TheChek News Flow

Maybe the pyramid structure of investment banks will start to look more like a diamond

AI technologies will change working practices across many white-collar professions, including investment banking. Tasks at the bottom of traditional career pyramids are likely to be automated first. Those towards the top will take much longer.

Many years ago, I started as a graduate trainee in the equity research department of a long-forgotten London stockbroker. My first jobs were to collect information, update models, format charts, prepare presentations and get the coffee. By doing these basics, I freed up my seniors to do more important tasks such as idea generation or meetings with companies and investors.

To climb from the bottom of this career pyramid, I had to prove competence at more advanced tasks — analyse information, set up models, create interesting charts, construct presentations and contribute to written research. I learnt these skills from colleagues further up the pyramid, more by osmosis than formal training.

I could even see a path to the top. To climb up there I needed to accumulate further analytical skills while learning how to originate ideas, present them to investors and build networks. This wasn’t easy. On the way up, I saw many talented colleagues fall by the wayside. Other white-collar careers such as consultancy, law, academia, even software coding, have similar pyramid structures.

My profession, sell-side equity research, peaked in the late 1990s bull market and has been fading ever since. Funding has been cut and headcounts reduced, but the job has changed remarkably little over my career. When I finally departed two years ago, I wasn’t doing anything much different to my very first boss. My juniors were doing a job similar to mine 34 years earlier. Sure the technology and access to data improved, allowing us to cover more stocks and markets with fewer staff, but the skills required to climb the pyramid were the same.

I was lucky to fall into a profession which changed little through my career. My hard-learnt skills did not become redundant overnight. But many think that artificial intelligence will change all that. Finally, here is a new technology that will disrupt the cosy white-collar professions. They could suffer the same fate as UK factory workers in the 1980s.

My default instinct is to be sceptical. So, having retired from equity research, I decided to see for myself. I’ve been working with a London-based start-up to identify opportunities to apply AI technology in the financial services industry. Apparently, I have useful “domain knowledge”.

There have been some wild claims about AI use cases in my old profession. Just ask ChatGPT to collect information, build company models, pick favourite stocks and write reports. Job done, no need for expensive analysts.

But that misunderstands the research career pyramid. The current technology is best suited to tasks which I performed in my beginner years. Information collection, model updates and presentation formatting can all be automated, even if getting the coffee cannot. However, as research tasks become increasingly complex and bespoke further up the pyramid, automation opportunities are more elusive. Right now, AI is better suited to being a research assistant than a research analyst. Or, to use the tech industry jargon, the scale opportunities are currently greater at the bottom of the pyramid than the top.

Of course, the AI will get better but it will have to prove itself before taking on more difficult tasks, just as I did all those years ago.

More focused work and AI models seem to be the way forward. One use case we have found is in content distribution. AI can allow customers to ask complex questions of a research department’s output, be it numbers or words — it’s almost like speaking to the actual analyst. This is a complex task, given the wide range of reports that analysts publish. For example, the shelf life of a two-page quarterly earnings preview is much shorter than an 80-pager published when an analyst initiates coverage of a stock. The generic LLMs struggle with these subtleties.

Given what I know about AI now, it’s tempting to say that I would never do my old job the way I used to. But it really depends on which old job. The new technology is most useful, and perhaps threatening, to those at the bottom of the pyramid. My old job at the top seems less vulnerable. AI will probably chip away at the traditional white-collar career path from the bottom upwards. Maybe the pyramid starts to look more like a diamond. But that raises an obvious question — where will the next leaders start their careers?

President Trump just made it a lot more expensive for companies to hire foreign workers through the H-1B program. The White House announced Friday that Trump signed a proclamation requiring employers to pay a hefty $100,000 fee for new H-1B visa applications, an enormous jump from the current $215 lottery registration fee.

H-1B visas allow U.S. companies to hire foreign workers in fields that typically require technical expertise like IT, engineering, mathematics, or medicine. The program is capped at 65,000 new visas annually, plus an additional 20,000 for foreign graduates with advanced degrees from U.S. universities. The visas are awarded through a lottery system and typically last three years, though holders can extend them or apply for green cards.

The administration’s new move is designed to crack down on what it calls widespread abuse of the program, which it blames for displacing American workers. According to the White House, the share of IT workers with H-1B visas has skyrocketed from 32% in 2003 to over 65% today, while unemployment among recent computer science graduates has hit 6.1%.

Silicon Valley will undoubtedly be up in arms over the initiative. The restrictions take aim at a program that helped create some of the region’s biggest success stories.

Elon Musk, Trump’s close ally for most of this year, initially worked in the U.S. on an H-1B after arriving as a student. In fact, Musk, taking issue with a perceived critic of the H-1B program in December on his platform X, tweeted to the individual that, “The reason I’m in America along with so many critical people who built SpaceX, Tesla and hundreds of other companies that made America strong is because of H1B. Take a big step back and F*** YOURSELF in the face. I will go to war on this issue the likes of which you cannot possibly comprehend.”

Mike Krieger, the Instagram co-founder turned Chief Product Officer at AI giant Anthropic, is one of those aforementioned success stories. The Brazilian-born Stanford grad worked at the early instant messaging platform Meebo on an H-1B visa.

Earlier this year, the National Venture Capital Association argued in a letter to the National Science Foundation that, “Raising the annual cap of H-1B visas issued each year to educated and highly skilled immigrants who work in jobs that require a substantial amount of technical and specialized training is fundamental to generating more successful immigrant-founded companies.”

The NVCA noted that while “H-1B visas are not ideal for immigrants who want to immediately found companies in the U.S., they are still critically important for the success of immigrant founded companies because they provide valuable work experience and widen the pipeline of potential immigrant startup founders.” (The H-1B’s employer-employee requirement makes it practically impossible for founders to obtain directly, forcing them to spend years tied to employers before getting green cards that let them launch their own startups. When Krieger wanted to co-found Instagram in 2010, transferring his visa took months, and he has said he almost abandoned the startup before it launched due to those complications.)

