TheChek News Flow

Microsoft has signed a 20-year deal to exclusively access 835 megawatts of energy from a nuclear plant.

Microsoft just signed a deal to revive the shuttered Three Mile Island nuclear power plant. If approved by regulators, the software maker would have exclusive rights to 100 percent of the output for its AI data center needs.

Constellation, the owner of the Three Mile Island plant, announced a power purchase agreement with Microsoft earlier today, which should see the site coming back online in 2028, assuming regulators approve it.

The reactor that Microsoft plans to source its energy from was retired in 2019 for economic reasons and is located next to a unit that was shut down in 1979 after the worst US nuclear accident in history. The plant that Constellation plans to reopen can generate 837 megawatts of energy, enough to power more than 800,000 homes — demonstrating the huge amount of power needed for data centers and Microsoft’s AI ambitions.

Microsoft has agreed to purchase power from the plant — which will be renamed to the Crane Clean Energy Center to honor the late Chris Crane, former CEO of Exelon — for 20 years in a first-of-its-kind deal for the software giant.

Microsoft’s own greenhouse gas emissions are growing with its focus on AI, putting its ambitious climate goals at risk. Bloomberg reports that this nuclear plant would help Microsoft’s plans to run its data centers on clean energy by 2025 and power data center expansions in Chicago, Virginia, Pennsylvania, and Ohio.

“This agreement is a major milestone in Microsoft’s efforts to help decarbonize the grid in support of our commitment to become carbon negative,” says Bobby Hollis, vice president of energy at Microsoft. “Microsoft continues to collaborate with energy providers to develop carbon-free energy sources to help meet the grids’ capacity and reliability needs.”

Microsoft has been betting on next-generation nuclear reactors to power its data center and AI plans recently, looking for someone who could roll out a plan for small modular reactors (SMR) last year. Microsoft cofounder Bill Gates is also a “big believer that nuclear energy can help us solve the climate problem.”

Constellation will invest $1.6 billion to revive the plant, and the company will need approval from the Nuclear Regulatory Commission to bring the site back online, alongside permits from state and local agencies. Constellation is also pursuing a license renewal to extend plant operations until at least 2054.

Lenders charged more for loans but kept interest payments to savers down, FT analysis finds

US banks made a $1tn windfall from the Federal Reserve’s two-and-a-half-year era of high interest rates, an analysis of official data by the Financial Times has found.

Lenders got higher yields for their deposits at the Fed but kept rates lower for many savers, the review of Federal Deposit Insurance Corporation data showed. The boost to the US’s more than 4,000 banks has helped pad out profit margins.

While rates on some savings accounts were raised in line with the Fed’s target of more than 5 per cent, the vast majority of depositors, especially those at the largest banks, such as JPMorgan Chase and Bank of America, got far less.

At the end of the second quarter, the average US bank was paying its depositors interest at the annual rate of just 2.2 per cent, according to regulatory data that includes accounts that do not pay interest at all. This is higher than the 0.2 per cent they paid two years ago but far lower than the Fed’s 5.5 per cent overnight rate that the banks themselves can get.

At JPMorgan and Bank of America, annual deposit costs were 1.5 per cent and 1.7 per cent, respectively, according to this data.

Those lower payments to depositors generated $1.1tn in excess interest revenue for the banks, or about half of the total dollars banks brought in during that time, according to the FT’s calculations.

This is in sharp contrast to Europe, where some governments imposed windfall taxes on banks which benefited from higher interest rates.

The Fed tightened its main policy rate this week, cutting by half a percentage point. Some US banks sought to pass the cuts on to depositors as quickly as possible, a move that would shore up their margins.

Hours before the Fed rate cut on Wednesday, Citi told its employees at its private bank, whose wealthy clients typically receive preferential rates, that if the US central bank were to cut rates by half a percentage point the bank would do the same to its rate on accounts paying 5 per cent or more, according to a person familiar with the matter.

At JPMorgan, bankers have been told that clients with $10mn in cash or above would see their savings rates cut by 50bp and future cuts would move in lockstep with the Fed’s actions, people familiar with the matter said.

Because of the Fed’s rate cut, banks will “certainly” have “the ability to reduce deposit costs”, said Chris McGratty, head of US bank research at KBW. “The degree of aggressiveness will, I think, vary bank to bank.”

JPMorgan said the bank aimed to ensure a fair and competitive rate. Citi declined to comment. Bank of America declined to comment.

