What You'll Learn
- Why Salesforce stock dropped 4% on June 2, 2026 and what the BofA Underperform rating actually flagged
- The three structural concerns that drove Bank of America's $160 price target and Tal Liani's full reversal from Buy
- How Agentforce's $800M ARR at 169% growth stacks against the seat-compression thesis management disputes
- The math behind Salesforce's $25B accelerated buyback and why 50 of 50 analysts still carry a $263 mean target
Salesforce (CRM) stock fell 4.18% on June 2, 2026, closing near $178.50 and dragging the Dow to one of its heaviest single-day losers alongside Nike and Microsoft, even as the S&P 500 printed a record 7,599 close. The catalyst was not a missed quarter, not a guidance cut, and not a regulatory action. It was a five-word sentence buried in a Bank of America research note from May 18: "AI-driven structural reset." When BofA analyst Tal Liani reinstated Salesforce at Underperform with a $160 price target, he did not just downgrade a stock. He opened a fault line in the central thesis that has carried the entire enterprise software complex for the last decade: the per-seat SaaS revenue model.
Salesforce's management spent the spring pushing back. At the Morgan Stanley Technology, Media and Telecom Conference in March, Chief Operating and Financial Officer Robin Washington told the audience, "We haven't seen seats year-over-year on quarter-on-quarter decline." President of Enterprise and AI Technology Joe Inzerillo argued that 26 years of customer data gives Salesforce a moat that newer agentic entrants cannot replicate. When the company delivered its Q1 FY2027 earnings on May 27, it launched a $25 billion accelerated share repurchase, the largest in its history, and raised full-year guidance to $45.9 billion to $46.2 billion. The Street did not blink. By the second trading day of June, the stock was 33% below where it started the year.
The deeper story is not about Salesforce alone. It is about whether the 2026 AI rally is masking a quiet revaluation of every enterprise software business that sold its customers on per-seat pricing. As the S&P 500 hit 7,599 with just 4% of stocks at new highs and the broader tape hid software's underperformance, the Salesforce story became a referendum on whether the agentic era is a tailwind or an extinction event for the seats economy. The BofA note landed 10 days before Bitcoin crashed below $70,000 and triggered $766M in liquidations, the same week that Tesla's Austin robotaxi fleet grew to 12 Model Ys, and roughly 14 days before the Fed's June 16-17 FOMC meeting. The macro and the micro collided.
The June 2 Drop: Anatomy of a 4% Selloff in a 7,599 Market
Salesforce's June 2, 2026 decline of 4.18% is best understood against the tape. The Dow Jones Industrial Average closed at 51,322, up 243 points or 0.47%, and the S&P 500 set a record at 7,599. Salesforce was one of three Dow stocks that lost more than 4% that day, alongside Nike at -4.79% and Microsoft at -4.18%. Cisco was the day's biggest gainer at +5.52%, followed by Caterpillar at +5.19% and Apple at +2.88%. The dispersion is the story: a market in which 96% of S&P 500 constituents are not at new highs cannot lift every name, and the software complex was the day's casualty.
Trading volume on June 2 ran roughly 1.6x the 30-day average for CRM, and intraday lows touched $176.40 before a late-session recovery to the $178.50 close. The move was not a single block trade; it was a steady grind lower that began at the open and accelerated into the 2:00 PM ET window when BofA's May 18 note circulated on a Bloomberg terminal refresh. The $160 price target from BofA implies a 10.4% downside from the close; the average Wall Street price target of $263 implies a 47% upside. That $103 gap is the single largest analyst dispersion on the stock in five years, and it explains why the tape punished the name so harshly: when 34 of 50 analysts are Buy-rated, the marginal seller is the one who believes BofA is right.
Salesforce has now lost roughly 41.8% from its April 2026 peak and 33% year to date. The market capitalization, around $156.5 billion, is back to mid-2024 levels. The decline has been driven by a mix of macro software de-rating, AI-monetization anxiety, and what BofA called "muted net new customer additions." For an enterprise SaaS franchise that grew sales 9% to 10% in the last three fiscal years, the slowdown in net new logos is the data point that cannot be dismissed as noise.
