Everything About AI
NextFin News - Thomson Reuters delivered an 11% jump in third-quarter revenue and accelerated organic growth to 10%, but investors sold the stock down more than 7% after the company disclosed that higher software amortization tied to its artificial-intelligence buildout is pressing on reported profit. The split verdict - strong growth, weaker earnings, punished shares - captures the tension now running through the information-services sector: the bill for winning the AI race is arriving before the payoff.
The Canadian company reported revenue of US$1.833 billion for the quarter ended September 30, up from US$1.658 billion a year earlier, driven by 9% growth in recurring revenue - which now makes up 85% of the total - and an 18% surge in transaction revenue. Organic revenue grew 10% across the company and an identical 10% across its "Big 3" segments: Legal Professionals, Corporates, and Tax, Audit & Accounting Professionals. Net earnings fell to US$339 million, or 78 cents a diluted share, from US$423 million, or 94 cents, though adjusted earnings rose to 81 cents from 71 cents. The year-ago quarter had been flattered by a one-time gain on the sale of its remaining stake in the Elite legal financial-management business.
By the close on November 3, Thomson Reuters shares were down US$8.93, or 7.31%, at US$113.31, in a session where the broader technology-heavy Nasdaq Composite slid more than 2% on fresh worries about AI valuations and spending. The message from the tape was unambiguous: investors are no longer willing to fund AI ambition on faith alone.
The Numbers Behind the Sell-Off
The growth figures were strong enough on their face. The 10% organic print marked an acceleration from the 8% organic growth delivered in the second quarter, and management said revenue growth and margins came in "ahead of our prior expectations." The company also used the release to update its full-year outlook to reflect the closure of the Global Print divestiture: total revenue growth of approximately 9.5% and organic revenue growth of 8.5% to 9.0%, with guidance for the Big 3 segments' revenue growth and adjusted EBITDA margin left unchanged from the August 5 outlook.
But the income statement told a different story underneath. Operating profit fell 22% to US$402 million from US$517 million. Thomson Reuters attributed the decline primarily to the prior-year Elite gain, and "to a lesser extent, higher amortization of software in the current period." The software amortization line rose to US$202 million from US$181 million - an 11.6% increase that is the accounting echo of capitalized development spending, and in 2026 is shorthand for the AI investment cycle: the company has been training and deploying its proprietary Thomson LLM, embedding generative features across Westlaw and CoCounsel, and launching the next generation of CoCounsel Legal and CoCounsel Tax, Audit & Accounting during the year.
Amortization is a non-cash charge, and adjusted earnings - which strip out these items - actually rose. Yet the market reacted to the GAAP profit decline and the trajectory it implies. When a company tells investors that its cost base is structurally stepping up to fund an arms race, the discount rate applied to future earnings tends to rise even as those future earnings are being promised.
Margin data reinforced the caution. Total adjusted EBITDA margin slipped 30 basis points to 35.5%. Within the Big 3, Legal Professionals - the largest segment - saw adjusted EBITDA rise 4% to US$368 million on revenue of US$788 million, but its margin compressed 240 basis points to 46.6%. Corporates posted adjusted EBITDA up 11% to US$197 million on revenue of US$544 million, with margin roughly flat at 36.2%. Tax, Audit & Accounting Professionals generated adjusted EBITDA up 10% to US$83 million on revenue of US$276 million, with margin down 40 basis points to 30.2%. The company noted that the tax business is its most seasonal, with about 60% of full-year revenue typically landing in the first and fourth quarters. Reuters News, by contrast, saw adjusted EBITDA margin expand 440 basis points to 24.3% on revenue of US$232 million.
"We are pleased with our third-quarter momentum, as revenue growth and margins came in ahead of our prior expectations," said Steve Hasker, President and CEO of Thomson Reuters. "Customer engagement with the next generation of CoCounsel Legal has been very encouraging, and we are excited about the recent launch of the next generation of CoCounsel Tax, Audit & Accounting. Customer discussions around commercial opportunities for Thomson LLM continue, and we remain optimistic about its potential. We are confident in the growing success of our innovation engine, which positions us well to deliver sustainable growth and long-term value for shareholders."
The divestiture backdrop matters to the read-through. On October 1, the company closed the sale of a 51% stake in its Global Print business to capital accounts advised by KKR for gross proceeds of approximately US$500 million. The business, now a joint venture operating as Westbridge Print, was the legacy, low-growth leg of the portfolio. The sale sharpens Thomson Reuters into a higher-growth, software-centric company - but it also removes a revenue base, which is why the updated full-year guidance is expressed excluding Global Print.
Why the Market Looked Past 10% Growth
The sell-off is not just about one quarter's amortization line. It is about the terms on which the market is now pricing AI. For much of 2024 and early 2025, capital-expenditure increases and software investment were rewarded as evidence of competitive positioning. By late 2026, the same signals are being read as margin risk until proven otherwise. Thomson Reuters is not a hyperscaler burning tens of billions on data centers, but the investor reaction shows that the skepticism has cascaded down the stack: any company whose growth story depends on AI monetization is now being asked to show the payoff, not just the pipeline.
There is also a compositional issue investors are learning to spot. Thomson Reuters's growth is now concentrated in the Big 3, which generated US$1.608 billion - roughly 88% - of the quarter's revenue and the overwhelming share of profit, while the Reuters News and Global Print legs are slower or being exited. That concentration is strategically coherent - it is exactly what a focused software company should look like - but it also means the valuation multiple rests almost entirely on whether legal and tax professionals keep paying for AI-enhanced workflow. The 10% organic growth shows they are, for now.
