Samer Choucair: SoftBank’s $6.3 Billion Bond Sale Reveals Who Is Financing the AI Race
Investment leader Samer Choucair said SoftBank Group Corp.’s announcement of the largest retail bond offering in the history of Japan’s market, valued at ¥1 trillion, or roughly $6.3 billion, reveals an important shift in how the artificial-intelligence investment race is being financed, with Japanese household savings increasingly becoming part of the funding base behind these investments.
Choucair said the seven-year bond, carrying an indicative yield of between 4.30% and 4.90%, is designed to help finance the group’s artificial-intelligence commitments, including its investments in OpenAI, while also refinancing outstanding bonds due in September.
According to Samer Choucair, the transaction should no longer be viewed simply as conventional corporate financing. Instead, it reflects the transfer of part of the risk associated with the AI investment cycle from institutional markets toward individual investors.
The Offering and Its Credit Profile
Samer Choucair said SoftBank plans to issue its 70th unsecured bond, known as the “Fukuoka SoftBank HAWKS Bond,” with a face value of ¥1 million per bond. Pricing is scheduled for September 4, subscriptions are expected to run from September 7 through September 16, and issuance is planned for September 17, with maturity on September 16, 2033.
Choucair noted that the company expects a domestic A rating from Japan Credit Rating Agency, while S&P Global Ratings has maintained SoftBank’s issuer rating at BB+, one notch below investment grade, with a stable outlook after revising it from negative in July.
He said the gap between the domestic and international ratings helps explain why the transaction is being directed toward retail investors rather than relying entirely on banks or international fixed-income funds constrained by investment-grade mandates.
A Yield Approaching 5%
Samer Choucair said the indicative yield of 4.30% to 4.90% is particularly significant in a Japanese financial environment that has changed substantially after years of near-zero interest rates.
The 10-year Japanese government bond yield has been trading near 2.88%, after reaching levels in August not seen for roughly three decades as markets priced the possibility of additional tightening by the Bank of Japan.
Choucair said the spread between government bonds and SoftBank’s retail offering compensates individual investors for credit risk, leverage, and exposure to volatile technology assets. At the same time, however, the transaction effectively channels part of Japanese household savings into indirect financing for SoftBank’s commitments to OpenAI and other AI-related investments.
¥1.68 Trillion Raised From Retail Investors
Choucair said the offering represents SoftBank’s third retail bond issuance of 2026, following ¥418 billion raised in April and another ¥260 billion in June.
If the current transaction is completed at its announced size, SoftBank will have raised approximately ¥1.68 trillion from Japanese retail investors in a single year.
Japanese reports indicate that the group intends to direct roughly ¥600 billion of the proceeds toward artificial-intelligence investments and around ¥400 billion toward refinancing existing bonds.
Bloomberg Intelligence has estimated SoftBank’s funding gap at more than $20 billion, suggesting that the group may need to return to international bond markets in the near future.
Debt Is Moving to the Center of the AI Cycle
Samer Choucair said developments in Tokyo reflect a broader repricing of how artificial intelligence is financed.
Capital commitments for data centers, semiconductors, electricity infrastructure, and large language models are increasingly exceeding the free cash flow available to many of the companies participating in the sector.
As a result, a growing portion of the burden is moving toward credit markets through bridge loans, bond issuance, equity-backed financing, and increasingly complex funding structures linking suppliers, investors, and customers within the same ecosystem.
Choucair said SoftBank has become one of the clearest examples of that transition, with market estimates placing its commitments to OpenAI at more than $60 billion while the group simultaneously accelerates spending on data centers, computing capacity, “physical AI,” and acquisitions in industrial robotics.
SoftBank’s Share Price Reflects the Cost of the Bet
Choucair said SoftBank shares falling approximately 5.3% on the day of the announcement, to their lowest level since late July, reflected equity-market concerns about the financing equation.
Borrowing increases SoftBank’s ability to continue investing aggressively in artificial intelligence, but it also increases its cost of capital and makes shareholder value more sensitive to any slowdown in the AI cycle.
