Two-thirds of Cboe’s Q3 revenue increase was a zero-margin regulatory pass-through as US equity fees fell

Cboe Global Markets reported that third-quarter revenue rose by $217.1 million to $1.359 billion, but $143.9 million—66.3% of the increase—came from a regulatory fee line with an equal cost. The fee was collected from customers and remitted under the US Securities and Exchange Commission’s Section 31 framework. It increased the top line and cost of revenue by the same amount, so it created no net revenue. The result is an accounting denominator problem: the headline growth rate describes cash passing through the exchange group as well as the economics Cboe retained.[1]

After subtracting liquidity payments, routing and clearing charges, Section 31 fees and other transaction-based expenses, Cboe’s net revenue rose from $605.5 million to $712.2 million. That 17.6% increase was still substantial, but materially below the gross revenue comparison. The distinction is important for brokers, venue customers and investors because transaction businesses often report gross fees that include sums economically owed to liquidity providers, regulators or other market infrastructure. A larger gross line does not automatically mean a larger spread or a richer exchange franchise.[1]

The filing makes the zero-margin nature of the regulatory item explicit. Cboe recorded $143.9 million of regulatory fee revenue and $143.9 million of regulatory fee expense in the quarter, compared with no corresponding amount in the prior-year quarter. Because both sides entered the income statement together, the addition explains most of the difference between the apparent 19.0% gross-revenue growth and the 17.6% increase in net revenue. It should not be described as new regulatory profit, a customer penalty retained by Cboe or evidence that trading economics improved by the full headline amount.[1]

A pass-through explained two thirds of the top-line increase

Dividing the new $143.9 million Section 31 line by the $217.1 million increase in gross revenue produces 66.3%. That is not a claim that two thirds of Cboe’s business was regulatory activity. It is a bridge explaining the year-on-year change in the chosen revenue denominator. The regulatory amount also exceeded the entire $106.7 million increase in net revenue. A press headline based only on gross revenue could therefore imply a scale of operating expansion that the retained-revenue comparison does not support.[1]

The remaining change was not one simple organic figure. Cboe’s results span options, North American and European equities, futures, global foreign exchange, data and access services. Different units have different fee schedules, market-share movements and expense treatment. The filing’s consolidated totals are useful for identifying the pass-through effect, but they cannot by themselves establish whether any customer paid more, whether spreads improved or whether a particular venue gained economically valuable flow. Those questions require the segment and product tables.[1]

This accounting treatment is legitimate and visible. Exchanges are often required to collect regulatory charges and can also pay substantial rebates to market makers or routing partners. The investigative finding is not that Cboe concealed an expense. It is that the gross headline blends economically different flows, and a newly appearing zero-margin item accounts for most of the reported dollar increase. The right comparison depends on the question: gross revenue measures billings entering the system; net revenue better approximates what remains after transaction-based pass-throughs.[1]

Cash-equities billings rose even as matched shares declined

The cash and spot markets segment supplies a second reason not to equate billings with trading strength. Its gross revenue rose from $415.0 million to $511.3 million, an increase of $96.3 million. Yet Cboe also reported that US equities transaction and clearing fees decreased because matched shares were 12% lower. The segment includes other geographies, products and fee lines, so the higher gross total does not contradict the decline. It shows that one consolidated revenue number can move in the opposite direction from a core activity measure.[1]

For a broker assessing venue economics, the relevant chain is more granular: eligible order flow, displayed or non-displayed execution, rebates, routing charges, clearing fees, data charges and regulatory assessments. A decline in matched shares can coexist with stronger revenue if price schedules, product mix, market data or foreign businesses change. Conversely, a high-volume quarter can produce limited incremental net revenue when rebates and other variable costs move with activity. Cboe’s filing supports that mechanism but does not provide a customer-by-customer cost reconstruction.[1]

The $96.3 million gross increase in cash and spot markets should therefore not be assigned to US equity executions without a proper segment bridge. Nor should the 12% decline in matched shares be treated as a 12% fall in the entire segment. The two measures have different perimeters. This is precisely the kind of denominator mismatch that produces misleading exchange comparisons: one side may include regulatory revenue, data and overseas venues while the other counts only matched US equity shares.[1]

What the evidence changes

The finding changes the interpretation of Cboe’s quarterly growth, not the validity of its accounts. On a gross basis, revenue increased 19.0%. On a net basis, after transaction-based costs, it increased 17.6%. Both are true, but only the first contains the new $143.9 million regulatory amount on both sides of the income statement. Analysts comparing Cboe with exchanges that present similar costs net, or comparing one quarter before and after the Section 31 line appeared, need to normalise the presentation before ranking growth.[1]

The same discipline applies to margins. Using gross revenue as the denominator makes cost of revenue appear larger and can mechanically change a gross-margin calculation without any change in underlying fee capture. Using net revenue removes the identified pass-throughs but still does not equal operating profit because compensation, technology, depreciation and other expenses remain below that line. Neither measure should be relabelled EBITDA, cash earnings or take rate. Each occupies a different level of the income statement.[1]

A stronger test of economic performance would track net revenue per matched contract or share within like-for-like products, alongside market share, pricing and rebate changes. It would also separate recurring data and access revenue from transaction income. Cboe publishes many of those components, but the evidence used here does not establish a complete cross-venue take-rate comparison. The conclusion could be refined if the company later restates the regulatory presentation, supplies a new historical bridge or changes whether Section 31 collections are shown gross.[1]

There is also no basis here to infer that customers were overcharged or that Cboe benefited improperly from the regulatory fee. The filing shows an equal expense and says the assessment is regulatory. The public-interest issue is clarity: market participants should know which revenue growth reflects retained economics and which reflects money collected for someone else. That distinction becomes more important as exchange groups expand across jurisdictions whose transaction taxes, regulatory levies and rebate structures are not uniform.[1]

A further control issue is the treatment of regulatory collections in forward estimates. If an analyst annualises the new $143.9 million quarterly line as though it were commercial revenue, the model would add roughly $576 million to annual sales without adding net revenue. That error could inflate enterprise-value-to-sales comparisons and make future growth appear dependent on a fee Cboe does not retain. The corrective is to forecast Section 31 revenue and expense together, then value the business on net revenue or another consistently normalised denominator.[1]

The same normalisation should be applied to historical charts. A series that begins showing the regulatory line in 2026 but compares it with periods containing no corresponding gross amount can create a structural break that looks like organic acceleration. Readers need either a recast history or a clear bridge. Cboe’s filing supplies the current-quarter bridge; it does not supply a full multi-year restatement.[1]

Report an error or submit an update