TheChek News Flow

Meta’s CEO has become one of the biggest landowners in Hawaii, growing his property’s footprint and erecting new mysterious buildings.

As a child, Julian Ako would visit his maternal great-grandfather’s home near Pilaa Beach in Kauai, Hawaii, where he and his family would gather edible fungi that grow on kukui trees and collect seaweed and fish from the reef.

For about a decade, that land has belonged to Meta CEO Mark Zuckerberg, who is constructing a massive compound at an estimated cost that exceeds $300 million. WIRED can now reveal that Zuckerberg’s property is atop a burial site: Ako’s great-grandmother and her brother were buried on the land.

After months of discussions with a Zuckerberg representative, Ako was successfully able to gain access to the property and identify and register the graves with the state Department of Land and Natural Resources, though he was not able to locate remains of other ancestors, who he believes could be buried on the property. In a report shared with WIRED, the state agency also confirmed “the probability (based on oral testimony) of additional burial sites.” Visits to Ako's family’s graves are coordinated by the team at the Zuckerberg ranch. Ako, who sits on the Oahu Island Burial Council, worries about what might happen if further burial sites are discovered, because of the extreme secrecy surrounding the compound.

FWhile NDAs are not unusual on billionaire construction projects, the scale of Zuckerberg’s compound has resulted in scores of local workers being forbidden from sharing what they’re doing and who they’re working for. “If all of the workers have signed these nondisclosure agreements, then basically they’re sworn to silence,” Ako says. “If they uncover iwi—or bones—it’s going to be a challenge for that to ever become public knowledge, because they’re putting their jobs in jeopardy.”

Asked about these burials, Zuckerberg representative Brandi Hoffine Barr acknowledged that the estate had been made aware of the family burial plot in 2015, which Hoffine Barr says they fenced off and maintained. She adds that their workers are bound by regulations that require reporting of inadvertent discoveries of iwi.

Meanwhile, Zuckerberg has quietly expanded his footprint on the island with a massive new land purchase, WIRED can reveal. Earlier this year, Zuckerberg purchased 962 acres of prime ranchland under a Hawaiian-sounding LLC across the road from the existing compound, which one person close to the sale estimated cost more than $65 million. This purchase, previously unreported, will increase his Kauai holdings from about 1,400 to more than 2,300 acres—placing him among the largest landowners in the state.

Development inside the ranch continues, as Zuckerberg has spent millions adding several new strange buildings to an already massive compound. Not far from Ako’s fishing spot, Zuckerberg has commissioned another three major buildings on previously purchased land. According to planning documents released to WIRED under a new public records request, they range in size from 7,820 to 11,152 square feet—nearly 10 times larger than the average home in Hawaii—and two are projected to cost between $3.5 and $4 million each.

These new buildings differ from the opulent mansions on the other side of the ranch, with few fun amenities and only one dedicated common space, a lanai larger than 1,300 square feet. Two of them seem designed to accommodate as many bedrooms and bathrooms as possible, and feature 16 of each between them, lined up like a motel or boarding house. As always, security is tight — with each new property featuring cameras, keypad locks, and motion detection devices. Hoffine Barr described these new buildings as short-term guest housing for family, friends, and staff.

This goes along with previous development across the ranch: two mansions with a total floor area comparable to the size of a football field, a gym, a tennis court, several guest houses, ranch operations buildings, a set of saucer-shaped treehouses, an elaborate water system, and a tunnel that branches off into an underground shelter about the size of an NBA basketball court, outfitted with blast-resistant doors and an escape hatch. Recent documents also show plans for a new water pump building, to go along with two existing pump buildings and an 18-foot-tall water tank. Satellite images of the property also show dozens of buildings that have not yet appeared in public records requests. Based on counting bedrooms in the planning documents we’ve seen alone, WIRED estimates that, when complete, the property could comfortably house more than 100 people.

The Meta CEO’s Kauai activities attracted international attention following a December 2023 WIRED investigation, based on planning documents and interviews with workers, that estimated the total cost of his compound development as at least $270 million, detailed strict enforcement of nondisclosure agreements, and described doomsday bunker-ish qualities of the project. According to some prepping companies, the report sparked an increase in bunker sales.

