The Part of S&P Global the Market Was Discounting Is Now Up for Review, and the Rest Just Raised Its Guide

Results vs. Consensus

Q3 2026 Scorecard

Year-over-Year Comparison

The company describes adjusted margin expansion as 60bps, and 120bps excluding the prior-year OSTTRA contribution. Its prior-year segment components sum to $1,822M of adjusted operating profit, which produces those figures; the filed prior-year total is $1,819M, which produces the 70bps and 130bps shown here. The gap is immaterial to the thesis and we use the filed totals throughout. The $200M GAAP gain on the sale of the upstream software portfolio is what separates the 26% GAAP EPS growth from the 9% adjusted figure.

Sequential Comparison

Revenue was flat to the dollar against the second quarter. The composition rotated: Ratings gave back $68M sequentially, and the three other divisions added $70M between them. That is the mix shift the thesis needs if the credit cycle softens, and this is the first quarter it has shown up.

Revenue. Seven percent is a step down from 11% in the second quarter, and it was supposed to be. The credit franchise was always going to slow against a third quarter of 2025 that had recovered from the tariff disruption, and Ratings grew 2.5% on the reported line against the flat-to-down cadence the July guide implied for the second half. What carried the print instead was breadth: Indices at 20%, Market Intelligence at 7% and Energy at 6% organic. Three of four divisions accelerated, and two of them are the ones the market has spent the year treating as structurally impaired. The revenue line no longer depends on the issuance cycle to the degree it did in the first half.

Margins. The enterprise margin is a weighted average of one division going backwards and three going forwards. Indices reached 71.8% on 20% revenue growth, Energy expanded 190bps to 48.1% on just 2% adjusted expense growth, and Market Intelligence reached 36.2%. Ratings went the other way, from 67.0% to 65.7%, because adjusted expenses grew 7% against 2% revenue growth. Management called that timing. The 10-Q attributes the expense growth to merit increases, additional headcount, technology costs and strategic investments, which are not timing items. A business with a fixed analytical cost base shows negative operating leverage when volumes flatten, and that is what this is.

EPS. The adjusted tax rate rose to 23.1% from 21.2% a year ago and 19.8% in the second quarter. In July we noted that the first-half rate of 21.5% sat below the 22% to 23% guide and that the second half would carry the difference. It has started to. The nine-month rate is now 22.0%, so the guide still requires a fourth-quarter rate somewhere between roughly 22% and 26%. The share count is doing the opposite work: diluted shares fell 4.2% year over year and 1.3% sequentially, and the $2.6B accelerated repurchase that drove that was completed at an average of $412.99 per share, 11.5% below the current price.

Segment Performance

Total adjusted operating profit is stated after $31M of adjusted corporate unallocated expense, which is why it is less than the $2,000M sum of the four divisions.

Revenue by Type

The table we track for the disruption debate. Subscription is the part a language model could in principle substitute; the other five lines are paid on issuance, surveillance, index assets, derivatives volumes and transaction counts.

The non-transaction line is shown net of the $(55)M intersegment elimination, following the filing's presentation.

Assessment: the gap that made this table so persuasive in July has narrowed. In the second quarter the non-subscription half grew 16% against 5% for subscription. This quarter it grew 7.7% against 6%, because transaction revenue slowed to 2% as issuance mix moved toward investment grade. The non-substitutable half is still the faster-growing half, and it is now growing on index assets and derivatives rather than on issuance. The review would remove a large part of the subscription book, which is the cleanest possible answer to the question this table asks.

Ratings (35% of revenue, +2.5%)

The deceleration management had signalled all year, delivered on schedule. Billed issuance rose 8%, led by investment grade up 28% on AI-related and M&A deals, while high yield fell 17% and the bank-loan, structured-finance and government bucket rose 1%. Transaction revenue grew only 1% because the mix moved toward investment grade, which carries lower fees per dollar than high yield and bank loans. Non-transaction revenue grew 4% on annual fees and CRISIL, against what the CFO called a much tougher comparison that continues into the fourth quarter.

"Ratings revenue grew 2% in the quarter. As expected, growth moderated from the first half of the year against tougher comparisons. Issuance was up 8% and transaction revenue increased 1%." — Eric Aboaf, CFO

Year to date, total billed issuance is up 15% to $3,725B, with investment grade up 32% and high yield down 1%. The margin result is the line to watch. Adjusted expenses grew 7% and adjusted operating profit rose 0.5%, so margin fell 130bps to 65.7%. Management pointed out that year-to-date Ratings margin is still up just over 100bps and guided the full year toward the lower end of the enterprise expansion range.

Assessment: better than guided, and the first quarter of a harder period. The July guide implied second-half Ratings revenue between down 4.4% and up 1.4%; the third quarter delivered up 2.5% and the guide was raised to 6.5% to 8% organic. But the fourth quarter is guided to a slight decline in transaction revenue, and 2027 laps a year in which the first three quarters were the strongest on record for billed issuance. The 68.5% margin we called the best result this division had produced was a peak, not a run rate.

Indices (15% of revenue, +20%)

The fourteenth consecutive record quarter. Asset-linked fees rose 24% on market appreciation and net inflows, with ending ETF assets under management up 26% to $6.517 trillion and average ETF assets up 32%. Exchange-traded derivatives revenue rose 19%, the fourth consecutive quarter of double-digit growth, led by SPX volumes. Data and custom subscriptions grew 7%, or 5% organically. Adjusted expenses grew 14%, and the margin still expanded 150bps to 71.8%.

