TheChek News Flow

Iron ore trader battling court claims from creditors around allegedly fraudulent documentation

Zurich and Allianz wrote policies covering losses on transactions linked to embattled iron ore trader Radiant World, potentially exposing trade credit insurers to yet another alleged fraud in the booming market for protection.

The unravelling of Radiant, which is battling court claims from creditors in the UK and Singapore around allegedly fraudulent documentation underpinning its trading, has left commodity traders and financial institutions rushing to manage their exposure. Radiant and its founder, Pinkesh Nahar, have denied wrongdoing. 

The debacle also threatens to spill over into the insurance market, where firms that specialise in writing policies covering the risk that trading counterparties do not settle their bills have potential exposure.

Zurich had insured Radiant against non-payment from multiple counterparties, one of which was the commodity trading giant Glencore, according to people familiar with the agreements, with a policy providing cover in the double-digit millions of dollars.

Glencore was one of Radiant’s biggest trading counterparties and has faced a lawsuit threat from the Singapore-based iron ore trader, which has alleged their relationship went beyond that of a counterparty and described the Anglo-Swiss commodities group as “a senior partner in the relationship”. Glencore has called these claims “meritless”.

Allianz Trade, a subsidiary of the large German insurer that specialises in trade credit insurance, had also insured Radiant against non-payment, though its total exposure is smaller than that of Zurich, according to people familiar with that policy. 

Insuring invoices underpinning global trade has become one of the most lucrative niches in the insurance market. Commodity traders and suppliers have long used insurance to shield themselves from the risk that counterparties default or pay late, but the product has grown in use as a hedging tool for financial institutions that lend against outstanding bills.

The trade credit insurance market has dealt with the fallout of several high-profile collapses amid fraud allegations in recent years, leading to messy contract disputes.

Trade credit insurance policies typically cover businesses against the risk that a customer does not pay, due to risks such as a customer’s insolvency or default, while excluding losses due to some forms of fraud. But these provisions are often litigated and courts have found that under certain policies insurance carriers must pay out even in cases involving fraud.

The collapse of invoice finance firm Greensill Capital in 2021 has resulted in years-long litigation, in which global insurers have alleged that billions of dollars’ worth of policies have been voided due to fraud. 

Zurich is defending itself against a $400mn claim in London’s High Court from the administrator of Greensill’s banking arm over insurance policies linked to the commodity trading arm of controversial metals magnate Sanjeev Gupta.

Zurich has refused to pay out on the insurance, contending that it is void because a key lending programme provided by Greensill to Gupta’s companies was a “sham” and relevant information provided to it was false. 

Allianz Trade was also among the groups exposed to potential claims stemming from the bankruptcy of US auto parts maker First Brands Group, which relied heavily on invoice financing and faced allegations of fraud after its collapse.

Zurich said in a statement that it “does not have any material exposure to Radiant World”. Allianz Trade and Radiant declined to comment.

Europe’s difficult choices on AI

Data is the one area where it can still be sovereign and it has the greatest potential to generate growth

Europe faces a difficult set of choices. It needs to grow, so that it can keep financing its social model and the new demands of a changed world, from defence to energy. It wants to remain sovereign, so that it can order its economy according to its own values. And it wants growth without the costs in terms of wealth inequality and environmental damage that have accompanied it elsewhere. Artificial intelligence bears on all three at once. 

Despite a solid start to the year, the trend in European growth has slowed materially since the pandemic. This is true relative both to its own pre-pandemic trajectory and to the US. By one estimate, the euro area-US productivity gap widened from $9 per hour in 2018 to $21 in 2025. Views differ on how productivity levels should be compared across countries, but the gap in growth rates is clear.

Europe has different routes to raise growth, such as removing the high barriers in its internal market. But widely adopting AI is perhaps the most promising one today. According to ECB scenarios, fast adoption of AI would add 0.3 to 0.4 percentage points a year to total factor productivity growth — the gains that come from working more efficiently rather than adding labour or capital — which has been roughly zero since 2022. 

