Four exchange groups reported $5.1 billion across data-labelled categories—but it is not a market-data league table

A mechanical addition of the most obvious “data” categories reported by CME Group, Cboe Global Markets, Nasdaq and Intercontinental Exchange produces an eye-catching result: $5.12 billion for the first nine months of 2026. It is also the wrong answer to a common question. The four numbers do not measure the same commercial activity. One is a relatively narrow market-data line, another combines access and capacity with data fees, a third includes index and workflow products, and the fourth includes fixed-income execution and clearing alongside information services. Calling the sum “market-data revenue” would create a league table that the underlying filings cannot support.[1][2][3][4]

FXBrokers.ai reconstructed each group’s disclosed category and its prior-year comparator. CME reported $705.3 million of market data and information services revenue, up $110.2 million or 18.52%. Cboe reported $349.2 million of access and capacity fees and $271.3 million of market-data fees; together those lines reached $620.5 million, up $73.7 million or 13.48%. Nasdaq’s Data & Listing Services, Index and Workflow & Insights businesses totalled $1.82 billion, up $268 million or 17.27%. ICE reported $1.976 billion for Fixed Income and Data Services, up $165 million or 9.11%. Every within-company growth calculation is valid. The cross-company ranking is not.[1][2][3][4]

CME publishes the narrowest label

CME’s category is the cleanest starting point because the filing separates “market data and information services” from clearing and transaction fees and from other revenue. For the nine months to 30 September 2026, the data line was $705.3 million, while clearing and transaction fees were $4.273 billion. Market data therefore represented 13.25% of CME’s reported $5.324 billion total revenue over the period. This does not mean the remaining business was unrelated to information. It means the accounting presentation places the sale and distribution of data in a distinct revenue line.[1]

CME attributed the increase largely to higher usage of certain products, demand from new customer segments and price increases. It also disclosed that its two largest market-data resellers generated about 25% of the line during the first nine months. That reseller concentration describes distribution, not necessarily concentration of end users: CME said some large institutional users report usage and remit fees directly, and customers could move to another reseller. The disclosure shows why a data business cannot be understood solely by counting terminals or direct subscriptions. The exchange may own the content while vendors meter, package or deliver it.[1]

A valid metric can therefore be constructed within CME: the data line grew 18.52%, faster than the group’s 9.27% total-revenue growth. Its share of total revenue rose from 12.21% in the first nine months of 2025 to 13.25% in 2026. That says something useful about CME’s mix. It does not say CME has less valuable data than a company reporting a larger, broader segment. A segment can be larger because it contains more kinds of service, not because the underlying market information commands a higher price.[1]

Cboe joins connectivity to content

Cboe’s presentation requires a different reconstruction. It reported $349.2 million of access and capacity fees and $271.3 million of market-data fees for the first nine months. Access and capacity includes the physical and logical connections, ports and related infrastructure that participants need to reach markets. Market-data fees concern information products. Both are recurring revenues generated around exchange participation, but one sells a route into the venue while the other sells information from it. Adding them is analytically defensible only when the label says exactly what has been added.[2]

The combined $620.5 million was 13.48% above the prior-year $546.8 million. Within that total, access and capacity grew 15.29%, from $302.9 million to $349.2 million, while market data grew 11.23%, from $243.9 million to $271.3 million. That decomposition matters. A broker, market maker or data vendor can face rising costs because it needs more connections, more capacity, more data products or all three. A single “data” growth figure would hide which part of the market-access stack is becoming more expensive.[2]

Cboe’s Data Vantage reporting architecture adds another complication. Some market-data fees also appear in Cash and Spot Markets or Derivatives Markets, while access and capacity sits within Data Vantage. The business-segment label and the revenue-category label are not identical dimensions. Analysts who pull only the Data Vantage segment can miss market-data revenue recorded elsewhere; analysts who pull every line containing “data” can double count. The filing is internally reconcilable, but only when the reader preserves the distinction between what was sold and the organisational segment that sold it.[2]

Nasdaq’s total includes intellectual property and workflow

Nasdaq’s Capital Access Platforms businesses are broader again. Data & Listing Services generated $655 million in the first nine months of 2026. Index generated $778 million. Workflow & Insights generated $387 million. The three businesses totalled $1.82 billion, compared with $1.552 billion a year earlier. The 17.27% growth is a reliable calculation for that defined group. Describing all $1.82 billion as exchange market data would not be. Listing services, index licensing, anti-financial-crime technology, investor-relations workflows and other analytics do not become the same product merely because they turn information into recurring revenue.[3]

Index revenue illustrates the distinction. An index provider creates and licenses intellectual property used by exchange-traded products, derivatives, structured products and asset managers. The economic driver can be assets under management, derivative volumes, licensing arrangements or subscriptions. A real-time quotation feed is sold and consumed differently. Both are data-intensive and both can be mission-critical, but comparing a dollar of index licensing with a dollar of quote-feed revenue does not reveal which exchange charges more for market data.[3]

Workflow products extend the perimeter further. Nasdaq’s software can sit inside surveillance, compliance, governance, investor relations or trading workflows. Revenue may be subscription-based and information-rich without originating from an exchange feed. The commercial strategy is important precisely because it reduces dependence on transaction volume. Yet that strategic advantage makes a simplistic peer comparison worse: the broader the information and technology portfolio, the larger the apparent “data” number becomes even when no additional market-data licence has been sold.[3]

