SaaSpocalypse deferred

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Good morning. If you are American, and you’re not already, you need to be cross about eggs. For months, eggflation was a hot political topic in the US — a point of obsession at times for Donald Trump. But it was also a macro issue. In January 2025, we were told avian flu had doubled the price of this key source of protein. Surging egg prices were a big reason for a pickup in US CPI inflation in February of that year. The later pullback in prices helped drag overall inflation back down, making it tricky for economists to figure out what was happening with tariff-led price increases. Now, big US egg producers have “agreed to settle accusations by US and state authorities that they colluded to manipulate a widely used industry benchmark to drive up the price of eggs”. We liked it better when the fraud and manipulation was confined to harmless stuff, like Libor.

Are you cross yet? Tell us: [email protected].

About the SaaSpocalypse

Way back in February, it looked like the end was nigh for software companies. The release of Anthropic’s AI assistant Claude Cowork, in January, crystallised fears about the future of a big sector of the economy. Suddenly it looked like everyone could vibe code their own enterprise software over a weekend, and pay no one for the privilege. The market reckoning was very bad. In less than a week nearly $1tn was wiped from the value of software companies.

Six months later, the sector has yet to recover. There have been a few attempts to claw back the losses, including a rally on the back of some strong first-quarter earnings reports, but they have not stuck. The S&P 500 software and services index remains down 21 per cent from a year ago.

Are the software catastrophists right? Opinions differ, and differences come down to one question: do the companies have a moat? Are their products so entrenched in customers’ businesses that replacing them is unthinkable? If so, the sell-off was badly overdone, and it might be a good time to grab a bargain.

Intuit, Workday, ServiceNow, Salesforce, Oracle are some of the big software companies that have been hit hardest. Forward price-to-earnings ratios have fallen, in some cases to the low teens, although profit growth remains in the double digits. That is: the stocks are pricing a moat that is draining fast.

The counter-argument is that these companies provide “systems of record” which gather and hold a company’s critical data. Their customers are often large, complex organisations operating across jurisdictions and subject to a doorstopper of regulatory requirements. SAP, for example, embeds those compliance and certification requirements, as well as tonnes of proprietary data about customer behaviour and the relationship between databases, directly into its products. There’s no denying AI coding agents mark a step change in how software is developed. But just because start-ups can build it faster doesn’t mean the customers will come. Mark Moerdler at Bernstein says:

Even if tomorrow you could snap your fingers and build an enterprise-class application, fully feature-rich, and it costs you zero — all the other costs and ramp times are still there, closing the clients, bringing the software on to market. That creates an opportunity, and time, for the incumbents to fight back. And I don’t think [Wall Street] is giving them enough credit.

The counter-counter-argument is that the incumbents are not innovating fast enough. Some analysts think they’ve been incumbents for too long, are saddled with legacy technical debt and products built for a different age — and some of them have simply got lazy. Start-ups can move at a pace that the incumbents can’t match. That leads to “value abstraction”: AI-native start-ups building new user interfaces, relegating the incumbents to the role of database managers. On this view, the incumbents that want to survive will need to acquire the best new technology through mergers 

That need not be catastrophic, though. Omar Sheikh at Rothschild & Co Redburn expects the frontier AI labs such as Anthropic and OpenAI to develop the “intelligence layer”, the new interfaces and assistants through which companies will interact with their enterprise data. But, he argues, the AI labs won’t be able to replicate the business logic — the compliance requirements and knowledge of customer behaviour — that the incumbents have built up. So, they’ll need access to the software companies’ APIs and the software-as-a-service business model will evolve to generate significant revenue from API tolls.

Some companies will face greater pressure to survive. While systems of record providers remain relatively well protected, more peripheral, “low or no code” companies that provide reporting tools and dashboards will face a “war for relevance”, according to Moerdler. Sheikh agrees: it’s hard to see the terminal value companies such as Monday.com or Asana.

The important point: the sell-off lacked granular discrimination between business models. The market panicked. The intervening months have provided perspective on the opportunities — not just the risks — that AI presents to the incumbents. The software company moats may be deeper than you think.

One good read

“Catching one’s thoughts as they arise is a lot more difficult than it sounds”

FT Unhedged podcast

Can’t get enough of Unhedged? Listen to our new podcast, for a 15-minute dive into the latest markets news and financial headlines, twice a week. Catch up on past editions of the newsletter here.