SoftBank AI strategy faces pressure as asset declines raise financing risks

Traders Union
All News
Business & Wealth
Companies
SoftBank AI strategy risks

Highlights

SoftBank’s shares and Arm have both dropped about 60 per cent since June, sharply increasing leverage risks before Tuesday’s earnings report.
Loan-to-value ratio climbed from 13 per cent in June to an estimated 18–22 per cent, leaving $10 billion to $26 billion in funding headroom but tightening financing options after a record $11 billion junk bond issue.
AI ambitions face delays as U.S. political uncertainty threatens SB Energy’s $50 billion IPO and a national security review blocks the ABB Beijing robotics deal.

Portfolio declines test funding capacity

As reported by Financial Times, SoftBank’s shares have fallen about 60 per cent since June, while the value of key holdings tied to its AI strategy has also dropped sharply. UK chip designer Arm, one of the group’s most important assets, is down about 60 per cent over the same period, increasing pressure on SoftBank ahead of its earnings report on Tuesday.

The sell-off is hitting a strategy that became increasingly concentrated around the AI ecosystem during the boom in the sector. Analysts say that concentration leaves the Japanese group more exposed now that sentiment has weakened, with fewer defensive assets to offset declines in technology and infrastructure-linked investments.

Falling asset values have also pushed up SoftBank’s loan-to-value ratio, a key measure of leverage. Analysts estimate it has risen to between 18 and 22 per cent, from 13 per cent at the end of June, although that remains below the group’s usual 25 per cent target and may still leave room for another $10 billion to $26 billion in funding capacity.

Even so, the company faces a double risk as weaker portfolio values coincide with tighter credit markets. Its five-year credit default swaps have risen sharply, and analysts say investor appetite to keep financing Son is limited, particularly after SoftBank’s record $11 billion junk bond issuance in September.

AI infrastructure and deal plans face broader scrutiny

Political and market uncertainty in the U.S. is also weighing on parts of SoftBank’s AI infrastructure buildout. The Democrats’ decisive victory in the U.S. midterm elections has heightened concerns around AI infrastructure projects, especially in Texas, where SoftBank is planning some of its largest data centre developments.

SB Energy, SoftBank’s U.S. energy infrastructure company, is seen as especially exposed because newer AI infrastructure businesses depend more heavily on external financing than established technology groups such as Meta and Google. According to people familiar with the situation cited in the report, plans for an SB Energy listing at a $50 billion valuation risk being put on hold.

Any delay to that flotation could have wider implications for the group’s AI ambitions. Progress on the IPO is important to SoftBank’s plan to list Roze, an AI and robotics business that Son wants to use in data centre development as part of a wider effort to demonstrate its value.

At the same time, SoftBank is dealing with geopolitical strain around its attempted acquisition of ABB’s robotics business in Beijing, where the deal has been blocked by a national security review. The group is assessing whether a remedy, including carving out the China business, could revive the transaction.

OpenAI remains the biggest element of Son’s AI thesis, with SoftBank having committed about $65 billion to the company. Investors are already concerned about any delay to OpenAI’s IPO until next year, and those worries are intensifying as rival Anthropic moves closer to its own potentially large listing.

Despite the market pressure, Son is still pursuing new deals, including talks to acquire security services group Allied Universal in a transaction said to be worth close to $30 billion including debt. He is also seeking backing from Gulf investors for a new vehicle provisionally called the Transformation Fund to help finance the effort.

Our earlier article on the U.S. AI data center buildout highlighted that rising funding costs, construction delays, and power constraints are undermining confidence in the sector’s $10 trillion expansion plans. We noted that only a minority of announced projects appear credible, while setbacks and cost overruns are forcing developers to seek additional capital and face more local and regulatory pushback.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.