FT : Musk’s xAI agrees Telegram tie-up as billionaire ‘bromance’ blooms

Musk’s xAI agrees Telegram tie-up as billionaire ‘bromance’ blooms
Messaging app’s 1bn users will gain access to Grok chatbot after Tesla chief met founder Pavel Durov

Elon Musk has agreed a $300mn deal with Pavel Durov, founder of Telegram, to distribute xAI’s Grok chatbot to the messaging app’s 1bn users, in a sign of blossoming partnership between the two mercurial billionaires. 

As part of a one-year deal, Telegram will receive $300mn in cash and equity from xAI, as well as half of the revenue from xAI subscriptions sold via the messaging app, Durov wrote on X and Telegram on Wednesday. Musk later wrote on X that a deal had not been “signed”, but Durov responded that they had “agreed in principal”.

Musk and Durov met in Paris recently ahead of the deal’s announcement, according to several people familiar with the matter, as their tech “bromance” grows. The Tesla and SpaceX chief executive shares with the Russian-born Telegram boss a passion for free expression and opposition to what they see as government censorship.

xAI’s Telegram tie-up marks the first big expansion for the artificial intelligence group into a new social media service beyond Musk’s X platform. It follows last week’s new alliance with Microsoft, which is making xAI available through its Azure cloud computing platform.

xAI bought X for $45bn in March. Telegram will own a stake in xAI as part of the two companies’ agreement.

The partnership will promote and integrate Musk’s chatbot in several ways inside the Telegram app, exposing xAI to new audiences and valuable data, as social media groups including Mark Zuckerberg’s Meta pour investment into the AI race. 

Last week, the Financial Times reported that Telegram was launching a bond offering of about $1.5bn in order to buy back existing debt. According to a presentation shared with investors and seen by the FT, the company made $1.4bn revenues in 2024, up from $343mn a year earlier, and made its first annual profit of $540mn. 

The rapid growth comes despite Durov being detained by French authorities last year. He was placed under formal investigation over Telegram’s alleged failure to address criminality including child abuse and terrorism on its platform. While the investigation continues, judges have ordered him to stay in France, although they have allowed him to go on business trips to Dubai.

Durov and Musk are aligned by their stances on free speech. In recent weeks, Durov has publicly criticised what he claims were French efforts to influence presidential elections in Romania where a far-right candidate was polling strongly. He said he had rejected a request by Nicolas Lerner, head of the French intelligence service, to close down channels of “conservative voices” in Romania — something the French agency flatly denied. 

Durov again criticised France while speaking remotely at the Oslo Freedom Forum after judges refused to let him travel there. He said authorities asked Telegram to shut down a channel of “far-left protesters and demonstrators” but he refused because it “seemed completely legitimate, and we refused to comply”.

He said he had done so “despite the obvious personal risks I’m taking with these decisions.”

The French interior minister did not respond to a request for comment.

Durov’s approach appears to borrow from the playbook of Musk, a self-declared free speech absolutist, who has increasingly used X to accuse certain foreign leaders of alleged censorship, challenging so-called “takedown” requests in countries including Brazil, India and Australia. He has also repeatedly accused both the EU and UK governments of censorship, and backed the far-right AfD party in Germany.

However, critics argue that Musk is selective in the challenges he takes up. 

On Sunday, the X owner shared Durov’s post about the French intelligence request on Romania to his 220mn followers writing: “Wow.”

xAI and Telegram declined to comment further.

FT : Letter: Don’t assume Ukraine is model for future wars

Letter: Don’t assume Ukraine is model for future wars
From Harlan Ullman, Chairman, The Killowen Group; Senior Adviser, The Atlantic Council, Washington, DC, US

Rana Foroohar’s “Defence budgets are being wasted” (Opinion, May 26), published on America’s Memorial Day, misses the strategic context for which military forces should be designed. Error number one is using the Ukraine war as the model for future conflict. The critical difference is that neither Ukraine nor Russia remotely possesses the military strike capability of the US.

When Vladimir Putin threatened the use of nuclear weapons in Ukraine, he was warned by the Biden administration that the US would and could exact a huge price by targeting Russian forces in Ukraine — something that the US was very capable of doing.

Clearly, Nato members should be amenable to adopting an alliance-wide strategy along the lines of a “porcupine defence” — using Ukraine’s example to put in place sufficient military might to make any Russian attack on the west too costly to contemplate. But they are not doing so. And, frankly, Russia lacks the capacity to launch an attack on the west and will be unlikely to do so for the better part of a decade.

