Quantinuum Inc. Class A Common Stock QNT

Key data

The business

Quantinuum builds quantum computers using trapped ions, individual atoms held in place by electromagnetic fields and manipulated with lasers, a different approach from the superconducting chips Google and IBM favor. Customers do not buy the machine, they rent time on it, mostly through Amazon, Microsoft and Google's clouds, to run experiments in chemistry, materials and cryptography. The company was formed by merging Honeywell's quantum unit with Cambridge Quantum, and Honeywell still controls it. The moat, if one exists yet, is a lead in qubit fidelity, the rate at which the machine computes without error, which trapped ions currently do better than most rivals. The customer today is not a corporation optimizing a supply chain, it is a national lab or a government agency funding research. The U.S. government alone was 32% of last quarter's revenue. That tells you what this is: a science program with a stock price, not a product with a market.

The numbers

Revenue is real but small, and growing fast off a tiny base. Last quarter's $8.0 million was almost quadruple the $2.1 million a year earlier, and consensus puts full-year 2026 at $30.1 million, doubling to $61.2 million in 2027. Set that against the cost of producing it.

The $555 million operating loss dwarfs revenue by nearly 70 times. Most of it is a one-time stock-compensation charge triggered by the IPO, not cash out the door. The reported net loss to public shareholders was $65.4 million, or $1.93 a share, because Quantinuum's structure sends roughly 87% of every loss to Honeywell. Strip the IPO charge and the business still spends about $680 million a year on research and overhead against maybe $13 million of trailing sales.

What the company has actually sold ahead is modest.

$74.2 million of contracted work, of which $29.4 million lands in the next year. For a $13 billion company, the entire book of signed business is smaller than a rounding error. And that book leans on a handful of names.

The three biggest customers were 71% of last quarter's revenue, and the mix reshuffles every period, a year ago one customer was 84% of the first half. Cash is the one number that is not small. Quantinuum holds $2.1 billion, up from $763 million at year-end after the IPO raise, against free-cash burn of about $168 million a year. That buys more than a decade of runway, which is the whole point, the company has bought time for the science to arrive. The point the price obscures: near-term financial risk here is close to zero, and the real question is not solvency but whether trapped-ion quantum computing ever earns its keep. The print that settles the direction is the 2027 revenue line against the $61 million consensus, and how much of it is government versus commercial.

Management

Honeywell controls Quantinuum, holding roughly 87% of the economics and the votes through Class B stock, and the public owns about 13%. That makes this a controlled company with a thin float, and the interests are not symmetric. A tax receivable agreement routes the bulk of future tax savings back to Honeywell and the former Cambridge Quantum holders, so public shareholders fund the losses now and share less of the eventual tax benefit. No insider bought or sold stock in the open market over the past year, which for a company months from its IPO tells you little. The compensation that matters is the equity granted at listing, the charge that produced the $555 million operating loss. Judge management on one thing: whether the $2.1 billion buys machines that win contracts or overhead that does not.

How it fails or surprises you

Customer concentration and the government checkbook. The U.S. government was 32% of revenue last quarter and 29% of the first half, and the top three names were 71%. Total signed backlog is $74.2 million. A federal research budget cut, or one agency pausing, halves the top line in a quarter, and there is almost no commercial base underneath to cushion the fall.

The loss does not shrink. Even setting aside the IPO charge, Quantinuum spends roughly $680 million a year to generate about $13 million of sales. Consensus sees the loss per share barely moving, from $1.64 this year to $1.65 next. The cash lasts over a decade, but a burn that climbs while revenue crawls turns a well-funded story into a slow, dilutive bleed and a fading multiple.

Error-corrected quantum arrives on schedule (right tail). If Quantinuum shows logical qubits reliable enough for a real commercial workload in the next year or two, demand stops being a research budget and becomes an IT budget. The market pays 988 times sales for the possibility but prices only $61 million of 2027 revenue. One large enterprise contract, visible first as a step-up in backlog, would break that estimate wide open.

Closing thoughts

Nothing in the next four quarters resolves whether quantum computing pays off, that timeline runs years, not quarters. What the next year does settle is narrower: whether the science is converting into commercial contracts or still leaning on government grants, and whether the burn is stabilizing. Survival is not the question, the $2.1 billion answers that for more than a decade. The payoff is a barbell. Most of the probability sits on quantum staying a lab curiosity, in which case the stock bleeds toward the cash as the story ages and the multiple compresses. A smaller slice sits on it working, in which case $13 billion looks cheap. The left tail is not bankruptcy but a long, dilutive fade, and the right tail is genuine and large. Which is fatter depends on physics I cannot underwrite, and neither can the market, which is why it is priced on hope.

The bet is still that quantum computing becomes a real business and Quantinuum's atom-trapping machines get there first, paid for by $2.1 billion in the bank and a government that keeps funding the research. What breaks it is the science failing to convert, revenue stalling near the $30 million consensus for this year while the operating loss stays vast. Watch two numbers together, backlog and quarterly cash burn. If the signed book climbs faster than the burn, the story is alive. If the burn climbs while the book stalls, the market is funding an experiment that never became a company.