Credit ratings, benchmarks and financial data. SPGI / MCO. Run 23 September 2026. Built from both companies' FY2023 to FY2025 10-Ks, Q2 2026 10-Qs, Q4 2025 and Q2 2026 earnings-call transcripts and 2026 8-Ks (the sources/ folders), plus the SEC Office of Credit Ratings 2025 staff report. Events swept through 23 September 2026; most recent events checked: S&P Global closed the sale of its upstream energy software to SLB (1 September 2026); Moody's added a director (8-K, 12 August 2026). Figures in US$ millions as reported; no FX conversion needed. S&P Global is shown ex Mobility (spun off 1 July 2026) unless stated. Not a valuation and not a recommendation.
Three cells carry the next five years. First, ratings in both regions, where growth comes from refinancing walls and new borrowers: S&P expects about $11 trillion of rated debt to refinance over the next four and a half years, and AI infrastructure is adding hyperscaler, data-centre and power issuance (S&P guides $250 to 300bn of hyperscaler issuance for 2026) (SPGI Q2 2026 call). Both agencies ride that equally, but Moody's gets more of it per dollar of revenue because 67% of its ratings revenue is transaction-based against 52% at S&P (FY2025 10-Ks). Second, the data and workflow cell, where Moody's Analytics grows ARR 8 to 9% on KYC (+13%), lending and insurance cloud migration, while S&P's Market Intelligence platforms (Capital IQ and similar) grow low single digits and only its Kensho data feeds grow high single to low double digits (Q2 2026 calls, both). Third, equity indices, where S&P alone collects asset-linked fees on $6.35tn of ETF assets that grow with markets and the shift to passive (SPGI Q2 2026 call). Weighting each cell's run-rate by its revenue share gives roughly 7.7% a year for Moody's against 7.2% for S&P ex Mobility (inferred), a gap inside the noise; S&P's 2026 organic guide is 6 to 8% (SPGI Q2 2026 call).
The consolidated statements point the wrong way: Moody's shows a 72.8% gross-margin proxy against 68.8% at S&P (3-year averages, 10-Ks). The segment data reverses it. On a like basis S&P Ratings earned 63.8% in FY2025 against 61.6% at Moody's MIS, and S&P led in every year since FY2023, though the lead has narrowed from 3.9 to 2.2 points (inferred basis: MIS adjusted operating income less its D&A). Market Intelligence earns about 32% before acquisition amortization against about 28% for Moody's Analytics on a comparable basis (MCO figure inferred). Mix then adds Indices at 69% and Energy at 41% GAAP. Cash confirms it: free cash flow was 35.6% of revenue at S&P and 33.4% at Moody's in FY2025, partly because Moody's capitalizes more software (capex 4.2% of revenue against 1.3%) (FY2025 10-Ks). The tension: Moody's grows a little faster, S&P keeps a little more of each dollar. S&P is the fortress with a slower data arm; Moody's is the more levered ratings compounder with a faster but thinner analytics arm.
Moody's single exposure is the issuance cycle: ratings are 53% of revenue and about 70% of adjusted operating income, and two-thirds of ratings revenue is transaction fees (FY2025 10-K). In 2022 MIS revenue fell 28% and adjusted operating income fell 40%, about 89 cents of profit per lost dollar (FY2023 10-K). S&P's ratings behaved the same way (revenue -26%, profit -36%), but ratings are now only 48% of S&P's segment profit and 52% transaction. Replaying 2022 on FY2025 mix costs Moody's about 27% of adjusted operating income and S&P about 16% of segment profit (inferred). S&P's own weak points are different: a bear market hits Indices asset-linked fees ($1,206m) almost one for one, and its balance sheet is now more stretched, with gross leverage guided to 2.8x at end-2026, above its 2.0 to 2.5x target, to fund more than $7bn of 2026 buybacks (SPGI Q2 2026 call). Moody's carried $6.9bn of debt against $3.2bn of equity at 30 June 2026 (MCO Q2 2026 10-Q). In an issuance downturn Moody's breaks first; in an equity bear market combined with tighter credit, S&P's leverage leaves less room.