Unmentioned and where Russia has a great advantage is in “active measures”, also known as “hybrid, grey zone and asymmetric war”, namely using cyber and infrastructure attacks along with dis- and misinformation and influencing operations to weaken the west. There, more money must be spent. But before we charge like the Light Brigade of old into drawing too many lessons from the Ukraine war, a little more introspection would not be a bad thing.

FT : Inside the Las Vegas bitcoin party: ‘Whales’, JD Vance and the Trump sons

Inside the Las Vegas bitcoin party: ‘Whales’, JD Vance and the Trump sons
The US president is a crypto booster — and the faithful went to America’s gambling capital to celebrate

Donald Trump’s two eldest sons came to America’s gambling capital to call for a new financial order — and the demise of the country’s biggest banks.

Eric and Donald Trump Jr took the Bitcoin conference by storm, promising a bonanza for digital asset holders, not least thanks to their father’s moves to lift the reins on their industry.

The brothers, stars in a firmament of crypto celebrities in Las Vegas this week, arrived to a stage named for Satoshi Nakamoto, the pseudonymous developer of bitcoin, after a string band’s rendition of techno anthem “Sandstorm”. They had Wall Street in their sights.

Eric Trump said to the massive, packed room at the Venetian hotel that “traditional finance was weaponised” against the crypto community and claimed that crypto transactions were “cheaper,” “faster”, “safer” and “more transparent.”

“I hate using the word hate, but honestly, I would love to see some of the big banks go extinct,” he said on Wednesday. “Honestly, they deserve it.”

It was an eye-popping comment matched only by the colour of the bitcoin faithful — orange — and the feverish mood of insurgent outsiders becoming rich.

The president’s sons have launched a series of crypto endeavours over the past year. World Liberty Financial created a stablecoin called USD1. Trump Media & Technology Group, which owns Donald Trump’s social media platform Truth Social, is raising $2.5bn to buy bitcoin. Eric Trump is taking public American Bitcoin, a crypto mining company. Their father’s memecoin — $TRUMP — raked in hundreds of millions of dollars.

“We’re very long crypto,” said Trump Jr. “It’s a huge part of everything that we do right now.”

The 2024 presidential election has changed the fortune of the crypto world once seen as a wild west of speculation and unregulated finance.

At last year’s Bitcoin conference in Nashville, Donald Trump promised to make the US the world’s “bitcoin superpower”. This year, his vice-president, crypto tsar, digital assets adviser and sons took over the event to pitch a new financial order.

Chris LaCivita, one of the managers of Trump’s 2024 campaign, said “so many friends” at the conference had been “instrumental in helping Donald Trump get elected president”. LaCivita joked that if Trump had lost, he and other supporters of the president would have needed to find refuge outside the country.

“I can say this: that had we not been successful, including myself, we would’ve been looking at a country with a non-extradition treaty,” said LaCivita.

Crypto enthusiasts thought the Biden administration was at “war” with their sector. Trump, a memecoin vendor himself, has been much kinder.

His administration has ended many of the Biden-era investigations into crypto companies, pushed legislation to pave the way for widespread adoption of digital assets — even in people’s pensions — and wants the US government to directly acquire bitcoin. 

Senior Trump officials pledged both a strong US dollar while pitching an alternative to the world’s reserve currency. JD Vance, Trump’s Silicon Valley-friendly vice-president, told the Las Vegas crowd that stablecoins were “a force multiplier of our economic might” rather than a threat.

“It’s only going to help the American dollar,” said Vance.

Bitcoin was created 16 years ago as a new, alternative, “decentralised” digital currency, born out of distrust for established currencies backed by governments and central banks. It quickly became an asset favoured by speculators, former gold bugs and criminals.

The cryptocurrency has been prone to sharp swings in value. But it has now come of age — endorsed not only by BlackRock, the world’s largest asset manager, but also the US president himself.

Last week, bitcoin reached a record high of more than $111,000. Eric and Donald Trump Jr said in Las Vegas that bitcoin would reach more than $170,000 within a year. 

The bitcoin enthusiasts in Vegas extolled its mission as a high-minded pursuit, while revelling in base tendencies. Vance said bitcoin would be “involved in the strategic future of our country” as the US competed with China. But he was not complimenting the crowd to “juice my own memecoins”, he joked.

Some crypto executives accused the president — and his wife, Melania — of doing just that when they launched their own memecoins earlier this year.

Near Nakamoto stage, the billionaire Winklevoss twins — co-founders of crypto platform Gemini — sped across the Venetian hotel’s cushioned carpets. A man in an orange bitcoin hat asked the bartender at Grand Lux Cafe about the credit card he left behind the previous night. Members of Congress wandered the pathways surrounding the indoor, faux-Venetian canal.

