AAs the FTSE 100 index hovers near all-time highs, it is easy to overlook a quiet share price decline in one corner of the market. It’s got a name – “cloud crash”, refers to the plug-in legal products added by AI firm Anthropic to its cloud cowork office assistant.
This launch, or as you might imagine from the nervous stock market reaction over the past few weeks, marks the moment when the AI revolution spills out into some of the UK’s biggest public companies – those in the lazy but successful “data” game, including Relax, London Stock Exchange Group, Experian, Sage and Informa.
Relx, formerly Reed Elsevier, whose brands include Lancet and LexisNexis, is the most interesting on that list. There are at least five words in the company’s description that surprise – “a global provider of information-based analytics and decision tools for professional and commercial clients” – but the surprise was the pre-cloud share price. It was £5 in 2012 and £41 in May last year, at which time Relx was valued at £70bn-ish and the fifth-largest company in the FTSE 100 index.
And now? Shares have halved from their highs and suffered the biggest decline since the cloud’s alleged panic-inducing plug-in appeared. The market has flipped from seeing Relx as an AI winner – because it’s using AI to boost existing products – to fear of its handsome 34% profit margin.
What does the company itself think? Well, Thursday’s full-year results weren’t headlined “Stock market gone crazy”, because that’s not Relx’s style. But the figures and statements reflected confidence: revenues rose 7% to £9.6bn and operating profits rose 9% to £3.3bn; Forecast of “another year of strong growth in 2026”; Dividend increased 7%; and a large share buyback of £2.25bn.
Chief executive Eric Engstrom also offered a coherent argument for why AI developments “will continue to be a key driver of customer value and growth in our business for many years to come”.
The short version: AI tools are not new; Recent launches that have caused an uproar are “workflow” products for storing, sorting, and reviewing documents for specific projects; Relax mostly seeks information needed in smaller markets that is comprehensive and can be relied upon in court.
Some of Relx’s data may be public, some may have been public in the past but is no longer so; Some are hard to find; Some are licensed; Something is proprietary; It all comes with “judgments, findings and interpretations collected over decades” that are useful to scientists, lawyers, insurance and city professionals and risk assessors; And AI can aid in that value-add process, or make goods easier to use for customers.
His other big point is that Relx is free to make limited licensing deals with AI companies if it wants, and can continue to launch its own “workflow” products, but it will never surrender its impossible-to-replicate proprietary information because that’s where the value of its business lies.
That was good for a 2% jump in the share price, which is better than nothing, but still suggests market concerns about where AI development goes next, and whether Relax’s “competitive moat,” as analysts say, is as deep as advertised. The issue, to some extent, is fear of the unknown.
But Relx’s strategic response should be clear. If he thinks growth can be relied upon “for many years to come,” as Engstrom says, and, if shares are available at half the price of a year ago, then simply keep buying back.
This year’s £2.25 billion version, up from £1.5 billion, is equivalent to 6% of the entire equity base. If it keeps this pace up for a few years, it will lead to serious earnings per share growth – assuming, of course, that projections for the business are actually solid.
There are also prescriptions from activist investor Elliott Management in LSEG, reportedly for a large buyback, where AI concerns and counterarguments are similar. If you’re really sure, it’s the right thing to do.
