TheChek News Flow

Anthropic offers Blackstone double-digit yield in latest chip financing

High borrowing costs complicate AI lab’s goal of achieving investment-grade status

Anthropic is offering Blackstone a high yield of more than 10 per cent to finance its latest chip deal, highlighting the start-up’s willingness to stomach steep borrowing costs to finance expensive AI infrastructure. 

The $60bn debt package, dubbed “Project North Star”, would help fund Anthropic’s procurement of about 1.5 gigawatts of custom semiconductors developed by Google, known as tensor processing units.

It is expected to be the largest chip financing to date and a crucial test of investor appetite in AI debt at a time when lenders are increasingly nervous about tech companies’ heavy capital spending and borrowing. 

Blackstone funds would invest $9bn in the debt backed by Anthropic’s five-year lease payments on the chips, accounting for about half of the $18bn junior loans, according to people familiar with the matter. 

The private capital giant was offered a coupon of 10 per cent and an even higher total return after accounting for new-issue discounts and other fees, the people said. A borrower with a similar credit profile typically pays a yield of about 7 per cent in the public bond market, according to the ICE BofA BB US High Yield Index. The rest of the $9bn would be sold to a broader group of investors in the coming months, potentially with a lower yield, the people added. 

Blackstone declined to comment. Anthropic did not immediately respond to a request for comment.

Just four months ago Anthropic sold about $4.4bn of debt in a similar deal known as “Project Big Sky” at an interest rate of 8.5 per cent, the FT reported at the time. 

The rise in borrowing costs underscores investors’ demand for higher risk premiums to finance the AI build-out, as they grapple with tech companies’ huge funding needs over a short period of time. Tech groups are also paying up to secure capital in mega-sized financings in order to attract anchor investors and early commitments. 

It also highlights the challenges facing the five-year-old AI lab as it aims to obtain an investment-grade rating. Blackstone told its investment partners that the latest chip deal would be a rare opportunity to earn high yield on Anthropic-linked debt, as it is expected to receive an investment-grade rating and vastly lower its borrowing costs once going public, the people added.

Anthropic was assigned a one-notch-below-investment-grade rating by KBRA in June, and the latest junior debt is expected to receive something similar. 

The new debt package also includes about $42bn of senior debt guaranteed by Broadcom, which helps develop the chips with Google. When the capital has been drawn down in 2027, the debt could become tradeable in the institutional 144A bond market. 

SoftBank in talks to buy US security group Allied Universal

Masayoshi Son looks to increase pace of his bets on robotics and AI

SoftBank is in talks to buy Allied Universal, the world’s largest private security group and owner of the UK’s G4S, as Masayoshi Son looks to accelerate his already sprawling bet on robotics and AI.

The potential deal for Allied could value the security company at close to $30bn, including debt, according to multiple people familiar with the matter. However, talks are still at an early stage and there was no guarantee that an agreement would be reached, they added.

A successful acquisition would allow Son to build up his growing bet on robotics. That includes the $5.4bn agreement to buy ABB Robotics — which is pending regulatory approval in China — as well as the planned listing of physical AI company Roze in the US.

Son has become increasingly interested in humanoid robots as part of his belief in the future of AI. He has long discussed with his advisers the idea of using autonomous robots to provide security services. Allied also provides security for data centres, another growing part of the SoftBank empire.

Allied, which is backed by private equity firms including Warburg Pincus and Canadian pension fund La Caisse, says it operates in more than 100 countries and employs close to 770,000 people.

It has been considering an initial public offering for several years, a possibility its chief executive Steve Jones has discussed publicly.

Warburg has been invested in Allied since 2016 before upping its investment three years later when La Caisse bought into the company. Collectively the pair own more than two-thirds of the company. Allied bought UK-based security group G4S in a $5.3bn deal in 2021.

Warburg, La Caisse and SoftBank declined to comment. Allied did not respond to a request for comment. The news of the potential deal was first reported by Bloomberg.