"As the world's largest index business by AUM, we achieved a new record quarter for ETF AUM of $6.5 trillion in the quarter, and our strategic initiatives are contributing to a number 1 ranking in total ETF market flow capture among index providers." — Eric Aboaf, CFO

The full-year guide moved to 17% to 18.5% organic from 12% to 14%, the largest revision in the print. It implies fourth-quarter growth of roughly 13% to 19%.

Assessment: the strongest division in the company, and it now has a contracted tailwind for 2027 from the Cboe extension discussed below. The one number worth watching is that asset-linked fees grew 24% against average ETF assets up 32%. Asset-linked fees include mutual fund and insurance products, so the comparison is approximate, but it suggests the fee rate per dollar of assets continues to drift lower as flows favour the cheapest products.

Energy (18% of revenue, +6%)

The reacceleration the July guide required arrived on time. Organic constant-currency growth rose to 6% from 2% in the second quarter. Platts grew 7% on price assessments, and Global Trading Services stabilised. CERA grew 5%, or 6% organically after adjusting for the upstream software divestiture, which closed on 1 September and produced the $200M GAAP gain. Sanctions cost about half a point of growth. Adjusted expenses grew 2%, and the margin expanded 190bps to 48.1%.

"We saw better traffic through the Strait of Hormuz, with flows reaching almost 2-thirds pre-war levels. That made it easier for customers to transact with confidence and drive subscription growth in the quarter." — Martina Cheung, President and CEO

Two caveats came with the result. The CFO said the quarter benefited from very strong one-time sales that may not recur in the fourth quarter. And the fourth-quarter margin will be held back by the loss of upstream software and the addition of datacenterHawk, which closed on 1 September.

Assessment: the division that was pricing below its franchise because of the Iran conflict is now pricing closer to it, and management says renewals are back at pre-war levels. The full-year guide of 4.5% to 6% organic is unchanged, and nine-month growth of 5.3% sits inside it. The more durable point is in the AI numbers: Energy is where AI customers grow their spend about three times faster than other customers, and Energy stays with the company whatever happens to Capital IQ.

Market Intelligence (34% of revenue, +7%)

The division that missed its guided acceleration in July delivered one in October. Organic constant-currency revenue grew 7%. Subscription grew 6% reported and 7% organic. Kensho Data & Platforms grew 10%, or 6% organically, up from 4% organic in the second quarter, with Kensho Data itself growing comfortably in double digits. Enterprise Solutions grew 2% reported and 8% organically on lending products such as WSO. Recurring variable revenue grew 11%. Adjusted expenses grew 5%, and the margin expanded 100bps to 36.2%.

"Sales cycles have normalized following a slight elongation last quarter. And we are no longer seeing a negative impact to revenue growth. In fact, we saw a small benefit in the quarter from some true ups on contracts that renewed after the second quarter ended." — Eric Aboaf, CFO

That last sentence matters for how the 7% is read. Contracts that slipped out of the second quarter renewed in the third with retroactive revenue attached, so part of the third quarter's growth belongs to the second. Averaging the two quarters gives an organic subscription rate of roughly 6.5%, which is also where annual contract value landed. The guide implies the fourth quarter runs below the 7% just printed.

Assessment: real improvement, smaller than the headline. Annual contract value moving to about 6.5% from the "6% range" is the forward indicator that did not move for two quarters, and it moved. But on the reported subscription line this is a third consecutive 6%, and the July trigger we set on that metric is graded in the Thesis Scorecard below. The larger point is that most of the subscription book in question is now inside the strategic review perimeter.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: confident and more strategic than any call this year, with the operating results presented briskly so the structural announcement could carry the hour. Management was specific on what is in and out of the review perimeter and deliberately unspecific on everything an investor would need to value it. The credit franchise's deceleration was owned rather than explained away, which is a change from a team that spent the first half telling investors not to extrapolate upside.

1. The Capital IQ Review: What Is In the Perimeter and What Stays

The announcement came in the release headline and opened the call. The company has started a review of alternatives, including separation, for Capital IQ and its associated data businesses, which together are the majority of Kensho Data & Platforms. In Q&A management listed financials data and Visible Alpha among the data businesses in scope. RatingsDirect, RatingsXpress and Enterprise Solutions, including the lending franchise built around WSO and ClearPar, stay with the company.

"We announced today that we have initiated a strategic review of alternatives for Capital IQ and its associated data businesses within Market Intelligence, including their potential separation. The businesses being considered for separation include the Capital IQ and the Kensho data and platforms businesses within Market Intelligence. While the exact perimeter remains under review, we expect to retain ratings direct, and ratings express as well as Enterprise Solutions." — Martina Cheung, President and CEO

The scale disclosed was customer scale only: more than 15,000 clients and more than 355,000 users. Press reports in early September had described a potential spinout worth billions of dollars, and the stock rose 11.7% in the thirty days before the print. Guidance assumes a full year of Market Intelligence as currently structured.

Assessment: this is the logical end of the argument management made in July, when it called the company "primarily a benchmark business" and disclosed that Platforms, the larger half of the data estate, grows low single digits organically. Keeping RatingsDirect and RatingsXpress keeps the distribution of the benchmark. Keeping Enterprise Solutions keeps the workflow tied to credit markets. What leaves is the desktop and the third-party data that a language model can most plausibly approximate. It is the right perimeter for the thesis we hold.