A report earlier this year from the Risk Management Association compared banks to petrol stations, which are typically quick to raise prices and slow to cut them. Banks, by contrast, are slow to raise the rates they offer on deposits and savings accounts but quick to cut them.

When the Fed began to tighten monetary policy in March 2022 many analysts predicted that competition from new financial technology companies and the growing ease with which consumers can move cash would force banks to dole out a greater share of the higher rates to their depositors.

But the FT’s calculations show that they were able to hold on to much of the benefit — although slightly less than in previous Fed tightening cycles.

The failure of Silicon Valley Bank and others in early 2023 forced many mid-sized and smaller banks to raise their rates in order to keep depositors from fleeing. Larger banks saw an influx of cash during the flight for safety, allowing them to delay the need to match higher rates elsewhere.

Overall US banks captured about two-thirds of the benefit of the Fed’s higher interest rates from March 2022 until the middle of this year, according to the FT’s calculations based on the latest data available. They paid depositors nearly $600bn in interest.

The last time the Fed raised interest rates, from early 2016 to until early 2019, US banks captured 77 per cent of the benefit.

Although the Fed has now begun to loosen monetary policy, bank stocks reacted positively on Thursday as investors bet that lower rates and a relatively healthy economy would create more demand for borrowing and boost investment banking dealmaking activity.

Nonetheless, the highest interest rates in more than a generation have pushed more money than ever, nearly $3tn, into certificates of deposit, which typically pay the highest rate of any bank deposits and also cannot be changed overnight.

As that money becomes unlocked, banks will be able to adjust their rates down, but not before, analysts said.

“It will be a slow grind down,” said Scott Hildenbrand, chief balance sheet strategist at Piper Sandler.

The whole sport was bracing for another season of Max Verstappen dominance. Instead, the field is deeper than it has been in years with seven different race winners this season.

Singapore

Charles Leclerc is one of only two men on the planet who gets to race a Ferrari Formula One car for a living—200 miles per hour, 24 weekends a year—and even he was convinced that this season would be boring.

All he had to do was look down the paddock. Red Bull’s car had been utterly dominant in 2023. Three-time world champion Max Verstappen had won 19 of the 22 races, and rolled that form straight into 2024. No matter how much money everyone else plowed into upgrading their technology, they were still playing catch-up.

“We all thought after the first two, three races that Red Bull would walk away with the title, the way they did last year,” Leclerc said.

Instead, this season has taken more sudden twists than a driver with a blindfold. Red Bull has hit the skids, McLaren has emerged as a contender, Mercedes experienced a brief revival, Ferrari is fast again, and Grands Prix have turned back into wide open, can’t-miss free-for-alls. The year that was supposed to deliver a string of dull processions has produced seven different race winners—the most in a single campaign since 2012. The latest was McLaren’s Lando Norris, who tore away from Verstappen to claim victory on Sunday in Singapore.

“We are really in a pack,” Ferrari team principal Frederic Vasseur said here, with six of the 24 races remaining. “I don’t remember when it was so tight in F1.”

All that on-track action has come with a healthy side of off-track lunacy. Two different teams have had to draft in rookies as emergency replacements. Ferrari’s Carlos Sainz had an appendectomy and came back to win a Grand Prix less than three weeks later. A practice session in Singapore was interrupted by an enormous lizard crawling across the track. And drivers have lately been up against a bizarre crackdown on their foul language.

Yet for the first few months, it looked as though all that excitement would be confined to a fight for second place. Verstappen won five of the first seven races and all was proceeding according to plan. But a sudden decline since the summer break combined with sweeping changes behind the scenes of Red Bull have made everything a little ragged for the sport’s all-conquering outfit.

Legendary F1 designer Adrian Newey and sporting director Jonathan Wheatley are both due to leave for other teams by next season. And the 2024 car hasn’t performed as reliably as its predecessors on some of the calendar’s more challenging street circuits.

Verstappen, a hothead who appeared to have mellowed in recent years, hasn’t been shy about letting his team know. He was so disappointed after the Azerbaijan Grand Prix that he declared in Singapore that he “knew the car was f—ed.” (That burst of obscenity earned Verstappen a reprimand from motorsport’s world governing body.)

But Red Bull’s loss has been the rest of Formula One’s gain.

Leclerc won on Ferrari’s home turf in Monza. Seven-time world champion Lewis Hamilton ended a 2.5-year dry spell with victory at Silverstone—and then picked up another win in Belgium for good measure. Somehow Verstappen hasn’t won a race since late June.