These six cells are 100% of both companies' continuing revenue (S&P ex Mobility); every score is argued in the tabs below.
| Cell (product x region) | S&P Global | Moody's | Why (one clause, sourced) |
|---|---|---|---|
| Credit ratings, US | 5 | 5 | Co-dominant: $2,711m vs $2,623m, both +22% a year FY23 to FY25; the big three held 93.4% of all NRSRO ratings at end-2024 (10-Ks; SEC OCR 2025 report) |
| Data, research & workflow, US | 3 | 4 | MI grows with the market (+6% a year) against MA +8% and ARR +9% with 95% retention (10-Ks; Q2 2026 calls) |
| Data, research & workflow, Intl | 3 | 4 | MA is larger outside the US ($2,051m vs $1,992m) and faster (+9% vs +6%), on Orbis entity data and KYC (10-Ks) |
| Credit ratings, Intl | 4 | 4 | Both +16% a year; S&P larger ($2,013m vs $1,496m) partly via Crisil; both buying African agencies (10-Ks; Q2 2026 calls) |
| Equity indices, US and Intl | 5 | 0 | $1,850m at 69% margin, 13 straight record quarters, first $1tn ETF tracks the S&P 500; Moody's has no index business (SPGI 10-K; Q2 2026 call) |
| Energy & commodity benchmarks, global | 4 | 0 | Platts assessments settle exchange contracts; 2026 slowed to +3% on Iran-related renewals and sanctions; Moody's absent (SPGI 10-K; Q2 2026 call) |
Scores are anchored to the exhibits in the three tabs; a score with no exhibit behind it does not ship. Scores are per cell and are not summed; the three lenses get their verdicts in the Three Answers above, not here. Mobility (S&P, $1,747m in FY2025) was spun off on 1 July 2026 and is excluded.
Normalization: both companies report two regions at segment level (US and non-US), so cells use that grain. S&P's Market Intelligence and Moody's Analytics are merged into one "data, research and workflow" label; Moody's splits it further into Decision Solutions, Research and Insights, and Data and Information, which S&P does not. S&P's regional splits are of segment revenue including intersegment royalties (Ratings $175m); Moody's are of external revenue. Growth is FY2023 to FY2025 CAGR on reported revenue, which includes small acquisitions and divestitures on both sides.
The table shows the whole duel in miniature. In ratings the two grow at the same rate in both regions, so the shared tide explains nearly all of both companies' recent acceleration. In data, Moody's is the faster grower in both regions. The remaining S&P cells, Indices and Energy, have no Moody's counterpart and together are 30% of S&P's revenue ex Mobility.
| Cell | S&P rev (% of total) | S&P CAGR | Moody's rev (% of total) | Moody's CAGR | Margin signal / leader |
|---|---|---|---|---|---|
| Ratings, US | 2,711 (20%) | 21.9% | 2,623 (34%) | 22.4% | Segment margins 64% vs 62% (like basis); tie on growth |
| Data & workflow, US | 2,924 (21%) | 6.0% | 1,548 (20%) | 8.3% | Moody's on growth, S&P on margin (32% vs ~28%) |
| Data & workflow, Intl | 1,992 (14%) | 5.9% | 2,051 (27%) | 8.7% | Moody's: KYC and Orbis entity data |
| Ratings, Intl | 2,013 (15%) | 15.5% | 1,496 (19%) | 16.1% | Tie on growth; S&P larger |
| Equity indices | 1,850 (13%) | 14.8% | 0 | ND | S&P only, 69% margin |
| Energy & commodities | 2,299 (17%) | 8.7% | 0 | ND | S&P only, 41% GAAP margin |
S&P % of total uses FY2025 segment revenue ex Mobility ($13,789m before $200m eliminations); Moody's uses external revenue ($7,718m). Sources: FY2023 and FY2025 10-Ks, both.
Ratings. Growth is issuance volume times price. Both disclose the same drivers for FY2025: tight spreads, refinancing, AI-related investment-grade issuance, CLOs and private credit (S&P private-market ratings +60% in Q2 2026; Moody's private-credit transactions +40%) (Q2 2026 calls). S&P's billed issuance was $4.3tn in FY2025 (+11%); Moody's rated more than $2tn in each of Q1 and Q2 2026. Price shows up in the annuity line at both: Moody's recurring revenue rose on "annual price increases and higher monitored credits", S&P's on "improved contract terms" (FY2025 10-Ks). Ratings growth is essentially organic at both.
Data and workflow. Moody's MA grows through embedding in regulated workflows: KYC ARR +15% in FY2025, insurance cat modelling migrating to its cloud platform with less than half of customers moved, and lending suites repriced on renewal (MCO FY2025 10-K; Q2 2026 call). S&P's MI is being split into a fast Kensho data business (about a third of the Kensho Data and Platforms line, growing high single to low double digits) and slower platforms (low single digit) to be consolidated for margin (SPGI Q2 2026 call). Both are partly bought: S&P added Visible Alpha (2024) and With Intelligence (2025); Moody's spent $227m on acquisitions in FY2025 and sold its Learning Solutions business in Q4 2025 (MCO FY2025 10-K).