A potential sale of Allied would mark a mammoth private equity exit when many financial sponsors are struggling to offload long-held investments.

Warburg has sold some of the biggest investments in its portfolio in recent months, including the sale of turbine parts manufacturer Consolidated Precision Products to GE Aerospace for $11.75bn and a majority stake in Ensemble Health Partners, valuing the healthcare technology group at $12bn.

Underpinning his plans for AI and robotics is SoftBank’s ownership of UK chip company Arm and the $65bn that Son has invested in OpenAI. Son’s investment in OpenAI has faced fresh uncertainty after the ChatGPT maker delayed plans for a stock market listing amid concerns about the risks and costs of developing increasingly powerful AI systems.

That delay could also hurt SoftBank’s plans for new investments in AI, with some investors questioning the rationale of further large-scale deals.

The company’s financing costs are already rising — last month it completed the largest junk bond offering on record, paying investors yields as high as 9.75 per cent, signalling the price of Son’s extreme ambition.

However, senior figures at SoftBank point to the group’s strong balance sheet as evidence that the company can finance Son’s plans for new acquisitions and investments.

In the search for more firepower, Son is also seeking to raise up to $100bn, including from Gulf investors, with the aim of creating a fund to buy whole companies that can be overhauled using AI. One of the people familiar with the matter suggested that such a fund would see Allied as a potential target.

SoftBank’s share price has fallen by close to 40 per cent since a peak in June but is still up 7 per cent so far this year. Japan’s benchmark Nikkei 225 index is up 26 per cent over the same period.

A trickle of sales to foreign buyers risks becoming a flood

The “for sale” sign at British property group Big Yellow is as hard to miss as the luminous warehouses that give the company its name. Two activist investors are pushing for a deal, arguing that it is under-appreciated on public markets. The worry, for those who care about London’s position as a listings hub, is that they could make a similar case for many of its peers.

Big Yellow has a portfolio worth a net £2.6bn, but a market capitalisation of just £1.8bn. It used to regularly trade above net asset value, but the stock has struggled as the UK grapples with high interest rates and low growth.

Its answer to the activists is, unsurprisingly, that it “continually evaluates all options”. Big Yellow has considered selling itself before: last year it reached out to potential buyers including Blackstone, which took a detailed look but could not agree on a price. 

A sale to a foreign buyer isn’t in itself a problem. Self-storage is useful but hardly a critical industry like defence. Nor is Big Yellow a much-beloved brand that would constitute a symbolic loss. If it was bought by, say, a large US self-storage specialist with ample access to financing, the UK economy could ultimately benefit as it would become easier for Big Yellow to grow.

But the risk is that a trickle of exits becomes a flood, and other companies find it harder to raise capital locally. Pushing for such deals is becoming a popular activist tactic. One in five European activist campaigns included M&A demands in the first five months of 2026, according to Alvarez & Marsal, up from 8 per cent in the same period last year.

Property groups could be particularly vulnerable. Many executives would identify with Big Yellow’s struggle to shrink a discount to net asset value. The number of London-listed real estate investment trusts already fell by a quarter in the three years to July. And it’s easier for activists to build an influential stake when even market leaders like Big Yellow are only worth a few billion pounds.

Management teams can’t do much about interest rates — that’s more of a question for Prime Minister Andy Burnham, whose first Budget is due later this month — but there are still some things they control. Even the largest remaining groups like Land Securities and British Land are small by international standards. Perhaps they could go on the offensive with deals of their own?

Last year investors in NHS landlord Assura rejected a take-private offer in favour of merging with domestic peer Primary Health Properties; the enlarged group currently trades closer to book value than most REITs. True, more deals like that means a smaller number of companies. But hosting fewer, larger groups would be better than losing the entire sector.

Qnity Electronics stock could gain 40% as investors improve their understanding of its products and their role in the supply chain.