In fact, the past nine Grands Prix have been won by six different drivers, blowing open the race for both available championships. Norris, a 24-year-old Briton, has given himself a chance to reel in Verstappen for the drivers’ title. And together with 23-year-old teammate Oscar Piastri, he has put McLaren ahead of Red Bull in the constructors’ standings.

“The competition has improved quite a bit,” Verstappen said. “We need to find a bit more performance and make our lives a bit easier.”

The natural state of F1 is to bounce from dynasty to dynasty. Every few years, the organizers make wholesale changes to the rulebook, radically altering the cars’ specifications. Whichever team does the best job of cracking the code tends to steal a march on the field and everyone else spends the next few years chasing them. It’s how Red Bull took back-to-back constructors’ titles in 2022 and 2023. And it was how Mercedes dominated much of the 2010s.

None of that prevented the sport’s Netflix-and-social-media-fueled revival in recent years. The Drive to Survive series sparked interest in teams and drivers up and down the paddock, regardless of results. The club of 20 F1 drivers all harnessed their inner influencers.

But in order for F1 to stay relevant, organizers knew that there was one thing no reality show or viral TikTok could ever replace: tense qualifying, and flat-out, wheel-to-wheel racing.

“It’s so competitive right now at the front,” Mercedes driver George Russell said.

Nothing makes that clearer than the number of tight finishes. Back when Verstappen was stomping the field in 2023, his average margin of victory across 19 victories was ​​13.4 seconds. This season, heading into Singapore, drivers had cut that by more than 40% to 7.6 seconds.

And recently, it had been getting even closer. Over the first 10 races of the year, drivers (mostly Verstappen) were winning by an average of 8.6 seconds. But between Verstappen’s last victory, eight Grands Prix ago, and the field touching down here, the margin was just 6.1. That’s why Sunday will go down as one of the most surprising results of the year. The margin of victory was 20.9 seconds—and Verstappen was on the wrong end of it.

No one in F1 had seen it coming at the start of the season. Which is also why no one in F1 wants to guess how this one ends.

“It makes no sense,” Vasseur said, “to imagine how the picture could be in two or three months.”

Show creators Mickey Down and Konrad Kay met at Oxford University and after leaving their day jobs in finance, created a show that’s full of sex, drugs the nuances of the British class system.

LONDON — HBO knows what makes a good watch: power dynamics, solid writing, sex and class systems as witnessed in “Succession,” “Euphoria” and “Game of Thrones.”

The network’s banking television series “Industry” — currently airing its third season and renewed for a fourth — takes it further with heaps of cocaine, male office tyrants, golden showers and fancy boats.

The show is to the finance world what “The Devil Wears Prada” was to fashion magazines. The show is set in the fictional bank of Pierpoint & Co, which loosely takes inspiration for its name from American financier, John Pierpont Morgan, who was head of the bank J.P. Morgan & Co. from 1895 to 1913.

Writers Mickey Down and Konrad Kay have won critical acclaim for their dramatization of real-life events they or their peers experienced while working in banking. They both had stints in the sector after graduating from Oxford University.

“People project their own experience onto [the show] — especially with assistants and junior people, as it reflects a cutthroat and fast paced [work environment], be it television, fashion [or finance],” Down says.

In the new season, the writers dig deeper into the nuances of the British class system, green tech energy, nepotism within finance and the British press rather than focusing purely on storylines.

For the first half of the series, the plot revolves around Henry Muck, played by Kit Harington, an English aristocrat who is chief executive officer of green tech energy company Lumi. Pierpoint is managing its IPO, which quickly fails in the marketplace after being oversold.

Down and Kay invited their friend from university, Pippa Lamb, an angel investor and partner at Sweet Capital, to star in the scene when the company goes public.

The show has had a polarizing reaction from those who work in finance, according to the creators.

“Half of them love it and can’t get enough of it, saying that it’s a reflection of their experience, meanwhile the other can’t watch it at all because it’s a reflection of their experience,” Down says.

“When I first started watching ‘Industry,’ I distinctly remember texting Mickey saying how it was almost too close for comfort and several friends who had also started out in banking said the same thing. It was too much for some of them to watch,” says Lamb, who spent five years at J.P. Morgan & Co. in London and Hong Kong.

“I went through my photos and pulled out a bunch of pictures from my time working in public equities — they were strikingly similar: my Bloomberg terminal, my managing director speaking on CNBC, my headset, the birthday cakes served at the desk and the office Christmas parties,” she adds.