Inside the conference entrance Gemini pitched “the bitcoin credit card” and offered a sweepstakes with two orange Tesla cybertrucks as the prize. 

There was an auction of Ross Ulbricht’s items from prison — ID cards, his paintings, sweatsuits. Trump pardoned the darknet Silk Road creator, who was sentenced to life in prison, in one of the first official acts of his second term. A copy of the first issue of Bitcoin magazine was on sale for $10,000.

A $21,000 conference ticket would make you a “whale” — a reference to the big investors that some people believe influence the price of digital assets — for the week. Privileges included skipping queues, finer food and entry to velvet rope parties for the chance to meet crypto billionaires such as Justin Sun and his colossal security detail. 

In a private, whales-only area called The Deep, crypto executives put questions to government officials such as Bo Hines, Trump’s 29-year-old crypto adviser, entered a sweepstake to win Louis Vuitton luggage, played pool and signed up to fly to outer space on Blue Origin’s New Shepard.

Advertisements for Geoship dome houses and “Rare”, a platinum- and diamond-encrusted sculpture in the shape of a steak, were on the high-top tables.

On stage, a series of speakers from “don’t die” investor Bryan Johnson to embattled New York mayor Eric Adams pitched the bitcoin acolytes.

White House crypto tsar David Sacks urged Treasury secretary Scott Bessent and commerce secretary Howard Lutnick to find budget-neutral ways for the US government to acquire bitcoin.

Sacks asked the audience, “Can you guys get their attention?”

FT : Kalshi, the prediction market encroaching on the $14bn US sports betting in

Kalshi, the prediction market encroaching on the $14bn US sports betting industry
Start-up advised by Donald Trump Jr emerges as new threat to sector’s dominant players Flutter and DraftKings

Flutter and DraftKings are weighing their responses to a new threat to their dominance of the $14bn US sports betting sector, as prediction markets led by politically connected Kalshi exploit a regulatory quirk to offer wagers on contests even where gambling groups cannot.

Prediction sites allow users to stake money on anything from economic data to the weather. Because their products are structured as peer-to-peer trades, providers are regulated not as betting groups but derivatives platforms, allowing them to bypass sports gambling bans in 11 states.

Kalshi counts Donald Trump Jr, son of the US president, as a strategic adviser. Board member Brian Quintenz is President Donald Trump’s nominee to chair the Commodity Futures Trading Commission, the derivatives regulator whose remit includes oversight of prediction markets, although he has pledged to step down from the start-up should he be confirmed.

The business made its first incursion into sports predictions in January with contracts on the following month’s Super Bowl that it said generated $27mn in trading volume, while subsequent trades on the March Madness college basketball tournament generated $500mn. Kalshi now offers prices on a vast array of events, from this summer’s Wimbledon tennis tournament to next year’s football World Cup.

The sudden arrival of a new type of competitor threatens to disrupt an industry that has grown rapidly since the Supreme Court overturned a near-nationwide ban on sports gambling in 2018. Traditional sports betting generated $14bn in US revenues last year, according to data provider Yield Sec.


DraftKings and Flutter — which account for more than 80 per cent of the US sports betting market, according to Barclays — are now considering how to fight back.

“This is something everybody is talking about right now . . . it is happening whether you want it or not,” said Jason Robins, chief executive of DraftKings, adding that the company was monitoring the development of prediction markets, which “warrant thoughtful consideration”.

Dublin-based Flutter, which is listed in New York and offers sports bets to Americans through its FanDuel business, has moved some of its engineers from UK division Betfair to the US to develop plans on how to respond to the new threat. Peter Jackson, chief executive, said the company was “looking very carefully” at the sports contract market and thinking of “how to attack the prediction market”.

Both companies this month lowered their full-year US forecasts after a second straight quarter in which sports betting profits declined.

Platforms offering or seeking to offer trades on sports events including Kalshi, Crypto.com, Sporttrade and Robinhood have “snuck up on the gambling companies . . . it just wasn’t on their radar”, according to Rick Arpin, head of US gaming at consultancy KPMG. “It is not like anything we have seen in a long time, it is unique . . . it could change how gambling companies operate in the US forever.”

Industry observers have speculated that traditional gambling groups will look to secure licences from the CFTC to offer prediction markets, or simply buy one of their small rivals. But some analysts say the newcomers have first-mover advantage — especially in states where legacy groups cannot operate such as California, Florida and Texas.


