Two Record Divisions, a Guidance Cut That Never Happened, and a Second Half That Asks for Almost Nothing

Results vs. Consensus

Q2 2026 Scorecard

Year-over-Year Comparison (pro forma)

Sequential Comparison

The first quarter figures below are the July 6 recast, not the numbers we published in April. Two things moved: Mobility came out, and $72M of first-quarter revenue moved from Market Intelligence into Energy with the transfer of 451 Research and Maritime & Trade. Energy's 14% sequential decline is CERAWeek seasonality rather than deterioration, since the conference lands in the first quarter.

Revenue. Eleven percent organic constant-currency growth from a company the market spent the first half of this year pricing as structurally impaired is the headline fact, and it is the second consecutive quarter of double-digit reported growth. What makes it more useful than the first quarter's is that it required no explanation: there was no acquisition contribution worth adjusting for, no currency tailwind worth naming, and no gap between the reported and organic numbers. Set against that, the concentration got worse rather than better. Ratings grew 17% and Indices 20%, while Market Intelligence grew 6% and Energy 3%. Those two slower divisions represent 51% of pro forma revenue and contributed 24% of the growth. A reader who believes the benchmark franchises are durable will read that as leverage. A reader who believes they are cyclical will read it as concentration risk that arrived precisely when the diversifying asset was spun off.

Margins. The 54.3% adjusted operating margin is the highest this business has printed and it expanded in all four divisions. Ratings reached 68.5% on 17% revenue growth against 6% adjusted expense growth, which is what operating leverage looks like in a business with a fixed analytical cost base. Indices reached 71.5% despite 16% expense growth, because 20% revenue growth outran it. The one genuinely notable line is Energy, which expanded 70bps to 47.5% on 1% adjusted expense growth in a quarter when its revenue grew 3%. That is a division defending profitability by not spending, in a period when management also chose not to push price on affected customers. It is the right decision and it is not a repeatable source of margin.

EPS. This is where the quarter is less impressive than the headline. Pro forma adjusted income before taxes grew 14%, and the reported 23% adjusted EPS growth required a 350bp reduction in the adjusted tax rate plus a 3.5% reduction in the share count to get there. The tax rate is the part that deserves scrutiny: at 19.8% it is 220bps below the low end of the company's own full-year guide, the first-half rate of 21.5% is also below it, and nobody on the call asked why. If the full-year guide of 22.0% to 23.0% is honest, the second half carries a tax headwind that neither the first-half run rate nor the call's discussion prepares anyone for. The share-count contribution, by contrast, is about to get much larger.

Segment Performance

Total adjusted operating profit is stated after adjusted corporate unallocated expense of $42M, which is why it does not equal the sum of the four divisions. Energy and Market Intelligence revenue is the adjusted as-recast line, which reflects the transfer of 451 Research and Maritime & Trade; the as-reported figures are $568M and $1,290M respectively.

Revenue by Type

This is the table the disruption debate should be conducted on, and it moved further in the company's favour this quarter. The subscription line is the piece a language model is theoretically capable of substituting. The other five lines are tied to issuance volumes, index assets, derivative trading volumes, sales royalties and physical price assessments, and none of them can be produced by a model.

The non-transaction line is shown net of the $(53)M intersegment elimination, following the filing's presentation.

Assessment: subscription is now 45.7% of pro forma revenue and grew 5%. The other 54.3% grew 16%. Removing Mobility, which was a high-single-digit subscription business, both shrank the substitutable share and raised the growth rate of the non-substitutable remainder. The bear case that treats S&P Global as a data vendor facing commoditisation is now arguing about slightly under half the company, and the half it is arguing about is the half that is not growing.

Ratings (36% of pro forma revenue, +17%)

A record quarter by revenue and the largest division post-spin. Billed issuance rose 25% year over year with strength across the risk spectrum. Transaction revenue rose 25% on high-20s growth in investment grade, driven by technology infrastructure and hyperscaler issuance plus M&A, with double-digit growth also in bank loans, high yield and structured finance. Private markets ratings revenue rose 60%. Non-transaction revenue rose 8% on higher annual fees, Ratings Evaluation Services and CRISIL. Adjusted expenses grew 6% against 17% revenue growth, producing the margin result.

"This quarter, Ratings reported a record quarter by revenue, which increased 17% year-over-year. We exceeded our internal expectations for both the quarter and for the first half of the year. Issuance in the quarter benefited from tighter spreads and favorable market conditions, though growth was also elevated due to a fairly soft compare in the year ago period, driven by last year's tariff uncertainty."
— Eric Aboaf, CFO

Two things about that sentence are worth holding onto. Management exceeded its own internal expectations for the half, and it volunteered the soft compare in the same breath, which is the second consecutive quarter it has told investors not to extrapolate the credit franchise. The full-year Ratings guide was raised only one percentage point to 5% to 8% after a first half that grew 15%.

Assessment: a 68.5% adjusted margin on 17% growth is the best result this division has produced, and the AI capital-expenditure cycle is reaching it as high-margin transaction revenue with no incremental capital, which is exactly the mechanism our initiation argued the market was not connecting. The catch is arithmetic: the raised full-year guide implies second-half Ratings revenue somewhere between down 4.4% and up 1.4%. Anyone modelling the first half forward is modelling against explicit guidance for the second consecutive quarter.

Indices (15% of pro forma revenue, +20%)

The thirteenth consecutive record quarter for revenue, and the division passed $2B of revenue on a trailing twelve-month basis. Asset-linked fees grew 22% in a fourth consecutive record quarter, on equity market appreciation and what management described as the best quarter of net inflows on record. Exchange-traded derivatives revenue grew 22% on strong volumes, particularly in SPX. Data and custom subscriptions grew 9% on new business in end-of-day contracts. Adjusted expenses grew 16%, the fastest of any division, on growth investments and compensation.