On Friday, tech leaders on X were already warning about talent fleeing to more welcoming countries. And Amazon, Google, and Microsoft have reportedly told employees with H-1B visas to avoid foreign travel and remain in the United states for now.

In the meantime, in its proclamation on Friday, the Trump administration went full bore on its criticism of the program, pointing to specific companies that approved thousands of H-1B workers while simultaneously laying off American employees. According to the White House fact sheet, one unnamed company received approval for 5,189 H-1B workers this fiscal year while cutting roughly 16,000 U.S. jobs.

The proclamation — which says it is partly rooted in an effort to “protect our national security” — includes wiggle room; case-by-case exemptions are possible if deemed in the national interest.

It also directs the Labor Secretary to revise wage requirements to prevent undercutting American salaries.

Another week, another review of giant AI-related financings. For this past week, the standout fundraiser was humanoid robot developer Figure, with its $1 billion Series C. Several other AI companies closed nine-figure rounds, and we also saw good-sized financings in biotech, manufacturing and reforestation.

1. Figure, $1B, robotics: San Jose, California-based Figure, the developer of general purpose humanoid robots, announced that it raised $1 billion in Series C financing. The round, led by Parkway Venture Capital, set a $39 billion post-money valuation for the company.

2. Groq, $750M, AI: Groq, a provider of cloud and on-premise AI compute centers that is also designing chips called LPUs, or language processing units, landed $750 million in a new round led by Disruptive. The financing sets a $6.9 billion post-money valuation for the Mountain View, California-based company.

3. Divergent, $290M, manufacturing technology: Torrance, California-based Divergent Technologies, developer of a digital manufacturing platform, closed a $290 million Series E financing at a $2.3 billion valuation. The round, led by Rochefort Asset Management, consisted of $250 million in equity and $40 million in debt financing.

4. Lila Sciences, $235M, AI: Lila Sciences, a developer of AI tools for use in science, locked up $235 million in Series A funding. Braidwell and Collective Global led the financing for the Cambridge, Massachusetts-based company.

5. (tied) Dyna Robotics, $120M, robotics: Redwood City, California-based Dyna Robotics, developer of a foundation model for robots to perform daily tasks, said it secured $120 million in a Series A round co-led by Robostrategy, CRV and First Round Capital.

5. (tied) Colossal Biosciences, $120M, de-extinction: Colossal Biosciences, the Austin, Texas-based startup known for its ambitious plans to de-extinct the mammoth and other species, said it raised $120 million from investors including filmmaker Peter Jackson. The company plans to use the money to help build a new avian research to further its work to bring back the dodo.

7. (tied) Invisible Technologies, $100M, AI: Invisible Technologies, an AI software platform for enterprises, raised $100 million in a growth round led by Vanara Capital. The San Francisco company said it reached $134 million in revenue last year.

7. (tied) Upscale AI, $100M, AI networking: Palo Alto, California-based Upscale AI, a developer of AI networking technology, emerged from stealth and announced it has raised over $100 million in seed funding co-led by Mayfield Fund and Maverick Silicon.

7. (tied) Ollin Biosciences, $100M, ophthalmology therapies: Ollin Biosciences, a clinical-stage biopharmaceutical company focused on acquiring and developing therapies for vision-threatening diseases, secured $100 million in initial funding. Arch Venture Partners, Mubadala and Monograph Capital led the financing for the Austin, Texas-based company.

10. Chestnut Carbon, $90M, forest restoration: New York-based Chestnut Carbon, which offers U.S. forest carbon offsets generated through reforestation and conservation projects, said it picked up $90 million in additional funds for its Series B financing. Canada Pension Plan Investment Board provided the funds, bringing total equity funding to date to $450 million.

Zak Dychtwald, founder of YCG Bridgeworks, on why every company needs a strategy to compete with China.

Zak Dychtwald, founder of the research and advisory firm YCG Bridgeworks, spends his time traveling the world to help companies understand China and the developing trends in his adopted country that will reshape global business.

Dychtwald grew up on the West Coast of the U.S. in a home of demographers and moved to Chengdu, China, after graduating from Columbia University. Initially he worked odd jobs such as tutoring and videogame translation, before diving more deeply into researching China’s equivalent of millennials. YCG stands for Young China Group.

He was on a work trip back in the U.S. when the Covid-19 pandemic hit China. Unable to get back, he eventually wound up in Mexico, getting a front-row seat to the rewiring of globalization as U.S. companies shifted production closer to home. As a sought-after advisor to industry executives in sectors ranging from travel to semiconductors, Dychtwald, now based in Shanghai, also has kept tabs on China’s growing cohort of entrepreneurs and the innovation feeding the U.S.-China strategic rivalry. The company is expanding its efforts in Southeast Asia, India, and Latin America, as well, a reflection of the continuing shifts in global trade.

Barron’s spoke with Dychtwald in early September about how Chinese consumers are adjusting to a prolonged economic downturn, why global companies should prepare for more Chinese competition, and how the government is keeping a lid on social unrest even as the country’s economic malaise continues.

An edited version of the conversation follows.

Barron’s: Why are young Chinese the key to understanding the country?

Zak Dychtwald: The whole boomers/Gen Z/millennial categorization is a U.S. and Western European–based construct for generations that doesn’t work for China because Chinese demographics change so rapidly. China generations are typically viewed on a 10-year basis, with savvy marketers taking even a narrower look.

People born after 1990 are the pivot generation. They were raised in multigeneration households. Like the parents of American baby boomers who grew up stashing money under the mattress through the Great Depression, the parents of China’s young adults clawed their way out of poverty after the Chinese famine and cultural revolution. China’s post-1990s cohort, however, was born into a growing economy. This is the first generation in modern China for whom “having stuff” wasn’t only possible but a statement of progress.

In the U.S., millennials and Gen Z make the trends, but the boomers move markets because the older generation has the money. In China, it is the young people who are moving consumer markets. It is a critical difference.

Yet there is double-digit unemployment among younger Chinese. There have been layoffs in the tech sector, and the economy’s growth rate is slowing. How have these developments affected this generation?

