Intercontinental Exchange, Inc. ICE

Key data

The business

ICE is three toll booths on the flow of capital. The Exchanges segment runs the energy and interest-rate contract markets institutions use to hedge, from Brent crude to global rate futures, plus the New York Stock Exchange and the clearing houses that stand behind the contracts. Fixed Income and Data Services prices nearly every bond in the market, sells the reference data and indices that asset managers build on, and clears credit default swaps. Mortgage Technology, built from the Ellie Mae and Black Knight purchases, is the workflow software a lender pushes a home loan through, from application to servicing.

The moat is that a regulated venue with the deepest pool of buyers and sellers is nearly impossible to unseat, and that once a bank runs its bond pricing or its loan pipeline on ICE's rails, ripping them out is expensive and risky. The concrete tell: the same Encompass origination system sits inside a large share of American mortgages, whether rates are high or low.

A read four weeks ago flagged an unexplained 4.3% pop and a missing MarketAxess picture; the August 21 financing 8-K settles the first, ICE is funding the MarketAxess purchase with new revolving commitments, while the deal's pro forma economics stay undisclosed.

The numbers

Revenue grinds higher across the cycle and earnings are lumpier than the top line because of one-time items. Read the quarters for the volume surge and the fiscal years for the compounding.

Q2 2026 revenue and profit rose about 11% and 13% over the year. The eye-catcher is Q1 2026: net income of $1.41B against operating income of $1.67B, a conversion rate far above the usual, meaning a large one-time gain flattered it. Strip that and the trailing multiple is nearer 25 times clean earnings than 23.

The 2021 and 2022 net income swings are mark-to-market and deal noise, not the engine. The engine compounded revenue about 8% a year from 2021 to 2025 and operating income about 9% to $4.9B, a steady machine that the messy bottom line hides. One caveat that widens the margins on paper: reported revenue is gross and includes transaction pass-throughs, so the net figure the company steers by is smaller and the true operating margin higher.

What I believe that the tape does not: the market prices ICE on cyclical volume fear and the coming MarketAxess debt, and underrates that a majority of revenue recurs monthly regardless of volumes. The print that settles it is mortgage recurring revenue turning back up alongside the MarketAxess pro forma.

Management

Jeff Sprecher built this from a single power exchange and still runs it, which is the reason to own it and the reason to watch the capital allocation now. Insiders sold about $62.7M over the last year across 45 sales and bought nothing, with Sprecher's own roughly $43M sale in February 2026 the bulk of it; plan status is not disclosed in the filings I have, so read it as neither reassuring nor alarming. Capital returns are real, $1.29B of stock bought back in 2025 and another $551M in Q1 2026, but the MarketAxess financing points the next stretch of cash at a deal on a balance sheet already carrying net debt near three times EBITDA. Buying back stock and buying a competitor at the same time is a bet on the cash machine that the record so far supports.

How it fails or surprises you

MarketAxess leverage and price. ICE is adding debt to a book already at about 2.8x net debt to EBITDA to buy a bond venue whose volumes have been under pressure from electronic competition. If the acquired flow keeps sliding, ICE overpaid into a weakening asset. The first tell is the disclosed purchase multiple and pro forma leverage when terms land.

The mortgage recurring stall. The Mortgage Technology recurring line barely grew, about 2% over the year, because high rates keep loan origination depressed. If rates hold, that segment stays a drag the multiple is not discounting. Watch for its recurring revenue reaccelerating, or not, next quarter.

Volatility that keeps compounding (right tail). Financial contract volumes ran up 43% and total volumes up 22% as rate and energy uncertainty pushed institutions to hedge. The market treats that as cyclical and non-recurring, so it pays nothing for it. If volatility persists, transaction revenue compounds well past the recurring story, and the first sign is ADV holding above 12M into the next print.

Closing thoughts

Two specific prints narrow this. The MarketAxess pro forma will convert the leverage question from rumor into math. Mortgage recurring revenue in the next quarter shows whether that segment escaped the rate freeze or not. The steady two-thirds of revenue that bills monthly is priced roughly right; the argument is whether the swing factors compound or cancel. The left tail is a leveraged deal into a fading asset at the wrong price; the right tail is a franchise whose volume engine and recurring base compound together while it looks cheap on cash flow. The fatter tail is the upside, because the downside is bounded by a balance sheet that has absorbed big deals before and a cash flow that covers the debt, while the cheapness on EV/EBITDA is genuine even after the one-time-flattered quarter.

The bet is still that one company keeps running the plumbing of modern finance, the oil and rate markets, the price of bonds, and the mortgage software, and gets paid for it every month whether or not volumes stay hot. What breaks it is the MarketAxess deal landing expensive on an already-levered book while mortgage stays frozen. The pair that tells you first: the disclosed pro forma leverage against mortgage recurring revenue turning up. If leverage climbs and that line stays flat, the read is wrong.