There are many ways to play the artificial-intelligence boom. From copper wires carrying electricity to servers, to data centers housing the servers, to the GPU, CPU, and memory chips inside the servers, to frontier AI model companies using the servers to build AI applications that, hopefully, transform the world for the better—all have been explored, some at length, in investing publications like this one.

These aren’t the only ways to play the trend, however. There are other ways for investors to profit, such as DuPont spinoff Qnity Electronics —and it’s a promising one. It benefits from the increasing complexity of semiconductors. Shares could rise 40% as investors better understand its growth potential.

The Wilmington, Del.-based Qnity supplies chip makers with things they need to make chips, including silicon polishing pads, photolithography materials, bonding materials, and heat-management solutions, among many others.

It was spun out of DuPont in October 2025, so the company is just about a year old. Corporations can spin off assets for several reasons, including improved capital allocation. Oftentimes, spinoffs are driven by valuation. Valuation is indeed a factor for Qnity. It competes with the likes of Entegris, which trades at about 33 times earnings expected over the coming 12 months. DuPont, at the time of the Qnity spin, traded at about 16 times. (Now it trades at 17 times.)

Qnity, for its part, trades at about 24 times earnings, despite similar growth expectations as Entegris. That’s better than DuPont, but a fairer multiple would be closer to 30 times, which implies a stock price of $160, up more than 30% from Wednesday’s close of $121.62.

Fairer in the stock market isn’t really a thing, though. Valuation boils down to earnings growth and investors’ confidence in it. That confidence may be lacking for now simply because Qnity is still new. “It’s among the most underfollowed names in the S&P 500,” says Fermium Research analyst Frank Mitsch. The stock has only 10 analysts covering it. The average for an S&P 500 stock is 25 analysts.

Being undercovered is very much part of the spinout process. Wall Street, after all, is a key source of new ideas for investment managers. “Qnity is in a great position to take advantage of the boom in AI and the advanced technologies…and yet, because it’s under the radar as of now, we believe that it’s suffering from a valuation discount,” Mitsch adds.

As investors start to tune in, they should like what they see. Qnity has reported only three quarters as a stand-alone company, but those reports have all topped Wall Street’s sales and earnings projections, with each posting at least 33% year-over-year earnings-per-share growth.

Mitsch’s target for the stock is $175, corresponding to a gain of more than 40% from recent levels. RBC Capital Markets analyst Arun Viswanathan has a higher price target of $189, up 55% from recent levels. Qnity has more exposure to “advanced node” chips, he says, which will result in above-average sales and earnings growth.

Chips are getting more complicated, with tiny, nanometer-scale transistors etched using extreme ultraviolet light and stacked on top of one another using some of the most complicated machines ever built by humanity. That complexity means more business per chip for Qnity, a big reason to expect stable, above-average sales growth for years to come.

M&A is another source of potential growth, according to Qnity CEO Jon Kemp. Qnity’s balance sheet can handle some deals. It carries about $3 billion in net debt, which is less than two times expected 2026 earnings before interest, taxes, depreciation, and amortization, or Ebitda. That isn’t an alarming ratio. The average net-debt-to-Ebitda ratio for S&P 500 industrial stocks is about 1.6 times.

Ebitda is expected to grow about 8% a year for the coming three years. Things can turn out better than that, though. Entegris’ Ebitda is expected to grow closer to 10%. Qnity’s estimates are also rising. Wall Street projects about $2 billion in 2027 Ebitda. That number was closer to $1.6 billion at the time of the spin.

To be sure, Qnity is a semiconductor-related AI play. That means geopolitics is a risk. Taiwan Semiconductor Manufacturing is one of its largest customers. Most advanced chips are still made in Taiwan. That, however, holds true for the entire semiconductor value chain.

Qnity Electronics stock could gain 40% as investors improve their understanding of its products and their role in the supply chain.

