{"id":31404,"date":"2026-09-29T04:00:00","date_gmt":"2026-09-29T02:00:00","guid":{"rendered":"https:\/\/www.thebrokernews.ch\/?p=31404"},"modified":"2026-09-28T16:51:18","modified_gmt":"2026-09-28T14:51:18","slug":"where-it-breaks-first","status":"publish","type":"post","link":"https:\/\/www.thebrokernews.ch\/en\/where-it-breaks-first\/","title":{"rendered":"Where It Cracks First"},"content":{"rendered":"
The U.S. market yields a single figure: the Magnificent Seven and three chip manufacturers.<\/span><\/div>\n\n\n\n

Ten stocks are now about 41 per cent of the American market, the exact level that capped every great bubble of the past century. The bond market has already cracked, quietly, to a twenty-year high, while the Treasury bails it with a teaspoon and its largest foreign lender turns for home. The AI keystone loses more than it earns and lives on the next cheque. The only real question left is which fault gives way first.<\/strong><\/p>\n\n\n\n

Every so often the market hands you a single number that ends an argument, and this week it produced one. On the broadest and longest measure we have of how top-heavy a stock market can become, ten companies, the Magnificent Seven<\/a> the three chip names that feed them, Broadcom, AMD and Micron, now make up, on Bank of America’s reckoning, about forty-one per cent of the entire value of the United States market. That is not a number you have to interpret. You have only to set it beside its predecessors. The utilities that led the market into 1929 topped out near thirty-six per cent; the Nifty Fifty<\/a> of 1972 at about forty; Japan, at the delirious height of its bubble, around forty-four; and the dot-com mania<\/a> of 2000 at almost exactly forty-one. Every one of those was, at the time, the obvious future, the thing a serious person could not afford to be underweight. And every one of them peaked in the same narrow band where we now sit. Concentration alone does not ring a bell; it never tells you the day. But it tells you, with the authority of a century of wreckage, that the air up here is thin, that the room is crowded at exactly the altitude where rooms have always emptied, and that the sensible posture is the one the flight attendant recommends. Belt on.<\/p>\n\n\n\n

Whether this is a bubble is, on the only measure that has ever counted in advance, not much of a question. The more useful one is the next. A bubble at its peak is not the same as a bubble that has burst, and between the two lies the only question worth a column: not whether, but where it cracks first. This week, as it happens, answered that question in two places, and the tell each time was that the crack came dressed as reassurance.<\/p>\n\n\n\n

The measure that should end the argument<\/h6>\n\n\n\n

Begin with the picture itself, because it deserves to be looked at slowly rather than waved away.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

The objection to a chart like this is always the same, and it is fair as far as it goes: concentration can persist, even climb, for longer than the bears can stay solvent, and the leaders of a mania are often genuinely the best businesses of their day, which is precisely why everyone crowds them. The Nifty Fifty really were great franchises. The point of the picture is not that the AI leaders are frauds; several are the most profitable enterprises in history. The point is narrower and harder to dismiss. When the market’s fate comes to rest on so few names, the index stops being a market and becomes a bet, and a bet of that size cannot be diversified away by the very people who most need to, because to be underweight the ten is to underperform, and to underperform is to be fired. So the money keeps flowing to the top ten not because anyone believes there is much left to gain, but because no professional can afford the career risk of stepping off the floor while the music plays. That is not investing. It is a game of musical chairs in which the chairs are forty-one per cent of everyone’s retirement, and the exits, when they are wanted, will be found to be the same width as the entrances were not.<\/p>\n\n\n\n

The crack that is already open<\/h6>\n\n\n\n

Here is the first place it has cracked, and the unsettling thing is how quietly it did so, while everyone watched the shares. The price of money, the thing against which every other price on earth is measured, broke this week to its highest in about two decades. For most of September the ten-year Treasury yield had sat, uneasily, in the high fours. Then, in a single week, it tore up into the fives, printing about five and a fifth per cent, a level not seen since 2007; the thirty-year went with it to five and a half, the most since 2004. That is not a slow drift you can narrate away. It is a crack arriving in public, in five sessions, and it did so in the teeth of a government trying, in full view, to prevent it.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Scott Bessent’s Treasury has spent weeks buying the government’s own long bonds in the open market to force their yields down, and enlarged the operation this month when the first rounds would not hold. The picture above is the result, and it needs no caption: the more he bought, the higher the yields went. Put the effort into arithmetic and the reason is almost a joke. On Wednesday the Treasury sold seventy billion dollars of five-year notes. On Thursday it bought back four billion of the long end. A few billion out, seventy billion in, against a market of more than thirty trillion that must swallow trillions of fresh borrowing every single year. That is not a bazooka, and it is not even a bucket. It is a teaspoon, and the thing he is bailing is the American deficit, which no teaspoon reaches. This is the crack that matters most and draws the least attention, because when the cost of money rises like this, everything priced against it is quietly repriced downward at the same time, and everything means everything: the mortgage the family can no longer refinance, the wall of corporate debt that must be rolled at the new rate, the private-credit loans still carried at prices the market no longer offers, and the long bonds sitting inside every pension fund that mistook 2023 for the high. A forty-one per cent market perched on the steepest discount rate in twenty years is a structure whose foundation is being drawn out from under it by the hour.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