2. Why Now, and What Happens to the AI Story

Management framed the timing as selling from strength, pointing to Kensho Data's acceleration. The awkward part is that Kensho, the engineering team behind the MCP connectors, the LLM-ready APIs and the adaptive retrieval agents, sits close to the perimeter. Asked what would remain, the CEO described Kensho as very important to the perimeter while promising the company would keep the AI capabilities it needs.

"And so, I think here, we would look at the Kensho team as being very important to the perimeter. We do not know exactly how that is going to shape up, but what I will absolutely say is that we would make sure that all of the momentum that is building up in KDP with the Kensho Labs team and the other Kensho engineers that is absolutely continuing at pace." — Martina Cheung, President and CEO

The integration answer was more reassuring than the AI answer. Ratings, Indices, Energy and Enterprise Solutions each run on their own technology stacks, and the main cross-divisional links are commercial: Ratings analysts use Capital IQ, and Market Intelligence pays Ratings an intersegment royalty, $47M in the quarter, to distribute its content.

Assessment: the review partly separates the company from the evidence it has used for three quarters to argue that AI is a monetisation channel. The ACV differentials, the call-volume multiples and the large strategic renewals were mostly Market Intelligence data. What stays is Energy's AI franchise and the Ratings content now available through MCP. That is a smaller AI story and a purer benchmark story, and we think the market will prefer the second.

3. Ratings: The Deceleration Arrives, and the Margin Gives Some Back

For three quarters management told investors the second half would be harder. It was. Revenue grew 2.5% with transaction revenue up 1%, and the variance between 8% billed issuance growth and 1% transaction growth came from mix: investment grade carries lower fees per dollar than the high-yield and bank-loan volumes that fell.

"While margins contracted approximately 130 basis points in the quarter, this can be impacted by the timing of expenses. Year to date, ratings margins have expanded just over 100 basis points." — Eric Aboaf, CFO

The guide still rose, to 6.5% to 8% organic from 5% to 8%, and the fourth quarter is guided to a slight decline in transaction revenue with modest non-transaction growth. Our arithmetic puts the implied fourth quarter between down about 3.4% and up about 2.5%. M&A-related issuance is now expected to grow 20% to 30% for the year.

Assessment: the franchise performed as guided and the guide was conservative again, the third time in three quarters. What is new is the cost side. Ratings expenses grew 7% on 2% revenue, and the explanations in the 10-Q are structural rather than timing. If 2027 issuance is flat to down, this is the margin profile of the division in a slower year, and investors should model it that way.

4. The Hyperscaler Ledger: $240 Billion Booked Against a $250 to $300 Billion Year

The quantified hyperscaler disclosure from July continued. Year-to-date hyperscaler billed issuance reached about $240B, which means roughly $71B in the third quarter after $169B in the first half. The full-year range of $250B to $300B was held.

"Hyperscaler issuance remained strong in the quarter, bringing the year to date total to approximately $240 billion, and we continue to expect between $250 billion and $300 billion for the full year. We have seen continued strength in investment grade issuance outside of the hyperscalers. Excluding hyperscaler issuance in the third quarter, investment grade issuance would have increased around 15%." — Martina Cheung, President and CEO

Holding the range implies $10B to $60B of hyperscaler issuance in the fourth quarter, a fraction of the third-quarter pace. Management also said its own team forecasts nearly $900B of capital expenditure this year for the six largest hyperscalers, nearly double 2025.

Assessment: the range is the conservatism pattern again. Management raised the full-year range in July after booking $169B in six months, booked another $71B in three, and kept the range. Either it expects a near-complete pause in the fourth quarter, or the range will be beaten for a second time. The more useful disclosure is that investment grade excluding hyperscalers grew about 15%, which means the AI cycle is not the only thing holding up investment-grade supply.

5. The Cboe Licence Is Fixed to 2051

In July the sharpest question of the call concerned the exclusive SPX options licensing relationship with Cboe, which the questioner sized at more than $200M a year. Management declined to comment. In September the extension was announced, and on this call the economics were described for the first time.

"Last month, as Martina mentioned, together with CBOE Global Markets, we announced the extension of our exclusive licensing agreement through 2051. The new agreement builds on more than 40 years of collaboration and positions us well to drive further innovation and growth. The contract extension will not impact 2026 financial results, but we do expect a couple percentage points of incremental total index revenue growth in 2027." — Eric Aboaf, CFO

The Indices CEO added that the extension accelerates the economics into the division by six years and provides a further tailwind after 2027. Detail comes with 2027 guidance in February.

Assessment: the single largest undiscussed asymmetry in Indices, as we described it in July, has been resolved in the company's favour and fixed for 25 years. A couple of points of incremental growth on a division above $2B of revenue at a 72% margin is worth a few cents of EPS in 2027. The value is less in the cents than in removing a renewal risk from a division that is now 15% of revenue and the fastest-growing part of the company. This is the main reason our target multiple rises by one turn.

6. Prediction Markets and Onchain Finance: New Adjacencies

The second half of management's strategy, expanding into high-growth adjacencies, now has names. A preferred, non-exclusive partnership with Kalshi was announced the day before the print, covering both index licensing for event resolution and new indices built from prediction-market data. The OpenZeppelin acquisition, a smart-contract security provider, closed on 26 October inside Ratings. The Agusto majority stake is expected to close in the fourth quarter.

"We are actively pursuing opportunities to expand the addressable market for our intellectual property in areas like prediction markets. And we continue to innovate and explore opportunities in partnerships that take advantage of the rapid innovation in this space." — Martina Cheung, President and CEO

Management said none of these announcements will materially affect 2026 results.