There are many ways to play the artificial-intelligence boom. From copper wires carrying electricity to servers, to data centers housing the servers, to the GPU, CPU, and memory chips inside the servers, to frontier AI model companies using the servers to build AI applications that, hopefully, transform the world for the better—all have been explored, some at length, in investing publications like this one.

These aren’t the only ways to play the trend, however. There are other ways for investors to profit, such as DuPont spinoff Qnity Electronics —and it’s a promising one. It benefits from the increasing complexity of semiconductors. Shares could rise 40% as investors better understand its growth potential.

The Wilmington, Del.-based Qnity supplies chip makers with things they need to make chips, including silicon polishing pads, photolithography materials, bonding materials, and heat-management solutions, among many others.

It was spun out of DuPont in October 2025, so the company is just about a year old. Corporations can spin off assets for several reasons, including improved capital allocation. Oftentimes, spinoffs are driven by valuation. Valuation is indeed a factor for Qnity. It competes with the likes of Entegris, which trades at about 33 times earnings expected over the coming 12 months. DuPont, at the time of the Qnity spin, traded at about 16 times. (Now it trades at 17 times.)

Qnity, for its part, trades at about 24 times earnings, despite similar growth expectations as Entegris. That’s better than DuPont, but a fairer multiple would be closer to 30 times, which implies a stock price of $160, up more than 30% from Wednesday’s close of $121.62.

LIVE Q&A

Our next live Q&A is coming up! Watch our reporters discuss recent stock picks, provide their input on the current market environment—and take your questions.

Register here.

Fairer in the stock market isn’t really a thing, though. Valuation boils down to earnings growth and investors’ confidence in it. That confidence may be lacking for now simply because Qnity is still new. “It’s among the most underfollowed names in the S&P 500,” says Fermium Research analyst Frank Mitsch. The stock has only 10 analysts covering it. The average for an S&P 500 stock is 25 analysts.

Being undercovered is very much part of the spinout process. Wall Street, after all, is a key source of new ideas for investment managers. “Qnity is in a great position to take advantage of the boom in AI and the advanced technologies…and yet, because it’s under the radar as of now, we believe that it’s suffering from a valuation discount,” Mitsch adds.

As investors start to tune in, they should like what they see. Qnity has reported only three quarters as a stand-alone company, but those reports have all topped Wall Street’s sales and earnings projections, with each posting at least 33% year-over-year earnings-per-share growth.

Mitsch’s target for the stock is $175, corresponding to a gain of more than 40% from recent levels. RBC Capital Markets analyst Arun Viswanathan has a higher price target of $189, up 55% from recent levels. Qnity has more exposure to “advanced node” chips, he says, which will result in above-average sales and earnings growth.

Chips are getting more complicated, with tiny, nanometer-scale transistors etched using extreme ultraviolet light and stacked on top of one another using some of the most complicated machines ever built by humanity. That complexity means more business per chip for Qnity, a big reason to expect stable, above-average sales growth for years to come.

M&A is another source of potential growth, according to Qnity CEO Jon Kemp. Qnity’s balance sheet can handle some deals. It carries about $3 billion in net debt, which is less than two times expected 2026 earnings before interest, taxes, depreciation, and amortization, or Ebitda. That isn’t an alarming ratio. The average net-debt-to-Ebitda ratio for S&P 500 industrial stocks is about 1.6 times.

Ebitda is expected to grow about 8% a year for the coming three years. Things can turn out better than that, though. Entegris’ Ebitda is expected to grow closer to 10%. Qnity’s estimates are also rising. Wall Street projects about $2 billion in 2027 Ebitda. That number was closer to $1.6 billion at the time of the spin.

To be sure, Qnity is a semiconductor-related AI play. That means geopolitics is a risk. Taiwan Semiconductor Manufacturing is one of its largest customers. Most advanced chips are still made in Taiwan. That, however, holds true for the entire semiconductor value chain.

Chinese officials agreed to limit hybrid vehicle exports to the European Union following a visit by EU Trade Commissioner Maros Sefkovic.