"June marked a significant milestone for the global markets as well as we saw the first-ever ETF surpass $1 trillion in AUM. We're incredibly proud to be part of that story as that ETF was based on the storied S&P 500."
— Martina Cheung, President and CEO

ETF assets under management tied to S&P Dow Jones Indices ended the quarter at $6.35 trillion, with more than $600 billion of net inflows over the trailing twelve months, and management said multiple asset managers switched benchmarks to S&P in the quarter, bringing tens of billions of additional AUM. Four of the five ETFs selected by the U.S. Treasury for inclusion in Trump accounts were linked to S&P Dow Jones Indices.

Assessment: the guide for this division was raised two percentage points to 12% to 14%, the largest upward revision on the call, and it still implies second-half growth of 6% to 10% against an 18% first half. The derivatives line remains the underappreciated asset: it grew 22% in a quarter of elevated volatility, which is the same regime that suppresses flows into higher-priced sector and thematic products. That internal hedge is why Indices revenue is more stable across market regimes than a pure beta read suggests, and it is not in how the division is valued.

Market Intelligence (34% of pro forma revenue, +6%)

The division at the centre of the disruption thesis, and the one that did not deliver what was promised. Subscription revenue grew 6% on both a reported and organic basis, the same rate as the first quarter, in a quarter where last quarter's call had guided to acceleration. Volume-driven revenue grew 9% on Corporate Actions, the Primary Markets Group and ClearPar usage. One-time revenue declined 2% on Consulting and Sustainable1. Under the new structure, Kensho Data & Platforms revenue grew 8%, of which 4 points were organic, and Enterprise Solutions grew 3% reported but 10% organically once the EDM and thinkFolio divestitures are stripped out. Adjusted expenses grew 4%, delivering 120bps of margin expansion to 36.0%.

"In the second quarter, subscription revenue increased a solid 6% on both a reported and organic basis, benefiting from growth across the subscription franchises and benefiting from some upfront revenue from a 10-year renewal in the quarter."
— Eric Aboaf, CFO

Pressed later on how much that renewal contributed, the CFO said growth "would have been slightly lower had it not been for that software renewal" and gave no number. The 50bp revenue-recognition headwind that management said in April would reverse in the back half was not mentioned once on this call.

Assessment: this is the second consecutive quarter at 6%, it was supposed to be higher, and the printed 6% flatters the underlying rate. The margin expansion is real and the annual contract value figure was again quoted at roughly 6%, so nothing here describes a franchise in retreat. But our bear case has always been that this is the metric the disruption debate gets settled on, and a management team that guides to acceleration and then delivers a flat number with a one-time item inside it has spent credibility it will need.

Energy (17% of pro forma revenue, +3%)

The most pressured division, and the one where the Iran conflict is transmitting most directly. Platts revenue grew 4% on strong price assessments, partially offset by declines in Global Trading Services, where extreme volatility pushes traders to the sidelines rather than into the market. CERA grew 1%, with growth in market insights and analytics largely offset by declines in upstream and in conference and training revenue, the latter hit by lower event attendance because of the conflict. The sanctions previously disclosed cost Platts 120bps and CERA 30bps of growth in the quarter. Adjusted expenses grew 1%, which is how a division with 3% revenue growth still expanded margin 70bps.

"Importantly, the Iran conflict has complicated contract renewals among some very large customers, and we have intentionally chosen to be flexible on price increases and other terms for affected customers during such a challenging time."
— Martina Cheung, President and CEO

That is a disclosed decision not to take price, in a division whose entire investment case is the pricing power of a benchmark. It is defensible commercially, and it means the reported growth rate understates the underlying franchise while also being the growth rate shareholders actually receive this year.

Assessment: the guide was left unchanged at 4.5% to 6.0% after a quarter that delivered 2% organic constant currency, which requires the second half to run at 4.3% to 7.3%. Management said growth normalises back to the 6% to 8% Investor Day range only after this year. Two structural changes are also in flight: the upstream software divestiture now expected to close in the third quarter, and the consolidation of the supply chain assets into this division. Neither is visible in a growth rate yet.

Key Topics & Management Commentary

Overall Management Tone: confident and noticeably more forward-leaning than the first quarter, where the posture was defensive and the prepared remarks were built around answering a de-rating. This call led with operating results and spent its structural airtime on how the four remaining divisions are organised rather than on why the data moat survives. Where the tone thinned was Market Intelligence, which was discussed in the language of portfolio management rather than growth, and where a specific commitment made in April was not addressed until an analyst raised it.

1. The Spin Arithmetic: A Beat That Printed as a Cut

Mobility Global separated on July 1, one day after the quarter ended, so this is the first print in which the company reports one set of results and guides to a different company. Full-year adjusted EPS guidance moved from $19.40 to $19.65 down to $17.50 to $17.75. Read as a single number that is a 9.7% reduction at the midpoint. Read correctly it is the removal of a business that contributed $1.98 of adjusted EPS in 2025.

"For the reasons I just mentioned, the consolidated guidance we are giving today is not directly comparable to the adjusted guidance we had issued previously, which assumed a full year contribution from Mobility. For all material purposes, however, our division guidance is still comparable to prior guidance."
— Eric Aboaf, CFO

The recast 2025 base published on July 6 is $15.85 of adjusted diluted EPS against the $17.83 originally reported. The new guidance range is therefore 10.4% to 12.0% growth, which is what the CFO meant by double-digit growth across the entire range. Underneath the headline, every component moved the right way: organic constant-currency revenue growth held at 6% to 8%, the ex-OSTTRA margin expansion target rose 25bps to 75 to 100bps, the Ratings guide rose a point, the Indices guide rose two points, and the repurchase target rose by nearly $3B.

Assessment: this is a documentation problem that cost shareholders roughly $4.6B of market value in a single session. It was foreseeable, it was disclosed in advance, and the company still let a $4.08 pro forma GAAP figure and a $17.50 guidance range go onto the wires without a comparison table that a headline writer could not misread. The underlying decision to spin was value-creating and the underlying guide was raised. Neither fact reached the tape before the open.

2. Ratings: The Company Beat Its Own Cadence Guidance and Told Investors Not to Notice

In April management said explicitly that it did not expect acceleration in Ratings revenue growth in the second quarter. Ratings grew 17% against 13% in the first quarter, and billed issuance grew 25% against 14%. The explanation offered was the soft prior-year compare created by last year's tariff disruption, which is true and which management had known about when it issued the April guidance.