Since then, Zuckerberg’s presence on the island has only continued to grow. Last January, Zuckerberg announced his intention to raise premier cattle on beer and macadamia nuts on the ranch, but it seems likely that he has bigger plans.

For locals, the question remains—what the hell is this guy up to?

Zuckerberg first bought into Kauai—the oldest and smallest of the four major Hawaiian islands—in 2014, when he grabbed 700 acres in a quiet oceanside stretch near the small town of Kilauea for roughly $100 million.

The purchase was incomplete however, as hundreds of locals maintained kuleana rights to four parcels within Zuckerberg’s property. These rights, afforded to descendants of previous landowners, would have allowed them to cross Zuckerberg’s land. In 2016, Zuckerberg moved to consolidate his holdings by filing “quiet title and partition” lawsuits against these kuleana descendants in order to clarify ownership of the land. He later abandoned these suits under public pressure, but the legal process continued under a kuleana descendant Carlos Andrade, whom Zuckerberg supported in an op-ed in the local newspaper. Andrade eventually won sole ownership of the land at auction, during which some believed he was backed financially by Zuckerberg. (In that 2017 op-ed, Zuckerberg wrote that Andrade, who died in 2022, could continue his quiet title action and pass down the kuleana rights because he had “lived on and cared for these lands for more than forty years.”)

By spring 2021 his compound had expanded further, with the addition of more than 560 acres of ranchland in total, some of it abutting Larsen’s, a nudist beach. Later that year, he added another 110 acres which contain the Kaloko Dam, an infamous earthen dam and reservoir that collapsed in 2006, killing seven people.

The 2025 land buy is Zuckerberg’s largest thus far in total acreage, situated on the mauka, or inland, side of the road across from his initial purchase. Public records list the Mary Lucas Trust Estate as the seller, descendants of an early British adviser to King Kamehameha I. The trust had leased its lands to sugar plantations before they were restored as pasture land for cattle in the 1970s by cousins and former trustees Jimmy Pflueger and Paul Cassiday. In recent years, the trust has been selling off significant chunks of its lands. Though the total purchase price is not listed, a source close to the sale estimated it was at least $65 million, and property records place the land’s market value at around $75 million. Hoffine Barr confirmed that the billionaire had purchased additional ranch land, but did not comment on the size or price. It’s unclear what the CEO intends to do with his new acquisition, but the source also described the parcel as “great cattle grass,” so it seems likely that Zuckerberg’s husky, beer-filled cows may soon be roaming its 962 acres.

“Mark and Priscilla continue to make a home for their family and grow their ranching, farming, and conservation efforts at Ko'olau Ranch,” says Hoffine Barr. “The vast majority of the land is dedicated to agriculture—including cattle ranching, organic ginger, macadamia nut, and turmeric farming, native plant restoration, and endangered species protection. After purchasing the ranch, they canceled the previous owner’s plans for 80 luxury homes.”

With the new buildings and new land, Zuckerberg’s total investment in his compound now exceeds the entirety of the $311 million fiscal year 2024 Kauai operating expenses budget.

This dramatic influx of wealth has led to inevitable changes in the community. On one hand, Zuckerberg has given millions to local nonprofits, including recent donations to build a charter school and an affordable housing nonprofit near the compound. His construction projects provide good-paying jobs. But there remains a lot of skepticism toward the recent trend of billionaires buying up Hawaiian lands.

As more of the defunct sugarcane plantations that own huge slices of Hawaiian land begin offloading their assets, more new-money billionaires have been buying in. In 2012, then Oracle CEO Larry Ellison bought almost the entirety of the smaller island of Lanai for $300 million, which he has been developing into a luxury resort destination. Amazon founder Jeff Bezos and former TV presenter and businesswoman Oprah Winfrey both have outposts on Maui. And Salesforce CEO Marc Benioff has been secretly buying up large swaths of the Big Island for unclear purposes. Billionaires pay top dollar, driving up property values.

Driven partially by an influx of wealthy mainlanders during the Covid-19 pandemic, Hawaii housing prices have skyrocketed, leaving home ownership out of reach for local renters. “If our island has any hope of remaining Hawaii, this kind of activity has got to stop,” professor of Native Hawaiian studies at the Kauai Community College Puali‘i Rossi tells me, when I mention the new Zuckerberg land buy. “Eventually Hawaii isn’t going to look like Hawaii anymore—it’s going to be a resort community. Are we really thinking about 100 years from now, what this island is going to look like?”