In 1990, gross domestic product per capita was $300, implying dire poverty. In their lifetime, young Chinese adults have seen a 33-fold increase. GDP per capita rose three times in the U.S., and six times in India in the same period.

This Chinese generation is conscious that such growth isn’t going to happen again. And they have now experienced their first recession [by China standards, as the economy’s annual growth rate has fallen from double to mid-single digits.] Consumer confidence is falling off a cliff, and there is underemployment alongside unemployment—that is, people taking worse jobs than those for which they are qualified. Yet, the government projects there are going to be 30 million unfilled factory jobs in 2025. Young people in China just don’t want to make the world’s shirts.

If they are unemployed and underemployed, how are they spending money and setting consumer trends now?

There was a crass attitude in China about what success looked like: Buy an apartment, buy a car, have a kid, then die. Even before the Covid-19 pandemic, younger people were questioning why they worked so hard and competed so feverishly, and what would make them happier.

In recent years, many of them have taken jobs in second- and third-tier cities and just gotten by, compromising on what they had believed was their dream life. Many are living off their parents in these less-expensive cities. This is the first generation for whom growth, progress, and more wealth aren’t guaranteed. People are integrating this new economic reality into their worldview and adjusting their spending habits and expectations.

But they are still a force because of their scale: 400 million people. Economic growth isn’t leaping as it did previously; things have slowed as much of China has already reached middle-class status. If Lululemon and Sam’s Club can thrive in China during an economic downturn, that means a swath of the population is spending at globally influential levels.

What are younger people spending on, beside yoga pants?

A quarter to a third of China’s population accounts for the Chinese consumer we think about. This consumer has three times the purchasing power of peers in India. These are demanding consumers, on whom new approaches to global retail, product design, and innovation are being tested. What this young generation wants—and the products made for them—will become what much of the world [eventually] buys.

Their parents were known as the “eat bitter generation”—eating difficult things for long periods in the expectation that the next generation could have a better life.

Younger Chinese want experiences, such as travel. There has been a good amount of downgrading to lower-priced goods, but high-value luxury—or high-quality goods with brand recognition that gives you a social bump—also still does well. China is still the largest growth market for most luxury companies. It is more challenging than in the past, as these consumers are more discerning and “foreign” no longer means “better,” but luxury companies may be more resilient than other global [consumer] companies.

Wellness is also a massive trend. Quality of food has become important, as well as supplements and exercise. Parks are full and people are engaged in more activities, often free, whether that is young people doing ballroom dancing or park Tai Chi.

How are Chinese companies dealing with tariffs and slowing economic growth?

Many companies in China have been preparing to live without the U.S. for a long time. Chinese companies excel in innovation and cost structure. When you ride in an electric vehicle in any price category in China, and then compare it to the limited EV options in the U.S. or classic autos in Europe that have been converted to EVs, there is no question the non-China products offer less quality and value.

There is an impulse among companies and policymakers to define success for a global company as having it arrive in your backyard. With China’s shifting or limited access to Western markets, many analysts risk missing Chinese companies’ success elsewhere. A lot of Chinese companies won’t stop just because they can’t enter or access Western markets. Plus, most of the world can’t afford [U.S.] products.

What is the risk to other companies from Chinese companies’ international expansion?

Chinese companies have to go global for growth, and now China’s incredibly cutthroat culture is being unleashed on the rest of the world.

We often confuse innovation with invention. BYD’s success isn’t due to inventing a new energy vehicle, but to innovating in the field—relentlessly, for a decade, and iterating with extraordinarily demanding customers. The company created something world-beating in quality, user experience, and value.

The easiest way to describe it is the difference between zero to one and one to 10. The U.S. thrives at zero to one, and China at one to 10. Every company in the world needs a China strategy, even if it doesn’t compete in China, because for the next decade it will either be competing with Chinese companies or, because of tariffs or other restrictions, with Chinese ideas.

BYD’s success and the dominance of China’s electric-vehicle industry are well known. What other sectors could see a similar transformation in China?

Chinese companies solve Chinese problems. There was brutal pollution in Chinese cities. New EVs were necessary for quality of life. For semiconductor chips and AI, you need power. China is energy poor, so [there was investment in] battery technology and alternative energy sources.

Agricultural technology is going to be massive. China has 20% of the world’s population and 10% of the world’s arable land. Life sciences, pharmaceuticals, and medical technology will also be enormous because China’s population over 65 years old will double from 2020 to 2050. If you make an affordable drug in China, why would someone in Mumbai, Mexico City, or Adelaide not take it? China solves its own problems first, then brings those solutions to the rest of the world.

What companies could be the next BYDs, in other industries?

Xiaomi is a great example of a company that makes technology that most of the world needs. It produces affordable smartphones, smart home devices, and connected electronics that bring high-end functionality to mass-market consumers. This is dependable, accessible tech with a blisteringly competitive value proposition for the everyman.

Shenzhen Mindray Bio-Medical Electronics and BeOne Medicines, which is focused on innovative cancer treatments, are making medtech and drugs under the price controls set in China to make healthcare accessible to a huge aging population. These companies are incentivized to solve China’s problems and are run through the product gauntlet in China.

Which countries benefit from companies shifting production away from China?

It is Mexico’s game to lose in the next decade. Chinese companies have started to invest in their former Mexican competitors to have access to the U.S. consumer. Vietnam is also going to be a winner.

Every manufacturer we have spoken with is trying to upgrade. If you go to industrial parks in Vietnam, the model they are working off of is China’s industrial-park model, which is based off Singapore’s industrial-park model. The Vietnamese are hiring Chinese managers at these factories who can pass on some of the best practices.

But we have also talked to American small businesses and manufacturers who are sticking with China [production], even with the tariffs. We often underestimate how good China has gotten at manufacturing skills and how much hard work has gone into the global management and ecosystem they have built. When you look at India, Mexico, or Vietnam, the gray area is still massive: Getting a cargo ship out of port might be as much about knowing the right guy or having a cousin who works in government as legal recourse. China used to operate that way, but those gray areas have largely closed over the past 10 to 15 years.