And there is a reason the teaspoon is failing that goes beyond the size of the spoon, and it is the mechanism I set out in these pages last week. The buyer who has gone missing is Japanese. For thirty years Japan was the world’s supplier of cheap money and the single largest foreign holder of American debt; its savers financed a good part of Washington’s deficit because their own bonds paid them nothing to stay home. Last week the Bank of Japan raised its rate to a thirty-one-year high and the yen fell anyway, which is the sound of the carry trade beginning to turn for home, and this week the American long end did precisely what a departing foreign buyer would make it do. So Bessent is not merely a man bailing a boat with a teaspoon. He is a man bailing with a teaspoon whose largest customer has just been handed a reason to stop buying and sail for home. The tide he is fighting is not only the deficit. It is the deficit at the very moment its biggest foreign financier loses the incentive to fund it.<\/p>\n\n\n\n

The keystone that loses more than it earns<\/h6>\n\n\n\n

The second fault is the one the crowd is actually standing on, and it runs through a single company that is not even listed. The whole edifice, the chips and the clouds and the forty-one per cent, rests to a startling degree on the health of OpenAI, and OpenAI, on its own leaked figures, loses more than it earns. In the first half of this year it booked around thirteen billion dollars of revenue and ran about twenty-one billion of losses. That is not a fast-growing company with a rounding error to explain away. It is a hole with a logo.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

I want to be precise here, because imprecision would hand the doubters an easy escape, and because the precise version is the worse one. OpenAI is not bankrupt; there is no filing and no locked door. Nor is its trouble that it cannot raise money. It has just raised the largest sum in the history of private enterprise, a hundred and twenty-two billion dollars in March, at a valuation of eight hundred and fifty-two billion, from Amazon, Nvidia and SoftBank. The trouble is that even that is not enough. By the company’s own projections that record cheque is expected to run dry around 2028, on the way to something like a quarter of a trillion dollars of cash burned before it promises, at last, to turn a profit near the end of the decade. It does not live on its revenue, which covers only a fraction of its costs. It lives on the certainty that a larger cheque will always arrive to fill the gap the last one could not. And so much of the rest of the complex has quietly staked itself on that same certainty that a single failed raise here would not be a private misfortune. It would do to the forty-one per cent what the failure of the first internet giants did to the dot-com index in 2000: not kill the technology, which lived on to remake the world, but vaporise the valuations that had run a full decade ahead of it.<\/p>\n\n\n\n

And there is a detail that turns the discomfort into something nearer vertigo. Follow the money and it runs in a circle. Nvidia takes an equity stake in OpenAI; OpenAI promises the cash onward to Oracle and the other cloud builders for computing power; and they, in turn, spend it ordering the chips back from Nvidia. The same dollar is an investment, a revenue and a purchase order at three stops on one loop, a loop now reckoned in the hundreds of billions. A keystone that is also its own investor, its own customer and its own supplier has stopped being a company and become a loop. And loops do not lean. They hold, and then they go. So if you ask me where the loudest first break will come, I do not point at the bond market, whose crack is deeper but quieter. I point here, at the keystone, because this is the fault the entire crowd is standing on, and the one whose failure would be impossible to look away from.<\/p>\n\n\n\n

The theatre staged over the fault<\/h6>\n\n\n\n

Over both of these faults the week laid a thick coat of reassurance, and the reassurance is itself a symptom. In New York the great and the good gathered at the United Nations to perform peace over the top of a war that has not stopped, the American president defending the very conflict whose front line still throttles the Gulf, the Iranian answering with defiance, everyone applauding on cue and settling nothing. Its only market function this year was to supply the headlines on which oil could be talked down, and talked down it duly was, by some six per cent on the week, on hopes of a truce and a reopened strait. Yet the tankers did not move with the price: throughput through Hormuz was essentially flat, the cost of freighting a barrel still carried a war premium near twenty dollars, and Europe’s diesel tanks sat below their five-year average on the doorstep of winter. The barrel fell on a story the plumbing refused to confirm, which is the same disease I described last week and no better for a week’s rest.<\/p>\n\n\n\n

