SoftBank Is Borrowing $11 Billion in Junk Bonds to Buy OpenAI Equity, While the Yield on Its Own Bonds Has Gone From 6.7% to 8.2%

SoftBank Group launched a speculative-grade bond offering worth more than $11 billion equivalent on September 21, comprising $10 billion of dollar senior unsecured notes across three tenors (3.5, 5.5 and 7.5 years) and €1 billion (about $1.1 billion) of euro debt across two maturities (4 and 6 years), with pricing set for September 24 and settlement on September 29. Proceeds go to SoftBank's $10 billion contribution to the third tranche of its OpenAI follow-on investment, a transaction expected to close October 1, plus general corporate purposes, and will also cancel a $10 billion bridge loan previously secured for the same investment. At the intended size it would be the largest non-financial corporate bond deal ever from Asia Pacific and Japan, topping the $10.93 billion raised by 7-Eleven Inc. in January 2021, and excluding distressed debt exchanges it would be the largest speculative-grade bond sale ever by a single corporate issuer. Fitch assigned the proposed notes BB+, its highest speculative-grade level, noting leverage is likely to increase as SoftBank meets its investment commitments while expecting the company to maintain sufficient liquidity and continued access to debt markets. On completion of the third tranche, SoftBank's cumulative investment in OpenAI is expected to reach approximately $64.6 billion for a stake of roughly 13%, which implies a pre-money valuation of about $730 billion. SoftBank has issued $15 billion equivalent of bonds in multiple currencies this year, making it the largest speculative-grade borrower in the bond market in 2026; it also raised the limit on a margin loan backed by Arm shares from $5 billion to $25 billion, expanded its bank credit facility by $450 million to $6.5 billion, and on September 15 repaid $25.9 billion of a $40 billion bridge loan. Financing costs are rising in parallel: the yield on its dollar bonds maturing in 2031 has climbed this month to 8.2% from a January low of 6.7%, and the cost of insuring against a default on SoftBank debt reached its highest level in three years. Citigroup is lead bookrunner on the dollar notes and JPMorgan on the euro portion, with Goldman Sachs and Morgan Stanley also on the books.

What This Deal Is: Speculative-Grade Debt Raised to Buy Equity in a Private Company

Start with the structure, because it is more worth recording than the headline number. The money does not come from operating cash flow. It is borrowed in the bond market, at speculative grade — junk. And it is not going toward equipment, plants or working capital. It is buying equity in a private company. At the intended size, **this would be the largest speculative-grade bond sale ever by a single corporate issuer, excluding distressed debt exchanges.** That is not a moral judgment, it is a description of risk structure. Debt service is a hard obligation with fixed dates; equity returns depend on a future liquidity event — an IPO or a sale — and on the valuation prevailing then. The two differ in both tenor and certainty, so a deal of this kind ultimately turns on one question: before the debt matures, can that equity be monetized or refinanced at no less than cost? The maturity ladder makes the same point. The dollar portion runs 3.5, 5.5 and 7.5 years; the euro portion 4 and 6. The shortest tranche comes due in three and a half years.

"Raised $11 Billion" Is Not "Has $11 Billion More to Spend"

An easy misreading: a substantial part of this money goes to cancel the $10 billion bridge loan previously taken for the same investment. Which makes this closer to a **debt swap** — short-term bridge financing converted into medium- and long-term bonds. The maturity profile improves, since nothing clusters in the near term, but total debt does not fall correspondingly. On September 15 SoftBank repaid $25.9 billion of a $40 billion bridge loan, and it recently closed a syndicated loan of roughly $11.87 billion. String those moves together and the through-line is rolling the same exposure across different instruments, not deleveraging. The $730 billion figure needs the same care. It is derived by working backward from "roughly $64.6 billion cumulative investment for about a 13% stake," not published by any party. Back-calculations like this are affected by differing prices across tranches, preference terms and more. **Useful as an order of magnitude; not usable as a transaction valuation.**

The Number to Watch Is Not $11 Billion — It Is 8.2%

The most informative figure here is the yield. The yield on SoftBank's dollar bonds maturing in 2031 has risen to 8.2% this month, against a January low of 6.7%, while the cost of insuring against a default on its debt has reached its highest level in three years. **Over the same period in which it is adding leverage, the market is pricing it more expensively.** Set that against Fitch's rating language and the condition becomes visible: Fitch assigned BB+, its top speculative-grade notch, said leverage is likely to rise as SoftBank meets its investment commitments, and expects the company to maintain sufficient liquidity and *continued access to debt markets*. The move from 6.7% to 8.2% is the evidence that this access is getting more expensive. The access has not disappeared — but its price is a variable, and the variable is moving the wrong way. The practical implication for a reader: do not judge deals like this by whether a single offering succeeds (some version of "oversubscribed" nearly always appears). Judge it by **how the same issuer's yield curve moves between successive offerings.** The actual coupon set at pricing on September 24 will say more about market sentiment than the $11 billion headline does.

Filling In the Other Half of the Money

Several recent pieces on this site covered the first half of the same story: on September 9, Nscale's contracted backlog rising from about $51 billion to about $103 billion in a month; on September 17, Anthropic signing roughly $517 billion in compute contracts over eleven months; on September 20, Crusoe's $3.9 billion Series F, with $140 billion of total contracted value against 1 gigawatt delivered. Those describe **how much the demand side has signed.** This one is where the money comes from. The answer is increasingly clear: **a substantial share of it comes from the bond market rather than operating cash flow.** That is not necessarily bad — heavy-asset expansion has always been debt-financed — but it changes how risk propagates. When one link in the funding chain is speculative-grade debt, the interest rate environment itself becomes a variable in the pace of AI investment, and that link has nothing to do with model capability or product demand. So the next two dates to follow are specific: the pricing result on September 24, and whether the third tranche closes on October 1 as planned.

via: The Japan Times, Yahoo Finance, GuruFocus, Quartz