BofA's $160 Underperform: The Three Concerns Behind the Downgrade
Bank of America's May 18 note, written by analyst Tal Liani, was unusual for two reasons. First, BofA was previously a Salesforce bull, making this a full reversal rather than a re-rating. Second, Liani did not call the company broken. He called it "a deeply entrenched platform" that is being repriced by AI, a critical distinction. The note identified three structural concerns that together drove the Underperform call.
Concern one: muted net new customer additions. Liani's data showed that Salesforce added fewer net new customers in its last two fiscal years than in any two-year window since going public in 2004. The seat model is mature. Existing customers are still expanding wallet share, but the greenfield flow that fueled the company's 20%-plus growth rates through 2022 has slowed. In a market where management pointed to "six of our top 10 deals" being Agentforce-driven, the offset is real, but it is not yet large enough to compensate for the slowdown at the logo level.
Concern two: limited upsell potential from existing seats. Liani's analysis of customer behavior suggested that the average Salesforce customer's wallet share has plateaued. The argument is that the addressable spend on sales, service, and marketing cloud has a ceiling, and CRM is bumping into it. BofA modeled structurally lower growth at approximately 10% annually, down from the 19% to 22% range that the company printed from FY2021 through FY2024. The 10% figure, if it holds, would reframe Salesforce from a growth stock to a value stock, and value stocks in 2026 do not trade at 3.4x revenue.
Concern three: an underwhelming AI monetization curve. This is the most contested pillar. Liani argued that even if Agentforce reaches scale, the revenue mix shifts toward consumption-based pricing, which is lower-margin and harder to forecast than seat-based ARR. The $25B buyback, he suggested, is a defensive admission that management cannot redeploy the cash at a higher return inside the business. Starboard Value's 13F filing the same week confirmed it exited its Salesforce position in Q1 2026 after pressing on capital allocation through 2025, a notable signal from one of the most respected activist shops in technology.
The $160 price target itself was built on roughly 9x estimated FY2027 free cash flow, well below the 10.5x peer average that Microsoft's enterprise software franchise and ServiceNow's NOW platform currently command. In BofA's framework, even a successful Agentforce rollout would not close that multiple gap, because the market will not pay a growth multiple for a company that grows at 10%.
The Agentforce Counter-Argument: $800M ARR at 169% Growth
Salesforce's defense starts with the numbers. As of the Q1 FY2027 report delivered on May 27, 2026, Agentforce has reached $800 million in annual recurring revenue, up 169% year over year, with combined AI-and-data ARR (Agentforce plus Data 360) exceeding $2.9 billion, up more than 200% year over year. The platform processed nearly 20 trillion tokens in the trailing year and delivered 2.4 billion agentic work units, or AWUs, a Salesforce-defined metric that measures completed AI-driven tasks rather than raw token consumption.
President Joe Inzerillo explained the AWU logic at Morgan Stanley. A task completed by Salesforce's deterministic reasoning engine may use fewer tokens than a pure large language model call but deliver better results, he said. Measuring only tokens would make efficiency improvements look like a decline. The AWU captures actual work delivered. The 2.4 billion figure is meant to demonstrate that the agentic layer is producing real economic output, not just API traffic.
On the customer side, Salesforce reported that 103 deals above $1 million in Q1 FY2027 included a ServiceNow component, evidence that the agentic platform is selling into the largest enterprises. Marc Benioff, the company's co-founder and CEO, told analysts on the Q1 call that "six of our top 10 deals in the quarter are now driven by companies that just want to transform with Agentforce." Premium SKU adoption, the highest-priced tier of the Agentforce product line, grew 300% quarter over quarter in Q4 FY2026, a leading indicator that enterprise buyers are willing to pay up for the AI functionality rather than treating it as a free add-on.