Assessment: options, not earnings. The logic is consistent with the rest of the strategy: a benchmark provider licenses the standard, whatever venue trades it. Choosing a non-exclusive arrangement while extending the exclusive Cboe licence is a sensible hedge in a market whose eventual winners are unknown. We give these nothing in estimates.

7. Market Intelligence: The Acceleration, and the Catch-Up Inside It

The July call left Market Intelligence as the division that had promised acceleration and not delivered. This call reported it, with organic subscription growth of 7%, Kensho Data & Platforms organic growth of 6% from 4%, and annual contract value at about 6.5%.

"In the third quarter, subscription revenue grew 6%, and 7% on an organic basis, benefiting from broad based growth across our subscription franchises. Kensho Data and Platforms revenue growth accelerated to 10% or 6% organically." — Eric Aboaf, CFO

The CFO also disclosed a small benefit from true-ups on contracts that renewed after the second quarter ended, and the 10-Q lists retroactive revenue from delayed renewals among the drivers. Management expects the division to finish the year slightly above the midpoint of its 5.5% to 7% range. With nine-month organic growth at 6.6%, that implies a fourth quarter closer to 6% than 7%.

Assessment: the honest reading is two quarters averaging about 6.5%, which is a modest improvement on the 6% we had been tracking and consistent with the contract-value figure. It is enough to say the drag is not getting worse. It is not enough to say the division has re-rated its own growth. The review makes the question less important for shareholders than it was three months ago.

8. AI Revenue Gets a Number, and It Is Small

For three quarters the company disclosed every AI metric except revenue. This call gave a growth contribution: discrete AI products contributed about 50bps of revenue growth in Energy and about 50bps in Kensho Data & Platforms. On Energy's prior-year revenue base, half a point is roughly $3M of incremental revenue in the quarter.

"In market intelligence, ACV growth is now 1.8x the growth rate of non AI customers. In energy, it remains approximately 3x. This is a strong signal of the value of our AI solutions to our customers. Over time, we believe growth rates will converge as more of our customers become AI customers." — Martina Cheung, President and CEO

LLM-ready API call volume grew nearly 50% sequentially, after growing more than fivefold in each of the two prior quarters. The customer count for MCP and LLM-ready APIs, more than 500 in July, was not updated.

Assessment: the first number confirms the direction and sets the scale. AI is raising price and retention across the base, which shows up in contract value, and it is producing a small amount of discrete revenue on top. That is a monetisation channel, as our thesis argued, not a growth engine. The deceleration in call volume from fivefold to 1.5x is the normal shape of adoption after a launch, and we would not read it as a warning.

9. AI Rights Negotiations Have Normalised

In July the CEO explained lengthening renewal cycles with the largest clients as the cost of writing intellectual-property terms for AI use. This call reported that the negotiations have settled into a standard template.

"Negotiations with customers on AI rights are normalizing, and we are signing deals with robust pricing, that protect our IP. Customers see the value in leveraging our AI-ready data while respecting our intellectual property. Sales cycles stabilized, following some temporary noise last quarter, contributing to the acceleration in subscription revenue growth in both market intelligence and energy." — Martina Cheung, President and CEO

In Q&A the CEO said clients have accepted what are now essentially standard terms and conditions across the business, and management described a large multiyear competitive displacement in Market Intelligence and a large energy deal combining an upstream win-back with AI-ready data.

Assessment: this was the most important operational disclosure of the July quarter, and it has resolved as management said it would. Customers accepted the company's IP terms and paid for them. That is the commercial evidence that matters most for the data moat, and it applies to Energy and the retained Ratings content as much as to the businesses under review.

10. Energy Reaccelerates as the Conflict Eases

The second quarter's Energy story was a division choosing not to push price on customers disrupted by the Iran conflict. The third quarter's is a division whose customers have adapted. Renewals are back to pre-war levels, Global Trading Services stabilised, and traders are returning to derivative products priced off Platts benchmarks.

"Energy revenue rebounded to 6% on both a reported and organic constant currency basis, benefiting from continued strong demand for Platts benchmarks. As well as our differentiated data research and thought leadership." — Eric Aboaf, CFO

Management expects oil to stay around $100 per barrel through the fourth quarter, says LNG, refined products and chemicals remain constrained, and continues to expect organic growth to return to the 6% to 8% Investor Day range in 2027.

Assessment: the July commitment, a second half running at 4.3% to 7.3%, has been met in its first quarter. The upstream transformation has also turned, with upstream data back to positive growth and early CERA Titan feedback described as encouraging. Energy is about to become a larger share of a smaller company, and this quarter is the right base from which to judge it.

11. Capital Return: The Accelerated Repurchase Bought at $412.99

The $2.6B accelerated share repurchase launched the day after the July print completed on 26 October. It retired 6.3M shares at an average of $412.99, against a 30 October close of $466.55. Year-to-date repurchases are $4.1B, and the company reiterated more than $7B for the year, which leaves more than $2.9B for the fourth quarter.

"We delivered 8% growth in adjusted operating profit and 9% growth in EPS, driven by more than $6 billion in share repurchases over the last 12 months, including another $2.6 billion in share buybacks launched in the third quarter, EPS growth was partly tempered by the timing of tax expenses." — Eric Aboaf, CFO

About $2B of new debt will be issued at prevailing rates to fund the remainder, consistent with the plan set out in July, and a new shelf registration was filed on 23 October. The adjusted interest guide rose to $405M to $415M, which implies a fourth-quarter adjusted interest charge of $120M to $130M against $93M in the third.