In new markets, “one of the most important ingredients for future success is being there from day one”, said Brandt Montour at Barclays. “Customers will download one app and go from there . . . they won’t download a new app every time some company launches their product.”

DraftKings and Flutter had little choice but to respond, he added, saying the newcomers are “a critical threat . . . It is something they can’t ignore.”

Trades on Kalshi are structured as short-term derivatives contracts on a binary outcome — for example whether the US will, or will not, win this year’s Ryder Cup golf competition against Europe. Users open a Yes or No position, priced at the time of writing in the golf example at $0.58 and $0.45 respectively, with all contracts paying out $1 if the choice proves correct. The platform takes a cut of each transaction.

Derivatives exchanges have to adhere to US financial rules but their overall compliance costs are lower than those of gambling groups, which are subject to additional state taxes and anti-addiction rules.

Privately owned Kalshi, which was last valued in 2022 at $787mn, according to PitchBook, does not disclose details such as revenues or user numbers.

Its backers include Alfred Lin, a partner at venture capitalist Sequoia Capital, and billionaire investor Charles Schwab, whose granddaughter Samantha Schwab was named deputy chief of staff at the US Treasury soon after Trump took office in January. The start-up’s former chief regulatory officer Eliezer Mishory left in March to join Elon Musk’s so-called Department of Government Efficiency and now leads its team seeking cost savings at the US regulator Securities and Exchange Commission, according to people familiar with the matter.

Another new entrant to the market, Philadelphia-based Sporttrade, this month sought the CFTC’s permission to offer its sports events contracts nationwide. Alexander Kane, its chief executive, said prediction markets would be “disruptive” and hit legacy operators’ earnings.


The disrupters may benefit from a perception among their users that prediction markets, which make money regardless of whether users win or lose, are fairer for gamblers than traditional betting sites, where the wager is against the house with the odds stacked against the customer.

Legacy operators “have brought this upon themselves”, said Matt Zarb-Cousin, a prominent campaigner against gambling harms. The traditional business model of online gambling, in which revenue is generated by users’ losses, is “commercially incentivised to take as much as it can from the consumer”.

However, the newcomers face legal challenges. Seven states have sent cease-and-desist letters to Kalshi, which they say did not comply with their laws. The company has responded with lawsuits against the gaming regulators of Nevada, New Jersey and Maryland over its right, “subject to the exclusive jurisdiction of our regulator, the CFTC”, to offer sports event contracts in their jurisdictions.

Kalshi last year won a battle with the CFTC to allow trades to be offered on the outcome of elections. Analysts say a similar victory on sports event contracts may open the door for other smaller players to enter the market.

The company argues that its federally regulated contracts supersede state laws. “We’re like, hey, you believe you have some legal theory? Let’s just prove it immediately in court,” Tarek Mansour, Kalshi’s co-founder and chief executive, told the Financial Times.


With federal courts in Nevada and New Jersey already having ruled in the company’s favour in preliminary injunctions, campaigners are turning their attention to ensuring the CFTC oversees the market to the same standards as state gaming regulators.

“There absolutely has to be recognition from the CFTC that gambling harm can arise” from prediction markets, said Zarb-Cousin.

The wins in Nevada and New Jersey have put fresh scrutiny on the stance of the CFTC. Robinhood had planned to offer contracts for the Super Bowl but withdrew them after a request from the regulator, which said it had “serious concerns” about event contracts and vowed to “exercise its oversight authority to the fullest extent as appropriate”.

The broker said it would follow regulatory guidelines and laws but that this may change if Kalshi wins its cases against the states and is permitted by the CFTC to continue offering sports events contracts.

Crypto.com suspended its Super Bowl contracts after being asked to do so by the CFTC. No such request was made of Kalshi, which in the view of US sports wagering lawyer Daniel Wallach operates “almost in a monopolistic vacuum” and had undertaken “a number of moves designed to stack the regulatory deck in its favour”.

While Kalshi has ties to the current administration, Mansour — who has posted pictures of himself with members of the Trump family on more than one occasion — insists it has no political affiliation.


Flutter and DraftKings declined to comment on Quintenz’s nomination as chair of the CFTC but one insider at a leading US gambling operator said their company was concerned about the potential conflict of interest.

Mansour said “anything that ethically he is going to have to do, he’s 100 per cent going to do and he is already doing it. I don’t have any kind of concerns around that.”

He added that Trump Jr’s views on prediction markets were “very” aligned with his, and that “Don has been a huge proponent of this . . . he’s helping us with go-to-market and general expansion”.

“This year our growth has been incredible — I believe we have the potential to build the largest financial market ever,” said Mansour, adding that “more people care about elections or sports or simple questions about the weather than they care about options trading”. 