On a damp Sunday afternoon during the February wet season, a few pickup trucks pass through the main gated entrance outside Zuckerberg’s compound on the Hawaiian island of Kauai. Backhoes and bulldozers rest in the red mud outside the tall stone walls, constructed in 2016 to keep out prying eyes.

It’s a relatively quiet day, but security is still alert. As I take a picture of the guard shack from the road, a woman’s head pops out. “Hoi, don’t do that,” she says. “They don’t like that. They’re very private.”

Design software company’s strong growth could be undercut by AI, competition and untested products.

After two blockbuster tech initial public offerings, investors are positively drooling over Figma’s upcoming listing. The design software company is expected to launch into a surprisingly hot market for tech offerings led by CoreWeave and Circle, which are up more than 200% and 600%, respectively, since they started trading.

Figma boasts blazing-fast sales growth: 46% in the first quarter, outshining the rest of the software industry. It has grown into a veritable cash machine. And Adobe’s failed bid to acquire Figma for $20 billion puts a big target out there that investors would love to eclipse. But I’d be wary of chasing the stock if it soars at the offering.

Adobe’s bid came during a period of high valuations for software companies. Adobe was willing to pay up for Figma because the deal eliminated a competitor, which is why regulators shot it down, and because of synergies between the companies. Figma won’t benefit from either of those as a stand-alone public company. Moreover, slowing revenue growth and pressure from artificial intelligence have driven down valuations since the Adobe deal.

If the $20 billion price is too high, there’s another less frothy price to consider. After the Adobe deal fell apart, Figma bought back a hefty amount of stock from its employees and investors in a tender offer valuing it at $12.5 billion.

A more realistic valuation for Figma in the offering is about $15 billion, based on comparable companies and taking into account the pressures Figma is facing.

Two software companies that closely resemble Figma are Datadog and Monday.com. All three companies have gross margins over 80%, free cash flow margins of around 30% and boast net revenue retention rates of over 100%, a sign customers are staying and increasing their spending. Datadog and Monday.com are among the fastest-growing publicly traded firms in the software business.

Monday.com trades at 10.4 times next year’s sales and Datadog at 13.9 times, one of the highest multiples among software companies. Assuming Figma grows its sales in 2025 at the same rate it did in 2024, on Datadog’s multiple, it would be worth $15 billion.

None of this analysis may make any difference, though, if investors see Figma as the next CoreWeave or Circle. Venture capital investor Tomasz Tunguz said the success of these two unusual companies, which he calls “a REIT masquerading as an AI company and a crypto stablecoin business,” is a good sign for a traditional fast-growing software company like Figma.

AI Agenda Live NYC: The Next Wave

Tue., September 30, 2025

New York, NY

Learn more

“If you think about the lack of IPOs and the massive outperformance of Circle and CoreWeave, it tells you that people are chasing growth of new stocks,” Tunguz said.

AI Brings Higher Competition, Costs

The trouble is that such rapid growth might not be sustainable. Like many other software firms founded before ChatGPT, Figma faces a major threat from AI. Keeping its technology relevant while also maintaining profitability will be tough.

The company is banking heavily on two unproven strategies to sustain its rapid expansion. The first is a set of brand-new products and features it announced at its annual conference, Config, in May, many of which are still in beta testing. These include tools such as Figma Draw, a kit of illustration tools, and Figma Sites, a website builder.

It’s too early to tell how Figma’s new products will be received. The company might lose customers to companies with overlapping products like AI-powered app builders Replit and Lovable. Such tools could obviate the need for some of Figma’s design features, which is one likely reason why Figma released its own version, Figma Make, in May. Tunguz says that with an AI coding assistant by their side, software engineers are now capable of being designers.

Figma is also feeling the heat when it comes to serving users who aren’t engineers. There, Adobe could pose a real threat. Adobe had suspended work on its own online collaboration tools while it was in talks with Figma, which “gave Figma a little bit of an opening to expand,” said D.A. Davidson equity analyst Gil Luria. However, he added, “Adobe will not just cede the market to Figma and will continue to invest resources and be more competitive there.”