Assessment: well executed. In July we noted that the second-quarter buyback had been halved in the quarter containing the 52-week low and that more than $5.5B had to be deployed at higher prices. The company launched $2.6B the day after the print, and the $412.99 average sits near the low end of the range the shares have traded in since. Diluted shares fell 4.2% year over year, and the share count is now the largest single contributor to EPS growth after operations.

Guidance & Outlook

Full-Year 2026 Enterprise Guidance (continuing operations)

Division Guidance (organic constant currency)

The July Ratings and Indices ranges were stated as revenue growth; the new ranges are organic constant currency, and the acquisition effects in both divisions are a few million dollars a quarter. Implied fourth-quarter figures are our derivations against recast 2025 divisional revenue and are directional.

The prepared remarks presented the raise as the product of outperformance in Ratings and Indices, with the margin raise coming from higher revenue and expense discipline, and the higher interest charge more than offset by lower corporate expense and a lower share count. The CEO framed the year as one of resilience against geopolitical conflict, shifting rates and commodity swings. The CFO was precise about what will not repeat: Energy's one-time sales, the upstream software revenue, and the Ratings transaction comparison.

Implied fourth-quarter ramp: recast 2025 revenue was $13,589M, of which $3,472M fell in the fourth quarter. The nine months delivered $11,074M. The 7.4% to 8.4% guide implies fourth-quarter revenue of $3,521M to $3,657M, growth of 1.4% to 5.3%. Recast 2025 adjusted EPS was $15.85, with $3.82 in the fourth quarter; the nine months delivered $13.86, so the $17.70 to $17.95 guide implies roughly $3.84 to $4.09, growth of 0.5% to 7.1%, in a quarter where the share count is about 5% lower than a year ago. The margin guide implies a fourth-quarter adjusted margin of about 47.3% to 48.5% against 48.3% a year ago. In other words, the guide again asks the last quarter for close to nothing.

Street at: third-quarter consensus was $4.43 of adjusted EPS on revenue of about $3.64B. The post-print full-year estimates we can see sit in the upper half of the new range, which means the Street has already concluded the fourth quarter is a low bar.

Guidance style: conservative, for the third consecutive quarter, and now measurable. The July guide implied second-half revenue growth of 1.1% to 5.1%; the first half of that half delivered 6.8%. It implied Ratings between down 4.4% and up 1.4%; Ratings delivered up 2.5%. It assumed $250B to $300B of hyperscaler issuance after $169B in the first half; the third quarter added about $71B. A team that guides this way is telling investors where the floor is. The same reading applies to the fourth quarter, with one caveat: the tax rate and the interest charge are both genuinely moving against it.

Analyst Q&A Highlights

Six of the fifteen questions concerned the review: why now, how big, what margin, how entangled, what happens to the AI capability, and what the remaining company looks like in three years. Management answered the structural questions fully and the numerical ones not at all. The rest went to AI and Market Intelligence, issuance, Indices and prediction markets, the bond offering and capital expenditure. Nobody asked about the Ratings margin decline, the tax rate, or year-end leverage.

Why the Review Is Happening Now, and What the Remaining Company Looks Like

The opening question asked the obvious two things together: the timing, and the effect on the financial profile of what remains. The CEO answered the first with a position-of-strength framing and the second with a strategic one. The CFO then declined to add numbers.

Q: "Why did you decide to launch the strategic review now? And also, as we think about the remainco, how does the strategic review enhance both the financial profile as well as the quality of the Remainco?"
— Ashish Sabadra, RBC Capital Markets

A: "This is really a function of long-term positioning. For both S and P Global as well as Cap IQ and the associated businesses. And think it is important to recognize that as we enter into this review, we are coming at it from a position of strength, really. ... And so, you know, while we have the potential to see if under a new owner, the KDP portions of the business that are in the perimeter could grow faster, could innovate faster, we also have an opportunity to look at a more focused portfolio within S and P Global and then make sure that we are actually focusing our investments in the markets where we are leading and where we are the strongest."
— Martina Cheung, President and CEO

Assessment: "could grow faster, could innovate faster" under a new owner is an unusually direct concession that the business is underinvested inside the current structure. That is a candid reason to separate it. It is also a reason a buyer will use when pricing it.

Sizing the Perimeter, and How Far the Process Has Already Gone

The most direct attempt to size the review. The question reminded management that Capital IQ alone has been described as roughly 6% of revenue and asked for a range including the associated data. No size was given. The CFO instead revealed that the process is further along than the word "review" suggests.

Q: "You have said in the past that Cap IQ is approximately 6% of revenue. Sounds like maybe there are some additional pieces of Kensho data and platforms that are being considered here. And so I was wondering if there was any way to give either a range or a approximate size and. Inclusive of this additional data."
— Toni Kaplan, Morgan Stanley

A: "We have also been front footed in terms of going out and exploring interest. We have gotten quite a bit of interest from a number of different potential-- potential firms on the other side. You know, we have gone ahead and signed NDAs with a number of them. You know, we have engaged with advisers."
— Eric Aboaf, CFO

Assessment: signed NDAs with several parties and engaged advisers, in a process announced that morning, describes a sale process already under way. The size question went unanswered, but the answer to the question that was not asked, whether this is a sale rather than a spin, leans clearly toward a sale.

The Margin Profile of What Leaves, and Why Enterprise Solutions Stays

The question that matters most for valuing the remaining company: what margin the perimeter earns once the high-margin Ratings distribution is carved out, and why the enterprise software stays. The margin question was declined. The Enterprise Solutions question received a full strategic answer about credit-market workflow.

