Thomson Reuters Corp (TRI.TO)

Believes it's making strategy mistakes on AI. On price, he'd be a tactical buyer today. The road ahead is significantly worse than the road behind. There's no hope it's going back to $300 or anything close.

Getting disrupted in the highest-margin part of its business (legal division).

Such a big run, now a huge amount's come off. Looks attractive. Pace of change in the AI space makes things uncertain. Hard to determine pricing power of a tool. The market's not stupid, there are serious concerns.

One thesis says to look through that and say the moat will be fine. For him, it's too risky.

He bought it earlier this year as shares declined. He owned it before. It always had a high valuation until recently when it came down. Likes it long term. The street misses the extent of their data moat; they have proprietary content, plus many lawyers curated that content. Also, TRI will benefit from AI to enhance their products. Good data means good products, and they have good product run by good people. It now trades around 20x PE and a free cash flow yield of 5%. A pristine balance sheet.

Frustrating. Collapsed, then a big rally. Rolling over again. Jury's out on whether AI will kill its legal and accounting software. For himself, he likes the law firm of "Claude, Copilot, and Gemini" ;) 

TRI's legal business has a bit of a moat around it. There's true value there. We'll have fewer lawyers, but the ones left will need access to the kind of data supplied by RTI. It'll survive, but the question is what do you pay for it?

Is -32% in Q1 and one of the worst performers on the S&P. The street expects AI to displace their business, compiling financial data, which he uses on the show. The street won't wait to find out if AI will indeed displace them and will sell now anyway.

Transitioned from newspapers to information. In the age of AI, information is much more commoditized. On AI concerns, many babies have been thrown out with much bathwater. If you have proprietary data, data will be king -- these companies will probably do better than the market fears.

Is wrestling with it. They have some proprietary data. A lot of what they do is data aggregation, which he will be able to do increasingly somewhere else at a fraction of the price. Large language models can't access their legal products, at least not yet. Will they be able to later? Doesn't know. Two years ago they traded at 50x PE, and now 20x which remains high given this AI pressure. TRI is intriguing but isn't convinced yet.

Software selloff is overdone. AI won't be that quick to replace a company like this. Already incorporates AI in its programs. Data is proprietary.

If you got in a couple of weeks ago, hold on. His firm is looking to get in.

Classic compounder, higher valuation. Narrative shifted around software. Lots of bad news priced in. Trades 21x forward PE, grows earnings at 10+% pretty consistently. 

AI fears overblown. Stock's finding support. Bounced on news that it's working with Anthropic. Don't expect a V-shaped bounce, be patient.

Valuation always excessive -- over 20x operating cashflow for single-digit growth. Now, that's changed with the pullback, so it might be more interesting down here.

Anthropic had announced an AI tool that could write legal briefs. There were worries this could displace TRI, which also was at a high PE before the drop. He hopes today is an inflection point where AI companies will partner with data providers which have unique, irreplaceable data. He hopes sentiment changes on the entire sector where they are beneficiaries and not disrupted by AI.

All of the Top Picks today are being tarred with the same software brush, with very little differentiation of what they do and how they do it.

Huge differentiator is its 3 end markets:  lawyers, compliance professionals, and accountants. The data in all of its core product offerings is proprietary. Employs more than 3k legal and accounting professionals to curate this information. 

Competitive advantage, including from Anthropic, is so difficult to replicate. In last few years, has already been incorporating AI into its products. Likely to maintain its moat, and perhaps deepen it further. He's never seen the valuation it's at, very attractive. Yield is 2.98%.

Piquing his curiosity in last weeks and months. Cut in half since last summer. Encapsulates all the consternation roiling around software vs. AI, and accelerated with the Anthropic release. Dominant franchise in law. Superficially, selloff is overdone.

While fears about AI disrupting are legitimate, companies can have proprietary moats on their data and networks. TRI owns 100 years of caselaw and subsequent analysis. Increasingly layering on AI tools. PE multiple not seen in years.