Luria’s view is supported by survey results released last month by another software analyst, Jackson Ader at KeyBanc Capital Markets. In a poll of more than 400 people who use Adobe Creative Cloud software regularly, 80% said they regularly used its AI tools—a higher proportion than Ader had expected.

There’s also mounting competition from Canva, another design software maker that has long been teasing the idea of going public. Dissatisfied Adobe customers whom Ader surveyed said they preferred Canva over other alternatives, including Figma.

Figma’s second strategy to boost growth is to increase prices, which it started to do in March. While that might help its revenue in the short term, it’s a risk that move could irritate some customers, hurting retention.

Plus, offering AI tools to customers is wildly expensive, partly because Figma has to incur the cost of routing each customer query to an AI model. Figma’s cost of revenue has been growing faster than its top line, and the company acknowledges that its investments in AI will negatively impact its margins.

One big point in Figma’s favor, though, is that it doesn’t have any substantial debt. It could potentially use the $1.5 billion in cash on its balance sheet and its IPO proceeds to buy up promising AI startups (though judging by recent AI funding rounds, doing so likely won’t be cheap).

These challenges aren’t unique to Figma—nor is it uniquely equipped to address them. That’s why Figma is probably not worth $20 billion, even if excited IPO investors are willing to pay that.

Boost to British businesses from contracts won through Safe project will require London to make financial contribution, diplomats say

Britain will have to pay the EU a percentage of the value of any weapons bought from UK companies via a Brussels-led defence fund, under plans being drawn up by the bloc to counter the growing threat from Russia.

Prime Minister Sir Keir Starmer in May said the UK would join the EU’s new €150bn Security Action for Europe (Safe) project to boost military spending across the continent, as part of a “reset” of bilateral relations.

Participating in Safe would provide “new opportunities for our defence industry, supporting British jobs and livelihoods”, Starmer said.

But London will be required to recompense Brussels for being allowed to take part in the EU-backed scheme aimed at procuring drones, missile defence systems and other capabilities, two EU diplomats told the Financial Times.

“What was written in the Safe regulation is that there shall be a fair balance as regards the contributions and the benefits” of outside countries such as the UK, said one of the diplomats.

The Safe fund is part of an effort by the bloc to mobilise €800bn in new defence spending by 2030, in the face of a growing threat from Russia and calls by US President Donald Trump for Europe to pay for more of its own security.

Launched in May, its loans will allow member states and designated “third countries”, such as the UK, to engage in joint procurement. 

The scheme is designed to leverage the EU’s credit strength to increase overall defence spending while better unifying Europe’s armed forces, which are saddled by inefficient and duplicative procurement processes. 

The exact figure that Britain will have to pay to access the fund is still under discussion by member states as they finalise their position on the deal with the UK, which is expected to be published this week.

But the diplomats said that since British businesses would receive EU money to create jobs and expand capacity under the scheme, London should recompense Brussels.

If UK companies won contracts funded by Safe money, the UK government must pay a percentage into the fund to help balance out the economic benefit of the contracts, the diplomat added.

The same mechanism would apply to Canada and any other countries that want their industry to access the money, they noted. 

In order for UK defence products to qualify, the value of their components from members of Safe must be at least 65 per cent. Members include the EU, Ukraine, Iceland, Liechtenstein, Norway, Switzerland and any third countries that join. 

Senior UK officials said France was driving a hard bargain in talks with other member states to agree an EU negotiating mandate this week. One described the situation as “tricky”. 

Mujtaba Rahman, managing director for Europe at the Eurasia Group consultancy, said both London and Brussels needed to keep the big picture in mind.

“It is really important that both sides don’t allow parochial national interests to get in the way of the bigger prize, which is strengthening collective European defence in light of the threat posed by Putin’s Russia,” he said.

France, which views Safe as a way to expand the EU arms industry, is pushing for a high UK contribution. But other countries, led by Germany, want to ensure the UK is not dissuaded from joining, a third diplomat said.