Q: "I was just curious if you could help us with the margin profile, especially once you move the high margin rating businesses? And then I also just wanted to understand why enterprise solutions you know, kind of is being retained and perhaps the margin profile there as well?"
— Manav Patnaik, Barclays

A: "So I think as we mentioned earlier, it is a little bit too early. it is premature really to comment on the growth and margin profile of the in scope products and services for the strategic review and that is really. We are going to have to just nail down the perimeter before we can do that."
— Martina Cheung, President and CEO

Assessment: the margin of the perimeter determines whether a separation is accretive or dilutive to the remaining company's earnings, and it is the one number management certainly knows. Market Intelligence runs at 36%; if the high-margin Ratings distribution stays behind, the perimeter's margin is probably below that. Selling a lower-margin, lower-growth business can lift the remaining company's multiple while reducing its EPS, and investors will need the number to judge which effect is larger.

How Entangled the Perimeter Is, Compared With Mobility

A practical question about separation risk, benchmarked against the spin that has just been completed. The answer named the commercial links and then made the case that the technology is already separate.

Q: "You talk about how integrated Cap IQ and the other businesses under review are with the rest of S and P? And maybe you could compare to how integrated they are versus, what we saw with mobility."
— Jason Haas, Wells Fargo

A: "I would say that there are cross-divisional commercial agreements. So for example, the ratings analyst actually use Cap IQ and then we have. Obviously, we have got a little bit of sort of a data foundation under CapIQ helps with some other parts of the business as well. I would not say that this is deeply complicated in the sense that for example, enterprise solutions has completely different technology underpinning their platforms."
— Martina Cheung, President and CEO

Assessment: credible, with one caveat. Separate technology stacks make a carve-out tractable. But "a little bit of sort of a data foundation" under Capital IQ that helps other parts of the business is precisely the kind of shared cost that became the unquantified stranded costs after the Mobility spin. Investors should expect the same treatment and ask for the number earlier this time.

Whether a Record Issuance Year Borrows From 2027

The most important question for the earnings of the company that remains. The question cited the company's own ratings research, which had just forecast slower market issuance in 2027. Management declined to give 2027 commentary, then made the medium-term secular case.

Q: "Martina, I believe your ratings research team just put out something in the last day or so, a forecast slower market issuance in 2027. ... And I know it is early but it would be great to hear your take on, if the strong issuance growth this year indicates perhaps slower growth next year and how we should think about the puts and takes under various macro and credit scenarios."
— Faiza Alwy, Deutsche Bank

A: "Yes, it is a little bit too early for 2027 commentary. Look, we look forward to providing more specifics on that in February. Now that said, I would say we look at this market overall over the medium term as being high growth with, very strong secular tailwinds. And it is an area where we have extremely strong refinancing walls around 11 trillion over the next 4 years."
— Martina Cheung, President and CEO

Assessment: the answer confirmed the forecast rather than disputing it. The medium-term case is sound and the refinancing wall is real, but medium term is not next year. The CEO also described the method for the 2027 hyperscaler assumption, discounting announced capital expenditure for the share that will be debt-financed and done in year, and cited sell-side estimates of just under $3 trillion of remaining performance obligations at four hyperscalers as a leading indicator.

Putting a Number on AI Revenue

A short question that produced the most quantified AI disclosure the company has made: the comparable figure for Market Intelligence, and the history behind the Energy figure.

Q: "So I think you commented that AI revenues drove like 50 bps of growth in energy. What was that in 2Q? And would you be able to give a comparable metric for MI, i.e., how much AI revenues contributed to growth in that segment?"
— Curtis Nagle, Bank of America

A: "It was about 50 basis points in the energy business. It was about 50 basis points of a tailwind in the market intelligence business and something that we see as a nice tailwind We did not we did not describe it in first quarter and second quarter because it is been building over time, and we are quite pleased with what we have seen."
— Eric Aboaf, CFO

Assessment: the second-quarter figure was not given, so there is no trend yet, only a level. In the prepared remarks the Market Intelligence figure was attributed to Kensho Data & Platforms rather than the division as a whole. The size matches what contract value has been telling investors all along: AI is lifting price and retention on the existing base, and discrete AI products are a small addition on top.

Market Intelligence Contract Value

A one-line housekeeping request that produced the forward indicator that matters most for the division. The answer came with a clear statement that it had moved up.

Q: "Could you provide us the organic constant currency revenue growth I am sorry, ACV growth for MI. The ACV growth for MI on a OCC basis."
— Andrew Steinerman, JPMorgan

A: "And then ACV came in nicely around 6.5 percent. You know, ticked up a bit from the 6% range."
— Eric Aboaf, CFO

Assessment: in July the same metric was quoted at roughly 6% for a second quarter, alongside subscription growth that had been guided to accelerate and did not. A half-point uptick is the first movement in the forward indicator this year. It also supports reading the 7% organic subscription figure as roughly 6.5% underneath.

The Cboe Extension and the Prediction-Market Opportunity

A multi-part question to the Indices CEO covering the derivative opportunity set, the Kalshi announcement and the risk of cannibalising the exchange partners. The answer added the one quantified detail about the Cboe deal that the prepared remarks lacked.

