Tokyo, 4 August 2026 – SoftBank Group’s coming quarterly results are set to place its financial capacity under renewed scrutiny as founder Masayoshi Son accelerates one of the technology industry’s most ambitious artificial intelligence investment strategies.
The Japanese investment group has committed tens of billions of dollars to OpenAI while also backing AI infrastructure, semiconductor design, robotics and data-centre development. These investments have strengthened SoftBank’s exposure to an expanding technology cycle, but they have also increased its reliance on borrowing and the valuations of largely unlisted assets.
Investor attention is likely to focus less on quarterly accounting gains and more on how SoftBank intends to finance its remaining commitments without placing excessive pressure on the balance sheet. Concerns have intensified alongside weakness in the company’s shares and higher costs for protecting against a potential default.
Under an agreement announced in February, SoftBank plans to invest an additional US$30 billion in OpenAI through three US$10 billion tranches scheduled for April, July and October 2026. Once completed, SoftBank’s cumulative investment in the AI company is expected to reach US$64.6 billion, giving the Japanese group an ownership interest of approximately 13%.
The investment is being made at a pre-money valuation of US$730 billion, meaning SoftBank’s future returns will depend heavily on OpenAI maintaining its growth trajectory and eventually achieving a substantially higher public or private valuation. OpenAI’s ability to generate sufficient commercial returns from enterprise products, consumer subscriptions and AI infrastructure will therefore carry significant implications for SoftBank shareholders.
SoftBank reported record annual earnings for the financial year ended March 2026, helped by valuation gains from its technology holdings. Those gains, however, remain sensitive to market conditions and do not provide the same liquidity as cash generated from operations or completed asset sales.
The group has increasingly used loans supported by holdings such as its controlling stake in semiconductor designer Arm Holdings. Its private OpenAI interest may also become an important source of collateral, although reliance on volatile technology assets could amplify financial pressure if AI valuations decline.
SoftBank maintains that it will continue managing its loan-to-value ratio below 25% under normal market conditions and no higher than 35% during periods of financial stress. It also intends to hold enough cash to cover at least two years of bond redemptions.
These safeguards provide a measure of protection, but the scale of the group’s AI programme means investors will be watching for additional borrowing, asset disposals or strategic partnerships. A significant correction in Arm’s share price or a downward revision to OpenAI’s valuation could reduce financing flexibility even if SoftBank remains within its formal leverage limits.
Beyond OpenAI, the group is helping develop the Stargate AI infrastructure programme in the United States. The initiative is intended to mobilise as much as US$500 billion for computing capacity, power infrastructure and data centres, although much of the capital is expected to be raised through partners and project-level financing rather than SoftBank’s balance sheet alone.
SoftBank has also expanded its position in AI hardware through Arm and its acquisition of chip designer Ampere Computing. Its proposed purchase of ABB’s robotics business reflects Son’s broader vision of combining advanced AI models, computing infrastructure and machines capable of operating in physical environments.
The Ledger Asia Insights
SoftBank’s strategy gives it exposure across several critical layers of the AI economy, from processors and data centres to software models and robotics. This integration could create substantial value if demand for computing capacity continues rising and OpenAI converts its technological leadership into sustainable earnings.
The accompanying financial risk is equally concentrated. SoftBank’s net asset value is increasingly influenced by a small number of AI-related holdings, while large funding commitments must be met before their long-term returns become clear. Investors will therefore need to distinguish between valuation gains, available liquidity and recurring cash generation.
The coming results will serve as a test of whether SoftBank can preserve financial discipline while pursuing Son’s expansive AI agenda. Strong asset values may keep the strategy moving, but the investment case increasingly depends on disciplined financing, credible commercial progress from OpenAI and sufficient diversification to absorb a downturn across the global AI market.