Under the EU’s terms, countries must bid for the loans by July 29, with a maximum pot of €150bn. They would then join another Safe member to buy weapons, aiming to lower prices by pooling demand. The UK would have to use national money to join such projects.

Third countries — other than those included by default, such as Ukraine — must first sign a security and defence partnership with the EU, and then a specific agreement to join Safe.

Britain signed the first of these at the reset summit in London in May, and will negotiate the second once the EU has agreed its mandate.

Time is tight because projects must be submitted by the end of November, with the European Commission, the EU’s executive, deciding which to approve. 

Thomas Regnier, commission defence spokesperson, said the EU-UK accord in May meant UK-based entities could provide up to 35 per cent of the value of a defence product procured through Safe.

To obtain a bigger share would require “an agreement with the EU on the precise modalities on aspects such as budget contribution and security of supply”, he added. 

The UK cabinet office said it would not pre-empt future discussions with the EU.

“It is in all our interests for the UK and EU to bring together our unique capabilities and expertise to make Europe a safer, more secure, and more prosperous place,” it added.

China will win AI race unless US cuts red tape, shale boss warns

Onerous permitting rules are holding back Trump’s ambitions for energy dominance, Toby Rice says

The head of the biggest US natural gas company has warned Congress it needs to cut project approval times to better compete with Russian gas exports and win the artificial intelligence race against China.

The comments by EQT chief executive Toby Rice add to growing concerns that America’s byzantine permitting process is driving up costs and project times for building infrastructure such as wells, pipelines and power plants — preventing the US from delivering the energy it needs to compete with its adversaries.

“Congress [needs] to step up and act,” said Rice in an interview with the Financial Times.

“The threat of not getting infrastructure built has only gotten larger — not only from bad actors getting rich by selling energy that could be replaced with American energy — it’s also the threat of China winning the AI race.”

Oil and gas companies have long bemoaned the pace of project approvals under US local, state and federal governments.

Of particular concern, Rice noted, is judicial review, which allows individuals and groups to legally challenge permit decisions for up to six years after an agency makes a decision. Previous attempts at permitting reform have suggested reducing that time window.

“We need to make sure we have judicial reform,” he said. “That would be incredibly impactful.”

Rice also blamed the fast-tracking of renewable sources, such as wind and solar, for driving up energy costs.

“When we spent the last 10 years ripping out coal, shutting down nuclear and making it more challenging to get natural gas infrastructure built, nobody should be questioning why prices are up and grid reliability is a major concern.

“[Congress should] get away from picking winners and losers,” he said.

President Donald Trump has made “unleashing American energy dominance” a priority for his second stint in the White House. In January, Trump lifted a Joe Biden-era ban on new liquefied natural gas export terminals along the US coastline and directed the Maritime Administration to expedite permits.

These actions come as the US races to meet growing domestic and global power demand caused by the data centres used to build and develop AI. The International Energy Agency predicts that electricity demand from data centres worldwide is set to double to 945TWh by 2030.

On Tuesday EQT signed an agreement in principle to provide gas to a 4.4GW plant that will power the Homer City Energy Campus, a 3,200 acre data centre in Pennsylvania.

Markets such as Europe have also been working to wean themselves off Russian gas since the country’s full-scale invasion of Ukraine in 2022, as well as turning to American imports to avoid tariffs.

On Wednesday one of Europe’s largest energy companies, ENI, signed a 20-year agreement to buy 2mn metric tonnes of LNG from Venture Global.

Trump’s flagship “big, beautiful bill” contained provisions capping the length of permit reviews, while the Department of Energy recently rolled back some environmental rules that it said were slowing down approvals.

Trump is believed to have revived the Constitution Pipeline, a 124-mile project that was cancelled five years ago after pressure from environmental activists, in exchange for lifting a stop work order on Equinor’s Empire Wind.

However, federal attempts at reform may still brush up against state-level restrictions.

“States that have been opposed to pipeline development, such as New York and California, can still use state level regulations to block projects,” said Eugene Kim, a research director on Wood Mackenzie’s Americas gas research team. “Despite all the Trump administration-led reform, we still need a lot more.”

The sector has also been hit by weak natural gas prices since the surge caused by Russia’s invasion in 2022. 