You have described a difficult backdrop for young Chinese, with the possibility that they don’t end up better than their parents’ generation. What is the risk of rising social unrest?

Investors focus on the top 25% [of the income distribution]—the people who can move companies such as Alibaba Group Holding and Tencent Holdings. The government legislates toward the middle and the bottom—the people who have the potential to cause political volatility.

The Chinese government is extraordinarily motivated to stay in power. The easiest way to do so is to keep the most people happy. I would bet on legislation aimed at pacifying a larger group of the Chinese population long before legislation stimulating that top quartile.

What would such legislation look like?

Health benefits? One easy example is [the property crackdown in the past few years] that caused a correction in the real estate market. The correction in real estate is one of the biggest signs of recognition [by the government] that if these tokens of success aren’t more accessible for average people, there is long-term political risk.

Thanks, Zak.

While Donald Trump’s U.S. slams the brakes on the electric-vehicle transition, the European Union is keeping its foot to the floor, probably.

An EU statute passed in 2022 will ban the sale of new gasoline-powered internal-combustion passenger cars by 2035.

EVs’ current share of the market is around 15%, says Guillaume Dejean, the industry analyst at Allianz Trade. Closing that gap over the next decade represents enormous risk or opportunity for an automotive sector that accounts for 7% of EU gross domestic product.

Captains of the industry see some of both. Their booths at the recent Munich Auto Show showed off new EV models from Volkswagen’s midmarket ID series to Bayerische Motoren Werke’s muscular iX3, aiming to outshine Chinese competitors. They decamped from there to a Sept. 12 Brussels sit-down with EU President Ursula von der Leyen, where they protested vehemently over the 2035 deadline. Eurocrats were forcing manufacturers onto “a crash course straight into a wall,” Mercedes-Benz CEO Ola Kallenius told German newspaper Handelsblatt.

“We’re getting kind of schizophrenic messaging from the auto makers,” says Germany-based industry analyst Matthias Schmidt.

The companies’ objections are easy to understand. Like their U.S. peers, they make much fatter profits, for now, on good old internal combustion vehicles. The term of art for EV sales is “margin dilutive,” Schmidt notes.

Brussels still won’t blink on phasing out the gas guzzlers, even with Friedrich Merz’s German government now pitching in on the industry’s side, thinks Michael Field, European equity strategist at Morningstar. “All the other countries that don’t manufacture autos are in favor of the deadline,” he says.

The EU is sweetening the pill substantially with tariffs on Chinese EVs that average 30%, by Field’s calculation. That has kept the Chinese share of budding battery electric vehicle sales in the bloc to around 10%.

“The tariffs have bought the European manufacturers some time,” says Eoin Drea, senior researcher at the Wilfried Martens Centre for European Studies.

U.S.-based Tesla, though still leading the European market, has sustained a 40% drop in sales, presumably due to founder Elon Musk’s political side hustles.

Bought time could be good enough to drive a rebound in European auto stocks that have been rocked by cratering sales in China, soaring power costs at home, and new U.S. tariffs.

The iShares Stoxx Europe 600 Automobiles & Parts exchange-traded fund has fallen nearly 30% from a peak 18 months ago. “European auto makers are gaining momentum,” Allianz’s Dejean says. “They will continue to dominate their home EV market.”

Much maligned Volkswagen could drive a “tipping point” starting next year as it rolls out new models driven by cheaper lithium iron phosphate batteries to challenge the Chinese on cost, Schmidt says. BMW and Mercedes maintain a strong grip on the premium end of the market. “Brand DNA will be a positive story,” Schmidt predicts.

For Morningstar’s Field, the whole industry looks cheap enough for a Buy rating. European auto makers are reaching a peak in their capital investments to shift to EVs, which should improve cash flow going forward, he adds. “Stellantis’ stock jumped 10% [on Sept. 11] just because the CEO shared some thoughts on what he might do,” he notes.

EU residents bought nearly 11 million cars last year. The epic transformation of this market to all-electric looked recently like it might kill the continent’s iconic industry. Now it has a shot at rebirth, too.

Andrew Left tries golf, TV appearances to get noticed

Andrew Left, facing securities fraud charges, hopes for leniency from the Trump administration.

Left’s legal troubles stem from allegations he manipulated stocks using his influence, a claim he denies.

Left, who has recently praised Trump’s economic policies, seeks a meeting with the Justice Department to dismiss the case.

Going to prison is a real risk for Andrew Left. Every day he thinks the Trump administration wouldn’t want him to be there.

The short seller, known for his feisty Citron Research reports and trades including shorting GameStop during the January 2021 meme-stock trading frenzy, is accused by prosecutors of exploiting his fame to manipulate stocks for quick profits. He is looking for a way to sway the administration to drop the criminal case prosecutors filed against him last year. If he is convicted on the 18 counts of securities fraud he faces, a court could sentence him to years behind bars. He has pleaded not guilty.

Left’s hopes have been stoked by President Trump’s role as an absolver of last resort for defendants and convicts complaining they were unfairly prosecuted, issuing a number of white-collar pardons and commutations.

Trump’s Justice Department has dropped marquee cases against New York City Mayor Eric Adams, a former Nebraska congressman accused of lying to the FBI, and a California restaurant-chain executive, Andrew Wiederhorn, charged with fraud and tax evasion. The president pardoned Nikola founder Trevor Milton, who was convicted of fraud in federal court for what prosecutors said were his lies to investors about his zero-emissions trucks. The Securities and Exchange Commission dropped its own civil fraud case against Milton last week. Adams, Wiederhorn and Milton have all maintained their innocence.

How to obtain this kind of mercy is something of a mystery, even to wealthy people like Left, who is based in Boca Raton, Fla. He and his lawyers say they will press the Justice Department in a meeting scheduled for this month to drop the case.

The Justice Department declined to comment.

Left had his picture taken with Trump in March at a member-guest golf tournament at Trump International Golf Club that Trump hosted (and won). He’s toured a condominium in Trump Tower in Manhattan, near where his son lives. Left, who says he avoided politics for years while living in California, has praised Trump’s economic policies in recent appearances on Fox Business.