"Transaction revenue increased 25%, partly driven by high 20% growth in investment grade, supported by tech infrastructure and hyperscaler issuance and M&A transactions in the second quarter. We also benefited from double-digit revenue growth in bank loans, high-yield and structured finance. Private Markets Ratings revenue increased 60% year-over-year."
— Eric Aboaf, CFO

The 60% growth in private markets ratings deserves more attention than it received. In the first quarter that figure was above 25%. A doubling of the growth rate in the business line most exposed to the private-credit scrutiny that dominated the first-quarter narrative is a direct rebuttal delivered in a subordinate clause.

Assessment: this management team has now guided conservatively and beaten its own cadence twice in a row. That is a pattern rather than an accident, and it is the single most useful thing to know when reading a second-half guide that implies Ratings revenue is flat to down.

3. The Hyperscaler Ledger, Now With Numbers

The most quantitatively useful disclosure of the call. Management sized the hyperscaler contribution to billed issuance for the first time and set a full-year range against it.

"In the first half, we saw approximately $169 billion in billed issuance from the hyperscalers while our initial outlook for the year assumed approximately $200 billion for the full year. Our updated financial guidance assumes billed issuance growth in the mid- to high single-digit range. We are now assuming $250 billion to $300 billion in hyperscaler issuance for the full year and double-digit growth in M&A-related issuance."
— Martina Cheung, President and CEO

That arithmetic implies $81B to $131B of hyperscaler issuance in the second half against $169B in the first, or 48% to 78% of the first-half run rate. The maturity picture behind it is the longer-dated argument: management said roughly $11 trillion of rated debt comes up for refinancing over the next four and a half years, and that billed issuance has grown at an average of roughly 20% over the trailing twelve months without depleting those walls.

Assessment: the deliberate conservatism on hyperscaler volumes is the right posture for a business that gets paid on volumes it cannot forecast, and it is also the reason the full-year Ratings guide looks disconnected from the first half. The strategically important claim is the second one: that a year of 20% issuance growth has not borrowed from the maturity wall. If that holds, the AI capital-expenditure cycle is an additive source of investment-grade supply rather than a pull-forward, and the duopoly monetises it at a 68.5% margin with no incremental capital.

4. Market Intelligence: The Acceleration That Was Promised and Did Not Arrive

On the April call, management said it expected some acceleration in Market Intelligence subscription revenue in the second quarter, citing customer traction and sales pipeline. Subscription revenue grew 6%, the same as the first quarter, and the CFO subsequently acknowledged that the figure included upfront revenue from a single ten-year renewal without which growth would have been lower. The 50bp revenue-recognition headwind that was supposed to reverse in the back half went unmentioned.

"Jason, it's Eric. As we covered in some of our prepared remarks, we've had solid performance in MI. We did have a couple of smaller products that are lagging, consulting, sustainability that we called out. And so those tend to have some impact on results. I think in general, we're pleased with what we've seen in terms of execution, whether it's OCC revenue growth in MI was at 6%. Subscription was at 6%. Net retention rates are continuing to tick up."
— Eric Aboaf, CFO

Annual contract value growth was again quoted at roughly 6%, in line with subscription revenue, which is the second consecutive quarter at that level against a range the company itself has described as running from 6% to the upper 6s.

Assessment: the answer to a missed forecast was a restatement of the number that was supposed to have changed. Net retention ticking up and a growing pipeline are real forward indicators and they have now been cited for two quarters without producing a change in the reported rate. This is the point at which a bear thesis stops being about narrative risk and starts being about a growth rate.

5. The Market Intelligence Reorganization and What It Disclosed by Accident

Announced on July 6 alongside the recast, the division is being rebuilt into two reported business lines: Kensho Data & Platforms, and Enterprise Solutions. The stated logic is simplification and proximity to the customer. The more consequential content was the growth disclosure attached to it.

"The Kensho Data component is roughly half the size of platforms by revenue, but is growing in the high single-digit to low double-digit range on an organic basis. The Platforms component includes Cap IQ, Consulting, Issuer Solutions, RatingsDirect, Visible Alpha and With Intelligence. Platforms is the larger component of the business line and in aggregate is growing low single digits on an organic basis."
— Martina Cheung, President and CEO

Read plainly: the desktop and workflow estate, including Capital IQ, is growing low single digits organically, and the data-feed and API estate is growing high single to low double digits at roughly half the size. Management's stated plan for Platforms is to consolidate redundant platforms onto unified infrastructure, simplify operations, and lift profitability using AI and conventional productivity while still growing revenue. For Kensho Data the plan is to accelerate revenue at strong incremental margins.

Assessment: this is the most important number disclosed on the call and it was not framed as news. The bear thesis on this company has always been specifically that the interface layer is what artificial intelligence eats first. Management has now quantified the interface layer as the larger half of its data business growing at low single digits, and has responded by declaring it a margin story rather than a growth story. That is the correct commercial response, and it is also a concession.

6. The Subscale Remainder, and Whether It Gets Fixed or Sold

Pressed on the parts of the portfolio facing headwinds, management named consulting services and sustainability products, drew a line around the strategic franchises, and left the disposal option open without sizing anything.

"The other point to make here is with respect to the softer products. These are smaller products, subscale, some consulting services, for example, some of the sustainability products. We're not talking about our strategic products like WSO or ClearPar, for example. And where there are cases where we may see cyclical impacts and growth could come back in those products, we're going to watch that very closely. And there may be opportunities otherwise if we don't see growth coming back, and we may think that we have to invest a little bit more, we may make some strategic choices there."
— Martina Cheung, President and CEO

Separately, management stated that small carve-outs are possible in the near term and that there remains no appetite for transformational M&A. The company has already divested EDM and thinkFolio in January and the upstream software portfolio is expected to close in the third quarter.

Assessment: a portfolio being actively pruned at the bottom is a healthier signal than one being defended in full, and the divestiture track record over the past two quarters is credible. What is missing is the size. Without a revenue figure for the lagging products it is impossible to know whether fixing them adds a point to divisional growth or a rounding error, which means it is also impossible to check the claim that Market Intelligence accelerates once they are gone.