While acknowledging that “prices are always going to be an issue and one of the bigger factors on our activity levels”, EQT says it has cut its unlevered cost structure from $3 to $2 per million British thermal units (MMBtu) by increasing rig efficiency and the acquisition of Equitrans Midstream. Rice predicts that prices will reach $4 per MMBtu in 2026.

“Just for perspective, $4 natural gas is the equivalent of $24 oil, so we’re still providing an amazingly affordable energy solution,” he said. “This will catalyse a lot of demand and growth prospects.”

Big Tech can beat political parties at their own game

Elon Musk does not need to launch the “America party” — he already owns the tools to herd individuals into a critical mass

One could be forgiven for being sceptical of Elon Musk’s recent announcement that he will launch the “America party” in order to “give you back your freedom”. The last time Musk claimed to act on such lofty impulses, he spent a mind-boggling $44bn to acquire Twitter, only to see between 50 and 80 per cent of that money temporarily evaporate as what used to be a lively public square turned into a cesspool where users have to sift through torrents of waste in order to retrieve anything of value.

In retrospect, however, we may have got the Twitter acquisition wrong. It is increasingly obvious that Musk wasn’t interested in curating a free marketplace of ideas. What he saw in the platform was a formidable reserve of raw political energy, which needed only to be tapped, mobilised, and channelled into the political cycle. By acquiring Twitter, Musk bought himself a political party — potentially the biggest in existence.

Of course, the “America party” may never materialise and remain as much of a stunt as the prospect of colonising Mars. But its creation would only officialise what is already reality. It is the latest development in the evolution of Big Tech from a position of neutrality at the margins of American politics to its decisive role in pushing it towards the alt-right.

The politicisation of Big Tech started in 2015, when Airbnb fought off Proposition F, a ballot initiative meant to cap short-term rentals to remedy the housing crisis in San Francisco. Airbnb didn’t just pour millions into the campaign against Proposition F: it tapped its trove of data to churn out “Airbnb voters” almost overnight. The company relied on tested canvassing techniques but turbocharged them with the razor-sharp targeting of voters made possible by the capillary information it had about its users. Inaugurating a script that would soon extend to crypto and AI, it claimed a “victory for the middle class” while defending the interests of venture capitalists, shareholders and property owners.

The defeat of Proposition F was a pivotal moment. Tech companies realised they could beat political parties at their own game. Traditionally parties aggregate social interests and mobilise them during election cycles. But as class cleavages align less and less with ideological commitments, they have become “catch-all” machines increasingly competing for the same electorate. This is precisely what platforms do better: they can directly reach out to millions of voters and organise them into a critical mass on any given issue.

The America party wouldn’t be the first political party dependent on a digital infrastructure. Until 2019, the populist Five Star Movement in Italy was powered by a digital platform that allowed party members to take part in institutional decisions such as selecting election candidates. The platform wasn’t superimposed on to the structures of the party: it was its backbone. Provided by a company that previously sold workforce-management software, it stood for the virtues of decentralised participation while allowing for unprecedented levels of centralised control. In the 2018 national elections, the Five Star movement got the most votes and defeated well-established parties.

The limitation of the Five Star platform was that it strictly overlapped with party membership. This is not the case with X. Musk could grow a party or a movement from a social network counting over 100mn users in the US alone. In effect, the America party already exists, with every X user a potential member. Trump is wrong to see it as a third party with bleak prospects (and, being its first successful candidate, he should know better). It could be capable of colonising entire sectors of the Republican and even Democratic parties, and of acting either through them or on its own. This is not politics as usual.

Traditional parties brought people together around collective interests and common values. Their decline predated the advent of Big Tech. But they are now being eviscerated by the direct control platforms exercise over political socialisation. Private companies characterised by an unprecedented concentration of wealth and power get to decide what is politicised, when and how.

The America party, or something like it, would be one further step in this privatisation of politics. Its pretence of giving power back to the people is a sham for herding isolated individuals through sophisticated algorithms. It will be democratic in the same way the old Stalinist parties practised “democratic centralism” — except this time it will be in the interest of an infinitesimally small and yet immensely powerful class of oligarchs.

L'institution n'investira plus dans les fonds alternatifs, jugés trop chers et trop peu performants. Elle critique les pratiques du secteur, qui rognent les rendements des investisseurs.