AI-led growth, however, creates a tension with Europe’s bid for sovereignty, because Europe controls little of the AI value chain. The technology is set to become completely pervasive: in the economy, in health systems, in education, in energy, in defence, to name just a few areas. This is no ordinary dependency. Being cut off from AI, once the economy runs on it, would be more like being cut off from the US financial system. The effects would be catastrophic.

That changes the terms of Europe’s relationship with partners it can no longer always rely on. Europe is not to the United States as Texas is to California. It is possible to imagine a future US administration making access to frontier AI models conditional on changes to Europe’s digital rules or digital taxes, or China using the licences on its open-weight models as leverage in a tariff dispute. Unless Europe controls some part of the value chain, growth through AI and sovereignty will pull against each other.

Europe’s potential zone of AI sovereignty is narrow. Its frontier labs cannot at present compete financially with their American and Chinese competitors, and leading-edge chip production is far behind. But data is the one area where Europe can still be sovereign, and it is also the area with the greatest potential to generate growth.

The continent sits on a wealth of public-sector data, such as decades of health records and data collection by statistical offices. Its highly automated manufacturing sector generates a deep well of machine-readable industrial information. European Commission estimates put Europe’s data economy at over €800bn, or more than 5 per cent of GDP, by 2030.

But this creates the next tension. To exploit these assets, Europe needs to have control over their storage and processing. This means it needs large-scale AI data centres.

Data centres, however, are fiercely contested. Local communities worry about their environmental costs and rising energy bills. It does not help that much of the capacity now being built in Europe is for the American tech giants.

Yet Europe also has to be realistic. The debate about overbuilding data centres is a US one. Europe’s problem is the opposite. The EU hosts under 5 per cent of the world’s AI compute versus 75 per cent for the US. Even for ordinary data centre capacity, the gap between demand and installed supply in Europe is estimated at around 3GW, roughly a quarter of what Europe currently has, and is expected to widen to 14GW by 2030.

As sovereignty comes to matter more, this lack of domestic capacity will start to bite. Compute could stop being fully fungible, and a large part of the infrastructure that processes and stores European data will have to sit on European soil. American operators can provide much of that capacity through what they market as sovereign cloud services, run from Europe but still under American ownership. The most sensitive uses, however, will have to run under European control, which is the tiered approach the Commission has proposed in its Cloud and AI Development Act. 

If the gap is not filled, two risks follow. Either European companies will not adopt AI because they cannot keep their data in Europe, a risk the Commission’s own impact assessment for the Act identifies. Or compute will stay expensive, slowing broad adoption for lower-value tasks. Amazon already charges 15 per cent more for its sovereign cloud in Germany than for its standard service.

The concerns about data centres can be met. Clean energy can be required to power them and waste heat reused where feasible. The facilities can be kept away from areas where energy and water are scarce. Data centres can be required to pay a fair share of the grid costs they create, and to share the benefits with local communities, whether through contributions to local budgets or lower energy bills.

So why is Europe not building more? Part of the answer lies in supply-side constraints. A data centre takes 24 months to build in the US but 42 months in Germany, owing to longer delays for permits and grid connections. More expensive electricity makes building and powering a facility in parts of Europe twice as costly as in China. But Europe does have regions where it is cheap to build: an AI data centre in Sweden, for example, costs only about a tenth more than one in China.

The main obstacle for Europe is the fragmentation of demand. Data centres require massive financial commitments, which investors will only make against bankable offtake contracts. In the US, hyperscalers and frontier labs provide them. In Europe, demand is spread across millions of companies, and even the largest companies tend to buy compute a year at a time due to uncertainty about how fast they will adopt AI. That is why companies operating neoclouds — data centres for AI workloads — in Europe, such as Nscale, sell most of their capacity to US buyers.

Europe must therefore create large, concentrated buyers by design. What is needed is for its largest companies to pool their midsized commitments into contracts large enough for investors to lend against.