“Everybody wants a meeting. Even the guilty guys want a meeting,” Left said in a recent interview at the Waldorf Astoria in Manhattan, where he was staying while apartment hunting. “‘Everybody thinks, ‘oh, Donald Trump doesn’t mind if I cheated on my taxes.’”

Left, 55 years old, grew up in south Florida, where his stepfather worked at a deli and his mother was an elementary-school secretary. His first job after graduating from Northeastern University and returning home to Florida was reading from a script to pitch silver and other investments for Universal Commodity Corp. In the mid-1990s, regulators sanctioned the company for fraud.

Left re-established himself in southern California in the late 1990s. Left liked that he could golf in the afternoons after the market closed. In the daytime, he flipped the newly issued stock of shaky companies promoted by small Wall Street brokerages. A friend urged him to start shorting, or betting against, the shares, since they were bound to fall in price.

Left’s bearish bets against China Evergrande in 2012 and Valeant Pharmaceuticals in 2015 made him a star of short sellers. Evergrande eventually defaulted on its debt and the company collapsed two years ago. Valeant paid a $45 million regulatory fine without admitting or denying fault, after Left drew attention to its use of an undisclosed mail-order pharmacy that helped boost its sales.

“There was a time when there was no one else in the market who could move a stock like he did,” said Nate Koppikar, a portfolio manager at hedge fund Orso Partners whose strategy focuses on shorting stocks.

Left’s troubles began in 2018, when the Justice Department and SEC launched a sweeping investigation of activist short sellers. Federal Bureau of Investigation agents searched Left’s Beverly Hills, Calif., house in early 2021, just after he had lost $20 million betting against videogame retailer GameStop. Left and others who wagered against GameStop were caught in a short squeeze driven by small traders who rallied online and bought up shares to push the stock higher in what became the poster child of the meme-stock craze.

Small traders vilified Left and other shorts. Some texted threats to his daughter on Instagram and ordered dozens of pizzas delivered to his house after midnight, he says. The ordeal was such a looking-glass experience—Left says he realized he was now seen as the “establishment,” not the contrarian outsider—that he announced in January 2021 that he was done with short activism.

After several years of investigation, prosecutors sent Left a second subpoena in June 2022 that asked for information on 95 different trades. Many were widely held stocks, such as Meta and Nvidia, that are generally considered too big and liquid to manipulate and which Left didn’t short. Instead, he had said they were worth buying.

After learning he had been indicted, Left told his wife there was good news and bad news, he said. The bad: prosecutors charged him with fraud. “What’s the good news?” his wife asked.

It wasn’t a short-and-distort case that accused him of lying about companies, he said. “I am being indicted for telling people to buy Nvidia, Facebook, Tesla, General Electric and to short American Airlines,” he said. “These are the stocks they chose.”

But prosecutors say his playbook was making bold, insincere statements about stocks and hoping his words would move prices in a direction that rewarded his short-term bets. They charged him with securities fraud and lying to federal investigators about coordinating with hedge funds on trades.

Some lawyers say the fraud charges against Left are unusual because they don’t allege he put out false information. Instead, prosecutors allege his tactic of commenting on stocks and then quickly trading was an illegal scheme. He didn’t give the whole story about his trading plans, making his commentary misleading, according to prosecutors.

“That is a murky area of law,” said Adam Pritchard, a law professor at the University of Michigan who worked at the SEC earlier in his career. “It matters that he is saying these things on social media, and any sensible person discounts significantly anything they see on social media.”

A federal judge declined Left’s request this summer for early dismissal of the case. Left has hired a new defense lawyer, Eric Rosen, who won a surprise dismissal last year in a case that targeted traders who promoted stocks on the Discord social-media platform and sold into the froth they created.

In early September, Left announced he had invested in Loan Depot, a lender and mortgage servicer, saying Trump’s push to lower interest rates would help to more than double its stock price.

“Trump will win the ‘war on housing’,” Left wrote on X. “LoanDepot is the top lender set to benefit—and the market is asleep.”

The shares have risen 60% since. Left has also made appearances on Fox Business, where he praised Trump’s approach to tariffs. “Trump knew what he was doing and people around Trump knew what they were doing and they would properly adjust,” he said. To take advantage, Left shorted an index of stock-market volatility, he said.

The only time Left puts down his phone and doesn’t look at his brokerage account is when he plays golf.

In March, a close friend invited him to play in a member-guest tournament at Trump International Golf Club in West Palm Beach, Left says. Trump’s team included Finland President Alexander Stubb and former pro Gary Player. They won.

Left says he was struck by how Trump mingled with people and got up to fetch his own hamburger from a buffet. Trump took a photo afterward with Left and his teammates, who finished second. Left flashed a thumbs-up sign like Trump.

Trump later shook Left’s hand and gave him a gift certificate to the pro shop for winning a contest where players compete to hit a ball closest to a pin on a par-three hole. There was no time to talk business, he said.

“What am I supposed to do, say ‘hey Mr. President, can I talk to you about my case?’” he said.

Left says he may hire a lobbyist to further his appeal within the administration. The odds of prosecutors dismissing their case now are long. Under Trump, the Justice Department unit that charged Left hasn’t dropped a case.

“It’s kind of crazy that Andrew is getting criminally charged for doing something that everybody does,” said Edwin Dorsey, who writes a newsletter, The Bear Cave, focused on the world of short sellers.

Building Nuclear Power in the U.S. Is Tough. NASA Wants to Do It on the Moon.

So do China and Russia. Here’s what you need to know about America’s new lunar-energy race

Call it a nuclear moonshot: The U.S. aims to deliver a reactor to the lunar surface and beat a push by China and Russia to do the same.

In August, NASA Acting Administrator Sean Duffy directed the agency to fast-track an effort to land a reactor on the moon by late 2029. The agency wants a 100-kilowatt system, about enough to power a small neighborhood—modest for Earth but unprecedented for space.