Mieux vaut un bon sommeil qu'un bon lit. C'est en essence la conclusion à laquelle est arrivée l'université de Californie (UCLA), rapportent le « Financial Times » et le média spécialisé PI Online. L'institution, qui dispose de presque 200 milliards de dollars d'actifs, ne confiera désormais plus un sou aux hedge funds.

Ces derniers n'ont pas joué leur rôle de couverture lors des retournements de marché en 2000, 2008 et 2020, a rappelé à l'occasion d'une réunion Jagdeep Singh Bachher, responsable des investissements au sein de l'université. Au contraire, en prenant des risques inconsidérés, les hedge funds ont aggravé les pertes au sein du portefeuille. Un comble, car le principal argument de vente de ces stratégies complexes, peu liquides et surtout onéreuses est d'offrir un rendement décorrélé des conditions de marché.

Au cours des vingt dernières années, si les investissements dans les stratégies alternatives avaient été à la place dirigés vers des actions et des obligations, « le rendement aurait été identique pour un risque comparable, mais sans toutes les inquiétudes, l'illiquidité et les coûts » associés à la gestion alternative, a regretté Jagdeep Singh Bachher.

Les fonds dégagés par l'université seront donc placés sur les actions, dont le poids au sein des portefeuilles de l'institution passera de 53 % à 57 %. En moyenne, au sein de ses différentes poches, l'université visait jusqu'à présent une allocation de plus de 4 % pour les stratégies alternatives. Même si les montants investis étaient en réalité déjà bien plus faibles.

La décision de supprimer complètement l'exposition aux fonds alternatifs avait en effet été prise dès 2020. Mais alors qu'il faut en général un mois et demi à l'institution pour faire évoluer ses positions classiques, la « beauté » des hedge funds explique que ce processus ait pris cinq ans, a ironisé le responsable.

En somme, selon le responsable de l'université, les stratégies alternatives profitent d'abord aux gérants qui les exécutent. « Les hedge funds sont un secteur fantastique si vous êtes à Wall Street : vous pouvez faire payer des frais élevés et vous acheter toutes les oeuvres d'art, les jets privés et les belles propriétés », a plaisanté Jagdeep Singh Bachher.

La critique est récurrente. Au cours des dernières années, le modèle de facturation des gérants s'est éloigné du traditionnel « two and twenty », c'est-à-dire 2 % de frais de gestion auxquels s'ajoutent 20 % ponctionnés sur les profits réalisés. Désormais, les fonds optent pour des modèles de « pass-through », pour lesquels la totalité des frais sont refacturés aux clients - salaires mais aussi loyer, coûts de maintenance ou de déplacement…

Cela en dépit d'une performance de plus en plus poussive. Entre début 2015 et aujourd'hui, la performance moyenne des hedge funds dépasse à peine 60 %. Contre plus de 200 % pour l'indice élargi S&P 500 au cours de la même période.

LCHI, filiale de sélection de hedge funds d'Edmond de Rothschild calcule ainsi que les frais perçus par les fonds alternatifs ont représenté plus de la moitié de leur performance totale au cours des dix dernières années. « Je regrette, non pas d'avoir investi dans les hedge funds, mais de n'être moi-même pas gérant de hedge fund », a logiquement conclu Jagdeep Singh Bachher.

Toyota Motors is one of the world’s most buttoned-up companies. But in 2017, an executive from the carmaker’s venture arm drove into the remote Santa Cruz, Calif., mountains to see a group of engineers living together in a 1960s-style commune. As he watched, they demonstrated a remote-controlled electric helicopter they predicted would usher in an age of air taxis for the masses.

Eight years later, Joby, now a publicly traded company, says it will launch commercial air taxi flights in 2026. The first flights will be in four Dubai “vertiports,” including at the city’s airport, downtown and marina. Later in the year, Joby plans to start operations in U.S. cities, such as New York and Los Angeles.

And behind the venture is Joby’s largest shareholder, the ordinarily ultracautious Toyota, which holds a 22% ownership stake and has invested nearly $1 billion in the startup.