He admitted there was “work to be done” on customer protections. “We don’t want them to lose the money . . . We have and should have the same types of safety measures” as those required of traditional gambling groups.

But he said he had the impression “a lot of people” did not want his company to succeed because it “may mean taking market share away from them”.

The CFTC, Trump Jr and Quintenz did not respond to requests for comment.

FT : The $1tn shadow bank lending boom

The $1tn shadow bank lending boom
‘This is a sensitive trade-off that has not always been well managed in the banking industry’

US banks are lending less and less to companies these days. But the business of lending to so-called shadow banks — such as private credit funds, insurers, asset managers and credit hedge funds — is booming.

On the face of it, bank lending to “commercial and industrial” companies (C&I loans) has merely stagnated in recent years, at about $2.8tn at the end of March.


However, as a percentage of bank assets and relative to the size of the US economy, C&I loans have now been shrinking for half a decade.

That’s not because US companies have suddenly discovered the virtues of resilient balance sheets and begun to borrow less. No, it’s because banks have begun to lend indirectly to many of the same companies by instead making loans to “non-bank financial institutions”.

NBFIs is currently preferred if obtuse term for what we used to call shadow banks — a huge and rapidly growing universe of funds, investment firms and financial companies of various shapes and sizes that raise money from other investors and then make loans with them.

According to a new report published by Barclays’ macro, credit and bank research analysts, US bank lending to these NBFIs has quintupled over the past decade to well over $1tn, and now accounts for more than 10 per cent of all US banking loans (and nearly 5 per cent of all assets)

In other words, as banks are making fewer direct loans to customers, their indirect lending via NBFIs has grown. Banks are now increasingly “lending to lenders,” as they replace C&I loans with NBFI loans. This is indirect lending because the funds provided to NBFIs ultimately flow to the same underlying borrowers (in the aggregate).

This is not a shocking revelation, even if the numbers are eye-popping. The ascendance of NBFIs is one of the most widely talked about subjects in finance, and has been for several years. And although they are often rivals, the fact that conventional banks are deeply enmeshed in the shadow banking ecosystem is well-known.

However, as the Barclays analysts note, it is becoming an ever bigger and broader trend, with no slowdown in sight. The larger US banks have generally been more active in lending to NBFIs, but mid-sized and smaller ones are now also entering the field.


Please use the sharing tools found via the share button at the top or side of articles. Copying articles to share with others is a breach of FT.com T&Cs and Copyright Policy. Email licensing@ft.com to buy additional rights. Subscribers may share up to 10 or 20 articles per month using the gift article service. More information can be found at https://www.ft.com/tour.

Why are banks suddenly so keen to finance what on the face of it would seem to be competitors? After all, becoming fully or partly cut off from corporate clients could hit a lot of other business areas. As Barclays points out: “Banks sell myriad financial products to their clients; often, a loan is the start of a relationship that can include many other fee-generating services.”

It’s pretty simple really. To paraphrase Method Man, regulations rule everything around us. Or, if you prefer, here is Barclays’ version of the same, with Alphaville’s emphasis in bold:

Banks are required to risk-weight their assets when calculating their capital requirements. In the US, risk weights are highly standardized, with most commercial exposures attracting 100% treatment. However, NBFI loans are eligible for much lower risk weights, as low as 20%. The lower RWA density of NBFI lending reflects a number of factors, including high collateral requirements (ie, low LTVs) and in some cases covenant protections. Many NBFI loans are structured, allowing banks to take senior exposure to a relatively diversified pool of assets. This lending format is similar to the standard in the investment grade securitized credit market. Collateralizing obligations in this fashion ensures that loans to NBFIs have a defined amount of equity beneath them to absorb loss. This generally results in NBFI loans having relatively low loan-to-values (LTVs). These creditor protections support lower risk-weighting of NBFI loans, which makes them less capital-intensive assets in the calculation of regulatory capital ratios (ie, CET1/RWA).

The corporate lending shift from banks to shadow banks might even be understated by the public data, as “synthetic risk transfers” — when a bank buys protection against credit losses from investment funds — are not included in the NBFI lending disclosures, even though they are functionally the same, Barclays notes.

The increase in these SRTs demonstrates the importance of the capital relief angle for banks. If the NBFI trend was being driven solely by increased competition among lenders, then banks would not be simultaneously executing synthetic versions of the same transaction. It also implies that the trend toward lending to lenders is larger than indicated using the official loan data, because the total volume of SRTs is large and growing globally.