The EU’s goods trade deficit with China has more than doubled since prepandemic days and could reach 400 billion euros this year.

French President Emmanuel Macron and German Chancellor Friedrich Merz called for a credible instrument to exclude trade distorters from the market.

A funny thing happened when European Union Trade Commissioner Maroš Šefčovič went to Beijing a week ago. He came back with a concession.

Chinese counterparts agreed to limit hybrid vehicle exports, which have quadrupled over the past 18 months as an end-run around EU tariffs on all-electric cars. That won’t automatically rescue Europe’s floundering auto makers, much less close the yawning bilateral trade gap. Shares in Volkswagen, BMW, and Mercedes-Benz Group, all down by a third or more this year, barely reacted.

Nevertheless, “the fact that they got anything is a success after years of the Chinese being totally dismissive,” says Andrew Small, director of the Asia program at the European Council on Foreign Relations.

The EU’s goods trade deficit with China has more than doubled since prepandemic days and could hit a record 400 billion euros ($445 billion) this year. European leaders are increasingly fed up with China’s penchant for keeping unprofitable companies alive, then selling their excess production at minimal margins abroad.

French President Emmanuel Macron and German Chancellor Friedrich Merz posted a joint letter to the European Commission on Oct. 5, calling for a “credible instrument up to and including exclusion from the single market” for unnamed trade distorters. The imprimatur of Germany, Europe’s top economy and leading China trade dove in recent years, marks a milestone, says Dimitar Lilkov, senior research officer at the Wilfried Martens Centre for European Studies.

Next steps remain unclear. Car imports are a symbolically resonant and easily defined target. In euro terms, though, Europe buys 10 times as much Chinese electrical equipment and seven times as much machinery as vehicles. “An approach to these more amorphous and bigger areas has to be invented,” Small observes.

The good news for auto makers is they are using the tariff reprieve to fight back, says industry analyst Matthias Schmidt. Chinese EV sales are “plateauing” at about 15% of that category as VW and French competitor Peugeot roll out hot-selling new models. EVs will exceed 25% of new EU car sales this year, with a boost from Iran war–fueled petrol prices, he estimates.

The bad news is that European car manufacturers won’t make much money. Chinese rivals at the gate plus the economics of EVs, where batteries account for much of the cost, are slashing margins from high-single digits to 3% to 4%, Schmidt reckons.

That leaves axing costs the key to survival. Investors were unimpressed this summer by VW CEO Oliver Blume’s taboo-smashing plan to cut up to 100,000 jobs and close four German factories, notes Michael Field, European equity strategist at Morningstar. “VW is getting huge pushback from their unions and local governments,” he says. “It’s hard to square that circle.”

VW shares are trading 80% below a theoretical fair value after crashing by three-quarters over the past five years, Morningstar estimates. Mercedes’ upside is 50%. Field sees “no single short-term catalyst” to reverse dismal market sentiment, however.

Brussels’ would-be tougher stance on China is constrained by at least two factors. One is the U.S., where Trump administration tariffs reduced EU exports by 16% in the first eight months of 2026. The second is Europe’s commitment to slashing its carbon output, which depends largely on Chinese equipment.

“The European Green Deal has caused a huge wealth transfer to China,” Lilkov comments.

Tokyo, Oct 16 (Reuters) - The next phase of Japanese corporate reform took center stage at the Sohn Conference in Tokyo, as hedge fund managers highlighted investment opportunities ranging from a financial conglomerate's capital ​allocation strategy to an X-ray scattering firm's AI solutions.

At the inaugural Sohn Conference in Japan, ‌which invited 14 funds to pitch their top investment ideas, more than half of the investors favored the theme of corporate reform.

Daisuke Nakayama, a portfolio manager at Hong Kong-headquartered Polymer Capital, picked Orix Corporation , a financial services firm, saying the change in management ​marks "a big turning point" for the company's transformation into a global investment firm.