Q: "First off, good to get the CBOE out of the way, but you also, I think, made some prepared remarks about just the opportunity set on the derivative side in general. ... And then just lastly, how do you think about growth in that area and how to not cannibalize existing relationships with, like, CBOE and CME?"
— Alex Kramm, UBS

A: "As Martina and Eric talked about, this really accelerates the economics into SPDJI by 6 years. And it also gives us a nice incremental tailwind of growth in the years following 2027. ... Now we are also going to be innovating in these markets with Cboe and CME, our long standing partners."
— Catherine R. Clay, CEO, S&P Dow Jones Indices

Assessment: in July the same questioner received a 28-word refusal on the same relationship. The answer this time confirms that the extension brings forward economics the company would otherwise have had to wait for, and the cannibalisation answer is the right one: license the benchmark to every venue rather than pick one.

What They're NOT Saying

1. Any financial figure for the review perimeter. No revenue, growth rate or margin for the businesses under review appears in the release, in the 10-Q filed the next day, or on the call, where both the sizing and the margin questions were declined. The 10-Q does not mention the review at all. The only scale given was 15,000 clients and 355,000 users.
2. The cost of separating it. Ratings analysts use Capital IQ, a shared data foundation supports other businesses, and Market Intelligence pays Ratings a $47M quarterly royalty to distribute ratings content. None of the transition costs, stranded costs or changes to that royalty were discussed. After Mobility, the stranded costs were folded into recast comparatives and never stated.
3. Year-end leverage. In July management guided gross leverage of 2.7x to 2.8x at year end, above its 2.0x to 2.5x target until the end of 2027. The release, the 10-Q and the call this quarter contain no leverage figure, although the company confirmed the roughly $2B bond issue that drives it.
4. The size of the retroactive revenue in Market Intelligence. The CFO called it a small benefit and the 10-Q lists it among the drivers of subscription growth. It was not quantified, which means the underlying organic rate cannot be calculated precisely. This is the second consecutive quarter in which a one-time item sat inside the division's subscription growth without a number.
5. Why Ratings expenses grew 7%. The CFO attributed the 130bp margin decline to the timing of expenses. The 10-Q attributes it to merit increases, additional headcount, technology costs and strategic investments, which recur. No analyst asked.
6. The fee rate in Indices. Asset-linked fees rose 24% while average ETF assets rose 32%. The categories are not identical, but the gap points to continuing fee compression per dollar of assets, which was not discussed in a quarter otherwise devoted to the division's records.
7. The AI customer count. In July the company reported more than 500 customers for its LLM-ready APIs and MCP-connected solutions, up more than 70% sequentially. That figure was not updated on the call or in the release, in the same quarter API call growth slowed to about 50% sequentially from more than fivefold.
8. Progress on Enterprise Data Office savings. In July the company said it had achieved nearly 60% of its $100M savings target, with the rest due before the end of 2027. The programme was not mentioned in the prepared remarks, the Q&A, the release or the 10-Q.

Market Reaction

• Pre-print setup: the shares closed at $450.65 on 26 October, up 11.7% over the trailing thirty days from $403.30. Press reports of a potential Capital IQ spinout had circulated since early September. The 52-week closing range was $373.75 to $546.35. Raw comparisons that cross 1 July exclude the Mobility Global shares distributed that day, so the headline figures of down 13.8% year to date and down 9.3% over twelve months overstate a holder's loss. The S&P 500 entered the print up 12.1% year to date.
• Print-day session (27 October, before the open): opened at $470.00, up 4.3%, traded as high as $474.61, up 5.3%, sold off through the day to a low of $446.62, and closed at $448.93, down 0.4% or $1.72.
• Volume: 3.7 million shares against a thirty-day average of 1.9 million, 1.9 times normal.
• Context: the S&P 500 rose 0.6% on the session. The information-services and exchange group fell broadly: Moody's down 2.2%, MSCI down 2.6%, FactSet down 4.3%, Intercontinental Exchange down 2.9%, CME Group down 4.7%, Thomson Reuters down 4.6% and Nasdaq down 0.3%.
• Following sessions: up 0.1% on 28 October, up 2.5% on 29 October and up 1.3% on 30 October, closing at $466.55. That is 3.5% above the pre-print close, over a period in which the S&P 500 fell 1.3%.

The intraday shape reads as a 4.5% sell-off from the open, and several headlines described it that way. Measured close to close it was a 0.4% decline on a day when every close peer we track fell, most of them by 2% to 5%. Relative to that group the shares outperformed by two to four points on the day. We found no single cause for the group's decline in the sources we checked, and we do not attribute one. What is clear is that the stock gave back its opening gain alongside the sector rather than on anything specific in the release.

The first reaction to the review was positive: a 4.3% gap up on a beat, a raise and an announcement the market had already been rehearsing for almost two months. The questions that followed are the ones the Q&A could not answer, about the size, margin and EPS effect of what leaves, and those are reasonable questions to hold a stock back on for a day. Over the following three sessions the shares added 3.9% while the index fell, which is the opposite of the July pattern, when a beat produced a 3.5% decline on a day the group rallied.

The larger picture is that results are now setting the price. Since our July note the shares have risen 9.9% against 2.0% for the S&P 500, through a quarter in which the company beat its own cadence guidance, raised the guide, fixed the Cboe licence and announced the review. Our third bear point argued that results did not reliably set the price for this stock. This quarter they did.

Street Perspective

Debate: Is separating Capital IQ value creation or an admission?

Bull view: the bull case circulating on the Street is that the review completes the transformation the Mobility spin started. The company that remains would be built on credit ratings, indices, commodity price benchmarks and credit-market workflow, all businesses that set standards rather than sell data. Such a company deserves the multiple of the pure benchmark providers rather than a blended one, and sale proceeds would add to an already large buyback.

