S&P Global Inc. SPGI

Key data

The business

S&P Global sells trust in numbers. Four businesses remain now that the automotive Mobility unit has been spun off: credit Ratings, the largest and a duopoly with Moody's; Market Intelligence, the subscription desktop data and analytics; Commodity Insights, the old Platts energy-price benchmarks; and S&P Dow Jones Indices, which licenses the S&P 500 name to fund managers who together run trillions. The moat is two-sided. Regulation and reputation make a third credit-rating agency nearly impossible to launch, and no sponsor can sell an "S&P 500 fund" without paying for the name. A bond issuer pays a few basis points to be rated because buyers demand the stamp, and an index fund pays a sliver of assets because the brand is the product. Most revenue now recurs through subscriptions and index license fees. The cyclical piece is Ratings transaction revenue, which rises and falls with how much debt companies issue, and it carries the fattest margin in the house. The pending datacenterHawk purchase, closing in the second half of 2026, pushes Commodity Insights into data-center supply intelligence, a genuinely hot corner.

The numbers

The most recent print, Q2 2026, put adjusted diluted EPS at $4.83 against a $4.81 estimate, the fourth beat in the last five quarters. The GAAP trajectory through Q1:

*Adjusted EPS; GAAP revenue and net income not in XBRL feed.

Revenue reaccelerated into 2026: Q1's $4.17B was up about 10% on the year, and net income jumped 28% as the margin widened. Both lines moved up together, not down, and the Mobility separation closed cleanly on July 1 through a 1-for-1 distribution, leaving the pure franchise you are underwriting.

*Q1 2026 only; Q2 2026 GAAP data not in XBRL feed.

The five-year arc hides a merger. The 2022 close of IHS Markit grew revenue by about a third but crushed GAAP EPS to $8.23 by 2023 under purchase amortization and integration cost. Since that trough, GAAP EPS compounded about 33% a year to $14.66, and revenue kept climbing double digits. That recovery is the operating leverage the model is built for:

What the market is not paying for is a cluster of open-market insider buying that says the people closest to the numbers think the post-separation company is worth more than 17.7x EV/EBITDA. The print that settles it is continuing-operations organic growth and the Ratings issuance trend over the next two quarters.

Management

The record backs the read. Over the trailing year insiders bought seven times for $3.6M and sold nothing, and the buyers were the decision-makers: CEO Martina Cheung put in about $1.0M in April, chairman Hubert Joly $0.9M in February, director Catherine Clay $1.1M in May, all open-market purchases with plan status not disclosed. Capital return is heavy and disciplined: $5.0B of buybacks in 2025 and $1.0B in Q1 2026, on top of the roughly $12B spent digesting IHS Markit in 2022. A team buying its own stock the same year it repurchases $5B is spending real conviction, not just spare cash flow.

How it fails or surprises you

Ratings freeze. The highest-margin revenue is transaction fees on new debt ratings, and it swings with issuance. A rate or credit shock that shuts high-yield and refinancing, as 2022 did, would cut that line fast and take the consolidated margin with it. The tell shows first in quarterly Ratings transaction revenue turning negative year over year.

Priced for perfection. You pay 27x GAAP earnings and 17.7x EV/EBITDA, above the roughly 14x peers fetch, for mid-teens growth. If the private-credit and AI-data demand story that justifies the premium stalls, the multiple compresses toward peer even with earnings intact. This is the fact the bull read explains least, and the reason the stock sits well below its $522 high.

Cleaner and faster than it looks (right tail). With Mobility gone, the remaining business is higher-margin, more recurring, and easier to value, and datacenterHawk plus private-credit ratings demand could push organic growth back to the high single digits. The market is discounting near-term issuance softness and separation dis-synergies. If post-separation growth reaccelerates while buybacks shrink the share count, earnings and the multiple can rise together. First sign: continuing-ops organic revenue growth above consensus in the first two clean quarters.

Closing thoughts

A specific print settles this. Under it sits a duopoly and an index franchise that neither a recession nor a competitor can easily take, wrapped in a multiple that assumes the growth continues. The fatter tail is the upside. The downside is a cyclical dip in one segment's transaction revenue that the recurring base and the buyback cushion, while the upside is a re-rating of a cleaner, faster-growing pure play whose own CEO and chairman are buying. What is genuinely at risk if issuance freezes is a quarter or two of flat earnings and a slide back toward $361. What the upside is worth is the gap between 17.7x and the low-20s multiple this asset has commanded before.

The bet is still that the world keeps borrowing and keeps paying S&P Global to grade the debt, and that fund managers keep paying to use the S&P 500 name. What breaks it is a durable freeze in debt issuance that the subscription and index lines cannot offset. The one pair of numbers that tells you first: Ratings transaction revenue growth against the consolidated operating margin, quarter by quarter. If issuance holds and the margin stays near 48%, the compounder is intact and cheaper than its history. If both roll over, the 27x was the whole story.