7. AI Monetization: The Gap Widened Again

Every AI metric the company disclosed in April moved in the same direction and by more. The gap in annual contract value growth between customers using AI solutions and everyone else widened in both divisions that report it.

"ACV growth in MI is now 60% faster in MI for AI customers and is approximately 3x in Energy. The demand signal from customers is incredibly strong. We continue to rapidly add customers to our LLM-ready APIs and MCP-connected solutions, with that number now sitting above 500 and increasing more than 70% quarter-over-quarter."
— Martina Cheung, President and CEO

In April those figures were 30% faster in Market Intelligence and roughly double in Energy, with more than 300 customers under contract or in trial. Call volume on the LLM-ready API was more than five times the first quarter's level, which was itself more than five times the fourth quarter's. Management also described a single large strategic renewal covering desktop, feeds, Visible Alpha and AI-ready data that produced an uplift of over 20% across several years, and said roughly 15% of clients licensed for the LLM-ready APIs are net new or returning customers.

Assessment: the direction is unambiguous and it has now been consistent for two quarters across four separate measures. What has not changed since April is that none of it appears as a disclosed revenue number, and the CFO's description of the monetization model as "a mix of some consumption pricing, some additional data set pricing" is the language of a pricing framework still being negotiated. The evidence keeps strengthening and the income-statement confirmation keeps not arriving. That gap is the whole argument.

8. The Token-Cost Conversation Is New, and It Cuts Both Ways

The genuinely new piece of customer intelligence on this call, and it was volunteered rather than extracted. Customers have started budgeting for artificial intelligence as a cost line rather than an experiment.

"Even as our AI solutions are seeing great demand, we are hearing more and more from customers that they are paying more attention to token costs and the overall expense of their own AI investments. Customers are looking for ways to minimize or manage token expenses, including building solutions in-house. Those customers want to build with S&P Global and with Kensho Labs."
— Martina Cheung, President and CEO

Management framed this as an opportunity, and paired it with the observation that customers are now rigorously evaluating AI budgets to prioritise solutions with positive return on investment. The related disclosure, from a separate exchange, is that renewal cycles with the largest and most sophisticated clients have lengthened because the negotiations now cover use of intellectual property in AI systems.

Assessment: a customer base that is cost-constrained on tokens is a customer base that will consolidate vendors, which favours the largest proprietary data estate. It is also a customer base that will scrutinise every incremental data licence, which is precisely the friction showing up as longer renewal cycles and a flat subscription growth rate. Both effects are real and they are working against each other inside the same division. This is the mechanism to watch for the next several quarters, because it will resolve one way or the other in the Market Intelligence subscription line rather than in a press release.

9. Capital Return: A $7 Billion Buyback Financed Past the Leverage Target

The largest single decision of the quarter. The repurchase target rose from roughly $4.5B to more than $7B, funded by the approximately $2B dividend received from Mobility Global on completion of the spin plus approximately $2B of new debt to be issued in the second half.

"In the second half, we expect to issue approximately $2 billion in additional debt to fund further buybacks with timing dependent on market conditions. We expect to end 2026 with gross leverage of approximately 2.7 to 2.8x EBITDA, modestly above our target leverage range of 2.0 to 2.5x."
— Eric Aboaf, CFO

At the current price that programme retires roughly 5.6% of the share count in a single year. The execution profile is the part that has not been discussed: $1.5B was repurchased in the first half, of which only $500M came in the second quarter, so more than $5.5B has to be deployed in the second half at roughly 3.7 times the first-half pace.

Assessment: retiring 5.6% of a business that converts nearly all of its adjusted net income to free cash flow, at a multiple the company itself considers depressed, is the highest-return use of capital available and management is right to lever modestly for it. Two caveats belong on the record. The buyback pace was cut in half in the second quarter, in the quarter that contained the stock's fifty-two-week closing low, and the second-half programme is now committed at a price roughly 14% higher than that low. And the leverage decision moves this company outside its stated target range for approximately eighteen months, which removes flexibility if the second half is worse than guided.

10. Portfolio Motion: datacenterHawk, Agusto, and the Upstream Exit

Two acquisitions were announced alongside the print, both small and both directionally consistent with the stated strategy of extending benchmarks into new asset classes and geographies.

"We also announced an agreement to purchase datacenterHawk and the majority stake in Agusto & Company. DatacenterHawk will combine with our 451 Research and energy forecasting assets to extend our leadership in the data center space. Agusto & Company is a leading credit rating agency with operations in Nigeria, Kenya, Rwanda and Ghana."
— Martina Cheung, President and CEO

Neither purchase price was disclosed. On the other side, the upstream software divestiture that management sized in April at roughly 25% of upstream revenue is now expected to close in the third quarter rather than the second half or early next year, and the division guidance has been restated on an organic constant-currency basis to accommodate it. CERA Titan, the AI-native upstream platform, remains on track to launch later this year.

Assessment: datacenterHawk is the more interesting of the two because it is the company positioning to sell data on the physical build-out it is already monetising through ratings. The pattern across the last three quarters is consistent: buy proprietary data, sell software. That is the correct direction of travel for a business whose durable advantage is content rather than interface, and it is a quiet answer to the disruption question that management has not framed as one.

11. Four Divisions, and What the Company Now Says It Is

The strategic reframing that ran through the prepared remarks is that this is not an information-services conglomerate but a benchmark business with adjacent data assets. Management put a number on it for the first time: benchmarks, defined as Ratings, Indices, the Platts business inside Energy, and the distribution of ratings content through Market Intelligence, account for nearly two thirds of revenue and more than 80% of operating profit, and grew 15% in the quarter.

"As I've shared with you before, we are primarily a benchmark business. Benchmarks include our Ratings division, our Indices division, the Platts business within our Energy division and the distribution of our Ratings content through Market Intelligence. These benchmark businesses account for nearly 2/3 of our revenue and now comprise more than 80% of our operating profits."
— Martina Cheung, President and CEO

Consistent with that, the Enterprise Data Office has achieved nearly 60% of its targeted $100M of annualised cost savings and remains on track to deliver the full amount, roughly 20% of that cost base, before the end of 2027.