Joby’s tight partnership with Toyota gives it access to the carmaker’s unparalleled manufacturing expertise—critical since the startup’s success depends on building its aircraft at scale. Elon Musk famously called this problem “production hell.” Numerous electric vehicle startups have failed to navigate this stage, so Joby is tying its future to Toyota, the avatar of modern manufacturing and the world’s leading car company by sales.

I visited Joby’s Marina, Calif., manufacturing facility and wrote a long feature about how electric air taxis are moving from the realm of science fiction to reality.

Investors have responded strongly to the arrival of these aircraft, known as electric vertical takeoff and landing aircraft, or eVTOLs. Joby shares have surged 120% year to date, including 48% last week on the news that the company doubled production capacity in Marina. Later this summer, Joby plans to start production at a new plant in Dayton, Ohio, which will eventually pump out 40 aircraft a month.

Toyota’s largesse has bankrolled much of this expansion.

The four-passenger Joby aircraft looks definitively sci-fi, an intimidating combination of the insectlike body you would expect, topped by six helicopter-style rotors and wings. What fundamentally distinguishes these from today’s helicopters is their quiet, battery-powered motors. There is no scary whoop-whoop, and no dipping your head as you approach in fear of a decapitation.

CEO JoeBen Bevirt says that at first these flights won’t be cheap—he expects fares running about the same as Uber Black, which can cost roughly $150 or $170 from John F. Kennedy Airport to Manhattan. The price will come down, he expects, and Joby flights will compete directly with UberX, which is about half the Uber Black fare.

For Toyota, the size and intensity of the bet on Joby may seem surprising. But it makes sense if you’re thinking about survival amid a massive shift in how people move.

The Japanese company deliberately took a pass on EVs the last few years. Instead, it doubled down on hybrids, which it argued would be good enough for consumers, and it has largely ignored driverless technology. Meanwhile, Chinese EV makers have gobbled up market share in Asia and elsewhere.

Other EV laggards in the West are doing the same: Stellantis pledged to spend $370 million to help Joby rival Archer Aviation ramp up its manufacturing capacity. Porsche is helping Brazil’s Eve Air Mobility build its air taxi.

Geopolitics are aligning with these moves: Last month, President Donald Trump signed an executive order declaring electric air taxis and drones strategic industries, and ordering pilot programs to “maintain United States leadership in eVTOL flight.”

At the state level, Michigan, pivoting off Trump’s order, last week carved out a 40-mile air corridor in which eVTOL startups can test their aircraft, including a vertiport and charging infrastructure at the University of Michigan—an apparent first for the country; eVTOL companies generally don’t have large spaces for testing their aircraft.

Trump’s move followed a similar directive by Beijing, which even gave the new industry a name—the “low-altitude economy.” China is engaged in an all-of-government effort to dominate this economy: Six Chinese universities have announced fall undergraduate programs in low-altitude technologies, which also include drones and low-orbit satellites, and businesses have formed the China Low-Altitude Economic Alliance. Chinese electric air taxi developer EHang plans to begin the world’s first commercial eVTOL flights later this summer, with service in the cities of Guangzhou and Hefei.

Toyota is everywhere inside Joby. About 50 of its engineers are embedded in nearly every facet of Joby’s facilities. They heavily influence its production decisions, down to how workers place their tools on tables, and weigh in on the startup’s choice of suppliers.

I walked around the Marina plant with Eric Allison, Joby’s chief product officer. Allison and other senior Joby executives spoke of Toyota with extraordinary deference, saying in almost every conversation that Toyota played a central role in their plans. They cite the Japanese company’s culture, including acronyms like “5S,” a system of rigorous workplace rules.

The citation of Toyota’s acronyms seemed a bit cultish, though I can see why the Joby crew is embracing them, given how central manufacturing problems have been in the demise of other EV companies.

The Toyota hand goes pretty far: Overhead video cameras capture all of the workers’ moves. Toyota engineers review the videos later in Japan, where the company has built a close mock-up of the Marina facility.

“From one workstation to another, we would videotape and time every function, and then assess the effectiveness of every action, from which direction you put the scissors that someone’s going to access to how a handoff happens from one station to another,” Didier Papadopoulos, president of Joby and head of manufacturing, told me. “This is where I think the strength of Toyota and their decades of expertise has been phenomenal.”