Nick Train, Portfolio Manager for Finsbury Growth & Income Trust shares his views on the recent AI/data sell off.
Like so many others, we have thought hard about the ramifications of emergent AI technology and looked for companies we believe are beneficiaries of its deployment. We have long held the view that among the biggest beneficiaries are companies that own and curate proprietary data, particularly when that data is vital to its users and already embedded in users’ workflows. AI needs data to generate insights and utility.
We have found and invested in a number of UK-listed companies with world-class data and digital assets, whose services are already highly valued by their customers. All of these available on the London stock market, which is not generally recognised as being home to such digital businesses. The companies provide services across a variety of important industries and across many geographies.
In our opinion, long-term holdings London Stock Exchange Group (LSEG), Sage and RELX, and newer holdings Autotrader, Experian, and Rightmove, have a credible opportunity to bring AI-enhanced services to their customers, an opportunity based on their ownership of data assets that are not available to emerging Large Language Models (LLMs), like ChatGPT or Anthropic. This is not a fantasy. All these companies are innovating and, crucially, growing more quickly as a result of their successful innovations.
We must acknowledge with disappointment that the share prices of the companies are saying we are wrong. An indiscriminate sell-off, arguably even a panic, has ensued. Take the announcement from Anthropic, reported in early February, that it has launched an AI-powered legal productivity tool for corporations.
RELX’s share price fell 14% on the day. Yet, by Anthropic’s own admission “AI-generated analysis should be reviewed by licensed attorneys before being relied on for legal decisions.” Those licensed attorneys are highly likely to be customers of RELX today and need RELX’s trusted data as much as ever.
I have personally experienced panic before in my career, most notably through the Great Financial Crisis, when there were true and imminent existential risks to financial institutions. I have to say, I do not register a similar sense of anxiety today, as regards our companies.
In fact, if one could ignore share prices, it is my view that recent developments for or announcements from them have increased our confidence in their strategic positioning and growth opportunity. Two of them reported trading updates in January, Experian and Sage. Below we discuss what we learned from those updates. The conclusions seem relevant for our other data companies too.
Sage’s share price was down 10% in January, despite the publication of what we and many viewed as a positive three month trading update, detailing organic and recurring revenue growth of 10%. This included 13% revenue growth in North America, now Sage’s biggest geography, over 45% of revenues.
Guidance for the full year of at least 9% revenue growth was reaffirmed, with further profit margin expansion to come too. This was an impressive set of results, yet clearly not enough to allay investors’ fears that AI might disrupt Sage’s business in the future.
Probably there will be casualties of AI – especially where software is simply summarising or producing basic reports, or building emails, based on publicly available data, for example. But Sage is not offering summaries or basic reports. It operates within the accounting industry, where trust, reliability and accuracy when dealing with complex work-flows are critical. Enterprises simply won’t trust an application that hallucinates and offers different answers each time.
We think that Sage’s deep domain expertise built on 40+ years of proprietary data gathering, accounting expertise and trust, as well as its ability to navigate complex regulatory environments, puts it in a strong position to leverage AI to improve its product offering and value proposition to its customers.
AI could become, as the company told us in late January, “an extraordinary tailwind for Sage”, precisely because of its ability to utilise its deep domain expertise. To summarise the company’s assertion about the opportunity presented to it, there are three propositions:
First, Sage is running Small Language Models (SLMs) on the data generated on its own platform, populated by millions of small and mid-cap companies worldwide. Sage Intacct alone is reported to handle 100 million application requests a day, with more than 50 billion financial records on the platform.
These models are generating insights and identifying efficiencies that benefit the companies within Sage’s eco-system. Of course, this data and Sage’s own domain expertise is not available to any external LLM. The company argues that the insights Sage can derive from its own data are at least as valuable for its clients as anything that could be derived by a general AI model.
Next, the company notes that Sage’s customer base is essentially made up of accountants – risk-averse and accuracy-fixated professionals. This reinforces our view that companies will not run their general ledgers on general AI.
Finally, the company insists that its existing AI services are “a real thing” (rather than a speculative possible product from AI in two years’ time, which is at the heart of the bear case). Sage’s CoPilot has been disseminated across the customer base and has resulted in actual price increases for its services, because the CoPilot has delivered actual efficiency gains for its adopters.