Assessment: the framing is accurate and it is the single best argument against the de-rating, because a benchmark is a standard rather than a dataset and standards are not substituted by models that were trained on them. The risk in leaning on it this hard is that it concedes the rest. If more than 80% of operating profit comes from the benchmark businesses, then the multiple the market applies to this company is really a multiple on issuance volumes and index assets, and the data business is being valued at close to nothing. That may be the correct answer. It is not the answer management was arguing for three months ago.

Guidance & Outlook

Full-Year 2026 Enterprise Guidance

The new guidance excludes Mobility for the entire year. Rows marked "not comparable" changed because the company changed, not because the outlook did.

Division Guidance

Management presents division guidance on a revenue-growth basis against recast financials, with Energy explicitly updated on an organic constant-currency basis to accommodate the upstream software divestiture. The prior Ratings and Indices ranges are derived from the CFO's statement that each was raised by one and two percentage points respectively.

Implied second-half figures are our dollar derivations against a percentage guide, using recast full-year 2025 divisional revenue and first-half actuals. They characterise cadence and are not company figures.

Implied second-half ramp. This is the most important arithmetic in the print and it was not performed on the call. Recast 2025 pro forma revenue was $13,589M, of which $6,674M fell in the first half, leaving $6,915M in the second. The first half of 2026 delivered $7,397M, up 10.8%. Applying the 5.9% to 7.9% full-year guide produces implied second-half revenue of $6,994M to $7,266M, or growth of 1.1% to 5.1%. On earnings the pattern is identical. Recast 2025 adjusted diluted EPS was $15.85, of which $7.82 fell in the first half. The first half of 2026 delivered $9.27, up 18.5%. The $17.50 to $17.75 guide leaves $8.23 to $8.48 for the second half, or growth of 2.5% to 5.6%. The full-year double-digit growth the company advertises was substantially banked before this call.

Where the Street sits. There is no clean comparison available, and that is the point. Pre-print consensus was scattered across two different companies: a recast estimate near $4.49 for the quarter, a partially updated $4.75, and a pre-recast $4.95 that still contained Mobility. Full-year estimates carry the same problem, so the only honest anchor is the recast 2025 base of $15.85 against which the new guide is 10.4% to 12.0% growth. Consensus will re-form on the new perimeter over the coming weeks, and until it does the reported surprise figures for this company should be treated as unreliable.

Guidance style. Conservative, consistently, and now demonstrably so. In April this team guided to no acceleration in Ratings in the second quarter and delivered acceleration from 13% to 17%. It guided second-quarter margin expansion below the enterprise range in Energy and within it for Market Intelligence, and delivered 70bps and 120bps against an enterprise ex-OSTTRA target of 50 to 75bps. It set an initial hyperscaler issuance assumption of roughly $200B for the year and booked $169B in six months. The second-half guide should be read the same way: as a floor built to be cleared rather than a forecast. The offsetting consideration is that the second half genuinely carries the hardest comparisons of the year in the two divisions that are now producing all of the growth.

Analyst Q&A Highlights

Eighteen questions, and half of them went to one division. Nine touched Market Intelligence in some form: the pricing model for data sold outside the desktop, the reorganization and what it implies for disclosure, the lagging products, the lengthening renewal cycles, annual contract value, the AI contribution to divisional margin, the ten-year renewal, vendor consolidation under MCP, and the acceleration that did not arrive. Three went to the credit franchise, two to private markets, two to Indices, and one each to Energy and to the reinvestment trade-off. Where the first-quarter call was an argument about whether the franchise survives the decade, this one was an interrogation of a single division. Nobody asked about the tax rate, the second-half revenue arithmetic, or why the second-quarter buyback was cut in half.

Modular Data Buying and Where AI Pricing Ultimately Lands

The opening question set the agenda, asking what happens to pricing power when the distribution and interface layer is stripped away and customers assemble data modularly. The answer separated the relationship layer from the delivery layer, then handed the economics question to the CFO, who described a monetization model that is still being constructed.

Q: "I wanted to ask about the Kensho Data & Platforms business. I guess as you focus more on the Kensho Data side of the business and maybe lose the distribution and interface layer in some cases, give us some context on how your thoughts have evolved around future pricing and AI monetization, especially if your clients are able to approach data acquisition in a more modular way and kind of limit some of the data sources."
— Faiza Alwy, Deutsche Bank

A: "Faiza, the monetization approach is multifaceted. First, what we've continued to see is just a dramatic uptick in data usage by our clients through their AI calls. The MCP Connectors are up to 500 clients. Data usage is up 5x relative to the prior quarter and the prior quarter was up 5x prior to that. And so there's a series of ongoing discussions that we have with clients to how do we economically share in that. They get enormous benefits. And really, it's a mix of some consumption pricing, some additional data set pricing. And we feel quite comfortable that the revenues are beginning to come through and along with higher retention, higher sales and so forth, bode well for our businesses."
— Eric Aboaf, CFO

Assessment: usage up twenty-five fold across two quarters, and a pricing model described as ongoing discussions. That is the honest state of play and it is also the reason this cannot yet be checked against the income statement. "Revenues are beginning to come through" is the most specific revenue statement the company has made about AI monetization in two quarters of talking about it.

Whether the Lagging Products Get Fixed or Sold

The most direct attempt to size the problem inside Market Intelligence, asking both for a cumulative figure on the businesses facing headwinds and for the intended remedy. The reorganization question was answered fully. The sizing question was not answered at all.

