The Daily Byte
Breaking
Welcome to The Daily Byte — real stories from across Pakistan.
Monday, 9 November 2026 · Pakistan

The Daily Byte

Stories from Pakistan that deserve to be heard
Minor Markets

SoftBank credit risk hits four-year high

By The Daily Byte ·09 Nov 2026 ·4 min read
SoftBank Group headquarters in Tokyo as the cost of insuring its debt against default hits a four-year high

SoftBank Group shares fell 10.8% on Friday and the cost of insuring the Japanese conglomerate's debt against default climbed to a four-year high, as doubts about the profitability of artificial intelligence investment spread from equity markets into credit markets.

The five-year credit default swap spread on SoftBank rose to as much as 466 basis points on Friday, the widest since October 2022, according to Markit data reported by Reuters. It stood at 457.50 basis points at 0653 GMT. That level corresponds to a modelled default probability of 30.6%, the data showed, although a credit default swap spread is a market price for protection, not a forecast that a company will fail.

SoftBank has become a bellwether for sentiment towards AI investment. It has put a cumulative $64.6 billion into OpenAI, the developer of ChatGPT, for a stake of about 13%, owns around 90% of Cambridge-based chip designer Arm, and is financing data-centre development through its subsidiary SB Energy.

Debt-funded bet under scrutiny

Investors have stepped up their scrutiny of SoftBank's finances this year as it has taken on more debt at record yields to fund its OpenAI stake. The final $10 billion of its additional $30 billion OpenAI investment, fully funded on 1 October, came from corporate bond proceeds. Returns on that investment are some way off: OpenAI has delayed its planned public listing to 2027, pushing back the point at which SoftBank could recover cash from the holding.

The same financing strain is visible elsewhere in the group. SB Energy filed in September for a public listing at a potential valuation of $50 billion, but media reports say the listing has been delayed until December or later. The filing showed the unit expects revenue of about $439 billion from its contracts with end users, yet it has no operational data centres at present and is substantially dependent on OpenAI.

Politics adds to the pressure

The immediate trigger for the sell-off was political. The Democratic Party's sweeping wins in this week's US midterm elections have raised fears of a slowdown in AI infrastructure development, with tougher regulation and slower approvals for data centres. One-third of US data-centre capacity projected for 2027 and 2028 is now seen as at risk, according to Rolf Bulk, head of semiconductors and infrastructure at the Futurum Group, because the planned sites are in states where governors may slow development. Bulk added, however, that most data centres due for completion in 2027 have already passed regulatory hurdles and secured power, chips and other components.

Amir Anvarzadeh, Japan equity market strategist at Asymmetric Advisors in Singapore, said there had been a steady decline in the credit standing of AI infrastructure firms and that equity markets were finally pricing in a slowdown in data-centre development. He added that he thought the market had hit the bottom in the near term, but warned that prices were likely to test the lows again once confirmation of the slowdown came through next year.

The other side of the balance sheet

The credit warning needs to be set against the scale of SoftBank's assets. At the end of June 2026 the group reported asset value of 83.11 trillion yen and net debt of 10.81 trillion yen, giving a loan-to-value ratio of 13.0%, below its normal management target of 25%. Its listed assets, above all the Arm stake, can be turned into cash if needed, although the amount that can be raised shrinks if share prices fall, and unlisted stakes such as OpenAI are harder to sell quickly.

SoftBank's shares have now fallen by almost half since the start of September. Whether the pressure eases will depend on how far the market's focus shifts from the promise of AI to the cost of funding it, and on how quickly the companies financing the build-out can show returns on the money already spent. This article reports market developments; it is not investment advice.