Indices. Asset-linked fees rose 22% in FY2024 and 15% in FY2025 as average ETF assets grew 33% and 24% (SPGI FY2025 10-K). Market levels do most of the work; net inflows of more than $600bn over the last year and asset-manager switches to S&P add the structural part (SPGI Q2 2026 call).
Both sell direct. In ratings the issuer pays and has no substitute for the big-agency ratings that index and mandate rules require, so both exercise annual price increases. The difference is how price is packaged: 48% of S&P Ratings revenue is non-transaction (surveillance, annual relationship fees, Crisil) against 33% recurring at Moody's (FY2025 10-Ks). Moody's also notes that frequent issuers on pricing programs carry lower average revenue yields, so higher volume from banks and data-centre issuers did not lift its 2026 revenue guide (MCO Q2 2026 call). In indices S&P prices by basis points on assets, a meter that runs with markets rather than with negotiations.
Neither company has physical supply risk; the inputs are analysts, engineers, data and cloud capacity. Compensation is the largest cost at Moody's ($2,574m of $3,776m operating and SG&A) (MCO FY2025 10-K); S&P does not split its cost base this way at group level. S&P's one partner dependency is CME Group, which co-owns S&P Dow Jones Indices and holds a put on at least 20% of its stake ($4.9bn redeemable non-controlling interest); Moody's depends on AWS and Microsoft as distribution partners for its AI and cloud products (SPGI FY2025 10-K; MCO Q2 2026 call).
Ratings: an issuance drought hits both, but Moody's more (see Answer 3). Data: AI tools that let clients combine feeds from several vendors threaten stand-alone platforms first, which is S&P's slow-growing Capital IQ-type estate more than Moody's workflow-embedded KYC and insurance software; an analyst raised exactly this on S&P's Q2 2026 call. Indices: a bear market cuts asset-linked fees directly, and switching by large asset managers works in both directions. Energy: geopolitics; S&P chose to be flexible on price with customers affected by the Iran conflict in 2026 (SPGI Q2 2026 call).
Ratings: named in bond-index and mandate rules alongside Moody's and Fitch; the SEC staff report identifies these guidelines, index rules and the QIB certification requirement as barriers to entry. Durability High. Indices: the S&P 500 underlies the first $1tn ETF, four of the five ETFs the US Treasury selected for Thrift accounts and SPX derivatives; replacing it means rebuilding a liquid market. Durability High. Platts: price assessments settle exchange contracts. Durability High, with 2026 geopolitical noise.
Ratings: the same mandate-driven moat, with an identical 2022 to 2025 path. Durability High. Analytics: Orbis ownership data on more than 630m entities, cat models for insurers, credit models embedded in bank lending workflows, 95% retention. Durability Medium: valuable and sticky, but money and time can buy a competitor; margins in the low 30s are the evidence the market does not treat it as a benchmark.
S&P: no customer above 10% of revenue; in ratings 48% of revenue is non-transaction; in indices it is paid by asset managers on their assets and by exchanges on volume, and management reports asset managers switching to S&P benchmarks in Q2 2026 (SPGI FY2025 10-K; Q2 2026 call). Moody's: no customer above 10% of receivables; ratings recurring share 33%; MA 96% to 99% recurring with 95% retention and upsell on platform migrations, but lower-yield pricing programs for frequent issuers dilute ratings price per dollar issued (MCO FY2025 10-K; Q2 2026 call). Both raise ratings prices annually; neither discloses the price contribution.
S&P: people and data; low capital intensity (capex 1.3% of revenue); CME partnership in indices. Moody's: people and cloud; capex 4.2% of revenue as it capitalizes software for MA's cloud shift, and it is sunsetting on-premise insurance models, which makes AWS a critical platform (FY2025 10-Ks; MCO Q2 2026 call). Pass-through: both held or raised margins through the 2022 to 2023 wage inflation once issuance recovered; neither shows input-cost stress.
| Cell | Route control | Pocket price | Continuity | Outcome (share / margin) | Confirming KPI |
|---|---|---|---|---|---|
| Ratings, US | Both direct to issuer; mandate-protected | S&P above (more annuity, higher margin) | Both high | Share flat both; margin up both, S&P lead narrowing | Ratings like-basis margin gap stays above 1.5pts in FY2026 |
| Data & workflow, US | Both direct; Moody's also via AWS and Microsoft | S&P above on margin, Moody's above on growth | Both high; Moody's 95% retention disclosed | Moody's share up, S&P margin up | MA ARR growth 2+ pts above MI organic growth |
| Data & workflow, Intl | Both direct | Parity (inferred) | Both high | Moody's share up | MA non-US revenue growth vs MI international growth |
The tension: Moody's grows fastest exactly where its pricing power is weakest (data at about 28 to 34% margins), while S&P's pricing power is concentrated in cells Moody's does not contest.