Q: "I wanted to go back to MI to talk about -- a little more about the reorganization. Should we expect any disruption from the realignment? And you talked about some parts of the portfolio facing headwinds. Are you able to cumulatively size them? And would we expect that you look to sell those or divest those? Or are you going to try to fix those areas?"
— Toni Kaplan, Morgan Stanley

A: "The other point to make here is with respect to the softer products. These are smaller products, subscale, some consulting services, for example, some of the sustainability products. We're not talking about our strategic products like WSO or ClearPar, for example. And where there are cases where we may see cyclical impacts and growth could come back in those products, we're going to watch that very closely. And there may be opportunities otherwise if we don't see growth coming back, and we may think that we have to invest a little bit more, we may make some strategic choices there."
— Martina Cheung, President and CEO

Assessment: the question asked for a number twice and did not get one either time. Naming consulting and sustainability while explicitly ring-fencing the strategic franchises is useful, and the willingness to divest is a genuine signal. But without a revenue figure there is no way to test whether removing the drag lifts divisional growth by a point or by nothing, which is exactly the claim investors are being asked to accept.

Why Renewal Cycles With the Largest Clients Are Lengthening

A short, well-aimed question about a phrase that had passed quickly in the prepared remarks. The answer reframed a sales-cycle problem as a negotiating position, and in doing so disclosed what the negotiations are actually about.

Q: "Great. Just one quick one for me. I think, Martina, you mentioned that you're seeing some elongated sales cycles than your largest -- larger clients. Maybe just unpack that, what's driving it? Maybe what changed? Is it budget environment? Or just what's behind that comment?"
— Curtis Nagle, Bank of America

A: "Curt, thanks much for the question. Look, we view it actually as a positive because it means that our clients want to do more with us. And as we've been going through these renewal cycles, we're having quite substantial conversations around use of AI and use of our IP and data for AI. And the conversations are getting a little bit complex in some cases, and that's because we are absolutely intent and focused on protecting our IP over the long term. And so we've seen some elongation as a result of that."
— Martina Cheung, President and CEO

Assessment: this is the most important operational disclosure of the quarter and it explains the subscription line better than anything else said on the call. The company is deliberately slowing renewals to write intellectual-property terms it can live with for a decade. That is the right long-run trade and it has a near-term cost that shows up as flat reported growth. It also means the 6% subscription rate is partly a choice, which is a materially better explanation than competitive loss and one management should be making explicitly rather than in response to a question.

The Index Licensing Relationship Management Would Not Discuss

The sharpest question of the call and the only one that was refused outright. It concerned a long-standing exchange licensing arrangement that the questioner sized at more than $200M a year, with a renewal approaching and an argument that the fee could be substantially repriced.

Q: "I actually want to ask about the Index business, and this may be a little bit in the weeds and detail, so bear with me. But you have that licensing HEG relationship with CBOE. And by my numbers, this is definitely more than $200 million a year. And there's been more investor concerns around more from their side about that renewal that's coming up in a few years. Now just wondering what your latest thoughts are on this? Because again, if you look at that relationship, some people would suggest that you can probably double those fees."
— Alex Kramm, UBS

A: "Alex, it's Martina. Thanks for the question. Look, we don't comment on our partners or the conversations that we would have with any of our partners. But thank you for the question."
— Martina Cheung, President and CEO

Assessment: a twenty-eight word refusal on a contract the questioner sized at more than $200M a year, roughly a tenth of a division that just passed $2B of trailing revenue, with an identified repricing option. The refusal is defensible on commercial grounds and it is also the single largest undiscussed asymmetry in the Indices franchise. Investors are being asked to value a division at seventy-one percent margins without any view on whether one of its largest licences reprices up, holds, or moves to a competitor.

The Acceleration That Was Guided and Did Not Arrive

The most direct challenge of the session, holding management to a specific forward statement made on the prior call. The response did not dispute the premise; it restated the outcome and moved on.

Q: "On the 1Q call, you said that you were expecting an acceleration in MI subscription growth, and it doesn't look like it showed up this quarter. So I'm just curious what was the reason why? And are you expecting that MI subscription growth to accelerate in future quarters?"
— Jason Haas, Wells Fargo

A: "Jason, it's Eric. As we covered in some of our prepared remarks, we've had solid performance in MI. We did have a couple of smaller products that are lagging, consulting, sustainability that we called out. And so those tend to have some impact on results. I think in general, we're pleased with what we've seen in terms of execution, whether it's OCC revenue growth in MI was at 6%. Subscription was at 6%. Net retention rates are continuing to tick up."
— Eric Aboaf, CFO

Assessment: the question was why the number did not change and the answer was that the number is 6%. The second half of the question, on whether acceleration is still expected, was not addressed. This is the same exchange that occurred in April on annual contract value, with the same forward indicators cited and the same reported rate. Two quarters of unmet forward guidance on the single metric the disruption thesis turns on is now a pattern rather than a data point.

Second-Half Issuance and the Pushback on Hyperscaler Capital Spending

A recurring line of questioning on whether the credit franchise's engine is durable, sharpened by a question that explicitly played devil's advocate on the market's growing scepticism toward hyperscaler capital expenditure. Management's answer was that the guidance already assumes the slowdown.

Q: "Just to continue the thought from the last question, let me just play devil's advocate here. The market seems to be pushing back on some of the hyperscalers in terms of the amount of cap spending that they might be doing. Are you expecting any potential slowdown from debt issuance because of that? And also, if we do see interest rates going up, would you be expecting the same thing?"
— Multiple analysts incl. Jeffrey Silber, BMO Capital Markets; Ashish Sabadra, RBC Capital Markets

A: "Jeff, it's Martina. Let me take that question. So maybe a couple of things. The first is when we came to you earlier this year with the full year assessment, we said, look, we look at the announced CapEx, and we discount that for what we think will get debt financed and then we make assumptions around that as well. And so we are already tracking well ahead of where we thought we would be for the full year through the first half. But we've also said to you that our range is about $250 billion to $300 billion for the full year, which basically says not too much more hyperscale issuance relative to the first half and the back half of the year."
— Martina Cheung, President and CEO

Assessment: the best answer on the call, because it converted a narrative risk into a disclosed assumption. The guide already embeds hyperscaler issuance falling to roughly half the first-half run rate, which means a genuine capital-expenditure slowdown is largely pre-funded in the outlook and a continuation of current volumes is upside. The company also disclosed that hyperscalers have been issuing outside the United States, which broadens the channel rather than narrowing it.