A model already exists. A group of European companies, including ASML, Capgemini and Amadeus, have committed to multiyear purchases of Mistral’s European Compute Units, which are intended to underwrite 1GW of capacity by 2030. The challenge now is to scale up this model with more neoclouds and larger consortiums, each committing to multiyear purchases of standardised capacity. 

The goal is a virtuous circle. Companies accelerate their AI transitions, providing the demand that finances European capacity. That capacity lets Europe exploit its own data on infrastructure it can control. Cheaper, sovereign compute then makes adoption faster and the next commitments larger, paid for out of the productivity gains the technology delivers. And with capacity of its own, Europe is better placed to build its own AI models.

That leaves the deepest fear: that AI hands wealth and power to a new tech oligarchy. But the approach set out here leans against it. Demand pooled by many consortiums lets more providers into the market, and competition among them keeps the rewards from concentrating in a few hands. Progressive taxation would ensure that the rest of society shares in them too. In this way Europe does not have to choose between growth, sovereignty and its values. It can have AI without oligarchs.

Virgin Media O2 owners target £600mn in cost cuts

Telefónica and Liberty Global seek job cuts and other reductions in a bid to calm bond investors

Virgin Media O2’s shareholders are targeting around £600mn in cost cuts at the heavily indebted telecoms group in a bid to calm investor concerns about the sustainability of the company’s debt pile. 

Liberty Global and Telefónica are proposing asking the UK telecoms group to find the reductions through a mixture of job cuts and reductions in operating and capital expenditure, according to two people familiar with the matter. 

The move comes after the company was rocked by a sell-off in its bonds over the summer, with investors worried about its ability to service a £22bn debt pile amid an influx of competition from dozens of small fibre network operators, known as “altnets”. 

The sell-off put pressure on the price of VMO2’s roughly £1.1bn of senior unsecured debt, pushing the price of a $925mn bond down to 61 cents on the dollar on Thursday. The bond was trading at about 78 cents at the start of July.

Some of VMO2’s safer senior secured bonds have also fallen below 80 cents on the dollar in recent months, having traded at 92 cents at the beginning of the year. The price drop turned VMO2 into a focal point for investors in Europe’s junk debt market.

The FT reported in July that the company’s two owners were weighing options to reduce the company’s debt, with a cut to its £200mn dividend also on the table. A person with knowledge of the matter said no formal proposal had yet been put to VMO2.

VMO2, which serves millions of UK customers with broadband and mobile services, has suffered heavily as a proliferation of “altnets” have raised £31bn to lure customers from VMO2 and BT-owned Openreach with cheap fibre broadband. 

Virgin shed 33,500 broadband customers in the first half of 2026, on top of the 138,400 it lost last year. It had a customer base of 5.42mn at the end of June. 

Telefónica and VMO2 declined to comment, Liberty Global did not respond to requests for comment.

Just the threat of market disruption might help ease bottlenecks, to the extent that it prods manufacturers to expand output

Elon Musk reckons he specialises in converting the impossible to the merely late. That neatly encapsulates Tesla and SpaceX’s pursuit of electric vehicles, humanoid robots, reusable rockets and Martian settlements. Now, however, the entrepreneur has turned his disruptive sights to the more prosaic gas turbine market: SpaceX is set to cast the blades and vanes for them in-house. 

The aim appears to have shifted to making the impossibly late more readily available. The data centre boom requires lashings of electricity. Developers struggling to connect to the overloaded electricity grid have turned to dedicated power generation facilities, mostly using gas as a fuel.

As a result, turbines, largely supplied by GE Vernova, Siemens Energy and Mitsubishi Power — which account for over three quarters of the installed heavy-duty base according to S&P Global — are not available for love or money before 2030.

Breaking into this market is tough, given the advanced engineering and prolonged testing that these devices require. And while Musk reckons that by producing his own blades and vanes he can accelerate the turbine delivery timeline by up to 18 months, there are lots of other bottlenecks to getting gas-fired power generation up and running, including specialised installers.