Sage reports efficiency savings of 5-10 hours a week. The next step will be the development and implementation of specialised agents, designed to solve specific tasks, scheduled to be offered first via Sage Intacct and then to the whole Sage Cloud Connected product portfolio by the end of 2026.
So far as external corroboration of Sage’s capabilities is that IDC MarketScape recognised Sage Intacct as “a Leader in ERP [enterprise resource planning] and PSA [professional services automation].” IDC further commented: “Sage is recognised for its unified AI strategy across PSA and ERP, strong roadmap for AI agents and focus on building trust and transparency into its AI experiences.” That is no trivial endorsement.
Clearly Sage is doing something right in terms of serving its customers with its existing products and early AI initiatives, as evidenced by the clear acceleration in Sage’s revenues as recently reported. And we can see how the next step, integration of AI with Sage’s unique ecosystem, would accelerate the business by further improving client retention and offering significant new product development and monetisation opportunities.
As Orlando Bravo, the founder of Thoma Bravo – one of the largest and most successful software-focused private equity investors globally – observed in a recent important interview: “Software is not all about the code or about the technology… Software is about your domain knowledge.
Most software companies know a specific vertical, a specific function so well that there are three to five companies in the world that know it… That is the franchise, that is the value. That is what you can’t replicate.”
When we look at Sage’s Cloud-Connected services, led by its Intacct product, being rolled out rapidly around the world and growing at 20% p.a. we conclude that Sage does offer the crucial domain expertise that Bravo extols.
Regrettably, Sage’s share price has not been the only casualty of AI disruption fears. The divergence between Experian’s company update and share price performance was perhaps even more stark. Its Q3 results detailed overall 8% organic growth – marginally ahead of consensus estimates – with solid performance across all divisions, yet during January the share price was down 18%.
Part of this weakness was due to the company-specific worry about the prospect of Bill Pulte imposing a 10% cap to credit card rates. Experian’s view is that there is a low probability of this happening and that even if it did, the demand for credit would simply shift to other lending verticals, all of which require as much risk management as credit cards. However, a much bigger fear is the prospect of wholesale AI disruption of Experian’s business model.
As yet, the bear case for Experian from an AI perspective is purely speculative. There appears to be much more evidence that AI will be a tailwind for future growth. Perhaps the most important figure in the recent results was Experian’s 9% growth in North America excluding mortgages, against a credit market growing at just 1%.
This indicates that there is no cyclical tailwind to that growth – in fact Experian is becoming less reliant on credit volumes and is instead driven by the sale of increasingly sophisticated software products which are firmly embedded in their customers’ workflows.
These products, in our view, could be boosted by the application of AI technologies and we recognise that Experian’s ability to build these kinds of products and tools is predicated on its ownership of a differentiated cache of credit data (the company processes 1.1 billion bits of unique data each month), as well as its positioning at the heart of the highly regulated global credit issuing ecosystem.
Experian’s growth speaks to the quality of this data and the product sets that the business has invested in over the past decade, and points to a real appetite amongst the customer base for new functionality and services.
Experian launched a new $1bn buyback at the end of January (c.3% of market cap). This seems rational given the strength of the results and the adverse share price reaction. Experian is not alone. We spoke recently with the senior executives of Autotrader.
As to its current depressed share price and the perceived threats of disintermediation to that business, the CEO said: “We think the best solution for Autotrader is to buy back as many of our shares as we can.” And, indeed, Autotrader has stepped up the pace of its existing buyback into share price weakness.
In addition, all our other UK-listed Data, Platform or Software companies – LSEG, RELX, Rightmove and Sage – are buying back shares, taking advantage of the current period of elevated volatility. Of course, they could be wrong to do so, but it is noteworthy that those who understand their businesses best see current developments in AI as making the companies more, rather than less valuable.”
Past performance is not a reliable guide to future returns. You may not get back the amount originally invested, and tax rules can change over time. The writer’s views are their own and do not constitute financial advice.
This information should not be relied upon by retail clients or investment professionals. Reference to any particular investment does not constitute a recommendation to buy or sell the investment.
Main image: luke-jones-tBvF46kmwBw-unsplash