“Industry” adds a sheen of glamour to the long hours of banking. Out of office, the Pierpoint team gallivants about town searching for their next big high.

“Everybody has the capacity for indulging their worst impulses and there’s a vicarious aspect of watching people indulge in all the things they would like if they were free from their wives, families and bank accounts. It’s the darker side of humanity and there’s a wish fulfillment aspect to it, which is just always very intoxicating,” Kay says.

He compares the audience’s intrigue to the show to a black box. “It’s an exclusive world and rarefied air, which people think they don’t really have access to,” he adds.

People have even messaged Down on LinkedIn telling him they got into the industry because of the show.

“The show does make that world look sexy, much sexier than it is,” he says.

“Industry” is also a cautionary tale of flying too close to the sun — and burning out.

In episode one of the first season, a death occurs that’s based on Moritz Erhardt, a Bank of America Merrill Lynch intern who died in 2013 after working 72 hours straight — an arduous rite of passage, otherwise known as the “magic roundabout,” where graduates work around the clock and get a taxi back to their homes in the early hours of the morning for a change of clothes and shower before returning back to their desks.

The shows actors are put through grueling character arcs.

Yasmin Kara-Hanani, played by Marisa Abela, grapples with losing her wealth and being in the press for her father’s financial mishaps; Harper Stern, played by Myha’la Herrold, gets caught lying on her résumé; Robert Spearing, played by Harry Lawtey, becomes Muck’s whipping boy, and Rishi Ramdani, played by Sagar Radia, confronts the realities of marriage and race.

It’s in the moments of humility that the audience catches a glimpse of who the characters really are behind the banking jargon, excessive swearing and calling each other c-nts.

“They’re toxic people in a toxic environment — they have to be toxic to succeed and survive — that’s the status quo of the show. Maybe at their core, they’re not like this. They’re constantly in conflict with themselves,” Down says.

“Industry” is a tragedy laced with comedy, cocaine and sex — and it’s moved into HBO’s prized Sunday slot, previously occupied by “House of the Dragon.”

‘Medium Earth orbit’ is a promising niche in space, but investors fear Elon Musk’s Starlink and the decline of linear TV

There seems to be no orbit where publicly listed satellite operators can escape the wrath of the market.

For the first time, the North Atlantic Treaty Organization will contract satellites in “medium Earth orbit” to fulfill some of its communications needs: Earlier this month, it awarded $200 million to Luxembourg’s SES, the only active commercial operator in this segment through its O3b mPOWER constellation.

The medium Earth orbit, or MEO, isn’t as famous as its neighbor, the low Earth orbit, or LEO, from which Elon Musk’s Starlink has revolutionized the space economy through mini-satellites that can communicate with Earth’s surface in 20 milliseconds. Nor is it as distant as the geostationary orbit used by traditional firms, where a single satellite can cover a third of the planet’s surface.

MEO sits somewhere in between, starting 1,200 miles above sea level and ending 21,000 miles above that.

Medium Earth orbit is the region of choice for navigation systems such as the Global Positioning System. It is also ideal for Earth observation, a booming area of business that spans agriculture, urban planning and environmental monitoring, and is expected to generate $700 billion in value added by 2030, according to the World Economic Forum.

From medium orbit, satellites can send signals to the surface and back in less than 150 milliseconds—five times as fast as from geostationary orbit. Apart from gaming and video chatting, this is usually sufficient for most use cases.

And MEO is still high enough to allow for global continuous coverage with six to eight satellites, if the poles are excluded. SES, the Luxembourg operator, currently does it with 26.

Meanwhile, Starlink has about 5,000 satellites in low orbit. Each sees only a small patch of the Earth, and briefly, because it circles the planet about 15 times a day. So it has to constantly hand off service to a neighboring node, which is complex and can lead to interruptions.

This is the reason MEO is often thought of as the sweet spot of space.

To reinforce congested areas and have backups, governments and commercial customers of satellite services often want to diversify. Traditional players such as California-based Viasat VSAT -1.62%decrease; red down pointing triangle, France’s Eutelsat ETL -2.15%decrease; red down pointing triangle and Canada’s Telesat are now trying to combat Starlink by combining LEO satellites with geostationary ones.

“Almost every ship we are in has Starlink too,” said Adel Al-Saleh, chief executive at SES, which provides broadband access to passengers and crew at the world’s top cruise companies.