Still, success would be a game-changer — not least for Musk himself. More electricity equals more data centre capacity and one of the ways in which SpaceX currently makes money is by renting it to the likes of Anthropic and Google: annual revenue from just these two deals amounts to roughly a quarter of next year’s total, according to S&P Capital IQ. It is also quite possible that advanced turbine parts, such as those Musk is proposing to cast, will double up for use in his rockets.

Musk has already been seeking creative ways round the gas turbine bottleneck, reportedly acquiring APR Energy, which operates a fleet of mobile gas and diesel turbines, for $1bn this year. SpaceX has been using mobile turbines in Mississippi and has agreed a plan to switch to a permanent power plant with local regulators.

Conversely, should Musk make convincing strides into the gas turbine market, that would put pressure on parts manufacturers including Howmet Aerospace. Its stock has fallen about a fifth since mid-August, but still trades at over 40 times this year’s earnings on S&P Capital IQ estimates. The impact on the big three turbine suppliers is less clear: while Musk’s arrival on their turf is a destabilising prospect, they may also benefit if more competition is introduced into their supply chains.

What’s more, Musk doesn’t have to succeed. Even just the credible threat might help alleviate the sector bottlenecks, to the extent that it prods existing manufacturers to expand production capacity. The idea here may simply be to put a rocket under their plans.

Michael Aiello’s exit comes after recent departures of other top lawyers

Weil, Gotshal & Manges is set to lose its top dealmaker and five other partners after a number of other top lawyers at the firm recently decamped for competitors.

Michael Aiello, who led Weil’s over 600-person corporate practice, is set to join Cravath along with other M&A-focused partners, people familiar with the matter said.

The Wall Street Journal reported last month that Weil was facing multiple partner defections, and was expected to lose several private-equity focused partners to rivals. Christopher Machera, the former co-head of Weil’s private-equity group and a former member of its management committee, has since joined Paul Weiss.

Weil is in the middle of a leadership transition, with Boston partner Ramona Nee set to succeed executive partner Barry Wolf after his more than 16-year run as the firm’s leader.

Aiello, who has worked on deals for clients including Dow Chemical and Sanofi, had become a towering figure within Weil as he was given increasing management duties. He was a close ally of Wolf, people familiar with the matter said.

“Mike Aiello and his team have informed us that they are leaving the firm for a smaller platform,” Weil said in a statement Thursday. “Weil has long believed that an ambitious growth strategy is in the firm’s best interest and intends to accelerate execution of that strategy going forward. We thank Mike and his team for their contributions.”

Top law firms are under pressure from all sides, with aggressive poaching of rainmakers, the pressure of artificial intelligence and less-loyal clients. Many of the biggest law firms have been intensely focused on courting and retaining private-equity clients, who tend to provide a steadier stream of work than big corporations.

Cravath has long been one of the most prestigious law firms on Wall Street, and a coveted landing spot for top law school graduates. But it has been threatened by megafirms like Kirkland Ellis and Latham & Watkins, and hasn’t heavily invested in private equity. Its leader for the last decade, Faiza Saeed, is also a top dealmaker and worked on deals including Disney’s acquisition of 21st Century Fox.

Cravath was one of the last firms to do away with a traditional lockstep model that pays equity partners equal shares of the revenue. Only a few still have this model, including Wachtell, which earlier this summer lost a team of six litigation partners in an almost-unheard-of defection to Gibson Dunn.

Austin-based Tern has won a contract from the U.S. Army to deploy the startup’s low-cost GPS alternative tech.

The $11.26 million contract will see the Army use Tern’s navigation tech in vehicles, though the startup was tight-lipped on the exact scale of the deployment of what it calls “Google Maps for the battlefield.”