So is Barclays as worried about this as the IMF, FSB, ECB, BoE, Fed, the pre-Trump FSOC and SEC etc etc all are? Kinda, yes!

Jeffrey Meli, Bradley Rogoff and Peter Troisi stress that while there are parallels with the pre-2008 era mortgage boom — when banks gleefully gamed capital rules and ended up ruing it, badly — there are also important differences.

First of all, the underlying corporate loans have historically had much higher recovery values than mortgages. Secondly, the NBFI really does have genuine skin in the game, which shields the bank from losses. Thirdly, capital requirements are higher across the board, so banks have a far bigger cushion to absorb blows than they did in 2007-08.

“That said, there are important concerns to consider as this trend grows,” Barclays dourly notes. Alphaville’s emphasis below:

The probability of a correlated and severe shock to borrowers is not a fixed number. It reflects the structure of the economy, including — maybe the most important of all factors — the amount of leverage in the system. More overall leverage means that less severe shocks can cause greater losses. As NBFI lending grows, so too does leverage. As stated above, banks are precluded from certain types of leverage; NBFIs are growing in part because they face no such constraints (and the banks are providing additional leverage through the NBFI loans themselves). There is also greater use of off-balance sheet leverage, particularly among investment grade companies.

Put differently, while the current systemic risks of NBFI lending may be small, they will grow with the size of that lending. Increased use of the leverage provided by NBFIs could eventually increase the probability of a severe shock that causes correlated losses (and corresponding low recovery rates), and thus affects banks, to a point where the capital is no longer sufficient. The capital charge may be adequate now, given the historical experience, but at some point the historical reference points are less useful. An analogy is that, prior to 2007, mortgage defaults had never caused a vicious cycle of declining home prices, leading banks, regulators, and investors to heavily discount the probability of such an event. But the leverage in the housing system in 2007 exceeded all historical experience and rendered that history irrelevant.

FT : How Wall Street offloaded $13bn of debt tied to Elon Musk’s Twitter deal

How Wall Street offloaded $13bn of debt tied to Elon Musk’s Twitter deal
Donald Trump’s election helped revive loans that once threatened to hit banks with big losses

Elon Musk promised his bankers they would not lose money on the $13bn they lent to finance his Twitter takeover in 2022. The world’s richest man has now largely made good on his vow after Donald Trump’s re-election helped rescue the debt deal.  

The sale by seven banks last month of the final slug of loans for Musk’s $44bn Twitter buyout marked an extraordinary turnaround for debt that had once appeared to be toxic. The lenders including Morgan Stanley and Bank of America faced a stark choice: offload the loans at a steep discount and lock in big losses, or trust guarantees Musk made in private. 

Their decision to hold the debt began to pay off in November, after Trump won the US election. Musk, one of the president’s closest allies, became a fixture in the White House — a shift that has elevated his influence and rippled across his business empire.

As bankers offloaded the final vestiges of the debt financing at the end of April, X took the unusual step of covering some losses that banks would usually absorb, people with knowledge of the matter said.

While the $1.2bn of loans were priced with a small discount, at 98 cents on the dollar, X made up the difference, the people said.

“I could do cartwheels,” one banker involved in the deal said. “It was a bet on the world’s richest man, and it paid off,” a second person said.

—

This account is based on more than a dozen interviews of people familiar with the financing of Musk’s Twitter purchase, which marked one of the most difficult debt deals for Wall Street since the 2008 financial crisis.

Morgan Stanley had rallied six other banks as it raced to provide financing to Musk in April 2022, with lenders tripping over each other to cement their ties with the billionaire.

The cohort, which included Bank of America, Barclays and MUFG, had just days to conduct due diligence and sign up to the deal as Musk pushed forward with his hostile takeover.


But just weeks later, with Musk attempting to back out of the buyout, it became apparent that they would have to provide roughly $12.7bn of loans themselves instead of selling the debt to large investors. They were “hung”, in Wall Street parlance.

As X’s business deteriorated, the banks were left holding billions of dollars of loans that were sinking in value. Federal Reserve rate rises in 2022 only made matters worse — as the banks had agreed to cap the interest rate on the debt they extended to X.

Investors, sensing the despair at the seven banks, began bidding on the debt. In 2023, some offered just 60 cents on the dollar, which would have implied losses of more than $4bn.

Morgan Stanley held the lenders together, betting on the relationship of one of its then top bankers — Michael Grimes — and his client Musk, people familiar with the matter said. The bank led weekly calls to update the other lenders on X’s performance. Working together as a bloc helped avoid one single bank abruptly selling the debt at a discount.  

