Salesforce Just Pulled the Biggest Buyback Lever in History

March 16, 2026 marked something genuinely unprecedented. Salesforce kicked off a $25 billion Accelerated Share Repurchase — the largest ASR ever executed — delivering roughly 103 million shares priced off the March 11 close. That's 80% of expected total repurchases, with the remaining 20% settling in Q3 or Q4 of FY2027. The stock jumped about 2.8% on the news. Not bad for a Monday.

Here's what makes this unusual: Salesforce didn't tap its treasury to fund it. The company issued $24.9 billion in senior notes — maturities running from 2028 all the way to 2066, its biggest bond offering ever — and funneled the proceeds straight into the ASR. On top of that, a new $6 billion five-year term loan refinanced earlier Informatica-related debt. In one trading day, Salesforce rewired its entire balance sheet.

CEO Marc Benioff called it "confidence in the future of Salesforce." COO and CFO Robin Washington pointed to the "durability of growth and cash flow trajectory." These aren't hollow talking points. FY2026 revenue hit $41.5 billion. Operating cash flow reached $15 billion, free cash flow $14.4 billion, and remaining performance obligations sit at $72.4 billion. That's not a company falling apart.

What the bond market whispered, though, was different.

Credit investors reportedly demanded wider spreads than comparable investment-grade issuers typically see. Rating agencies are growing visibly uneasy with the leverage climb. That's the part bulls seem eager to skip past.

Salesforce is levering up into a transition, not after one. EPS growth expectations have already compressed from a projected 27.8% CAGR down to roughly 15%. The ASR mechanically flatters EPS and FCF-per-share optics, and it does signal management genuinely believes the stock is underpriced. What it can't do is answer the harder question lurking underneath: what does Salesforce's competitive moat actually look like when AI agents increasingly handle CRM workflows outside the traditional seat-based model?

Agentforce is the company's answer — reporting $800 million in ARR, with Agentforce and Data Cloud combined crossing $2.9 billion ARR. Against a $41.5 billion revenue base, that's directionally promising but not yet architecturally decisive. Key questions stay open. How much of that Agentforce ARR represents genuinely new spend versus reclassified bundle revenue? And can organic growth reaccelerate without acquisition fuel?

Oracle's recent surge sharpens the contrast uncomfortably.

Oracle rallied roughly 10% the prior week, driven by AI infrastructure momentum — OCI hyperscale demand, cloud backlog visibility, the whole compute-scarcity trade. Meanwhile, Salesforce needed a record buyback to move the needle. That contrast is telling, though the comparison demands precision. Oracle's rerating isn't about winning CRM share. The market is rewarding names sitting on the compute bottleneck side of AI and demanding proof of ROI from names on the application monetization side. Oracle shows backlog. Salesforce still has to show returns.

SAP, quietly, is the cleaner story here.

The January 29 selloff — SAP's worst single-day drop since 2020, around 15% — looked dramatic. The underlying numbers told a calmer story. Cloud revenue grew 26% constant currency. Cloud ERP suite grew 32%. Total cloud backlog grew 30%. Current cloud backlog grew 25%. The selloff reflected consensus expectations getting reset, not a thesis falling apart. SAP's separate €10 billion two-year buyback, announced alongside Q4 results in February 2026, is textbook shareholder-return optimization — not a debt-funded emergency play.

The structural difference matters enormously. ERP sits closer to a company's financial and operational core than CRM does. In an AI-driven architecture, governed master data and mission-critical process controls become more valuable over time. SAP's real risk is that AI compresses the application layer unless it can own the orchestration layer sitting above core data — a genuine challenge, but a stronger starting position than most bears currently credit.

Salesforce's ASR is defensible capital allocation and a credible valuation signal. It's not a growth catalyst. SAP's multiple reset looks temporary rather than structural. Between the two, SAP offers the cleaner risk-reward setup. Neither buyback, however, resolves the central question both companies must eventually answer — what does enterprise software authority look like when the agent, not the seat, becomes the fundamental unit of value?

not investment advice