"The new leadership has a ​greater focus on capital allocation and ROE," Nakayama said, adding the firm is now using ⁠overseas capital, including from Qatar Investment Authority, to bid for companies.

The number of activist campaigns in Japan hit ​a record last year, according to brokerage Jefferies, after years of pressure from regulators and the Tokyo Stock Exchange ​on companies to improve shareholder returns and invest in growth.

Fund managers said that the next chapter is not just about companies increasing dividends and buybacks but also about focusing on those deploying capital to expand businesses and drive long-term growth.

One ​example is Rigaku (268A.T), opens new tab, an X-ray equipment supplier pivoting to semiconductor metrology, measuring and inspecting chips.

New York-based Indus Capital ​Partners said it had invested in the company and had engaged with the firm to sell a stake to a US ‌semiconductor firm ⁠to drive revenue.

As semiconductor companies are going into vertical structures to scale, X-ray is needed to detect defects deeper in the 3D chips, said Byron Gill, managing partner at Indus.

Rigaku might be one of the few companies that "has the key to unlocking scale for the semiconductor industry in the middle of a super cycle", he added, forecasting ​a 100% rise in ​its valuation by 2029.

The corporate reform theme still has a long way to run, market participants said.

"Japan changes slowly, but once the change starts, it does not reverse," said Naohide Une, ​partner and CEO at Investment Lab, a former senior managing director at Japan ​Post Bank.

He picked ⁠Japanese chemical company Ube Corporation (4208.T), opens new tab for its corporate reform efforts.

In a separate Q&A session, Seth Fischer, founder of activist Oasis Management, said his fund may substantially increase its stake in Japanese electric motor maker Nidec (6594.T), opens new tab.

AI ⁠was another ​key investment theme.

Investment firms Ruane Cunniff and Japan Catalyst picked software ​companies Money Forward (3994.T), opens new tab and OBIC Business Consultants (4733.T), opens new tab, respectively.

They argued that back-office software operators are AI winners rather than losers, as clients require accuracy.

And these ​businesses are helping Japan overcome labor tightness, they said.

I am sending you an updated version of our Musk Empire Consolidation note, incorporating a significant development published yesterday.

On October 16, RBC analyst Tom Narayan published explicit deal mathematics for a Tesla/SpaceX combination: a $480 per share offer to Tesla shareholders “could be compelling,” implying a 52/48 exchange ratio tilted toward Tesla at current SpaceX prices of approximately $150. At RBC’s own $225 SpaceX price target, that ratio shifts to 68/32 in favour of SpaceX holders. On the same day, Tesla launched its new Roadster at Starbase — a deliberate choreographic signal. Barclays had flagged deal news as “possible this year” two days prior.

Our read differs from the street consensus on timing. We believe a transaction, if it occurs, happens before SpaceX reaches $225 — not after. At $150, Tesla holders capture the most favourable exchange ratio they will ever be offered. Every $10 of SpaceX re-rating narrows that window materially. Musk’s incentive is clear: lock in Tesla’s EV, FSD and energy cash flows before the SpaceX multiple eliminates the strategic need for them.

This creates a distinct trade structure: long Tesla as a SpaceX acquisition proxy at a discount, with a long TSLA / short SPCX spread to capture convergence. The key risk is a deal that falls through, or one announced only after SpaceX has already re-rated — at which point Tesla holders lose their negotiating leverage permanently.

We have integrated this analysis as a fourth vector in the Trade Framework section of the attached note, alongside our updated Starship Flight 14 section reflecting the confirmed September 28, 2026 orbital launch and 26 V3 satellites now operational.

The full note covers: the 7 structural gaps in consensus, the 3-leg trade framework, the SpaceX flywheel, the Starship/V3/spectrum moat update, the lock-up mechanics and flow calendar through August 2026, and the risk matrix.

We remain available to discuss any aspect of the above.

Kind regards,

Laurent Chekroun