Sizing the Ten-Year Renewal Inside the Subscription Line

The most practically useful modelling question of the call, asking for the contribution of a single contract to the quarter's most contested growth rate. The number was not given, and the answer redirected to what the contract's duration signals.

Q: "Just had a question on the Market Intelligence subscription revenue growth. Eric, I believe you called out some upfront revenue from a 10-year renewal that benefited this quarter. I was wondering if you could size how much that helped the MI subscription revenue growth? And then just how you're feeling on that building from that level throughout the rest of the year?"
— David Motemaden, Evercore ISI

A: "Subscription revenue growth, and saw a good performance in the 6% range the first 2 quarters in a row in MI. And obviously, we track that in our other divisions as well, with -- growth would have been slightly lower had it not been for that software renewal. But if you remember, there's always going to be lumpiness in software renewals. And what's particularly striking in this case is we have a client who wanted to sign a 10-year contract, right, not a 3-year, not a 5-year, but a 10-year contract."
— Eric Aboaf, CFO

Assessment: "slightly lower" is not a number, and on a metric this contested the absence of one matters. The redirection is nonetheless the strongest single piece of evidence against the disruption thesis produced on this call: a client signing a ten-year commitment to a workflow software product in the exact quarter the market has decided workflow software is obsolete. Both readings are correct. The reported growth rate is worse than 6% and the customer behaviour underneath it is better than the narrative.

What They're NOT Saying

1. The adjusted tax rate, at all. It fell to 19.8% from 23.3%, contributed $0.22 of the $0.91 increase in adjusted EPS, and sits 220bps below the low end of the company's own full-year guide. It was not mentioned in the prepared remarks and not asked about in Q&A. If the 22.0% to 23.0% full-year range holds, the second half carries a tax headwind that is nowhere in the discussion.
2. Stranded costs, still never quantified. The post-spin guidance reset arrived, and the number that was promised with it did not. The recast release folds them into a line called "allocation changes" alongside methodology enhancements, so the figure exists inside the comparatives but has never been stated. It was the largest unquantified item in the 2026 outlook in April and it remains one.
3. The 50bp revenue-recognition headwind that was supposed to reverse. In April management said Market Intelligence subscription growth carried a 50bp timing headwind that would reverse in the back half of the year. It went unmentioned on this call, in the same quarter that the promised acceleration did not appear. Either it reversed and was offset by something worse, or it did not.
4. Why the second-quarter buyback was halved. Repurchases fell to $500M from $1.0B in the first quarter and from $651M a year ago, in the quarter containing the stock's fifty-two-week closing low. More than $5.5B now has to be executed in the second half at a materially higher average price. The reduction was not explained and was not asked about.
5. What the full-year guide implies for the second half. Neither management nor any questioner stated that the enterprise guide implies second-half revenue growth of 1.1% to 5.1% and adjusted EPS growth of 2.5% to 5.6%, after 10.8% and 18.5% in the first half. The full-year figure was presented as double-digit growth, which is true and which the first half already delivered.
6. The dollar size of the ten-year renewal, or of the lagging products. Two direct requests for sizing on the two items that most affect how Market Intelligence's 6% should be read, and two answers without figures. On a division representing 34% of pro forma revenue and the entirety of the disruption debate, "slightly lower" and "smaller products, subscale" are not adequate disclosure.
7. Any revenue figure attached to artificial intelligence. Third consecutive quarter. Retention deltas, contract-value differentials, customer counts, call-volume multiples and a renewal uplift, and still no attach rate in dollars, no share of the installed base repriced, and no AI revenue line. The strongest counter to the de-rating remains uncheckable against the income statement.

Market Reaction

• Pre-print setup: the shares closed at $439.83 on 27 July, up 13.9% over the trailing thirty days and 17.7% above the fifty-two-week closing low of $373.75 set on 25 June. Raw price comparisons that cross 1 July exclude the Mobility Global share distributed that day, so the headline figures of down 15.8% year to date and down 17.1% over twelve months overstate the loss to a holder. The S&P 500 entered the print up 8.3% year to date.
• Premarket: the stock indicated down 5.2% at $416.94 within minutes of the release, before the call began.
• Print-day session (28 July, before the open): opened at $418.72, down 4.8%, traded as low as $406.00, down 7.7%, recovered to a high of $435.96, and closed at $424.36, down 3.5% or $15.47.
• Volume: 4.2 million shares against a thirty-day average of 2.3 million, or 1.8 times normal, though six sessions since the February guidance reset were heavier.
• Context: the S&P 500 rose 0.2% on the session. Across the closest information-services comparables, FactSet rose 6.8%, Thomson Reuters rose 5.7%, Intercontinental Exchange rose 2.7%, CME Group rose 1.8%, MSCI rose 1.5% and Moody's fell 0.4%.

The peer column is the whole story. This was a session in which the entire information-services complex rallied, several names sharply, and S&P Global fell 3.5% on a quarter that beat on revenue, beat on adjusted earnings, raised two division guides, raised the margin guide and raised the buyback by nearly $3B. Whatever moved the stock was specific to how this company's own numbers were presented, because nothing else in the group traded like it.

The mechanism is documented rather than inferred. The company reported a pro forma GAAP diluted EPS of $4.08 and a pro forma adjusted diluted EPS of $4.83, and multiple wire services picked up the first figure, labelled it as adjusted, and compared it to consensus estimates ranging from $4.75 to $4.95. One published the result as a shortfall of about 14.1%. Another headlined that the company had missed profit estimates. Simultaneously, a full-year adjusted EPS guidance range of $17.50 to $17.75 was set against a prior range of $19.40 to $19.65 that had assumed a full year of a business that no longer belongs to the company, producing a headline guidance cut of 9.7% at the midpoint. Both errors point the same way and both were available to any reader within seconds of the release.

The intraday shape is consistent with a market that worked it out slowly. The stock opened down 4.8%, made its low of down 7.7%, and then spent the session recovering to close down 3.5%, roughly half way back. That is the signature of an initial algorithmic and headline-driven reaction being partially corrected by readers who reached the reconciliation tables. It is not the signature of a considered verdict on the quarter.