But the reassurance that should trouble you most is the one the stock market told itself. In the same five sessions that the cost of money broke to a twenty-year high, the S&P 500 and the Nasdaq closed the week at fresh records, carried there by the very ten names that are the forty-one per cent.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Set the two lines side by side and you have the whole column in a single frame. In the exact week the price of money tore to its highest since 2007, the index made a new all-time high, on the narrowest leadership of the cycle, its breadth by some measures the thinnest since the year 2000. A market that sets records on ten stocks while the ground beneath all of them is being repriced is not showing strength. It is showing the last and most dangerous phase of a mania, the one in which the story narrows to a handful of names precisely because they are the only story still standing, and the crowd mistakes the narrowing for leadership. The new high is not the refutation of everything above it. It is the final exhibit for it.<\/p>\n\n\n\n

Where it cracks first<\/h6>\n\n\n\n

So put the week together. A market sitting at the exact altitude where every great bubble in a century has run out of air. A bond market that has already begun to lean, with a treasury secretary bailing the tide with a spoon while its largest foreign financier turns for home. An AI complex whose keystone loses more than it earns and lives entirely on the next cheque, spinning in a loop with its own suppliers. And over all of it a fog of reassurance, a peace conference above a war, a barrel marked down on a rumour, an index setting records on ten names. It is worth saying, in fairness, where the fire has not yet reached, because that is the difference between a warning and an obituary. Gold has run to fresh records, which is the old money quietly voting with its feet while the paper is repriced. But corporate credit has not yet broken: high-yield spreads are still near their tights, around two hundred and seventy basis points, which is the one gauge in this whole account that is not yet flashing. A crack that has not reached the credit market is a different and less advanced animal than one that has, and the honest reader should watch that spread more closely than the index, because it is where the next real news will come from.<\/p>\n\n\n\n

So where, then, does it give first? The deepest crack is in the bond market, and it is already open, and it is the tide that will in the end pull everything else down with it, because it reprices the whole structure at once and needs no catalyst beyond itself. But the first crack you will actually see, the one that makes the headlines and empties the floor, will run through the keystone, through the single unlisted company that the entire forty-one per cent has silently agreed to carry, on the day one of its ever-larger cheques fails to clear.<\/p>\n\n\n\n

The cautious response is not to name that day, because no one can, and the ones now crowding the microphone to call the top are as unreliable as the ones who spent the last year calling it a new paradigm. It is simply to accept the seatbelt for what it is. To hold the real things, the ones a paper claim is only ever a claim upon, and to keep enough of one’s balance off the ten crowded chairs to be able to stand when the music stops, rather than to be found, like everyone else, lunging for a seat that was never really there. We are, on the best measure history offers, at the top. That does not mean tomorrow. It means that the reward for staying fully in the dance from here is small and the penalty for being caught in it is total, which is the precise arithmetic at which a prudent person stops adding and starts, quietly and without drama, edging his chair a little nearer the door. Belt on. The sign has been lit for a while now. This week it simply stopped blinking.<\/p>\n\n\n\n

Eric Lefebvre<\/p>\n\n\n\n

See also: Dancing on the Volcano<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Ten stocks are now about 41 per cent of the American market, the exact level that capped every great bubble of the past century. The bond market has already cracked, […]<\/p>\n","protected":false},"author":2,"featured_media":31056,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"_price":"","_stock":"","_tribe_ticket_header":"","_tribe_default_ticket_provider":"","_tribe_ticket_capacity":"0","_ticket_start_date":"","_ticket_end_date":"","_tribe_ticket_show_description":"","_tribe_ticket_show_not_going":false,"_tribe_ticket_use_global_stock":"","_tribe_ticket_global_stock_level":"","_global_stock_mode":"","_global_stock_cap":"","_tribe_rsvp_for_event":"","_tribe_ticket_going_count":"","_tribe_ticket_not_going_count":"","_tribe_tickets_list":"[]","_tribe_ticket_has_attendee_info_fields":false,"footnotes":""},"categories":[5100,12135,5134],"tags":[15324,14585,15326,15329,15328,15332,15322,7959,15327,12934,14102,15333,15321,15323,15330,15325,15331,15320],"class_list":["post-31404","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-current","category-columns","category-general","tag-15324","tag-bond-market","tag-chip-manufacturer-2","tag-crack","tag-fate","tag-fault","tag-iran","tag-market","tag-nifty-fifty","tag-nvidia","tag-oil-price","tag-staging","tag-stocks","tag-strait-of-hormuz-2","tag-supporting-element","tag-the-magnificent-seven","tag-theater","tag-war","ownarticle"],"acf":[],"cc_featured_image_caption":{"caption_text":"The U.S. market yields a single figure: the Magnificent Seven and three chip manufacturers.","source_text":"","source_url":""},"_links":{"self":[{"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/posts\/31404","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/comments?post=31404"}],"version-history":[{"count":2,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/posts\/31404\/revisions"}],"predecessor-version":[{"id":31407,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/posts\/31404\/revisions\/31407"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/media\/31056"}],"wp:attachment":[{"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/media?parent=31404"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/categories?post=31404"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.thebrokernews.ch\/en\/wp-json\/wp\/v2\/tags?post=31404"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}