Bear view: the bear camp contends that the company is selling the asset at the centre of the AI-disruption debate precisely because that debate has lowered its value, so a buyer will price the disruption in. Selling a business at a lower multiple than the company's own and buying back stock at the higher multiple dilutes EPS. The data relationships with Ratings will also leave stranded costs that, on the evidence of Mobility, will never be fully disclosed.

Our take: both are right about different things, and the bull view is right about the thing that matters more. The EPS arithmetic is probably modestly dilutive, and we will not know by how much until the February update at the earliest. But our thesis has always been that the market was valuing the benchmark franchises at a discount because of the data business attached to them. Removing the data business is the most direct way to test that, and we think the result will be a higher multiple on lower but faster-growing earnings.

Debate: Does Ratings have a 2027 problem?

Bull view: the constructive case rests on duration. Roughly $11 trillion of rated debt matures over the next four years, private credit and project finance are growing, investment grade excluding hyperscalers grew about 15% this quarter, and hyperscaler remaining performance obligations point to continued capital spending into 2027. Management has beaten its own Ratings cadence three quarters running.

Bear view: the sceptical case is arithmetic. The first three quarters of 2026 were the three strongest on record for billed issuance, the fourth is guided to a transaction decline, and the company's own research arm has forecast slower market issuance next year. High yield fell 17% in the quarter, and the division's margin fell 130bps on 2% growth. A flat 2027 in the largest division is a meaningful drag on a company whose other divisions are growing mid to high single digits.

Our take: the bears have the near-term argument and the bulls the medium-term one, and the near-term argument is the one that will set 2027 estimates. Our FY27 estimate assumes low-single-digit Ratings growth. A February guide that calls for a decline would sit outside our range, and we have added it as a downgrade trigger below.

Debate: Has Market Intelligence turned, or was the third quarter a catch-up?

Bull view: the optimistic camp points to every forward indicator moving at once: organic subscription growth of 7%, contract value up to about 6.5%, Kensho Data & Platforms organic growth up to 6% from 4%, sales cycles normalised and AI-rights terms standardised with robust pricing.

Bear view: the sceptics note that part of the acceleration is retroactive revenue from renewals that slipped out of the second quarter, that the reported subscription line printed 6% for a third straight quarter, and that the full-year guide implies a fourth quarter closer to 6% than 7%. Two quarters averaged give about 6.5%, a small improvement dressed as a turn.

Our take: the bears' arithmetic is right and the bulls' indicators are right, and the reconciliation is that the drag has stopped worsening without yet reversing. The review changes how much this matters. If Capital IQ and its data businesses are separated, what remains in Market Intelligence is Enterprise Solutions, growing 8% organically, and the retained Ratings distribution. The division that has been the thesis's main bear point would then be a smaller and faster-growing part of the company.

Model Update Needed

Valuation impact: at the 30 October close of $466.55 the shares trade at 26.2x the midpoint of the new FY26 guide and 23.3x our FY27 estimate of $20.00. On the same bases, the price at our July note was 23.8x the new FY26 midpoint and 21.4x our then FY27 estimate, so the stock has re-rated by about two turns in three months. The market capitalisation is about $135B. Our price target rises to $515, which is 25.75x our FY27 estimate and implies 10.4% upside before a dividend yielding about 0.8%. At our prior 24.75x multiple the target would be $495 and the upside 6.1%, which on its own would not support an Outperform rating. The extra turn reflects the Cboe extension, which converts the largest undiscussed risk in the fastest-growing division into a contracted tailwind through 2051. We attribute nothing to the Capital IQ review until its perimeter and economics are known. The $412.99 average price of the accelerated repurchase completed last week is a reasonable marker of where the company itself saw value.

Thesis Scorecard Post-Earnings

The pillars below are the standing thesis from our April initiation and July update, graded against this quarter. One new bear point is added and marked as new.

The July triggers, graded as published. In July we wrote that the two things that would change the rating were "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." On the first, reported subscription growth was 6.4%, a third consecutive 6%, while organic subscription growth, the basis on which management guides the division and on which both prior quarters also printed 6%, was 7%. We grade the trigger as not fired, narrowly. Part of the 7% is retroactive revenue, and contract value, the forward indicator the trigger was meant to capture, moved up for the first time this year. The trigger stays in force unchanged: a fourth-quarter organic subscription rate of 6% or lower, once the retroactive revenue drops out, fires it. The second trigger cannot be tested until February. One trigger is added: a February 2027 outlook that guides Ratings revenue to decline, which sits outside our estimate of low-single-digit growth. If that and the first trigger both fire, we would downgrade. If only one fires, the 2027 Indices tailwind from the Cboe extension is the offset we would weigh against it.

Overall: strengthened. All four bull pillars held, two of three standing bear points eased by a full status level, and a fourth bear point now replaces the second-half concern as the main near-term risk. The thesis has also become more concentrated in a way we think is favourable. If the review proceeds, the company we own becomes almost entirely the benchmark franchise our initiation argued the market was undervaluing, plus a buyback retiring about 5% of the shares a year.

Action: hold the position and add on weakness below the $450 level where the shares sat before the print. This is a smaller and more conditional edge than at initiation: most of the July target has been paid, the 2027 Ratings comparison is a real headwind, and the review's EPS effect is unknown. What keeps the rating at Outperform is a management team that has guided conservatively for three consecutive quarters, a capital-return programme bought well below the current price, and a portfolio decision that, if completed, does the re-rating work the market has so far declined to do.