SES satellites could have a groundbreaking role to play in “hybrid” constellations: They sit closer to the LEO layer below, and could serve as “space relays” that aggregate and transmit data from the minisatellites to ground stations or other parts of the constellation. SES successfully tested this capability in June.

And yet, SES stock has lost more than 80% of its market value over the past decade, roughly in lockstep with Viasat, Eutelsat and Telesat.

Investors are so down on the sector that those who buy telecommunications firms will often bet against satellite firms as a way to hedge risk, brokers say.

Previously, these businesses would typically spend a few hundred million dollars on a big satellite that could spend 20 years providing lackluster service, and enjoy captive demand from governments and consumers. Suddenly, advances of semiconductor technology and the LEO revolution have led to minisatellites being built for $1 million and replaced in five years’ time.

Also, the popularity of streaming has led to “cord-cutting” of satellite-TV services. SES’s video division has lost a quarter of its revenue over the past five years. Its other services have grown 53%, but they don’t generate the same massive profit margins.

Incumbents are making big investments. However, the resulting capacity war with Starlink—run by privately owned SpaceX, which can cross-subsidize satellites with its lucrative rocket-launch operation—is depressing prices.

Satellite firms are also gaining scale, as shown by the recent Eutelsat-OneWeb and Viasat-Inmarsat mergers. SES is swimming in cash, which it got from selling a portion of its C-band spectrum in the U.S. to make it available for 5G. In April, it announced that it will use the funds to purchase Virginia-based Intelsat, which recently emerged from bankruptcy. Investors, who wanted payouts instead, have reacted badly.

There is still potential upside: SES trades at a lower valuation than its peers, despite having the largest operating margin and being the only one with a significant dividend yield. Also, it doesn’t directly compete in consumer broadband, where Starlink is unbeatable.

But investors might not jump in until the MEO approach proves that it can live up to its promise. Power issues with six of SES’s mPOWER satellites, built by Boeing BA -0.84%decrease; red down pointing triangle, have significantly affected their operational life and broadband capacity. It underscores a key difference with Starlink: MEO satellites aren’t as disposable when something goes wrong; they are costly, sophisticated machines—more in the vein of the traditional satellite business model.

“Rather than being a sweet spot, it has a bit of the advantages and disadvantages of both,” said Bernstein analyst Aleksander Peterc.

Einstein’s equivalence principle states that floating in space and free fall are indistinguishable. So far, this remains true for listed satellite companies.

MILAN —Sabato De Sarno’s latest fashion show on Friday marked a slight improvement in terms of coolness and impact thanks to a Gucci collection brimming with elegant and glamorous outfits harking back to the 1960s. It was a natural evolution of the Italian designer’s style, oozing more maturity and self-confidence. But it did not get the fashion crowd so excited that it’s likely to change the overall downward trajectory of the Italian luxury brand, industry sources predicted.

Gucci, Kering’s biggest business and source of profit, continues to struggle in major markets such as China, industry and internal sources said. Key managers are leaving the company and many positions remain vacant, they said.

Federico Turconi, CEO and President of Gucci U.S., left this month as did Alexis Katana, the company’s vice president in charge of global media. Their exits follows the departures of many other important staff including Gucci’s chief marketing officer Jonathan Kiman – now at Burberry. There’s been a high turnover at Gucci’s marketing, merchandising and public relations teams in the past six months, as Miss Tweed reported. Many positions – including that of chief marketing officer – remain vacant. Gucci declined to confirm or comment on the departures.

Stefano Cantino, let go in April by LVMH’s Louis Vuitton where he was in charge of communications and events, started as Gucci’s deputy chief executive in early May. Staff say Cantino is expected to replace CEO Jean-François Palus, possibly as early as January. Since he arrived a year ago, Palus has not succeeded in building strong ties with teams, staff say. “Palus does not appear to be very involved with the brand,” one internal source said. “We rarely see him or interact with him.”

Palus, a senior member of Kering’s executive committee and longtime friend of CEO François-Henri Pinault, left the group’s board last year and is preparing to move on from Gucci, internal sources at the group say.

The mood at Gucci is somber. “People know that it will take time for the brand’s sales to improve and with such high turnover, they have the feeling that they can be sacked any moment,” another internal source said.

Some staff who left say the management structure of Gucci is unclear and the decision-making process is not very fluid. Who’s really in charge at Gucci? De Sarno? Palus? Cantino? Or Francesca Bellettini, Kering’s co-CEO who oversees all of the group’s luxury brands? The brand’s previous boss, Marco Bizzarri, was difficult to work with, staff say. But at least he federated teams around clear objectives and drove them to deliver results, industry sources say.

