The €10 Billion That Didn't Leave

On the morning of 24 June, the German defense ministry did something that looked, on the tape, like a catastrophe. It killed the F126 frigateF126 frigateGermany's planned class of six custom-designed anti-submarine warships, its largest surface-warship order since World War II, cancelled on 24 June 2026 after costs ballooned. program, six anti-submarine warships, the largest surface-combatant order Germany had placed since the Second World War, after the projected cost to finish crept toward €18 billion against an original budget near €10 billion.

A frigatefrigateA mid-sized warship used mainly to escort other ships and hunt submarines, smaller than a destroyer, larger than a corvette. is a mid-sized warship built to escort other vessels and hunt submarines. Six of them, custom-designed, had become Germany's flagship naval bet. And the company the market had penciled in to lead a follow-on order, Rheinmetall, fell as much as 18% intraday, one of its sharpest single-day falls in years. On the order of €10 billion of its market value evaporated before lunch.

The selling did not stay put. Renk dropped 7.2%. Leonardo closed down 4.7%. Hensoldt fell 3.3%, Saab 2.8%. None of them had anything to do with the F126.

So the question this article answers is narrow and testable. When €10 billion vanished from the screens, did investors reprice a real loss of cash flow, or how they value the whole business of building custom warships? The evidence points hard at the second. And it matters this week, because the NATO Summit Defence Industry Forum opening in Ankara today, an integral part of the 2026 NATO Summit, is the first live test of whether Europe's spending pledges become signed steel.

Start with the falsifiable version of the claim, because a claim you can't kill isn't worth printing. If the selloff were a rational repricing of lost contracts, each firm's one-day drop should scale with its exposureexposureHow much a company's revenue or profit actually depends on a particular contract or event, the test in this story is whether each firm's drop matched its F126 exposure. to the F126. Firms with a lot of F126 revenue should fall a lot; firms with none should barely move. That is what a cash-flow event looks like. That is not what happened.

Here is the cleanest way to read that chart. Five of these firms, BAE, Saab, Hensoldt, Leonardo and Renk, had exactly the same F126 exposure: zero. If exposure drove the tape, their drops should have been identical too. Instead they scatter from BAE roughly flat to Renk down 7.2%. Renk, a maker of tank gearboxes and transmissions with no stake whatsoever in the F126, fell the second-hardest of anyone. And the one company that actually won the reallocated order, TKMS, went up.

When the input is identical, no exposure, and the outputs run from roughly 0% to -7.2%, you are not watching the market price a contract. You are watching it price a category.

One more control rules out a market-wide risk-off dayrisk-off dayA day when investors dump risky assets across the board out of caution, ruled out here because the wider European market barely moved., when investors dump risky assets across the board. The pan-European Stoxx 600 barely moved on 24 June, slipping about 0.1%. This was a defense-sector event sitting inside a calm tape. The panic was specific, and it was about defense as a theme, not about anyone's confirmed loss of cash.

Even Rheinmetall, the one firm with any real F126 connection, lost far less than the tape implied. Its actual role on the program was as a subcontractorsubcontractorA company hired by the prime to supply one part of a project, here Rheinmetall's role was to supply the ships' gun systems, not to build the ships.: it supplied the MLG27 gun systems, with the Dutch group Damen as the prime. That gun work is a small line, not the ship. The €12.8 billion the market appeared to mourn was something else entirely: an unsigned bid for a hoped-for continuation program in which Rheinmetall would have moved up to prime. Unsigned. A wish, not a contract.

Put a real number on the loss. Rheinmetall itself now puts the hit to this year's sales at about €300 million, and Jefferies reached the same order of magnitude, calling the F126 under 3% of the group's long-run revenue forecast. Jefferies cut its price target by 31% to €1,300, and kept a Buy rating, writing that the market value wiped out in the selloff far exceeded the profit value of the contract at stake. Against that, the market erased something close to €10 billion. You can reconstruct that figure independently: Rheinmetall has roughly 45.5 million shares outstanding, and a swing of about €227 a share lands near €10.3 billion.