1. The index franchise (Major). S&P: $1,850m, 69% margin; Moody's: none. Not closable: it would take decades and a liquid derivatives market. 2. Ratings annuity mix (Moderate). S&P 48% non-transaction vs Moody's 33% recurring. Partly closable over 3 to 5 years through relationship pricing and first-time mandates (Moody's expects 750 to 850 in 2026), but it is also a legacy of how each agency priced for decades. 3. Data arm economics (Moderate, and moving toward Moody's). MI is larger and more profitable; MA grows faster and targets mid to high 30s margins by end-2027 (MCO Q2 2026 call). If MA gets there while holding 9% ARR growth, this cause flips to Moody's within about two years.
Both present costs by function under US GAAP, with depreciation and amortization on separate lines, so the mapping is clean: "operating-related expenses" (S&P) and "operating" (Moody's) are used as the cost-of-revenue proxy; "selling and general" and "SG&A" as SG&A. Neither discloses R&D, so that row is ND for both. S&P's reported EBIT carries $1.0 to 1.1bn a year of IHS Markit-related intangible amortization and disposal gains or losses; Moody's carries $198 to 215m of acquired amortization plus restructuring. A second EBIT line removes acquired amortization (and S&P's disposal gains) from both. S&P figures include Mobility (not restated); fiscal years both end 31 December.
| % of sales, 3y avg FY23 to FY25 | S&P Global | Moody's | Gap | What drives it |
|---|---|---|---|---|
| COGS (operating expense proxy) | 31.2 | 27.2 | +4.0 | Mix (Energy, Mobility, MI content costs) plus more software capitalized at Moody's |
| R&D | ND | ND | ND | Not disclosed by either |
| SG&A | 23.4 | 25.1 | -1.7 | S&P's larger scale spreads corporate cost |
| Gross margin proxy | 68.8 | 72.8 | -4.0 | Moody's mix is 53% ratings |
| EBIT margin, reported | 37.9 | 40.0 | -2.1 | IHS Markit amortization at S&P |
| EBIT ex acquired amortization | 45.0 | 43.0 | +2.0 | Higher segment margins and Indices at S&P |
| By year | S&P FY23 | FY24 | FY25 | Moody's FY23 | FY24 | FY25 |
|---|---|---|---|---|---|---|
| Revenue ($m) | 12,497 | 14,208 | 15,336 | 5,916 | 7,088 | 7,718 |
| COGS proxy % | 33.1 | 30.7 | 29.8 | 28.5 | 27.4 | 25.6 |
| SG&A % | 25.3 | 22.5 | 22.3 | 27.6 | 24.5 | 23.4 |
| Gross margin proxy % | 66.9 | 69.3 | 70.2 | 71.5 | 72.6 | 74.4 |
| EBIT margin % | 32.2 | 39.3 | 42.2 | 36.1 | 40.6 | 43.4 |
| EBIT ex acquired amort. % | 41.1 | 46.4 | 47.4 | 39.5 | 43.4 | 46.2 |
| Capex % | 1.1 | 0.9 | 1.3 | 4.6 | 4.5 | 4.2 |
Sources: SPGI FY2025 10-K (income statement, cash flow statement); MCO FY2025 10-K (statement of operations, cash flow statement, Note on acquired intangible amortization).
The most persistent gap is the operating-expense ratio: Moody's ran 3 to 5 points lower in each of the three years. It is not a cost-discipline story. It comes from mix, since Moody's is 53% ratings while S&P's cost base carried Energy, Mobility and Market Intelligence content operations, and from accounting, since Moody's capitalizes about three times as much of its revenue as capex, moving software cost from operating expense into amortization (inferred). At segment level the gap reverses: S&P Ratings out-earns MIS by 2 to 4 points every year and MI out-earns MA by about 4 points on a comparable basis. That confirms the power map: the company with more benchmark pricing power shows it where the businesses overlap, and in free cash flow (35.6% vs 33.4%). Moody's lead in consolidated gross margin is a mix effect a reader should not mistake for a cheaper engine.