What the session does not settle is whether the residual 3.5% decline reflects the two genuine disappointments in the print. Market Intelligence subscription growth did not accelerate as guided, and the full-year outlook implies a second half that contributes almost nothing to growth. A market that had read the release correctly would have found both, and might well have marked the stock lower on them. The problem is that there is no way to separate the part of the move that was analysis from the part that was a misread line item, and the burden for that ambiguity sits with the company's disclosure choices rather than with the tape.

Street Perspective

Debate: Has the re-rating already happened?

Bull view: the bull case being made on the Street is that at 24.1x the midpoint of a guide this management has twice proven conservative, against a business compounding at low double digits with 5.6% of the share count being retired this year, the multiple still sits well below where an unimpaired benchmark franchise trades. The de-rating that took the stock to 22x in April has only partly reversed, and it reversed on the spin rather than on any resolution of the disruption question.

Bear view: the bear camp contends that the easy part of the re-rating is done. The stock has rallied 13.5% off its June low, the multiple has expanded 190bps since April, and the expansion came from removing a lower-multiple asset rather than from anything the remaining business proved. At 24x with a second-half guide implying low-single-digit growth and the two slowest divisions representing half the revenue, the risk is now symmetric rather than skewed.

Our take: the bears have the better of the valuation point and the bulls have the better of the earnings point. Our initiation argued the market was applying a disruption discount to a business its own customers contradicted, and roughly half that gap has closed without the argument being settled either way. What replaces it is narrower and more checkable: whether a company guiding its second half to near-zero growth, having beaten its own cadence guidance twice in a row, is describing the year it expects or the year it can promise. We think the latter, which is why the rating holds, but this is a smaller and more conditional edge than the one we initiated on.

Debate: Does the growth concentration make the company better or more cyclical?

Bull view: the argument for concentration is that Ratings and Indices are the two highest-margin, lowest-capital, most defensible franchises in financial information, that they now produce more than 80% of operating profit, and that a company earning its returns from standards rather than from software is precisely what an investor should want to own into an AI transition. The spin removed a good but structurally different business and made the remaining company easier to value.

Bear view: some sell-side desks argue the reverse, that a business deriving 77% of its revenue growth from debt issuance volumes and equity market levels has become a cyclical dressed as a compounder, that both drivers are at cycle highs, and that the diversifying asset was disposed of at the top of the cycle in the businesses that remain. The fourth quarter, which management has guided to negative growth in Ratings, is the first test.

Our take: the concentration is real and the cyclicality argument understates the duration of the driver. The AI capital-expenditure cycle is producing investment-grade supply on a multi-year schedule with roughly $11 trillion of rated debt refinancing over the next four and a half years behind it, and it reaches this company as transaction revenue at a 68.5% margin with no incremental capital. Where the bears are right is that this leaves nothing to absorb a shock. If issuance and index assets turn together, there is no longer a Mobility to cushion it, and Market Intelligence at 6% will not do that job.

Debate: Is Market Intelligence fixable or melting?

Bull view: the constructive case is that the division has been reorganised around exactly the right distinction, that Kensho Data is growing high single to low double digits and is where the AI monetization lands, that margin expanded 120bps on 4% expense growth, and that the lengthening renewal cycles are a deliberate choice to write better intellectual-property terms rather than evidence of competitive loss.

Bear view: the sceptical case is that the company just disclosed that the larger half of its data business is growing at low single digits organically, that subscription growth has been flat at 6% for two quarters after management guided to acceleration, that the printed 6% contains one-time upfront revenue, and that the strategic response is a margin programme rather than a growth plan.

Our take: the bears are winning this one on the evidence and the outcome is still genuinely open. What tips it is that management's own remedy concedes the premise: Platforms is being managed for profitability and consolidation while Kensho Data is being managed for growth. That is the correct allocation of effort and it is an admission that the workflow estate is a mature asset. The division can still be a good business on that basis. It cannot be the growth engine the market once paid for, and at 34% of pro forma revenue that matters to the whole.

Model Update Needed

The Mobility separation resets every full-year figure, so the table below revises our April initiation onto the new perimeter rather than adjusting it at the margin.

Valuation. At the 28 July close of $424.36 the shares trade at 24.1x the midpoint of the new full-year guidance and 26.8x recast 2025 adjusted EPS of $15.85, against 22.2x on the old perimeter at the time of our initiation. The market capitalisation is $125.1B. Our price target moves to $490, which is 24.75x our FY27 estimate of $19.80 and implies +15.5% from the 28 July close, before a dividend yielding roughly 0.9%. That target is our April target of $510 less the value distributed as Mobility Global, which is to say it is unchanged in substance: the case still requires no acceleration, no estimate revision and no resolution of the disruption debate. It requires the multiple to hold while the company compounds earnings at low double digits and retires 5.6% of its shares. The clearest downside marker is the fifty-two-week closing low of $373.75 set five weeks before this print, which is 12% below the current price and was reached without any of this quarter's news.

Thesis Scorecard Post-Earnings

The pillars below are the ones established at initiation in April and carried in our standing thesis. They are graded against what this quarter's print and call revealed, not re-derived.

Overall: the thesis is intact and its composition has shifted. Three of four bull pillars strengthened materially, one strengthened in size while acquiring a leverage cost, and the two bear points that carry real weight both moved against us. The version of this thesis we initiated on rested on a multiple that was too low. The version that survives this quarter rests on benchmark earnings power the guide does not credit and on 5.6% of the share count disappearing, with Market Intelligence now a genuine drag on the story rather than a debate about one.

Action: hold the position and add on weakness toward the June low. The quarter did not change what this company is and it did change what the case depends on. Buying at 24.1x a guide from a team that has beaten its own cadence twice in a row, into a $7B repurchase programme executing over the next five months, is a reasonable risk. Buying it as a re-rating trade is no longer the argument, because half the re-rating has been paid. The two things that would change the rating are a third consecutive quarter of 6% Market Intelligence subscription growth, which would make the drag structural, or a fourth-quarter Ratings decline materially worse than the guide implies.