More importantly, it’s not clear how long De Sarno will stay at Gucci’s creative helm.

FASHION SHOW

The designer’s latest fashion show was a continuation of what he presented before– nice clothes but nothing to write home about, fashion critics and retailers said. “Top clients liked it but what is the message? What does the brand stand for now?” asked one multi-brand retailer.

De Sarno’s collection included shiny leather jackets and matching skirts and see-through lace dresses in his trademark Ancore burgundy red. There were also a few Greek goddess toga-style dresses slit open on the side, worn with a tubular golden necklace and arm bracelets. The theme was “casual grandeur,” De Sarno said in his notes about the show. Not everyone understood what it meant and reactions on social media were mixed, as they have been before.

Some loved his shimmering dresses and oversized bright colored coats worn with Jackie Kennedy Onassis-style headscarf and oversized sunglasses. Others found that there were a lot of generic clothes that could have been designed by any brand. Why would a customer want to spend vast amounts on an outfit if you cannot tell that it’s Gucci?

Kering wants Gucci to be about timeless elegance, but critics say the brand is not taking a strong enough stance in terms of design. As a result, it does not stand out in the ultra-competitive global fashion market. In the current tough climate, customers have become reticent to spend on luxury goods. They need a compelling reason to open their wallet and pull out their credit card. They want to be wowed and surprised. Gucci has not managed to achieve that yet.

Since De Sarno started in May last year, Gucci’s marketing campaigns have been relatively bland, often featuring a model posing on a white backdrop with Gucci printed in white. The brand’s unexciting campaigns could be symptomatic of the lack of clear direction among the brand’s communication and top management teams. No one is ready to take any risk or say anything too strong or provocative.

DARING

At a Mediobanca luxury goods conference in Milan this week, former Gucci CEO Marco Bizzarri complained of how boring and monotonous many big luxury brands had become. This is partly because their creative directors were not given enough freedom to try out new things. “The sector is also suffering because there is a lack of innovation… no one dares,” he said.

De Sarno’s looks are elegant, chic and wearable, critics say, but they’ve not been flying off the shelves. Kering reported in July that Gucci’s turnover fell 20 percent in the first half and suffered an even bigger decline in China, one of its biggest markets. Its performance in China has worsened in recent months, several industry sources have said, which does not augur well in terms of profitability.

Kering cut its full-year financial performance expectations for the third time in July, issuing a profit warning and publishing a 19 percent drop in second-quarter sales for Gucci. Kering reported a 42 percent drop in first-half operating profit and said it would be down 30 percent in the second half compared with the same period last year.

Kering shares have lost more than 42 percent since the beginning of the year and are now trading at the same level they were back in 2017. The shares of rivals like Richemont and LVMH have also suffered, but they are still much higher than they were seven or eight years ago.

Kering investors have lost patience and many of them have given up on the stock, which partly explains why the shares have fallen so much. Luxury and fashion are about selling dreams. De Sarno’s dreams do not appear to have captured customers’ imagination as much as Kering hoped.

Several industry sources said Cantino was thinking of alternatives to De Sarno as creative director for Gucci. He’s been in contact with Fabio Zambernardi, Prada and Miu Miu’s former designer director who retired a year ago after four decades working alongside Miuccia Prada. But fashion veterans say Zambernardi made so much money at Prada, he does not need to work anymore. It’s also far from certain that his Prada-imprinted style would fit Gucci. “Zambernardi is very talented and highly regarded but I’m not sure he would want to come back to the fashion circus,” one Prada source said.

A surge in wind and solar power means many businesses and consumers around Europe can get paid for plugging in. The U.S. could be next.

KERKDRIEL, the Netherlands—For much of the spring and summer, Jeroen van Diesen got paid for using electricity.

Sometimes his neighbors came over to power up too, generating even more cash.

Van Diesen’s situation reflects the strange, new dynamics of electricity that could soon become the norm in many parts of the world: A big increase in wind and solar power has pushed wholesale prices to zero or below for many hours of the year, spurring a sea change in the way people use power—based on whether the sun is shining or the wind is blowing.

Most people pay a fixed price for each kilowatt of electricity they consume throughout the day. The price is set by their power company and only changes at infrequent intervals—once a week, a month or even only once a year.

Van Diesen, a software salesman, recently signed up to receive electricity from two providers that charge him the hourly price on the Dutch wholesale power market, rather than a fixed price that resets monthly or annually. When the price of electricity falls low enough, smart meters in his house begin charging his two electric cars.