“The banks knew if someone dumped a portion at 50 cents on the dollar they’d all have to take a massive hit,” one debt investor who did due diligence on the loans said.

Everything changed on election night 2024, which featured scenes of Musk side-by-side with a victorious Trump. The banks spotted their exit and Morgan Stanley finally began leading serious conversations with investors. Bids that were just 70 cents on the dollar before the election had shot up 75 to 80 cents on the dollar.

Morgan Stanley declined to comment, while X did not respond to requests for comment.

The importance X played in the election was not lost on the bankers and a rally in debt markets was only helping their book. Musk’s decision to give X a stake in his fast-growing artificial intelligence company xAI also burnished the social media company’s prospects.

“He’ll never let Twitter [X] fail,” a banker involved in the deal said. “I don’t think people even did their credit work. They just trusted the Musk halo.” 

In January, the banks quietly sold the first chunk of debt they had been holding for years to two investors, Diameter Capital Partners and Darsana Capital Partners. The firms bought $1bn of the loan at 93 cents on the dollar, according to three people familiar with transactions. 

The sales to Diameter and Darsana helped “build interest” from other investors, one person who bought the debt said of Diameter’s role.

Bankers were enthusiastic with the result — and began contacting the world’s largest credit shops. Before they would share any of X’s financial information and performance, they wanted to know that an investor could buy $250mn or more of the debt.

Some investors who looked at the debt were aghast with the information they were given. Financials were heavily adjusted and while the company’s revenues looked to have bottomed, the data on offer lacked many of the key figures most investors would need to conduct due diligence. Several told the Financial Times the business had shed cash at a rapid rate.

But buyers including Apollo Global and Citadel lined up and Morgan Stanley successfully sold $5.5bn of term loans in early February, this time at 97 cents on the dollar. The price was important. Banks generally break-even on their underwritten debt deals at 96 or 97 cents on the dollar, as the fees they earn offset the discounts provided to entice buyers.

“They were able to sell it off some of the thinnest financials,” one investor who passed on the deal said. “It was surprising, you could do it as a trader — but to buy $1bn, which they want us to do, we don’t do that . . . It was f*cking wild.”

“Twitter was not an investment, it was a trade,” a second investor who bought some of the loans said. The person added it was “too early to know if they’re through the trough”.

Just days later the banks returned and sold another $4.74bn of fixed-rate term loans at par — or 100 cents on the dollar — earning their fees on that portion of the sale. Demand was so strong that the size of the sale was increased, leaving the banks with roughly $1.2bn junior loans — the riskiest part of the financing — they had agreed to provide.


In April, the banking cohort sold the final $1.2bn of debt, winning a reprieve from lenders to convert those junior loans into more senior debt, slashing the company’s interest costs in the process.

Morgan Stanley’s first-quarter earnings were bolstered by the debt sales, with the US lender reporting nearly $700mn of “other” revenues within its investment bank, up from $242mn a year earlier. A person familiar with the matter told the FT that much of the boost was related to selling X debt. The bank had previously taken mark-to-market losses on the loan. 

“Our investment thesis was correct,” a banker involved in the deal said. “We backed Musk more than we were backing Twitter itself.”

Still, another banker complained that the offloading of the hung loans was not reflected in their compensation.

“We made fees. We wrote the loan back up. But we didn’t see that reflected in bonuses,” one banker said. But they acknowledged: “You can’t ring the bell on something you had to hold for two years.”

>>> US After Hours Summary: NVDA +5%, VEEV +15.7%, AI +10.7% higher on earnings;

After Hours Summary: NVDA +5%, VEEV +15.7%, AI +10.7% higher on earnings; HPQ -13.3%, S -12.2%, PSTG -3.7% lower on earnings

After Hours Gainers:

Companies trading higher in after hours in reaction to earnings/guidance: VEEV +15.7%, AI +10.7% (also announces renewal and expansion of JV with BKR), A +5%, NVDA +5% (also files mixed securities shelf offering), SNPS +3.2%, NDSN +3%, NCNO +0.8%, ELF +0.6% (also to acquire rhode in $1 bln deal), CRM +0.5%

Companies trading higher in after hours in reaction to news: ORIC +21.4% (data from ongoing phase 1b Trial of ORIC-944; also announces $125 mln private placement financing), TSHA +13.5% (announces pivotal part B Trial Design Details for TSHA-102; also stock offering), POWW +7.1% (names new Chairman/CEO), ALC +4% (FDA Approval of TRYPTYR), MCY +2.2% (COO bought 15000 shares), GOGL +1.8% (CMBT and GOGL to merge), ETON +1.8% (FDA approves KHINDIVI), BGM +1.5% (acquires Xingdao Intelligent and YD Network), TSLA +1.3% (has targeted June 12 launch of Robotaxi service in Austin, according to Bloomberg), JACK +0.5% (names new CFO), SDHC +0.3% (authorizes new $50 mln share repurchase program), LMT +0.2% (awarded a $509.8 mln modification to a previously awarded US Air Force contract)