The question is revenue at risk against revenue gained, bucket by bucket, not the headline AI growth rates. AI hurts seat-based screens and analyst research, because an AI front end can pull the same answer from several vendors. It helps data sold as a licence to feed a customer's own models. It barely touches software built into regulated processes. Sorting each data arm that way gives a first-cut exposure map.
| Bucket | S&P Market Intelligence ($4,916m, FY25) | Moody's Analytics ($3,599m, FY25) |
|---|---|---|
| Exposed: seats and research | Platforms (Capital IQ, RatingsDirect, Visible Alpha, With Intelligence): growing low single digits; size ND, the largest part of MI (inferred) | Research & Insights: $995m, 28% of MA, recurring revenue +6% (Q2 2026) |
| Beneficiary: data licences | Kensho Data (feeds, AI-ready APIs, RatingsXpress): about half the size of Platforms, high single to low double digits | Data & Information (Orbis): $912m, recurring +7 to 8% |
| Protected: embedded workflow | Enterprise Solutions (lending suite, ClearPar, WSO): +10% organic in Q2 2026; size ND | Decision Solutions (banking, insurance, KYC): $1,692m, 47% of MA, ARR +10% |
Sources: SPGI Q2 2026 call (new MI two-line structure, bucket growth rates); MCO FY2025 10-K (revenue by line of business) and Q2 2026 call (ARR by line). S&P has not yet published dollar revenue for these buckets; its recast reporting from Q3 2026 should.
Net AI effect on growth is roughly the AI customers' share of recurring revenue times their extra growth, minus the exposed bucket's share times its slowdown. Neither company discloses the AI-customer share, so neither net number can be computed yet; the pieces below are what exists.
| Lever | S&P Market Intelligence | Moody's Analytics |
|---|---|---|
| AI-customer growth | ACV growth "60% faster" for AI customers against about 6% overall, implying about 9 to 10% (inferred); cohort share ND | Customers with a standalone GenAI or agentic product grow "roughly twice" the rest and renew at 97%; cohort share ND |
| Price realized from AI | ND in the filings and calls reviewed | About two-thirds of eligible lending renewals moved to the AI-enabled Lending Suite in 2025 at an average uplift of about 67%; CreditLens growth near 20% |
| Retention | ND ("net retention rates are continuing to tick up") | 95% trailing 12 months |
| Cost savings | $100m target in the enterprise data organisation, about 20% of its cost base, about 60% delivered; up to about 2 points of MI margin if all of it lands in MI (inferred) | Not quantified; MA margin guided to 34 to 35% in 2026 and mid to high 30s by end-2027 |
| Usage (leading, not revenue) | More than 500 clients on Kensho LLM-ready APIs and MCP connectors, +70% quarter on quarter; API call volume 5x Q1 | More than 100 MCP and Smart API connections in use or trial; 170+ exclusive or semi-exclusive data sources |
Sources: SPGI Q2 2026 call; MCO Q4 2025 call (GenAI cohort, Lending Suite), Q2 2026 call (retention, connections, margin guide), 8 June 2026 investor update (data sources).
Usage counts only matter if they show up in the data-licence bucket. If S&P's API calls keep compounding while Kensho Data growth stays in the high single digits and Platforms slows, the AI traffic is substituting for seats. The two disclosures that would turn this lens from qualitative to quantitative are each company's AI-cohort share of recurring revenue and S&P's MI net retention.
| Metric | Threshold | By when | If it hits, it favors | Where published |
|---|---|---|---|---|
| MA ARR growth minus MI organic revenue growth | Gap of 2 pts or more, four quarters running | Q2 2027 | Moody's | Quarterly 10-Q / earnings calls |
| MA adjusted operating margin | 36% or higher | FY2027 | Moody's | MCO 10-K segment MD&A |
| Ratings like-basis margin gap (S&P Ratings minus MIS adj. OI less D&A) | Stays above 1.5 pts (2.2 pts in FY2025) | FY2026 | S&P Global | Both 10-Ks, segment notes |
| S&P asset-linked fees / average ETF AUM | Holds at 2.4bps or above | FY2027 | S&P Global | SPGI 10-K, Indices MD&A |
| S&P MI Platforms growth minus Kensho Data growth (first disclosed with Q3 2026 recast) | Gap wider than -8 pts for two quarters | Q2 2027 | Moody's | SPGI earnings calls / 10-Q |
| AI-cohort share of recurring revenue, both companies | Disclosed and above 25% with growth premium intact | FY2027 | Whichever discloses the higher net effect | Earnings calls, investor days |
| MIS recurring revenue share of MIS external revenue | Rises above 38% (33% in FY2025) | FY2028 | Moody's | MCO 10-K revenue note |