The Army is tapping Tern at a time when GPS disruptions are on the rise, especially in the Middle East and in the skies above Ukraine and Russia. The U.S. military itself has reportedly been experimenting with GPS jamming, which may have even contributed to the deadly crash of a medevac plane earlier this year. There have been multiple efforts from Congress and President Trump (during his first term) to develop resilient, non-GPS navigation technologies for military use.

Modern vehicles generate all kinds of data while driving, and Tern’s whole tech stack is oriented around using some of that information to navigate without GPS.

“We’ve got a device that plugs into the diagnostics board of the vehicle, wired to a tablet, and then that tablet is the interface for the soldier to use, and we provide both the position and navigation and routing there for the soldier,” Tern co-founder and CEO Shaun Moore said in an interview.

Moore said Tern is “passively listening to” this information and not disrupting its flow in any way. Tern then uses “proprietary methods” to make the information usable for navigation. (He declined to go into specifics.)

“There is quite a bit of work that was done to get it to the point where we can just receive that information and then output latitude / longitude, so big effort by our team,” he said.

All this allows a vehicle to essentially determine its own position independently, using edge processing, which is helpful both on- and off-road, in tunnels or in deserts. Moore’s co-founder Brett Harrison even recently used it to navigate a 1,300-mile drive between Austin and Laguna Beach, California. He said Tern’s tech never lost his vehicle’s place during the whole trip, while traditional GPS cut out dozens of times.

Another aspect of Tern’s tech that may be just as important as accurate positioning is that it’s a closed system. This means it is not just a backup, it’s actually harder for adversaries to disrupt it.

“I don’t know that there is broad awareness on the fragility of the [GPS] system from a consumer standpoint,” Moore said. “I think people just see the blue dot on their phone and say it will always be there because it’s there now, right? Why wouldn’t it be there?”

As a military vet himself, Harrison said the Army contract is validating after Tern spent years proving that its tech works. He said his time in Afghanistan as part of a special operations task force is a big part of where the desire to solve this problem came from.

“From the very foundation of when we decided to embark upon this effort, we recognized it has to be actually deployable, actually scalable, and not a science project,” he said. “We weren’t interested in just seeing what we could prove. We wanted to make sure it actually was something that could get into the wild and serve a purpose”

Fluidstack, an AI infrastructure company, is being advised on the loan by the bank started by Palmer Luckey, an early supporter of Donald Trump

The Pentagon is in talks to lend roughly $5 billion to AI cloud-computing startup Fluidstack, according to people familiar with the matter, adding the U.S. military to the growing list of financiers of artificial-intelligence infrastructure.

The money would come from the Pentagon’s Office of Strategic Capital, which provides loans to companies working in areas deemed critical to U.S. national security. A $5 billion loan would be, by far, the office’s largest to date.

Fluidstack would use the loan to shore up the U.S. supply chain and manufacturing capacity for certain components related to data centers, rather than funding a new AI facility outright, one of the people said. It has been working on securing the loan for several months and is being advised on its application by Erebor Bank, a newly chartered national bank founded by entrepreneur Palmer Luckey that has focused on tech and defense clients, the people said.

Late last month, President Trump issued an executive order declaring a national emergency over the country’s reliance on foreign-made electrical equipment, which is used in AI data centers.

“Compute capacity is a strategic national asset,” Gary Wu, Fluidstack’s co-founder and chief executive, said late last year.

The deal hasn’t been finalized, and the terms, including the interest rate that Fluidstack would pay, could change.

“The Department does not comment on transactions that are either pending or in the middle of negotiations until the press release comes out,” a Pentagon spokesman said.

From public offerings of stocks and bonds to bank balance sheets and private-credit funds, the AI build-out is consuming every source of capital it can find. Lenders are expected to supply $4.1 trillion in financing related to AI infrastructure through 2030, with companies expected to “tap every single capital market to support their growth needs,” analysts at JPMorgan wrote in a June report.

