expansion; otherwise next quarter’s EPS will mechanically step down as private marks normalize. In other words, the stock is now partly trading on Anthropic exposure, which is a volatile asset-price proxy, not a cash-flow stream.

The more important read-through is competitive: Anthropic embedded inside Salesforce’s workflow stack is a defense against the “AI kills SaaS” narrative, and it modestly improves CRM’s negotiating power versus point solutions. The second-order winners are not just CRM customers but the model and infra layer that sits behind enterprise deployment—NVDA and cloud capacity providers should see incremental demand if enterprise copilots move from demos to usage. But the primary economic beneficiary of this quarter is still buyback-driven EPS, which is a slower, less durable lever than product-led acceleration.

Near term, the stock is vulnerable to digestion after a gap move because the catalyst mix is noisy: part fundamental, part paper gains. Over 1-3 months, the key test is whether ex-gain EPS growth stays in the mid-teens and whether AI ARR continues compounding at a rate that supports a higher terminal multiple. Over 6-18 months, the thesis breaks if the partnership proves decorative, Agentforce decelerates, or the private-mark windfall reverses enough to expose the underlying growth rate as merely solid rather than exceptional.

Key Decisions for Investors

• Avoid chasing the gap: wait for a 5-8% pullback or a close back below the post-earnings breakout zone before adding long CRM; use a mental stop if price fully retraces the move and closes below the pre-report range.
• If building exposure, prefer a 3-6 month CRM call spread over outright stock: upside is a retest of the prior high, while downside is capped if the market re-focuses on ex-gain operating EPS.
• Pair trade idea: long CRM / short IGV for 1-3 months to isolate idiosyncratic AI distribution optionality while hedging broad software multiple compression.
• Watch next quarter’s ex-gain adjusted EPS growth and AI ARR growth as the real falsifiers; if ex-gain EPS slips below ~15% growth or AI ARR slows materially, fade strength and reduce exposure.
• Small thematic watchlist: treat NVDA as a second-order beneficiary only if enterprise AI usage data corroborates the narrative; otherwise don’t extrapolate CRM’s mark-to-market gains into infra demand.