After Hours Losers:

Companies trading lower in after hours in reaction to earnings/guidance: HPQ -13.3%, S -12.2%, PSTG -3.7% (also CFO to step down), NTNX -0.9%, ATS -0.1%

Companies trading lower in after hours in reaction to news: TATT -5.3% (commences 4.15 mln share offering, comprised of 1.625 mln shares being offered by TATT and 2.525 mln shares being offered by selling shareholders), HESM -2.4% (stock offering by selling shareholders), AS -0.9% (announces commencement of 35 mln share offering by selling shareholder), CMBT -0.8% (CMBT and GOGL to merge), VSCO -0.4% (has stopped some operations amid a "security incident", according to Bloomberg), KAI -0.2% (awarded $18 mln in orders), MELI -0.1% (Mercado Pago will apply for a banking license in Argentina)

FT : Finance is ready for a blockchain reset

Finance is ready for a blockchain reset
This is not about replacing national currencies or eliminating banks but creating interoperable new infrastructure

The modern financial system is undergoing a foundational stress test. Decades of globalisation combined with increasingly fragile institutions have given way to a period of volatility marked by inflation shocks, debt overhangs and declining confidence in centralised authorities. Cross-border payments remain inefficient, sovereign currencies face growing scrutiny, and trust — long held together by central banks and legal regimes — grows ever more fragile in a fragmented world.

This is not a transient crisis, but a signal of architectural fatigue.

A restructuring is required. In the 1990s, global information systems underwent their own architectural reset. New protocols like HTTP acted as rules that determined how computers communicated across networks. They created a common foundation that enabled co-ordination. The result was the internet: an open network that no one owns and everyone can use.

Financial systems have not yet experienced this sort of revolution. But blockchain-based systems, a new category of financial infrastructure, could facilitate one.

At their core, blockchain networks like ethereum and bitcoin enable the movement of value in the same way the internet enables the movement of information. You can store, transmit and manipulate real world value in a global digital context, sending it through blockchain transactions as easily as sending an email.

The newer class of blockchain-based networks, including smart contract application platforms, allow for the movement and management of digital assets like cryptocurrency, as well as proof of identity and contract agreements, without relying on traditional intermediaries.

Unlike payment networks, they are not run by individual corporations or governments, but by decentralised networks that use cryptography to come to a consensus on the veracity of the entries, and which then ensure that the transaction history recorded is tamper proof.

Institutions as varied as BlackRock, Apollo Global Management, Franklin Templeton and JPMorgan are already offering tokenised assets and settlement processes on blockchain. The technology is no longer speculative. It is operational.

Of course, early adopters engaging with blockchain infrastructure know that it requires ongoing technical evolution to scale and support global throughput and usability. The upgrades are complex. Still, ethereum has undergone over a dozen major upgrades since inception nearly 10 years ago without downtime or compromise of on-chain assets. The result is a platform that, while still evolving, has proven technically resilient and thus increasingly credible to institutions.

But beyond the technical design, the broader philosophical shift is noteworthy too. Decentralised systems reframe trust as something that can be embedded in infrastructure, not granted by institutions.

Trust can therefore be understood as a new kind of commodity, and decentralised trust is the highest octane, gold standard of that commodity. In a world where global co-ordination is increasingly difficult and political consensus is brittle, systems that minimise counterparty risk by design become more compelling.

This is not about replacing national currencies or eliminating banks. Rather, it’s about creating interoperable layers of financial infrastructure that can coexist with existing systems and offer a path to lower friction, broader access and stronger resilience for financial systems.

The use cases extend beyond capital markets. Digital identity, intellectual property rights, payment rails for emerging economies, even machine-to-machine transactions by autonomous AI agents will all require infrastructure that can operate beyond the constraints of national borders. Much of the world would not function without the internet; much of the future economy will not function without these blockchain-based networks.

Some perceptions of the cryptocurrency industry have been marred by speculative excess, price volatility and high-profile failures. But separating speculative assets from the infrastructure behind them is critical. The underlying protocols are defined by the quality of their design and their capacity to enable co-ordination.

We are entering a multi-polar world with contested governance and overlapping regulatory regimes. In that environment, neutral, programmable infrastructure is no longer a luxury. It is a necessity.