Wholesale prices swing wildly each hour of the day, and even more so as a larger share of electricity flows from wind and solar installations. Because the generation costs of wind or solar farms are negligible, market prices will be near zero when there is enough renewable power to cover most of a region’s electricity demand.

Electricity market dynamics get weirder when renewable-energy producers don’t have an incentive to stop feeding power into the grid, usually because of government subsidies. Then grids can be flooded with excess power, pushing prices into negative territory.

Van Diesen said he’s made 30 euros, equivalent to around $34, over the past five months charging his car, enough to cover the service fee from his power supplier, a Norwegian company called Tibber.

“I’m charging the car for free,” said van Diesen, who is part of a group of clean-energy enthusiasts in the Netherlands who call themselves green nerds. “To me it’s also like a hobby and a game—how far can I go?”

Doing laundry in the evening? The electricity could be free a few hours later when demand dies down and the wind picks up. Likewise, in regions with lots of solar power, charging an electric vehicle in the morning is usually far more expensive than powering up under the midday sun—or whenever the price is right.

In the U.S., most states don’t currently allow such real-time pricing, but many think that will change. Already, in some of the world’s biggest economies from Western Europe to California, the occurrence of zero and negative wholesale power prices is growing fast.

Negative prices

Wholesale prices across continental Europe have fallen to zero or below in 6% of all hours this year, up sharply from 2.2% in 2023 and just 0.3% in 2022, according to data collected by Entso-E, the group of European transmission system operators. In markets with lots of renewable capacity, this year’s figure was higher: 8% in the Netherlands, 11% in Finland and 12% in Spain. Analysts expect those numbers will grow as more solar panels and wind turbines are installed.

The changes sweeping Europe’s electricity markets, which were accelerated by the energy crisis brought on by the war in Ukraine, show what could happen in the U.S. in a few years when renewable capacity reaches a similar scale. In 2023, 44% of EU electricity was generated by renewables, compared with 21% in the U.S.

In some U.S. markets—sunny California, the wind-swept Great Plains, and Texas—zero and negative prices are already common. The wholesale price in Southern California was negative nearly 20% of all hours this year because of the region’s boom in solar-panel installations. That compares to around 5% last year, according to data collected by the U.S. Energy Information Administration.

U.S. regulators have been cautious about allowing households and companies to sign up for electricity plans that charge them wholesale prices, fearing consumers could be hit with big bills if prices jump. Texas consumers who signed such contracts were walloped with huge bills in 2021 when a rare winter storm sent prices soaring.

States’ reluctance, however, may now be waning as policymakers increasingly see real-time pricing as a way to lower peak demand, reduce the need for costly infrastructure and integrate more renewables into the grid.

California regulators this year ordered the state’s utilities to expand dynamic price pilot programs that have only been available for a select group of customers.

Your overall power bill still won’t be zero in a clean-energy future. Generation costs comprised around 60% of customer bills on average in the U.S. in 2023. Transmission and distribution costs account for most of the rest—and are expected to grow sharply in the coming decade to reinforce the grid for electric heating, electric transport and data centers.

Negative prices could also be reined in over the next few years as governments from Europe to California pare back renewable-energy subsidies. Governments are particularly focused on trimming subsidies for solar power, which is driving negative prices in a number of markets.

Green nerds

In Europe, energy-hungry manufacturers are shifting their operating strategies to maximize energy consumption when prices are close to zero or negative, while throttling back when prices are high.

Linde, a U.K.-based engineering company, is building a new generation of industrial gas plants that can be quickly ramped up and down depending on the wholesale price of power.

When solar and wind power drive prices down, Linde’s plants fire up and send the output to large tanks. When electricity prices shoot up again, the plants can ramp back down and supply customers out of the gases stored in the tanks.

“The tank functions like a virtual battery,” said Klaus Ohlig, a research and development executive at Linde Engineering.

Trimet, an aluminum producer that is one of Germany’s single-largest power consumers, is overhauling its smelters to vary their power consumption depending on the availability of renewable energy on the grid.

A new European Union law requires dynamic-price power contracts be made available to consumers across the 27-nation bloc. Tibber, a power retailer based in Norway that charges its customers the wholesale hourly price, has signed up more than one million households across the Nordic countries, Germany and the Netherlands.

Edgeir Aksnes, Tibber’s co-founder and chief executive, says he doesn’t expect customers to constantly track the hourly price before deciding when to charge their car or run appliances.