That gives the National Aeronautics and Space Administration a tight deadline to turn a wildly complex idea into something real. The agency expects to lean on U.S. industry to design a reactor, get it to the moon and operate it. But any companies that sign on will face steep engineering hurdles and financial risks.

All the while, NASA and its contractors will be racing against China and Russia. Those countries have been exploring the idea of jointly deploying nuclear energy on the moon by the middle of the 2030s, possibly using the power for a lunar base China is developing. NASA leaders worry a rival could potentially use its own station to establish a keep-out zone on the lunar surface, limiting U.S. efforts to build up its own presence and gaining a strategic advantage in space.

Here is what you need to know about U.S. plans to get a reactor to the moon.

Why do it?

The Apollo program successfully delivered astronauts to the moon a half-dozen times. What it didn’t do was create a lasting U.S. presence there.

A permanent nuclear reactor could open up new possibilities—including a science-fiction future where moon outposts spark new scientific and economic activities around research, mining, tourism and more. A nuclear reactor offers consistency that solar power with batteries can’t manage—given the two-week lunar night—and would generate about a decade of electricity.

Bhavya Lal, a former high-ranking technology official at NASA, says fission energy would allow government and private companies to start developing permanent settlements.

“It isn’t about doing the same missions better. It’s about doing fundamentally new missions,” Lal says.

Steven Sinacore, NASA’s program executive for fission surface power, says that nuclear energy is a crucial technology for establishing continuous U.S. operations on the lunar surface and beyond. “We want to go to the moon. We want to go to Mars,” he says. “This is the enabling technology.”

How would it work?

It is difficult to build and operate nuclear-power systems on Earth. Making one work on the moon won’t be any easier.

“I think there are a lot of folks that think we can just put lander legs on a terrestrial micro reactor and fly it to the moon and we’re good to go. And that couldn’t be further from the truth,” says Bill Pratt, director of in-space infrastructure at Lockheed Martin.

One challenge is cooling. On Earth, reactors are usually located near bodies of water for cooling the reactor core, and they can dissipate heat in the atmosphere. But there is no water or air in space. Instead, large radiator panels must be built to shed heat, adding to the overall mass of the project. Heavy radiation shielding, too, would be needed to protect the lunar environment and astronauts.

How would a reactor get to the moon?

A powerful rocket will be needed to get the reactor into space in the first place. Elon Musk’s SpaceX, Jeff Bezos’ Blue Origin or United Launch Alliance, owned by Boeing and Lockheed Martin, could potentially handle the flight. The project will also need a landing vehicle capable of taking the reactor to the lunar surface, which presents many complications: Since the moon pretty much doesn’t have an atmosphere, a landing vehicle will need a propulsion system to slow its descent, stores of fuel for that operation and a sophisticated guidance system to pull off a soft landing.

NASA’s directive assumes the existence of a lunar lander capable of delivering up to 15 metric tons to the surface. That is a huge amount. By contrast, Texas-based Firefly Aerospace landed about one-tenth of a metric ton on the moon earlier this year.

Can a reactor be launched safely?

Before a rocket goes up, regulators will want to know what would happen in the event of an explosion, says James Walker, chief executive at reactor developer Nano Nuclear Energy. “How far would fuel be dispersed? How would you pick it up?” he asks.

John Kennedy, director of emerging technologies at reactor developer X-energy, says that a lunar reactor wouldn’t be operating at takeoff, and control mechanisms would be locked to prevent the fission process from starting. “You literally can’t turn the reactor on until you’ve landed on the moon’s surface,” he says.

Meanwhile, some reactors under development would use coated pebbles of uranium—which could be helpful in case of an accident. The coating could help the pebbles stay intact and help with containment, says Walker.

Once on the moon, Walker says, “your risks kind of diminish.”

Are there regulations to cover lunar reactors?

On Earth, there are norms for safety zones around reactors. Those don’t exist on the moon, and establishing zones would have technical and geopolitical implications, says Lal, the former NASA official.

“Should the safety zone be half a kilometer? Ten kilometers? More?” she says. “Are we going to tell China not to put their reactor close to ours, and how will that happen?”

The 1967 Outer Space Treaty, signed by the U.S., China, Russia and other countries, provides a framework for how nations interact in space. It is light on many specifics, meaning countries will need to discuss details, according to Christopher Johnson, director of legal affairs and space law at the Secure World Foundation.

Meanwhile, in 2019 President Trump signed a memorandum that outlines how a reactor in space would be licensed—but the process hasn’t been tried yet for a fission launch, says Lal, who helped write the memo.

Then there’s the matter of liability. The domestic nuclear-power industry relies on the Price-Anderson Act, which limits liability as a way to enable investment. Something similar will be needed in space, or companies won’t be able to get private insurance to cover their efforts.

Can nuclear scale up in space?

Nuclear energy in space isn’t new, but efforts so far have been small scale, such as powering space probes. In 2018, as a step toward a lunar or Mars reactor, NASA and the Energy Department demonstrated a reactor that generated enough power for a small household for a little more than a day. A fully realized lunar reactor would need to generate 100 times as much. “There’s going to be some work done,” says Koroush Shirvan, an MIT professor who specializes in advanced nuclear reactor technology. “That power system is going to be a very massive part of the whole technology.”

European airports hit by cyber attack on check-in provider

Heathrow, Brussels and Berlin suffer delays and cancellations after Collins Aerospace is targeted

Airports including London’s Heathrow, Berlin Brandenburg and Brussels reported flight delays and cancellations on Saturday after a cyber attack on the company providing check-in technology. 

Heathrow said that Collins Aerospace, “which provides check-in and boarding systems for several airlines across multiple airports globally, is experiencing a technical issue that may cause delays for departing passengers”. 

The airport, which shut down for 24 hours in March because of a fire at an electricity substation, advised passengers to check their flights before travelling to Heathrow, and not to arrive more than three hours before their scheduled departure for long-haul flights or two hours for shorter trips.

Around a dozen inbound flights were cancelled, and a similar number of outbound flights were also affected, a spokesperson said.

British Airways, the largest airline at Heathrow, said it was not affected because it had a backup system that allowed it to continue processing customers. Gatwick and Luton airports were not affected. 

