Invest1 publisher3 min readPublished
Fortune's SaaS debt-trap thesis rests on an 83% collapse in revenue multiples
A Fortune column says SaaS incumbents answered collapsed multiples with record debt and buybacks presented as AI strategy. It quantifies the de-rating precisely and leaves the debt unsized at the five companies it names.
The Investor · Invest desk
What happened
- A Fortune column argues that when SaaS multiples collapsed, the incumbents took on record debt, bought back their own stock, and presented the whole exercise as an AI strategy.
- It names Salesforce as running the most extreme version of that response, with HubSpot, Workday, ServiceNow and Adobe running variations of it.
- Its central price evidence is that median SaaS revenue multiples went from 18 times in 2021 to about 3 times.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- constraint With the category's median revenue multiple near 3 times, issuing equity is an expensive way for an incumbent to fund a repurchase or a deal, so the cheaper money is borrowed and it arrives with a coupon.
- exposure Salesforce, Adobe, ServiceNow, Workday and HubSpot are named in a distress case whose debt figures sit outside the record, so their own next filings are where the claim gets tested.
- decision Anyone weighing an incumbent at 3 times revenue against a $965 billion private mark has to decide which of those two prices it could actually transact at.
- precedent If the endpoint the column names holds, the exit for a de-rated incumbent is a takeover offer priced off its debt schedule.
Start with what the revenue has to do. Anthropic booked $11.5 billion in the second quarter of 2026 [10], which annualizes to $46 billion [1], about $19 billion below the run rate Bloomberg reported for the end of July [10]. To top $100 billion annualized this year, the figure Bloomberg attributed to New York Times reporting [9], a quarter has to come in near $25 billion, roughly 2.2 times the second quarter [2]. Growth of 14 times year over year puts the comparable quarter a year earlier at about $820 million [3].
The market-cap comparison in the column sets two different kinds of price side by side. Anthropic's $965 billion [11] is a private mark set by whoever bought the last round. The roughly $544 billion across Salesforce, Adobe, ServiceNow, Workday and HubSpot [12] is what five public companies traded at. The private mark is about 1.8 times the five combined [4], and adding OpenAI's $852 billion [13] takes the two labs to $1.817 trillion, about 3.3 times [5].
The de-rating is the best-evidenced part of the case. The median revenue multiple has fallen about 83% from its 2021 level [6], and in 2025 the SaaS index underperformed the S&P 500 by 24.1 points [7]. The column does not give debt balances, maturities, buyback amounts or interest coverage for any of the five companies it names [17], and those are what would turn "a debt-driven death loop that ends in a Bending Spoons offer letter" [3] into a calculation.
One piece of its own evidence complicates the timing. The column's author, quoted by Steve Rosenbush in the WSJ's CIO Journal, said of AI buyers: "Most of them are not using the end of the frontier. A version from two years ago would be perfectly fine. There are audiences that barely can prompt. So a delay wouldn't make a big difference." [14] Buyers who barely can prompt do not swap a renewal for an AI-native build this year. The displacement evidence on offer is IBM's 13.2% fall on February 23, its worst single day in more than 25 years, after Anthropic showed Claude Code modernizing COBOL [6], and Jasper going from $120 million to $55 million of revenue in a year [7]. One is a one-day share price move. The other is a single company.
In my view the compression is established and the leverage claim is a hypothesis with the five most obvious names attached to it. Two other readings fit the same figures. Anthropic's run rate may be mostly new budget instead of seats cancelled at the incumbents, in which case those incumbents de-rate on slowing growth and service their debt out of operating cash. Or the private marks move first, and the correction lands on the expected IPO price of about $1.93 trillion [9]. The thesis fails if the five hold flat or growing net revenue retention through the next two reporting cycles and refinance at spreads no wider than their existing paper.
What to watch
- An Anthropic listing would show how much of the reported run rate converts into collected revenue in a filed set of accounts.
- Salesforce's next repurchase and borrowing disclosure would put a figure on the playbook the column describes.
- A takeover bid for one of the five named incumbents would test whether the endpoint the column predicts is arriving.