“We can automate all of this for you. You don’t have to think about it,” he said.

Some enthusiasts, however, like to get into the weeds.

Wouter van Embden, a 49-year-old Dutch entrepreneur and one of the country’s so-called green nerds, switched to Tibber earlier this year. On a recent summer Sunday, the battery in his home began charging as solar power flooded the Dutch grid and the wholesale power price fell to zero. He also charged his two electric cars and programmed his heat pump to make the water in the house tank extra hot.

Toward the evening, as prices rose with the drop-off in solar, van Embden’s battery—which he and his son built at home—would power his home as well as feed into the Dutch grid.

“I have to be honest, when I started building the battery I had so many outages. There was a lot of testing to do,” he said. “But now it’s working pretty stable.”

GPUs will always have a critical role to play in AI computing, but as models mature, there’s a growing opportunity for more custom chips

Future computer chips may be able to help ease the alarming energy demands of generative artificial intelligence, but chip makers say they need something from AI first: a slowdown in the sizzling pace of change.

Graphics processing units so far have dominated the bulk of training and running large-scale AI models. The chips, originally built for gaming graphics, offer a unique blend of high performance with the flexibility and programmability required to keep up with today’s constantly shifting swirl of AI models.

Nvidia’s dominance in the GPU market has propelled it to a trillion-dollar valuation, but others, including Advanced Micro Devices, also make the chips.

As the industry coalesces around more standardized model designs, however, there will be an opportunity to build more custom chips that don’t require as much programmability and flexibility, said Lisa Su, chief executive at AMD. That will make them more energy-efficient, smaller and cheaper.

“GPUs right now are the architecture of choice for large language models, because they’re very, very efficient for parallel processing, but they give you just a little bit of programmability,” Su said. “Do I believe that that’s going to be the architecture of choice in five-plus years? I think it will change.”

What Su expects in five or seven years’ time isn’t a shift away from GPUs, but rather a broadening beyond GPUs.

Nvidia and AMD haven’t been vocal around specific plans here. Nvidia declined to comment for this article.

Some custom chips are already hard at work handling aspects of AI.

Large cloud providers like Amazon.com and Google have developed their own custom AI chips for internal use, such as Amazon’s AWS Trainium and AWS Inferentia, and Google’s tensor processing units, or TPUs. These are built to execute only specific functions: Trainium can only train models, for example, while Inferentia can only run inference, a less intensive process than training in which models process new information and respond.

Broadcom CEO Hock Tan said in an internal address this year that his company’s custom chip division, which mostly helped Google make AI chips, was bringing in over $1 billion in operating profit a quarter.

Custom chips can be far more energy efficient, cheaper and smaller because they can be hard-wired to a given degree: they can perform one specific function, run one specific type of model or even one specific model itself, said Shane Rau, research vice president for computing semiconductors at market intelligence firm International Data Corp.

But the market for commercially selling these super-custom, application-specific chips is still immature, Rau said, a symptom of how much innovation is happening in AI models.

Highly customized chips also present a challenging lack of flexibility and interoperability, said Chirag Dekate, a vice president analyst at research firm Gartner. To the extent that they are programmable, they’re very difficult to program, typically requiring custom software stacks, and it can be difficult to make them work with other kinds of chips.

Many chip offerings today exist on a continuum, however, with some GPUs that can be more customized and some specialized chips that provide a level of programmability. That gives chip makers an opportunity, even before generative AI becomes more standardized. It can also be a conundrum.

“That’s a big thing we’ve struggled with here,” said Gavin Uberti, co-founder and CEO of Etched. The startup makes chips that do inference only on the transformer architecture, developed by Google in 2017, that has since become the standard for large language models. Despite customizing up to a certain point, the chips also have to be flexible enough to adapt to smaller operations that vary model to model.

“Right now, the models have plateaued enough that I think making a bet on the transformer makes sense, but I don’t think making a bet on say, Llama 3.1 405B makes sense yet,” Uberti said, referring to an AI model from Facebook owner Meta Platforms. “Transformers are going to stick around, but they’re going to keep getting bigger and evolving.” He added, “You do have to be careful to not overspecialize.”

There is also no one-size-fits-all when it comes to computing, said Su, the AMD CEO. AI models in the future will use a combination of different types of chips, including today’s dominant GPUs and also more specialized chips still to be developed, for various functions.

“There will be other architectures,” she said. “It’s just that it’ll depend on the evolution of the models.”