major exchanges tracked by The Economist. Hong Kong’s predicament also serves as a warning sign for America, which is enjoying an IPO surge of its own. Amos is in a long queue of companies waiting to list in Hong Kong. The exchange is encouraging them by extending the window for firms to complete their paperwork and allowing them to file it in secret. In the first seven months of 2026 it hosted 104 IPOs, which together raised about $42bn. It offered a refuge to Shein, a pedlar of superfast-fashion which debuted on August 31st, after geopolitical concerns deterred the company from listing in New York or London. The sum is more than double the haul from the same months of 2025, itself a bumper year in which Hong Kong debuts raised more money than any other exchange. If markets for stocks were like ordinary ones for cars or other goods, Hong Kong’s would be unquestionably booming. If motorists were snapping up lots of new vehicles at healthy prices, no one would worry that prices for used rides were flat or falling. But with equities, the market for “second- hand” securities overshadows that for newly minted shares. Hence Hong Kong’s conflicting sensations: zingy and downbeat all at once. Some of the dissonance is caused by unique Chinese characteristics. Mainland regulators have become stricter about share offerings on the Shanghai or Shenzhen exchanges. In response, some Chinese companies have turned to Hong Kong instead, even if conditions are not otherwise inviting. America’s increased hostility to its superpower rival has also persuaded a number of Chinese companies on American bourses to seek secondary listings in Hong Kong. Their arrival has perhaps crowded out investor demand for other shares. Like someone who has gorged on too much confectionery, Hong Kong’s market is suffering from indigestion. It also echoes a familiar tune. A study of American markets from 1928 to 1997 by Malcolm Baker of Harvard University and Jeffrey Wurgler, then of Yale, finds that firms issue more equity just before periods of low market returns. Another paper from 2005 by Lubos Pastor and Pietro Veronesi of the University of Chicago similarly finds that IPO waves follow good times and precede bad times.

In keeping with this score, Hong Kong’s IPO boom began amid a dramatic stock-market rally: the Hang Seng rose by over 40% in the year after China’s government announced fresh monetary stimulus in September 2024. The index’s subsequent performance is jarring only in that the low returns have arrived before the IPO wave is over. Some scholars think the phenomenon reflects markets’ irrationality. Owners sell shares when they know the market is overvalued and outside investors do not. But Mr Pastor and Mr Veronesi have a different explanation. The decision to go public is made under uncertainty—corporate insiders do not know anything the market doesn’t about future conditions—and hard to reverse. Since markets always fluctuate, it makes sense not to list when times are bad, in the hope that conditions will improve. Waiting thus carries what economists call an “option value”. As a consequence, many firms list after valuations improve, but long before valuations peak. What does this mean for Hong Kong investors? Though the secondary market as a whole is struggling, the debutantes have mostly done well at least in their first days of trading (even shares in the ill-starred Shein ended their opening day roughly flat). Those lucky enough to get an allocation of shares in an upcoming IPO can, then, look forward to three sensations: a tingle on the tongue, a jingle in the ear, and the clink of money in their pocket. ■ This article was downloaded by zlibrary from https://www.economist.com/finance-and-economics/2026/09/01/ipo-booms-can-spell- trouble-for-the-markets

Finance & economics | Free Exchange Central banking has a forbidding future But Kevin Warsh has bought himself time Sep 3rd 2026 Bears, bulls, hawks, doves: each animal in the macroeconomic menagerie can be found between the peaks, prairies and ranches of Jackson Hole, Wyoming. During the Kansas City Federal Reserve’s annual central-banker bash on August 27th-29th your columnist spotted a black bear and learned the rudiments of cattle-roping. (The trick to roping a bull, it turns out, is knowing when to let go.) But it was a hawk that was the talk of the town. Kevin Warsh, the new-ish Fed chair, arrived in the Grand Tetons after a frustrating start. Markets had been baffled by his gnomic, occasionally contradictory press conferences. Worries lingered that he lacked the backbone to defy Donald Trump. A week earlier Scott Bessent, the treasury