ICE’s segment contains trading as well as data

ICE reported $1.976 billion of Fixed Income and Data Services revenue, the largest of the four headline categories used in this reconstruction. The segment provides fixed-income pricing, reference data, indices, analytics, execution services, credit-default-swap clearing and multi-asset data-delivery technology. Its name accurately reflects that breadth. It does not invite the entire amount to be treated as pure information revenue. A fixed-income execution fee and a reference-data subscription solve different customer problems and can respond to different market conditions.[4]

ICE supplies a narrower clue elsewhere in the filing. Revenue transferred over time included $856 million of data services in the Exchanges segment and $1.583 billion of data services in Fixed Income and Data Services. Together, those disclosed data-services components reached $2.439 billion, up 10.16% from $2.214 billion. This total is more data-specific than the $1.976 billion segment figure, yet it crosses two reportable segments and is defined by timing of revenue recognition. It still should not be assumed to match CME’s line item or Cboe’s combination exactly.[4]

The apparent paradox—ICE’s data-services components exceeding the revenue of its Fixed Income and Data Services segment—has a straightforward explanation. Some data revenue belongs to ICE’s Exchanges segment. This is not duplication when the filing is read as intended: one table presents segment revenue, another identifies over-time revenue components inside multiple segments. It becomes duplication only if an analyst adds the segment total and the component total together. The example shows how easily a cross-company dataset can become internally inconsistent when labels are collected without their table context.[4]

The $5.12 billion sum is a stress test, not a market size

The mechanical $5.12 billion result adds CME’s $705.3 million, Cboe’s $620.5 million, Nasdaq’s $1.82 billion and ICE’s $1.976 billion. It is useful as a stress test because it exposes the taxonomy problem. The total contains market-data subscriptions, connectivity and capacity, listings, index licensing, workflow software, fixed-income execution and other services. It omits ICE exchange-segment data services from the headline segment number, while including activities that peers place in other categories. It therefore cannot measure the size of the exchange-data market, customer spending on feeds or the price burden facing brokers.[1][2][3][4]

Nor can the growth rates be ranked without qualification. CME’s 18.52% increase occurred in a narrower category than Nasdaq’s 17.27% increase. ICE’s 9.11% segment growth may conceal faster or slower movement in individual data products. Cboe’s 13.48% combined result partly reflects connectivity. The correct comparison is first longitudinal—how each company’s consistently defined line changed—and only then functional, using subcategories that genuinely overlap. A defensible peer table should retain the issuer’s original label, list included activities, show whether revenue is gross or net, and flag acquisitions or reclassifications.[1][2][3][4]

Why brokers and investors should care

For brokers, the taxonomy maps onto real cost centres. Exchange connections and capacity determine how orders and market data reach infrastructure. Data licences govern what can be displayed, stored, redistributed or used in algorithms. Index licences can affect the economics of products. Workflow and surveillance tools sit in operating expenses. A broker describing all of these as “data costs” may be commercially understandable, but it obscures which supplier, contract and usage rule drives the bill. That makes it harder to compare execution venues or explain a new customer fee.[1][2][3][4]

For investors, the distinction changes the quality of recurring revenue. A connection fee may be sticky because a participant cannot trade efficiently without it. A real-time feed may be exposed to changes in user counts, distribution arrangements or regulation. An index licence can scale with assets or volumes. Workflow software may carry conventional subscription economics and switching costs. Two dollars reported under information-oriented labels can therefore have different growth drivers, margins, customer concentration and regulatory sensitivity.[1][2][3][4]

For regulators and market users, comparability matters because exchange-data pricing is contested. A large segment total does not establish that core market-data charges are excessive, just as a small line does not establish they are cheap. Answering a pricing question requires the relevant product, user class, distribution route, licence terms and cost allocation. Consolidated financial statements are valuable evidence of scale and growth, but they are not a tariff schedule and cannot substitute for contract-level analysis.[1][2][3][4]

A better comparison framework

A high-quality exchange comparison should use at least four layers. First, preserve each issuer’s audited or filed category and calculate its year-on-year change. Second, split recurring access, connectivity, market data, index, analytics and workflow products where disclosures permit. Third, identify whether revenue is recognised gross or net and whether the segment includes transaction-based activity. Fourth, connect the revenue to operational metrics such as subscribers, devices, connections, assets linked to indices or annualised subscription value only when the company defines those metrics consistently.[1][2][3][4]

On that basis, this investigation establishes a narrower but more reliable conclusion. All four groups reported material growth in information- or access-oriented revenue during the first nine months of 2026. CME’s disclosed data line grew fastest in percentage terms among the four headline categories used here; Nasdaq generated the largest absolute increase; ICE reported the largest broad segment; and Cboe’s combined access-and-data lines rose by $73.7 million. Those statements can coexist because they answer different questions. None identifies a universal market-data leader.[1][2][3][4]

The unresolved issue is not a missing number but a missing common definition. Until exchange groups report a standardised breakdown—or an analyst rebuilds one product by product—the safest course is to resist the attractive total. The $5.12 billion figure demonstrates the scale of businesses built around market access and information. It does not tell customers what they paid for a feed, what they paid to connect, or which company has the largest comparable data franchise. The accounting labels are evidence. They are not interchangeable facts.[1][2][3][4]

Report an error or submit an update