SoftBank's November Reckoning: Can the Balance Sheet Carry the AI Vision?

SoftBank's grand AI narrative now hinges on a more prosaic question: whether the value of its holdings leaves enough room to fund the buildout. The answer arrives with the earnings report expected on 10 November, and credit markets are already signalling impatience.

The cost of insuring against a SoftBank default tells the story. The group's five-year credit default swap spread climbed to 462 basis points today, according to Markit. That is a steep hurdle for a company asking investors to look past delayed paydays and trust in a long-term vision.

SB Energy: A $50 Billion Valuation Without a Working Data Centre

Perhaps nowhere is the tension between promise and execution sharper than at SB Energy. The unit filed in September for an initial public offering at a potential valuation of $50 billion — yet it has no operational data centres to show for it.

Media reports say the listing has slipped to December or later. Reuters notes the company leans heavily on OpenAI as a customer, which ties the infrastructure arm's fortunes tightly to another core SoftBank bet. Until facilities are running and generating revenue, the valuation rests largely on the ability to synchronise construction, financing and customer uptake. Operational assets and the income they produce would make a far more persuasive case than the sheer scale of long-term expectations.

The Numbers Behind the Balancing Act

The metric that binds SoftBank's holdings to its borrowings is loan-to-value — adjusted net debt measured against adjusted holdings. The last reference point stands at 13.0%, and the question is how much of that cushion survives the recent turbulence.

Concentration cuts both ways. Arm accounted for roughly 60% of adjusted holdings at that reference point, and SoftBank has committed a cumulative $64.6 billion to OpenAI while holding about 13% of the AI developer and 90% of Arm. Whether those stakes are valued generously — and when cash can flow back from them — shapes any assessment of the capital structure.

A Longer Wait for the Payday

Timing has turned against SoftBank. Reuters reports that OpenAI's planned IPO has been pushed back to 2027, stretching the wait for what could be a key route to realising value. Should valuations of major holdings weaken at the same time, the need for fresh funding would bite harder.

The operating backdrop adds friction. After Democratic gains in the US midterms, concerns are mounting over stricter AI regulation and a slower pace of data-centre construction. Projects could face later delivery or delayed earnings.

An unconfirmed claim from Elon Musk adds another layer of noise. He asserted yesterday on X that SoftBank is failing to honour turbine orders, offering no evidence or detail. SoftBank has not confirmed the allegation. Concrete disclosure on payment obligations would narrow that uncertainty; a confirmed liability would sharpen the financing question considerably.

The Infrastructure Bottleneck

The Futurum Group sees roughly a third of projected US data-centre capacity for 2027 and 2028 at risk. Analyst Rolf Bulk says near-term demand is unaffected. That distinction matters for SoftBank: demand can underpin its opportunity only if the necessary capacity actually comes online. For a company with large AI commitments, the timetable can weigh as heavily as the long-term market opportunity.

What to Watch on 10 November

A stable capital structure would be the starting point for any brighter reading. If SoftBank reports a still-viable LTV and explains its funding plausibly, fears of added capital strain could ease, and the holdings would regain their role as an asset base — provided their valuations withstand a more sceptical market. The substance is there: 90% of Arm and around 13% of OpenAI offer considerable value potential if the businesses meet expectations. What matters is whether SoftBank can finance that potential without squeezing its own room to manoeuvre.

Investors should therefore scrutinise three things in the upcoming report: the updated leverage ratio, the trajectory of holdings values, and any commentary on financing further investments. A convincing reassessment requires more than optimism about artificial intelligence. It requires evidence that SoftBank can bridge the gap until returns arrive — and that is the yardstick against which confidence in the holdings strategy will be measured.