SoftBank suffers a massive $7.1 billion loss, stock price plunges 11%! Masayoshi Son’s AI ambitions challenged by power supply and financing costs
After SoftBank Group reported its first net loss in nearly two years, its share price fell as much as 11%. The loss was attributed to a $3.7 billion write-down on its U.S. data center project and a decline in AI valuations. On Wednesday, the stock fell as low as 3,465 yen during Tokyo trading. The company reported a loss of 1.12 trillion yen (approximately $7.1 billion) for the second quarter.
According to Zhitong Finance APP, SoftBank Group Corp., led by legendary investor Masayoshi Son, saw its share price drop as much as 11% during early trading in the Japanese stock market before significantly narrowing the decline to about 5% at the time of writing, as bargain hunters stepped in. This comes after the Japanese technology investment company reported its first net loss in nearly two years the previous day, due to a $3.7 billion impairment related to a US data center project and a decline in AI asset valuations.
On Wednesday, SoftBank's share price in Tokyo fell as low as 3,465 yen. The company swung from a significant profit to a net loss of 1.12 trillion yen (equivalent to $7.1 billion) in the second quarter of the fiscal year. Previously, the company wrote down non-refundable prepayments made for securing power generation equipment for its Ohio 8-gigawatt project and other US data centers.
In a report, the Jefferies analyst team led by Atul Goyal wrote that while part of the loss stems from accounting treatment, “the real issue lies in the 597 billion yen write-off related to power equipment, which raises questions beyond what management can address,” as well as financing costs rising to 2.6 times the same period a year ago. “Management is reacting passively to the progress of data center construction.”

As shown in the image above, SoftBank Group’s stock plunged as its recent performance deeply disappointed the market.
SoftBank shares have fallen nearly 60% from their all-time high in June, making them one of the worst-performing constituents of the Nikkei 225 index over the same period. Meanwhile, as investors try to evaluate the risks and opportunities embedded in its AI investments, stock price volatility has increased.
After a rebound, another hit: Masayoshi Son’s AI ambitions face a capital test
The renewed plunge in SoftBank's share price highlights the market's reassessment of the comprehensive procurement costs of its global AI application and AI computing resource expansion plans, AI infrastructure financing burdens, substantial delays in power supply, and the realization timeline of investment returns. On November 11, SoftBank Group’s Tokyo share price fell as much as 11% intraday, hitting a low of 3,465 yen, a nearly 60% drop from the historical peak in June.
SoftBank previously announced a net loss of about 1.12 trillion yen (equivalent to $7.1 billion) attributable to shareholders for the fiscal second quarter of July–September 2026, compared to a profit of about 2.5 trillion yen the same period last year. Adjustments in the procurement of power equipment for US data centers and significantly increased financing costs have become the immediate triggers for investors to reassess this AI investment giant.
This decline occurred after a recent strong rebound: before the earnings announcement on November 10, SoftBank shares had just gained about 13.9%. Previously, the US tech stock selloff had also exerted pressure on SoftBank; on November 6, its share price dropped about 11%, and its five-year CDS spread widened to around 466 basis points, a four-year high.
The simultaneous pressure in equity and credit markets is enough to show that investors are now concerned about whether SoftBank can continue to finance at bearable costs when asset values decline. After the Democratic Party won control of both chambers of the US Congress, uncertainty around the regulation and approval of data centers further affected the outlook for future projects, although this cannot be retroactively blamed for losses confirmed as of the end of September.
Masayoshi Son’s investment blueprint covers not only AI models, AI chips, and large-scale data centers but also cutting-edge technology fields such as humanoid robots. This means SoftBank is exposed to both the valuation changes of technology assets connected to AI applications and to the funding pressures of infrastructure construction. When power equipment procurement and financing expenses occur before projects come online, any adjustment in procurement plans or delays in income realization may widen the gap between capital investment and cash recovery. These dynamics are the underlying investment issues behind the stock’s volatility—more enduring than any single-quarter earnings swing.
Securing power is not enough: equipment, funding, and customer delivery must be synchronized
The roughly $3.7 billion equipment-related loss reveals the cost of locking in scarce supply in advance. The official financial report shows related losses totaling about 597.2 billion yen, including 457 billion yen in non-refundable prepayment losses and 140.2 billion yen in contract modification and settlement losses. The former refers to payments made for equipment and reserved capacity not included in the revised procurement plan, while the latter represents cash payments expected from the fiscal third quarter onwards. Therefore, this loss cannot be entirely viewed as new cash outflows for the quarter, nor dismissed as a costless accounting adjustment.
From a project execution perspective, pre-ordering long lead-time power equipment can reduce the risk of having no equipment available in the future; however, if project specifications, equipment configuration, or deployment plans change, non-refundable prepayments reduce flexibility for adjustments.
SoftBank still retains up to $2.1 billion in conditional rebate rights for future purchases, but these have yet to be recognized as assets. Management has also clarified that there have been no discussions about scaling down the Ohio PORTS-Pike project. That project corresponds to 8GW of IT capacity, already has a 20-year lease from OpenAI, and is scheduled to be phased into operation starting in 2028. Customer contracts provide long-term demand support, but construction, power supply, and financing must still be implemented one by one before commissioning.
Financing costs make the waiting period for returns even more tangible. Second-quarter financing costs were about 461.9 billion yen, 2.64 times that of the same period last year. This increase is related both to higher borrowings and bond balances and to one-off recognition of unamortized financing expenses after early repayment of bridge loans, so not all of the cost increase should be mechanically projected into future quarters.
This is why the Jefferies analyst team highlighted equipment write-offs and rising financing costs as core questions and believe management is being forced to give reactive explanations. Meanwhile, OpenAI's valuation was not written down this quarter, so the market has not attributed SoftBank’s massive loss to an OpenAI investment impairment.
Masayoshi Son remains proactive in advancing large AI projects. On November 9, SoftBank, Grab, and PETROS announced plans to jointly invest $1.85 billion of initial capital to develop AI infrastructure in Malaysia, with SoftBank intending to contribute $1 billion. The long-term goal is to reach at least 5GW of computing power. For investors focused on the AI investment theme, the next key indicators will be whether new projects can receive phased investment in line with client contracts and construction progress, and whether post-commissioning cash returns can cover financing and ongoing construction costs. AI demand growth provides investment direction, but procurement discipline, delivery capability, and financing arrangements will ultimately determine how much return shareholders can achieve.
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
On's Q3 revenue increased by 17.2%, CEO says communicating with multiple parties regarding the Mbappé incident
Swiss sports brand On responded for the first time to the controversy sparked by football star Mbappé. On November 10, On's Co-CEO Cas...

As tech stocks fluctuate violently, hedge fund mogul warns: crowded "dispersion trades" may become a "magnifier" of market volatility
A hedge fund manager stated that dispersion trading is like a "time bomb"; if the market suddenly shifts, it will cause pain. Traditional dispersion trading involves betting that a basket of stocks or an index will remain stable, while individual stocks will have greater and more varied volatility. The hedge fund manager warned that if correlations between stocks increase, dispersion trading will become "very painful."

CoreWeave (CRWV.US) Q3 Conference Call: CEO Addresses High Debt Concerns, Says "Will Not Use the Unsecured Debt Market"; 75% of Committed Capital Expenditures Tied to Investment-grade Clients
AI cloud computing service provider CoreWeave held an earnings call after releasing its third-quarter 2026 results (as of September 30) following the close of trading on November 10, Eastern Time.