Launched under the Biden administration in 2022, the OSC helps finance companies with national-security applications through loans and loan guarantees. President Trump and last year’s One Big Beautiful Bill Act expanded the program, allowing it to hire Wall Street veterans with big salaries and giving it the authority to provide more than $200 billion in financing over the coming years.

The OSC has previously lent money to companies in rare-earth supplies and drone making. Some federal financing deals have also included warrants or equity stakes, giving the government an ownership interest in the firms receiving support. The OSC is also involved in Trump’s plans to give the U.S. government a stake in Venezuela’s oil production.

Founded in London in 2017, Fluidstack last year pivoted to building AI infrastructure in the U.S. and has signed multiyear leases for data-center sites in states including New York, Indiana, Louisiana and Texas.

Fluidstack relocated its headquarters to New York. Its founders moved to Austin, Texas, and the firm’s 700 employees are also now based in the U.S.

Earlier this year, Fluidstack announced that it raised $830 million in a round led by Leopold Aschenbrenner’s Situational Awareness before that fund tumbled. More recently, the firm raised $1.5 billion in a round led by Jane Street at an $18 billion valuation, one of the people said.

Much of Fluidstack’s demand comes from a broad partnership it struck with Google and Anthropic to help the startup access Google’s tensor processing units. Google has backstopped Fluidstack’s data-center leases, helping bring down those projects’ borrowing costs.

Erebor caters to startups working in sectors like defense and industrial tech. Luckey, who co-founded defense company Anduril Industries, supported Trump during the 2016 election, one of the few Silicon Valley figures to do so publicly, and donated to Trump’s 2024 re-election campaign.

The bank has been in discussions with prospective investors to raise at least another $1 billion round in venture capital at an $8 billion valuation, people familiar with the matter said.

Erebor has told some of them that it sees a business opportunity in using its relationships and expertise to advise startups on how to access federal financing under different Trump administration programs, including the OSC, some of the people said.

Long a maker of low-end generic drugs, the country is now home to biotech firms that are innovating fast

SHANGHAI—Chinese biotech companies, rapidly evolving from manufacturers of low-end generic drugs to serious competitors to American pharma giants, are training their sights on one of the most lucrative medicines: GLP-1s.

Today Eli Lilly and Novo Nordisk, with drugs like Mounjaro and Ozempic, are the undisputed leaders of the $80-billion global GLP-1 market, which is dominated by U.S. sales.

But Chinese biotechs are pouring in, hoping to steal a slice of the weight-loss drug market. There are currently nearly 250 Chinese competitors in the research pipeline, according to data provider PharmCube. More than 20 of those treatments are already licensed overseas.

One example is Vincentage Pharma, a five-year-old company that is on track to become one of the first oral GLP-1s developed in China. In late 2024, the company, which has just 50 employees, secured a licensing deal to sell the drug, called VCT220, to overseas markets like the U.S.

“It’s like a 100-meter sprint: From the first step, we have gone very quickly,” said Feng Chen, Vincentage’s chief medical officer.

It caught the eye of Corxel Pharmaceuticals, the biotech firm that licensed VCT220, partly because the drug causes fewer side effects like nausea and vomiting than other GLP-1s.

“It’s a very tiny company,” said Sandy Mou, chief executive of Corxel, which completed a Phase 2 trial for the drug in the U.S. this summer and aims to seek Food and Drug Administration approval for the treatment. “They can still deliver so much.”

In the last two decades, China has transformed from a basic manufacturer of generics and precursor chemicals into a hotbed of scientific discovery, boosted by government support and low operating costs. Global pharmaceutical companies are now trawling for promising drugs from labs in biotech centers such as Shanghai and Suzhou. Last year, China licensed more than $130 billion in new drugs overseas, across cancer, weight-loss and immunology therapies, according to PharmCube.

“The growth has been remarkable,” said Todd Liao, a partner at Morgan Lewis in Shanghai who has worked on biotech licensing deals. “Deal volume has gone from essentially negligible prior to 2020 to well over $100 billion…with GLP-1 and obesity among the hottest areas within that.”

