Morningstar, Inc. MORN

Key data

The business

Morningstar sells the plumbing behind other people's investment decisions. An advisor pulls up a fund's star rating and a Morningstar Direct screen before recommending it; a private-equity associate lives inside a PitchBook terminal; a bond desk relies on a DBRS Morningstar rating to price a securitization. Most of that revenue is subscription: multi-year contracts for data feeds, research, indexes and workflow software, renewed year after year because ripping the data out of a firm's process is expensive and disruptive. The credit-ratings arm is the one cyclical piece, earning fees when new bonds are issued. The moat is two things stacked: curated data nobody wants to rebuild, and workflows people are trained on and reluctant to leave. The live question, and the reason the stock is where it is, is whether AI-native tools let clients skip the terminal and get the same numbers somewhere cheaper. Management's answer in August was that on-platform IP and consumption-aligned pricing defend the value, and that AI accelerates rather than drains the data flywheel.

The numbers

The sequence shows a business growing revenue high-single-digits while profit grows twice as fast.

Q2 revenue rose 9.6% and first-half revenue 10.2%, with operating margin near 24% against about 21% two years earlier. Adjusted earnings have run ahead of what analysts penciled in for four straight quarters. Under the hood the mix is healthy: within the credit-ratings segment, transaction-based recurring revenue grew 8.7%, license-based revenue 24.8%, and transaction-based new-issuance ratings jumped 31.4% as credit markets stayed busy.

Revenue compounded about 9.5% a year from 2021 to 2025. Operating income more than doubled over the same stretch, roughly 20% a year, once the 2022 trough, PitchBook integration cost and a market drawdown that pulled EPS to $1.64, gave way to margin recovery. That operating leverage is the whole story: the top line grows steadily, the profit grows faster, and the stock is nonetheless valued at EV/EBITDA of about 10x, below its own 13.6x-to-25.3x usual band and below peers near 14.6x. The market is pricing Morningstar as if the data moat is about to leak; the number that settles it is recurring-revenue growth, which in credit was 8.7% last quarter and has not yet cracked.

Management

Founder Joe Mansueto remains a significant insider and large shareholder, and over the past twelve months he sold about $26.2M across 92 transactions, the largest a $1.5M sale on August 10, 2026, with no purchases and plan status not disclosed on the filings. Against a multi-billion holding these are trims, not an exit. Capital allocation is more pointed: the company repurchased $787M of stock in 2025 and another $300M in Q1 2026, and funded it by roughly doubling long-term debt to $1.70B from about $700M a year earlier. The buyback has continued, but it is now bought with borrowed money rather than internal cash, at prices around today's. Net debt sits near 1.7x EBITDA with interest covered about 11 times, comfortable, though the lever is now pulled.

How it fails or surprises you

Pricing power cracks. If Perplexity, MCP connectors and AI-native rivals let clients get equivalent data and research without the subscription, price per account stalls and recurring growth drifts below 8%. The company was asked this directly in August and answered with capped token usage and consumption pricing. The first tell is recurring-revenue growth slipping toward mid-single-digits for two straight quarters.

The valuation gap that shouldn't exist. The fact this read explains least is the stock at a twelve-year-low multiple while operating income compounds 20%. If the market is right and the moat is quietly eroding, the cheapness is a warning, not a gift, and recurring growth decelerating would prove the bearish read. That single line, held above 8%, is what says the market is wrong.

Consumption pricing lands (right tail). If usage-based pricing turns AI into more billable events per account rather than fewer, revenue per account reaccelerates, and a sustained credit new-issuance cycle keeps that 31% transaction line hot. Recurring growth back above 12%, on unchanged margins, would pull the multiple toward its 13.6-to-25.3 historical band, a re-rate on flat earnings.

Closing thoughts

This turns on recurring-revenue growth: above 8% says the moat holds, below it for two quarters says the market is right. The left tail is real but slow: AI erodes pricing over years, not a quarter, and you would see it in the recurring line long before it hit the whole business. The right tail is a re-rate back toward the company's own history on earnings that are already compounding. What is at risk if pricing cracks is that 20% profit growth flattening to high-single-digit revenue growth while the multiple stays depressed; what is at stake if it holds is a multiple reverting from 10x toward the mid-teens on unchanged numbers. The fatter tail, on the evidence in hand, is the upside, because the deceleration the price implies has not shown up in the data yet.

The bet is still that advisors, asset managers and banks keep paying Morningstar every year and keep paying a bit more, and that AI does not learn to give the same data away free. Terminals have survived every prior "this kills the terminal" cycle, and the switching cost is still there. What breaks it is the recurring line, and the pair that tells you first is recurring-revenue growth against revenue per account. If recurring growth slips below high-single-digits for two straight quarters while new debt still funds the buyback, the read is wrong.