The peer context sharpens the point. In the same earnings season, companies across the technology complex were penalized for AI spending without visible returns: shares of AI chip leaders and high-multiple software names fell on the same session despite solid results, as investors rotated toward businesses that can demonstrate cash payback. Thomson Reuters's 7.31% decline fits that pattern - a fundamentally sound print caught in a repricing of the AI premium.
Cyclical Backdrop or Structural Shift?
The central question for investors is whether this is a cyclical bout of AI skepticism or a structural reset in how software businesses are valued. The evidence points to a structural shift with a cyclical overlay.
The cyclical leg is real but familiar. Markets periodically rotate out of growth and into quality when rate expectations shift or when a crowded trade - in this case, AI - becomes over-owned. Similar episodes occurred during the 2022 software de-rating and the 2018 growth scare, and in each case the recovery was led by companies that could show margin discipline alongside growth. Thomson Reuters still carries that profile: recurring revenue remains the dominant share of the base, adjusted earnings are rising, and the company continues to generate strong cash flow.
But the structural leg is the more important call. The market is not merely rotating sectors; it is changing the unit of measurement. The question is no longer "how fast can you grow with AI?" but "what is the incremental margin on AI revenue, and how much capital does it consume?" That is a regime change in valuation logic, and it will not revert on its own. Companies that built business plans on the assumption that AI spend would be rewarded indefinitely - regardless of payback period - are now being forced to re-underwrite those plans. Thomson Reuters is better positioned than most because its AI features sit on top of entrenched, mission-critical workflows where switching costs are high and pricing power is real. But even well-positioned companies must now prove the arithmetic.
This is where the second-order effect bites. The first-order story is simple: higher software amortization reduces reported profit, and investors dislike it. The second-order story is that the cost of capital for AI-intensive businesses has risen, which compresses the present value of the very AI-driven growth those companies are promising. If Thomson LLM and the next-generation CoCounsel suite are to justify their cost, they must do more than retain customers - they must expand revenue per user faster than the amortization and compute costs attached to serving them. That is a higher bar than the one the market applied two years ago.
The Counter-Case: Why the Sell-Off May Be Overdone
The strongest argument against the bearish read is that the market is penalizing an accounting charge for an investment that is already producing growth. Adjusted earnings rose 14% on a per-share basis, organic revenue accelerated to 10%, and the company raised its full-year organic growth outlook to 8.5%-9.0%. The Elite gain distortion is a one-time comparator that will fall out of the numbers after this year. And the KKR divestiture, by removing a low-growth asset, improves the quality of every future dollar of revenue.
The bull case also rests on positioning. Thomson Reuters is not selling AI as a standalone product; it is embedding AI into workflows that professionals already pay for, with content moats - Westlaw, Practical Law, Checkpoint - that competitors cannot replicate. CoCounsel reached one million users across 107 countries earlier in 2026, and management cited encouraging engagement with the next generation of the product. If those users convert into expanded contracts rather than mere usage, the amortization charge becomes a bridge to higher margins, not a permanent drag.
But the counter-thesis has a clear falsifying signal. If organic revenue growth for the Big 3 segments fails to hold above 8% over the next two quarters while software amortization continues to climb, the market's concern shifts from timing to substance - and the multiple will stay compressed. Conversely, if the company reports Big 3 organic growth at or above 9.5% alongside stable or expanding adjusted EBITDA margins in the fourth quarter of 2026 and the first quarter of 2027, the sell-off will look like a buying opportunity in hindsight. Watch that specific pairing: growth above 8% plus margin stability is the threshold that separates a cyclical dip from a structural de-rating.
What Comes Next
Short term (next quarter): Sentiment will be driven by the fourth-quarter print and, specifically, by the amortization line and any updated capital-expenditure commentary. The base case is for continued solid Big 3 growth with margins under mild pressure as AI investment ramps. The upside case is that CoCounsel conversion accelerates and amortization peaks; the downside case is that cost growth outpaces revenue growth and guidance is tempered.
Medium term (6-12 months): The key variable is monetization. Investors need evidence that AI features are expanding revenue per user - through seat expansion, module attach rates, or price - rather than simply defending retention. The Thomson LLM commercial discussions management referenced are the swing factor: a signed enterprise rollout would be the clearest signal that the investment is converting.
Long term (structural): The thesis rests on whether Thomson Reuters can remain a toll-road on professional work as AI changes how that work is done. If AI makes professionals more productive, the risk is that fewer seats are needed; the offset is that each seat becomes more valuable. The company's bet is that workflow integration and trusted content keep it at the center of the process regardless of productivity gains. That is a defensible position, but it is being tested in real time.
For now, the market has drawn a line: growth is welcome, but not at any cost. Thomson Reuters delivered the growth; the next few quarters must deliver the proof that the cost is temporary. Shares closed at US$113.31, down 7.31% on the day, with analyst price targets averaging US$123.81 - implying the street still sees value, but only if management executes on the very AI payoff it is asking investors to trust.
The irony of the moment is sharp: Thomson Reuters is being punished for investing in the same technology that its own newsroom and professional products exist to explain. The market's verdict is not that AI is worthless - it is that the era of funding it on promise alone is over.
Explore more exclusive insights at nextfin.ai.