The hybrid model that Robin Washington described at Morgan Stanley is the linchpin of the bull case. "We're seeing the great adoption, the momentum metrics around our agentic products, but we haven't seen seats year-over-year on quarter-on-quarter decline," she said. The argument is that AI agents and human seats are complementary, not substitutive. A salesperson using an AI co-pilot still needs a seat. A customer service agent using a triage bot still needs a seat. The revenue per seat goes up, even if the seat count stays flat. Marvell Technology's 25% surge on Jensen Huang's trillion-dollar call on June 2 showed that the AI infrastructure trade is alive; the question is whether the AI capex arms race produces enough enterprise revenue at the application layer to justify Salesforce's multiple.
The $25B Buyback, Starboard Exit, and Capital Return Math
On the same day Salesforce delivered its Q1 FY2027 results, the company announced a $25 billion accelerated share repurchase, its largest in history, executed under a broader $50 billion authorization. The buyback is structured as an ASR, meaning a large block is retired immediately at a discount to the prevailing market price, with the remainder settled over time. The buyback represents roughly 16% of Salesforce's $156.5 billion market cap and is funded entirely from the company's balance sheet, which ended Q1 with more than $14 billion in cash and short-term investments.
Management's framing is that returning capital is the highest-return use of cash. Robin Washington said at Morgan Stanley that Salesforce returned roughly 99% of its free cash flow to shareholders in FY2026 across buybacks and dividends. "We see no better investment right now than Salesforce," she said. The buyback is, in effect, a statement that the company cannot redeploy the cash into M&A, R&D, or international expansion at a return higher than buying back its own stock at a 30% discount to the 52-week high.
The Starboard exit cuts the other way. Starboard Value, one of the most active and successful tech-focused activist investors of the last decade, disclosed in its Q1 2026 13F filing that it had exited its Salesforce position. Starboard had pressed the company on capital allocation through 2025 and, per the BofA note, the activist concluded that the buyback-heavy strategy was the best outcome they could extract. The implication is that Starboard did not see a path to a strategic re-rating, only a financial-engineering one, and chose to take the 2025-vintage profits rather than wait for an operational catalyst that may not arrive.
For income investors, Salesforce paid its most recent dividend on May 28, 2026, and the dividend yield is roughly 0.7% annualized. That is well below the S&P 500's 1.4% average yield but consistent with where most large-cap software franchises trade. The buyback, not the dividend, is the capital return story. If Salesforce retires 1.5% to 2% of its float per year through the ASR, and the buyback is executed at an average price below $200, the math supports a mid-single-digit EPS tailwind even before the company's own operating performance improves.
The Valuation Gap: 3.37x Revenue vs Microsoft's 8.57x
Salesforce now trades at 3.37x next-twelve-months enterprise value to revenue and 8.50x NTM EV/EBITDA, the lowest forward multiple in the company's history. The peer comparison makes the discount vivid. Microsoft trades at 8.57x NTM EV/Revenue and 13.74x NTM EV/EBITDA. ServiceNow, with similar gross margins to Salesforce but considerably faster expected top-line growth in FY2027, trades at 5.90x NTM EV/Revenue and 15.89x NTM EV/EBITDA. On every forward multiple, Salesforce is the cheapest of the three.
| Metric | Salesforce (CRM) | Microsoft (MSFT) | ServiceNow (NOW) |
|---|---|---|---|
| NTM EV/Revenue | 3.37x | 8.57x | 5.90x |
| NTM EV/EBITDA | 8.50x | 13.74x | 15.89x |
| YTD 2026 Return | -33% | -7% (approx.) | -18% (approx.) |
| 2026 AI ARR Growth | +169% (Agentforce) | Copilot $20B+ run rate | AI Agents 30%+ of new deals |
| Buyback Authorization | $50B (incl. $25B ASR) | $60B (in flight) | $3.5B |
The table is the central tension in the Salesforce story. Bulls argue that the discount is a margin of safety: if the business stabilizes and Agentforce scales, the multiple can re-rate toward ServiceNow's 5.90x, generating 75% in multiple expansion alone. Bears argue that the discount is deserved because the business is structurally lower-growth than the peers. Both interpretations cannot be right; the next two earnings reports will determine which one is.
