SoftBank markets jumbo AI bonds at record yields

SoftBank Group has begun marketing more than $11 billion of dollar and euro bonds at record prospective yields for comparable company maturities, as Masayoshi Son’s investment group turns again to debt markets to finance its expanding artificial intelligence commitments.

The Tokyo-based group is seeking $10 billion through three dollar tranches and €1 billion through two euro tranches. Price discussions for the dollar bonds point to yields of 8.75% to 8.875% for $1 billion of 3½-year notes, 9.375% to 9.5% for $4.5 billion of 5½-year securities and 9.75% to 9.875% for $4.5 billion of 7½-year debt.

Those levels, if maintained when the bonds are priced, would represent SoftBank’s highest yields for dollar securities with those maturities. The figures remain preliminary and can change before pricing, which is scheduled for September 24, with settlement expected on September 29.

The offering would become the largest non-financial corporate bond transaction on record in the Asia-Pacific and Japan region if completed at the planned size. The dollar and euro notes are senior unsecured obligations, and the euro portion is being marketed in four- and six-year maturities.

Proceeds are intended principally to finance a $10 billion payment connected with the third tranche of SoftBank’s follow-on investment in OpenAI, expected to close on October 1, while also supporting general corporate purposes. The financing is part of a wider effort by Son to increase SoftBank’s exposure to artificial intelligence companies and infrastructure.

SoftBank has committed close to $65 billion to OpenAI and related artificial intelligence investments, placing substantial demands on the group’s balance sheet and funding programme. The company has used a mixture of asset-backed borrowing, bridge facilities, bond issuance and other financing to meet those commitments while maintaining access to liquidity.

Earlier this month, SoftBank said it would repay the entire $25.9 billion outstanding balance of a $40 billion bridge facility established primarily for follow-on investments in OpenAI. The group had drawn $30 billion under that facility before deciding to make the early repayment on September 15.

The latest bonds are also intended to replace a separate $10 billion bridge loan linked to the OpenAI investment. That shift from shorter-term bank financing towards longer-dated capital-market debt reduces refinancing pressure from bridge facilities but locks in borrowing costs that are considerably higher than those paid on several previous SoftBank bond issues.

SoftBank’s April foreign-currency bond sale illustrates the change in pricing. Its dollar notes issued then carried coupons of 7.625% for 3½ years, 8.25% for 5½ years and 8.5% for 10 years. Euro securities from the same transaction carried coupons ranging from 6.375% to 7.375%.

The group also raised ¥1 trillion through a seven-year domestic unsecured bond this month at an annual coupon of 4.75%, underscoring the scale and frequency of its funding activity.

Fitch Ratings has assigned the proposed foreign-currency notes a BB+ rating, placing them at the highest level below investment grade. The agency has said SoftBank’s debt is expected to increase as committed investments are funded, while adding that the group should retain adequate liquidity and capital-market access.

The elevated yields arrive as credit investors become more selective about debt linked to artificial intelligence.

The pricing discussion also highlights the gap between SoftBank’s proposed borrowing costs and broader high-yield benchmarks. Dollar bonds rated in the BB category have been yielding about 6.55% on average, while B-rated securities have averaged roughly 7.5%. Goldman Sachs credit strategists estimate global AI-related debt issuance has already exceeded $575 billion in 2026, reflecting the enormous financing needs created by data-centre construction, semiconductor purchases and other infrastructure. That supply has made investors increasingly sensitive to new-issue premiums and issuer-specific leverage across global markets.

Global borrowing to finance data centres, chips and other AI infrastructure has surged, prompting investors to focus more closely on leverage, cash-generation prospects and the timing of returns from heavy capital expenditure.