Global giants like AstraZeneca, Merck and Novo Nordisk have all landed licensing agreements worth billions to secure the next big Chinese obesity drug. No China-origin GLP-1s have received FDA approval so far, but several are undergoing clinical trials in the U.S. and elsewhere.

In April, Kailera Therapeutics, whose business is based on GLP-1 drugs from China-based Jiangsu Hengrui Pharmaceuticals, raised more than $700 million in a Nasdaq initial public offering. Kailera is currently conducting global clinical trials for the Hengrui-originated treatments.

The recent GLP-1 breakthroughs in China are largely incremental rather than revolutionary, experts say. Still, for patients in the U.S. seeking to shed pounds, the potential influx of new weight-loss therapies from China may push down drug prices and provide a wider array of treatment options.

Chinese biotechs are pushing overseas in part because the domestic market is far less lucrative. China has a large—and growing—obesity problem and counts nearly 150 million diabetic adults. But the slice of the population taking GLP-1s is small compared with the U.S., while drug pricing is far lower in China. The national drug reimbursement system also covers the medications only for diabetes, not obesity, forcing most patients to pay out of pocket.

“Making money can be very hard” in the Chinese biotech industry, said Leon Tang, founder of InScienceWeTrust BioAdvisory, which helps broker connections between Chinese and international biotech firms.

That pricing squeeze in China is also hitting Novo Nordisk, the Danish pharmaceutical giant that developed Ozempic and now sells GLP-1s in China for diabetes and obesity. Under pressure from Chinese upstarts, the company saw an 8% drop in net GLP-1 sales in China in 2025 and lowered the list price of two such products in that market earlier this year. The company is soon set to face a flood of generic alternatives, after the patent for semaglutide, the key component of Ozempic, expired in China earlier this year.

“The competition is quite intense,” said Hai Pan, the CEO of Sciwind Biosciences, a Hangzhou-based drugmaker developing GLP-1s. “But competition makes you better.”

London-based Verdiva Bio signed a deal with Sciwind to license three of its metabolic drugs last January for the global market, in a deal potentially worth $2.4 billion. Verdiva is conducting Phase 2 clinical trials in the U.S. for one of Sciwind’s oral therapies.

Just like in other cutting-edge fields in China, biotech companies benefit from state support. Beijing has poured billions into the industry as part of its self-sufficiency drive and sought to attract legions of scientists working overseas to do pioneering research in Chinese labs.

Speed and cost are also important factors. The clinical trial process in China is more efficient, biotech executives say. Trials are concentrated at a few large public hospitals with vast patient populations.

China’s biotech hubs also have a dense network of specialized contract research organizations, or CROs—which are critical for running clinical tests—allowing researchers to quickly transform scientific hypotheses into results. Executives at Sciwind, for instance, say that the company is located nearby its CRO, allowing them to check on research progress daily.

More than 30% of all global clinical trial launches were in China last year, up from 1% in 2009, according to the Cure Innovation Index. Phase 1 trials in China cost half what they do in the U.S.

“This is our competitive advantage in the global market: It is much more efficient and cheaper,” said Matthew Qiu, head of public affairs for Innogen Pharmaceutical, a Shanghai-based biotech firm that received Beijing’s approval for its weekly injected GLP-1 last year.

Biotech experts expect that GLP-1s from China will gradually upgrade from fast-follower drugs to groundbreaking treatments. Cancer therapies, for instance, are already trying to make this shift.

“You must develop step by step—you cannot jump to first-in-class drug when you just start,” Sciwind’s Pan said. “You should learn how to make generics first, then how to make ‘me too’ drugs, then ‘me better’ drugs, then innovative drugs. It’s a process.”

Selling treatments overseas is still difficult for small Chinese biotech firms, whose drugs need to go through U.S. clinical trials, FDA approval and commercialization in an increasingly crowded weight-loss market.

International licensing agreements make the process of marketing and distributing drugs easier and can help Chinese firms conduct clinical trials outside of China, which typically leads to smoother FDA approval.