A word of caution on that €10 billion, because honest arithmetic demands it. The -18% is Rheinmetall's intraday low; the stock closed off a smaller amount, so the round €10 billion is best read as the shock at its worst point, not a settled loss. But even taken at face value it makes the mismatch starker, not weaker: a worst-case market-value swing of €10 billion sat on top of a worst-case cash hit near €300 million, a gap of more than thirty to one between what the screens said and what the ledger said.

If this had been a cash-flow event, the money would have drained out of naval procurement altogether. It did the opposite. The euros were re-pointed, not withdrawn, and they bought more steel, not less.

Read the ledger plainly. Germany walked away from six hulls whose completion cost was heading past €18 billion and committed instead to up to eight hulls for about €11.6 billion, from a domestic yard, on a design that already exists and already floats. That is two more ships for roughly €6.4 billion less in forward-committed money. For the German navy it is arguably a better deal; for the defense sector as a whole, the order book did not shrink. It changed hands. The only way to see a €10 billion hole is to look at one company's share price and ignore the €11.6 billion order sitting on the other side of the trade.

There is a simple test to separate a mood from a loss: wait a week and see whether it reverses. Cash flow that is genuinely gone does not come back on its own. Sentiment does.

Rheinmetall clawed back roughly 15% from its 24 June trough into the opening of the Ankara summit. Nothing about the F126 changed in those two weeks, no contract was un-cancelled, no cash returned. What changed was the story investors were telling themselves about defense, helped along by Rheinmetall's CEO Armin Papperger buying about €7 million of his own company's shares on 22 and 25 June. A rebound that fast, with no new cash, is the signature of an overshoot correcting itself, the emotional half of the re-ratingre-ratingWhen investors change what they will pay for a stock or a whole sector because their view of its future has shifted, not because of a single confirmed cash loss. unwinding while the rational half, the harder question about margins, stays open.

The overshoot is not the whole story, and it would be dishonest to pretend it is. There is a sober reading of the same tape that also fits the evidence, and it is more interesting than panic. Germany, faced with a bespoke warship designed from scratch for a single buyer, chose a cheaper design that already exists. That decision is a data point about every prime that earns fat margins on custom, one-off programs, the bespoke-prime premium.

The market did not decide Rheinmetall lost €10 billion of cash. It briefly wondered whether the whole business of designing warships from scratch is worth what everyone had been paying for it.

That is a re-rating, not a write-off. It is the market marking down, across a category, the confidence that governments flush with new defense budgets will keep buying the most expensive version of everything. It can be right or wrong, but it is a judgment about future margins, not a receipt for lost cash. And it is exactly why the two readings, overshoot and re-rating, both belong in the answer.

None of this means nobody lost. Somebody did. The F126 prime, Damen, and the German yards building the hulls lost a program. And German taxpayers lost the roughly €2.3 billion already sunk into ships that will now not be finished, money the switch to MEKO frigates does not refund. The honest claim is precise: the €10 billion did not leave the defense sector, not that no one lost anything.

• The one-day moves did not scale with F126 exposure, so the selloff was not the market pricing lost cash flow.
• Rheinmetall's real one-time hit was about €300m against a market-value swing near €10bn, a gap of more than thirty to one.
• The procurement euros were re-pointed to TKMS, not withdrawn; the sector's order book did not shrink.
• The roughly 15% rebound in two weeks marks the emotional overshoot correcting, while the sober re-rating of bespoke-prime margins is a separate, still-open question.

Which brings the story to today. The NATO Summit Defence Industry Forum opens in Ankara this morning, alongside the 2026 NATO leaders' summit, and it is the first live test of the thing the selloff was really about: whether Europe's headline spending pledges, up to 5% of GDP, actually convert into signed contracts. Watch what gets signed there, not what the tape did two weeks ago. The €10 billion never left. The question is where, and on whose terms, it lands next.