Brussels Airport said that 17 flights had been cancelled and four diverted after it was forced to manually enter details for passengers who checked in at the airport, or had hold luggage. It asked airlines to cancel half their scheduled departure flights on Sunday.

“There was a cyber attack on Friday night, 19 September against the service provider for the check-in and boarding systems affecting several European airports including Brussels Airport,” the airport said. “This means that at the moment only manual check-in and boarding is possible.”

Berlin Airport said it was experiencing longer wait times for customers checking in.

RTX, the US aerospace and defence company that owns Collins, said: “We have become aware of a cyber-related disruption to our MUSE software in select airports. We are actively working to resolve the issue and restore full functionality to our customers as quickly as possible.” 

It added: “The impact is limited to electronic customer check-in and baggage drop and can be mitigated with manual check-in operations.”

The breach appears to be the second such attack on the group. 

In 2023, the names, addresses and contact details of pilots and other staff affiliated with Collins’ partner companies were allegedly leaked following an attack claimed by ransomware group BianLian. 

Can Germany afford its €1.35tn welfare state?

Chancellor Friedrich Merz wants benefit cuts, but his coalition partners are pushing back

Germany’s conservative Chancellor Friedrich Merz has set up a battle with his centre-left coalition partners by declaring that Europe’s largest economy can no longer afford its generous welfare state.

“The welfare state as we know it today can no longer be financed by our economy — and that is why we have to change it,” he told members of his Christian Democrats (CDU) party in August.

A reform of Germany’s social security, which is part of the coalition’s agreement, comes as the country is launching a €1tn debt-funded spending programme to modernise its armed forces and repair its infrastructure.

Attempts elsewhere in Europe to slash welfare spending have proved politically fraught. In the UK, Sir Keir Starmer was forced into a climbdown over plans to rein in Britain’s spiralling benefits bill. In France, fierce opposition to pensions reform contributed to the downfall of prime minister François Bayrou’s government.

Merz’s own labour minister Bärbel Bas, who is also the co-chair of his junior coalition partners, the Social Democrats (SPD), called the chancellor’s declaration “bullshit”.

While Germany is no outlier in Europe, its social spending has crept up in recent years and its demographic time bomb threatens to undermine the welfare state’s funding — just as its economy struggles to emerge from its longest postwar phase of stagnation.

Sprawling and rising

Germany’s welfare state dates back to the days of Chancellor Otto von Bismarck, who first introduced statutory health and pension insurance from 1881 to counter socialism.

Enshrined in the country’s constitution since 1949, the welfare state has grown into one of the world’s largest, accounting for 31.2 per cent of the country’s GDP — the highest level on record when excluding the Covid-19 pandemic, when the data was distorted by emergency government spending.

Healthcare and old-age care costs have risen at faster rates than overall inflation. At the same time, the country has been stuck in economic stagnation for more than three years.

“The system is creaking,” Germany’s President Frank-Walter Steinmeier said in a landmark speech this week in which he argued that the country’s welfare state was a “treasure” but needed to be made “fit for the future.”

Together with civil servants’ pensions, the pay-as-you-go pension insurance and the health insurance account for three quarters of total welfare expenditures of €1.35tn, according to data from the German labour ministry.

The demographic time bomb

Germany’s pension system has survived two world wars, hyperinflation and several currency reforms. But the rapidly ageing workforce will pose a structural challenge as 16.5mn baby boomers — people born between 1954 and 1969 — will retire by 2036 while only 12.5mn young workers will join the workforce, according to the Cologne Institute for Economic Research.

As retirees live longer, the financial pressure on the pension system will grow over time.

On top of €306bn of contributions from employers and workers, the government in 2024 paid another €118bn to fund deficits in the system, or about a quarter of the total federal budget.

This share will double to about 50 per cent over the next two decades, according to Axel Börsch-Supan, head of the Munich Center for the Economics of Ageing at the Max Planck Institute.

“The big elephant in the room is the pension system,” said Jens Südekum, an economist advising SPD finance minister Lars Klingbeil.

Merz’s flagship measure so far on pensions is a plan to incentivise those willing to work beyond retirement age by exempting the first €2,000 they earn monthly — on top of their regular pensions — from income tax.

But he has refrained from being specific about how he intends to overhaul the overall pensions system, and has kept incentives for early retirement untouched.

Critics point out that Merz’s own camp is simultaneously pushing for measures that would worsen the problem. The coalition has pledged to maintain the current pension level at 48 per cent until 2031. And the agreement includes a plan championed by the CDU’s conservative Bavarian sister party CSU to lift pension perks for non-working mothers from 2027 — at an additional cost of €5bn per year.

It will be down to a commission, due to be set up early next year, to make reform proposals.

Persistent poverty

Germany’s welfare system has had limited success in reducing the average poverty rate.

According to think-tank WSI, which is funded by Germany’s trade unions, this is mainly driven by below-inflation increases in minimum payments to the non-working poor and a falling level in state pensions. The average state pension, which in 2005 was equivalent to 52.6 per cent of the average wage, has fallen to 48 per cent.

Reform of the Bürgergeld

Merz has vowed to overhaul the so-called Bürgergeld, a basic income for those outside the labour force. While it accounts for just 3.5 per cent of Germany’s total welfare payments, it has become a lightning rod for the far right.

Introduced by the previous coalition in 2023, it is claimed by 5.5mn people — three-quarters of them deemed able to work — and costs €47bn, according to government data.

Close to 50 per cent of the payments go to foreigners, including €6.3bn to Ukrainian refugees. Critics say the system is too generous and disincentivises work, especially after a 25 per cent increase in monthly allowances during the high-inflation years of 2023-24.

For instance, a non-working couple with two children aged five and 14 can claim €2,754 a month, only €660 less than the amount they would earn if one of the parents took on a full-time job at the minimum wage, according to WSI calculations.

A commission set up by the labour ministry is due to outline reform proposals by the end of the year, but Merz has already said he wants to extract €5bn in annual savings from the Bürgergeld.