Bond investors, by contrast, are being offered a higher coupon in exchange for assuming credit risk and exposure to a company whose investment portfolio is heavily concentrated in technology.
Redistributing AI Risk
Samer Choucair described the bond sale not merely as a temporary liquidity exercise but as “a redistribution of risk within the Japanese financial system.”
He said the difficulty of absorbing a seven-year maturity and a global below-investment-grade rating entirely through conventional banking or institutional fixed-income channels is helping push part of that risk onto household balance sheets.
Institutional investors in the Gulf and Asia, Choucair added, may therefore find the deal more useful as a benchmark for the cost of capital than as a direct subscription opportunity.
A financing cost approaching 5% for AI-linked commitments can become an important reference point for comparable investments ranging from data centers to computing infrastructure.
What the Transaction Tells Investors
According to Samer Choucair, the issuance carries three simultaneous messages for institutional investors.
The first concerns credit quality, particularly the gap between SoftBank’s domestic A rating and its global BB+ rating. The second is macroeconomic, reflecting the return of meaningful yields in Japan and the broader repricing of risk assets across Asia. The third is sector-specific: artificial-intelligence investment is gradually shifting from an equity-funded expansion toward a cycle increasingly dependent on debt.
Choucair said continued capital expenditure without matching cash-flow generation increases dependence on refinancing and makes the AI investment cycle more sensitive to interest rates and to any slowdown in the conversion of technological capacity into sustainable revenue.
Financing and Technology-Cycle Risks
Choucair said the principal risks include issuer concentration, the ratings gap, and the possibility of further increases in Japanese interest rates and government-bond yields.
Higher benchmark yields could reduce demand for future corporate issues and increase refinancing costs as SoftBank approaches the bond’s 2033 maturity.
Technology-cycle risks are equally important.
Record spending on artificial intelligence has already generated warnings about excess capacity, circular financing structures, and uncertainty over how rapidly computing infrastructure can be converted into sustainable profits.
Choucair said any significant reassessment of OpenAI or the data-center infrastructure linked to the broader AI ecosystem could move beyond private-equity valuations and public equities into credit spreads and eventually influence the willingness of Japanese retail investors to continue financing similar transactions.
Reassessing the Cost of Capital
Samer Choucair said the most important opportunity for institutional investors may not lie in the bond itself, but in what its pricing reveals about the relationship between savings, yield, and risk in 2026.
Japan is rediscovering fixed income after years of monetary suppression, while companies with large financing requirements are attempting to take advantage of that transition before the yield curve is fully repriced.
Choucair said the lesson is also relevant for investment funds in Saudi Arabia and the wider Gulf, where institutions are expanding exposure to the digital economy and artificial intelligence as part of diversification strategies and Vision 2030.
In that environment, he argued, liability engineering can be just as important as asset selection.
The Capital Outlook for the AI Race
Samer Choucair said: “The technology asset may be attractive, but the financing structure determines who retains the value at the end of the cycle.”
Institutions participating entirely through equity bear the full volatility of changing valuations. Investors that also understand how debt is priced around those assets, however, can potentially construct more disciplined exposures.
Choucair said the offering could succeed at its targeted size because the yield remains attractive relative to deposits and Japanese government bonds. Even so, the transaction is unlikely to eliminate SoftBank’s broader financing gap as commitments to OpenAI, computing infrastructure, and outstanding debt maturities continue to require substantial capital.
Samer Choucair concluded that investors in 2026 need to evaluate artificial intelligence across the entire financing chain, from a retail bond coupon in Tokyo to credit spreads, private-market valuations, and data-center capital expenditure.
“The capital structure will become a decisive factor in determining the winners of the artificial-intelligence race,” Choucair said.
The true test, he added, will not be whether a single bond offering succeeds. It will be whether companies can ultimately service their commitments through genuine operating cash flows rather than relying on successive rounds of financing.