secretary, had meddled in America’s bond market, which smelled like pre- midterm politicking. Despite those efforts, yields continued to surge. Fortunately for Mr Warsh, the upside of low expectations is that they are easier to surpass. By that standard, the hawk soared. After reiterating his dislike of explaining himself to markets, Mr Warsh did just that. He walked through inflation (troubling) and unemployment (not so): the sort of “reaction function” he’d disavowed, and a hawkish one, too. He emphasised the Fed’s main inflation gauge over funky alternatives. Insiders saw the clawprints of career staffers. The hawkish talk half-worked. That day markets shortened the odds of a rate hike but, reassuringly, without a further rise in long-term yields. Still, tough talk can do only so much: yields resumed their upward march the following week. Has Mr Warsh now backed himself into a corner? The crowd at Jackson Hole was split. One side reckoned that after such a firm steer he has no choice but to raise interest rates. If not, questions about his seriousness and political independence would deepen. Others noted that even a pre-Warsh Fed would be cautious about big moves so close to an election. Besides, raising rates in September versus December makes little economic difference. Markets now put the odds of a September hike at a tad better than even. Whatever leeway Mr Warsh bought himself came from going back to central-banking basics. But the official theme of Jackson Hole was on the future: “Financial Innovation: Implications for Payments and Policy”. Panellists debated whether stablecoins and tokenised deposits might revolutionise financial plumbing. Darrell Duffie of Stanford University suggested that programmable money may better co-ordinate complex multi- step transactions. Perhaps it could smooth cross-border payments. Wenxin Du of Harvard Business School countered that conventional financial technology could do the same, often more simply and cheaply. Kenneth Rogoff, also of Harvard, spoke for many when he decided to ignore stablecoins (a dry topic, even by central-banking standards) and focus on bond yields and debt. Channelling Alexandria Ocasio-Cortez, a leftie standard-bearer who recently described the social-justice frenzy of “woke 1.0” in the early 2020s as “crazy”, Mr Rogoff denounced “macro woke 1.0”:

arguments from the same era that interest rates would stay low, justifying more public borrowing. He added swipes at old adversaries: “I wish I was as sure of anything as Paul Krugman is about everything.” Now interest rates are up and public debt is higher than ever: a woeful combination, especially for central bankers. One risk is that government bonds stop behaving like a safe asset. Eswar Prasad of Cornell University presented a paper noting that the “convenience yield” investors are willing to forgo for safety has vanished. Still, he found, the dollar’s position remains strong and financial innovations like stablecoins may bolster that dominance. Any boost for American assets from stablecoin demand would be dwarfed by a second risk of high borrowing: that debt pushes politicians to meddle with central banks. Mr Rogoff framed the worry around Elizabeth Warren, another leftie icon (whose leftier advisers “make her look like Ronald Reagan”). But the problem is structural. When debts are high, co-opting central banks into financing the government can become appealing. Left unsaid was that greater risks today, as Mr Bessent’s moves show, come from the opposite side of politics to Ms Warren. If “macro woke 1.0” is buried, what about “macro woke 2.0”? That, Mr Rogoff suggested, might be the hope that artificial intelligence solves economic woes. Mr Warsh was as AI-woke as anyone while vying to become Fed chair; in his speech in Wyoming, less so. His remarks on AI were framed mainly as questions, and he hinted that the initial effect of the technology would probably be to raise inflation by supercharging data-centre investment. AI poses more direct risks, too. Markus Brunnermeier of Princeton University presented a paper arguing that markets may be hit early by advanced AI, endangering financial stability. Robot traders have “asymmetric understanding”, he argued: they get humans perfectly but are themselves perfectly inscrutable. That observation led Mr Brunnermeier to some radical ideas: human-only markets off limits to AI, and randomised central-bank decisions that AIs have no edge in predicting. Those conclusions were controversial. Raghuram Rajan of the University of Chicago observed that markets are already full of humans with conflicting,

often opaque motivations; AI may itself help humans understand AI. As for markets, they are already a bit of an alien superintelligence. Next year will be the Jackson Hole pow-wow’s 50th birthday. The Tetons will still be full of bears and hawks. Many of the human faces will be the same, too: central banking is not exactly a field with a high staff turnover. Debt problems and AI risks will also persist, and probably mount. If Mr Warsh keeps up his new regular-central-banker act, he may at least have built more credibility to find a mountain path through them. ■ This article was downloaded by zlibrary from https://www.economist.com/finance-and-economics/2026/09/02/central-banking-has-a- forbidding-future

· Science & technology

Humble exoskeletons work better than combat supersuits A suspicious signal in a dark-matter detector has physicists excited NASA’s latest mission to map the cosmos takes off A horde of AI agents conspired against their creators How to combine exercise with video games

Science & technology | Power dressing Humble exoskeletons work better than combat supersuits Wearable machines that assist just one part of the body are proving surprisingly useful Sep 3rd 2026 AFTER HIS hands were crushed two years ago in a machine that bends sheet metal, Remco Letschert feared he would never work again. Yet a year later Mr Letschert returned to his construction job in Aarhus, Denmark— even though his dominant left hand is too weak for such work. A powered glove he affectionately calls his “robot hand” allows him to carry materials, climb scaffolding, squeeze tin snippers and operate a drill. Using the system, branded “Ironhand”, requires no special thought, he says. When sensors in the glove’s fingertips and palm detect pressure, five little motors instantly whir into action, pulling cables that tighten his grip.