Many say such a target is unrealistic. “The German constitution only provides a limited scope to cut welfare benefits,” says Eckhard Janeba, professor at Mannheim university and chair of the economy ministry’s academic advisory council.

Burden on labour costs and competitiveness

Any increase in welfare costs automatically leads to an increase in non-wage labour costs for employers: Under German law, employers are obliged to cover half of their employees’ insurance contributions.

Since the end of the pandemic, non-wage labour costs have risen at a faster rate than total wages, eating into companies’ profits and reducing the room for wage increases.

Total social security contributions, which for decades hovered below 40 per cent of total salaries, have increased to 42.5 per cent this year. They are expected to rise to close to 50 per cent over the next decade, according to IGES, a Berlin-based think-tank.

But Germany’s bigger structural problem remains its shrinking workforce.

Both encouraging people to work longer and less generous welfare benefits would “improve labour supply, which is hit by the baby boomers’ retirement and low fertility rates in Germany,” Janeba says.

Merz’s decision to set up commissions to make proposals raises concerns that he will delay action, Janeba warns: “The intentions are all there, but we know from past experiences that reforms implemented early in an election cycle are more likely to be successful.”

‘Stargate of China’ plan emerges to challenge US as AI superpower

Beijing increases oversight of its data centres to make better use of limited computing resources

On a 760-acre island on the Yangtze River, rice fields are being turned into a series of huge server farms as part of China’s effort to consolidate its position as an artificial intelligence superpower.

The building work in the farming city of Wuhu is an effort to “build the Stargate of China”, said an executive at a supplier for one of the projects, referring to the $500bn plan by OpenAI, Oracle and SoftBank to build the world’s largest AI data centre in Texas.

The Wuhu “mega-cluster” will not match the scale of the US project. It forms only one part of Beijing’s greater oversight of the country’s fragmented data centres to make them better equipped to handle booming AI demand from consumers.

This move comes in response to the strong lead held by the US over access to AI computing capacity, with research group Epoch AI estimating America has around three-quarters of the global computing power, compared to 15 per cent in China.

In March, Beijing unveiled a plan to have existing data centres in remote regions in the west focus on training large language models.

Meanwhile, new server farms are being built in locations closer to key population centres. These will focus on “inference” — the process through which AI tools such as chatbots generate responses, with closer physical proximity to users designed to allow for faster AI applications.

“China is starting to triage scarce compute for maximum economic output,” Ryan Fedasiuk, a fellow at the American Enterprise Institute and former state department adviser on China. “Beijing is now planning data centre infrastructure with this in mind.”

One example is Wuhu’s so-called “Data Island” — home to four new AI data centres, operated by Huawei, China Telecom, China Unicom and China Mobile.

Wuhu’s data centres will cater to the wealthy Yangtze River Delta cities of Shanghai, Hangzhou, Nanjing and Suzhou, while in the north, Ulanqab in Inner Mongolia will feed Beijing and Tianjin.

In the south, Guizhou will supply Guangzhou, and the central city of Qingyang in Gansu will serve Chengdu and Chongqing.

So far, 15 companies have built data centres across the city, according to a local government notice, totalling an investment of Rmb270bn ($37bn).

An executive at a state-owned cloud operator with a Wuhu project said the local government is offering subsidies that cover up to 30 per cent of AI chip procurement costs — more generous than in other regions.

The greater co-ordination is designed to help offset China’s disadvantages to its geopolitical rival. US export controls have also prevented Chinese groups from accessing the best processors and hardware built by Nvidia, the leading AI chipmaker.

Domestic chipmakers such as Huawei and Cambricon have struggled to fill the void in part due to limited manufacturing capacity in China. The US has also prohibited TSMC and Samsung from making advanced AI chips for Chinese clients.

By contrast, American groups such as Meta, Google, and X.ai are racing ahead to deploy tens of thousands of Nvidia’s latest chips. The US Stargate project alone is planning complexes that can house as many as 400,000 processors.

Chinese AI data centres have had to depend on less powerful processors or cobble together advanced hardware from the black market.

However, a network of intermediaries has sprung up across China to secure Nvidia GPUs that are banned for export to China, said multiple people familiar with the dealmaking structure.

One such supplier is the Wuhu-based Gate of the Era, which the FT previously reported has secured large quantities of Nvidia data centre servers that deploy its advanced Blackwell chips which are banned for export to China. The company declined to comment.

Nvidia said: “Trying to cobble together data centres from smuggled products is a non-starter, both technically and economically. Data centres are massive and complex systems, making smuggling extremely difficult and risky, and we do not provide any support or repairs for restricted products.”

China is also seeking to make better use of existing AI processors lying unused in faraway facilities in remote regions.

A boom in construction from 2022 onwards concentrated these facilities in energy-rich but distant provinces such as Gansu and Inner Mongolia.

But a lack of technical expertise and customer demand meant they were underutilised with valuable processors left idle even as demand soared elsewhere.

In many cases, procurement of these AI chips has been financed by local governments, which are unwilling to part with the assets, as they want to boost local GDP.

Instead of physically moving the servers, Edison Lee, an analyst at Jefferies, said, “A technical solution has to be found. That is connecting data centres.”

Beijing has directed the use of networking technology from China Telecom and Huawei to link disparate processors scattered across multiple sites into a centralised computing cluster.

The technology is already used by cloud service providers to connect multiple sites together to create redundancy in case one goes offline.

Chinese telecoms giants, including China Telecom and Huawei, are using the same combination of transponders, switches, routers and software solutions to move computing power from data centres in the west to the east.

However, there are issues with this approach. “Using multiple smaller and older data centres is less efficient than using one bigger and modern one,” said Edward Galvin, founder of data centre research consultancy DC Byte. “It’s about economies of scale.”

Huawei is working on a solution to solve this efficiency problem. The tech conglomerate is leveraging its dual expertise in telecoms and AI hardware to pioneer a new networking technology called UB-Mesh, which claims to double the training efficiency of LLMs across multiple computing clusters by better allocating tasks across the network.

Lee said the move to network the chips together is “an important method of consolidation in this fragmented market”.