The earnings cadence matters. Salesforce reports Q1 FY2027 on May 27, 2026, and the report has now been delivered. The next two prints are Q2 FY2027 in early September 2026 and Q3 FY2027 in early December. Between those, the macro will weigh on the multiple: the June 16-17 FOMC meeting will set the rate path, and any signal that the Fed is on hold for longer than markets expect will compress the present value of all software cash flows. The Coinbase-style risk that an entire sector re-rates on a single FOMC dot plot is real, and Salesforce's 33% drawdown reflects part of that exposure.
The Q1 FY2027 Print: Six Numbers That Decide the Debate
With the Q1 FY2027 results now in the rearview mirror, the debate has moved to the next inflection points. The six numbers that will decide whether the BofA $160 target or the Street $263 mean target is closer to reality are these. The data, sourced from Salesforce's investor relations filings and the Yahoo Finance CRM quote page, gives the cleanest read on whether the BofA thesis or the management rebuttal is winning.
Number 1: Q2 FY2027 revenue guide of $11.27B to $11.35B. The mid-point of $11.31B missed the $11.4B consensus by roughly $90 million, a 0.8% shortfall. The market reaction was muted because the miss is small, but the directional signal is what BofA flagged: a company that prints a record Q1 and then guides the next quarter below Street expectations is telegraphing slower growth.
Number 2: Agentforce ARR progression from $800M toward $1B. The TIKR mid-case model assumes Agentforce approaches $1 billion in Q1 FY2027 ARR. The actual Q1 print showed $800M, slightly below the bull case but well ahead of the bear case. The Q2 print in September will show whether the trajectory is $900M, $1.0B, or $1.1B; that single data point is the most important forward indicator in the Salesforce story.
Number 3: Current remaining performance obligations, or cRPO, the most forward-looking bookings indicator. Q4 FY2026 cRPO grew 16% year over year to $35.1 billion. If Q1 FY2027 cRPO holds at 14% to 16%, the bookings story is intact. If it slips below 12%, the seat-compression thesis gains weight.
Number 4: Operating margin trajectory from 34.1% in FY2026. Salesforce's EBIT margin expanded from 30.5% in FY2024 to 34.1% in FY2026, even as it absorbed Agentforce build-out costs. The bear case requires margin compression as AI build-out costs outpace consumption-pricing growth. A Q1 FY2027 operating margin above 33% would confirm the leverage story; below 31% would be the first hard evidence of margin pressure.
Number 5: Net new customer additions. Salesforce stopped reporting exact logo counts in FY2025, but management cited "strong" net new adds on the Q4 FY2026 call without quantification. BofA's downgrade rested on the claim that net new adds are muted. The Q1 FY2027 transcript's tone around logo growth will be parsed carefully.
Number 6: Buyback execution pace. The $25B ASR retires a portion of the float immediately. The remaining $25B under the broader authorization will be executed over the next 12 to 18 months. If the buyback is executed at an average price below $200, EPS accretion runs at 4% to 5% per year; above $220, the accretion is closer to 2% and the multiple expansion argument weakens.
For a final external read, CNBC's May 18 coverage of the BofA downgrade captures the Street sentiment succinctly: the buy-side is split, the sell-side is range-bound, and the catalyst that closes the $103 gap between BofA and the Street will come from Agentforce numbers, not analyst upgrades. The June 16-17 FOMC meeting and the September Q2 FY2027 print are the two most likely moments of resolution.
Salesforce (CRM) is the cleanest single-stock referendum on whether the 2026 AI rally masks a structural revaluation of the per-seat SaaS model. Bank of America says the model is breaking under AI; Salesforce says the model is being augmented by AI. The $103 spread between BofA's $160 target and the $263 Street mean is the most analyst-dispersed name in software, and that dispersion is itself a tradable signal. The June 16-17 FOMC will set the macro rate backdrop, the Q2 FY2027 print in September will resolve the Agentforce trajectory, and the $25B buyback will compress the float regardless of which side is right. For investors, the right framework is not to pick a side. It is to wait for the data and trade the volatility that comes when it lands.