But the biotech licensing boom is facing mounting U.S. regulatory scrutiny as politicians have grown more concerned about China’s biotech rise. U.S. Congress is considering a bill to screen outbound investment in the biotech industry, and lawmakers have also proposed banning any FDA consideration of Chinese clinical trials due to concerns about data reliability and transfer of scientific know-how to Beijing.

Platform known for prediction markets has seen perpetual futures surge in popularity

A risky type of trade that has exploded in popularity overseas this year is poised to become more mainstream in the U.S.

Prediction markets platform Kalshi is planning to seek regulatory approval to start offering the country’s first regulated perpetual futures tied to single stocks including Tesla, Apple and Nvidia, according to people familiar with the matter.

The contracts—known as perps—let traders bet around the clock and pile on leverage that amplifies their potential gains and losses, with the lack of expiration date further heightening the risks. The bets gained attention earlier this year when conflicts in the Middle East prompted a wave of trading in oil-linked perpetual futures on the decentralized crypto exchange Hyperliquid. The site, which is based in Singapore, isn’t technically supposed to be accessible to U.S. residents, but users find ways around its geofencing.

The push into single-stock perps is part of a race among Wall Street’s new-age powerhouses to serve investors with all manner of exotic trading opportunities, including bets critics say are risky and hard to understand. Kalshi plans to launch around 60 perps tied to popular exchange-traded funds and single stocks of companies with a market value of at least $100 billion, according to one of the people familiar with the matter.

Coinbase Global, the largest U.S. crypto exchange, has also filed paperwork that could help pave the way for the exchange to offer single-stock perps.

In late May, Kalshi and Coinbase received approval to launch the first true perpetual futures in the U.S.—but those were tied to cryptocurrencies. The contracts quickly rose in volume, and Kalshi on Thursday launched perpetual futures linked to gold and silver, also the first of their kind in the U.S.

Some exchange operators say that perpetual futures are riskier than garden-variety futures because of their so-called funding rates, or the payments that traders must make to keep the contract’s price aligned with the underlying asset’s price. And because perpetual futures allow leverage, skeptics warn they can intensify a downward spiral during periods of market stress.

Benjamin Schiffrin, director of securities policy for Better Markets, a group that pushes for tighter financial regulation, compares single-stock perpetual futures to leveraged single-stock exchange-traded funds, which have triggered massive losses for individual investors from the U.S. to South Korea.

“I think there’s just the potential for huge losses, especially amongst individual investors,” said Schiffrin. “You’re talking about 24/7 trading, so you’re going to have the ability to place these bets around the clock. And perpetual futures usually have much more leverage than you have with respect to single stock ETFs.”

Kalshi has said that the leverage it offers for perps is comparable to the extra exposure available for traditional futures, and lower than the leverage offered on offshore perps. The platform has also said that perps don’t have fees that can be incurred from rolling over contracts like traditional futures.

On Hyperliquid, the total trading volume (including leverage) in single-stock perpetual futures surged to $212 billion, up from $4 billion at the start of the year, according to data from Blockworks Research.

Since individual equity futures are legally categorized as security futures, companies like Kalshi that want to issue them need approval from both the Securities and Exchange Commission and Commodity Futures Trading Commission.

The trading of perps in the U.S. is already being contested.

CME Group in June sued the CFTC over its approval of perps on Kalshi in the U.S., alleging that the regulator violated U.S. law by classifying Kalshi’s perps as futures rather than swaps. Kalshi, Coinbase and the CFTC said that they view the lawsuit as an attack on market competition. The CFTC recently moved to dismiss the lawsuit.

Kris Marszalek, chief executive of Crypto.com, said single-stock perpetual futures are already impacting U.S. markets but trade overseas on unregulated venues.

“It makes sense for both regulators to come together and allow this to happen onshore in a more controlled environment where the consumers are protected,” he said in a recent interview.