Masayoshi Son: The Billion-Dollar Conviction Machine Behind SoftBank

There is a particular kind of investor who does not diversify, does not hedge, and does not apologize. Masayoshi Son is that investor, scaled to a size that bends markets. The founder of SoftBank has spent four decades converting an unshakable belief in his own foresight into one of the largest technology conglomerates on earth — and into a personal fortune measured in the tens of billions. His career is a study in the economics of conviction: how a single mind, armed with capital and audacity, can move billions of dollars and thousands of companies in a direction only he can see.

Son does not present himself as a steward of capital. He presents himself as a man with a three-hundred-year plan. That framing — equal parts visionary manifesto and negotiating tactic — is the engine of the entire SoftBank empire.

From a Kyushu Classroom to the World’s Boldest Bet

Masayoshi Son was born in 1957 in Tosu, a small city on the island of Kyushu. His family was Korean-Japanese, and he grew up in a Japan where that identity carried stigma. He has said the experience of being treated as an outsider sharpened his ambition rather than diminished it. At sixteen he left for California, talked his way into a high school, and eventually enrolled at UC Berkeley, where he studied economics and computer science.

The origin story that SoftBank itself likes to tell involves Son, still a student, inventing a pocket translator and selling it to Sharp for roughly a million dollars — an early proof of concept that technology plus audacity equals money. He returned to Japan in 1981 and founded SoftBank, initially a software distributor. The name was a statement: soft (software) plus bank (a reservoir of value). For a decade the company was a scrappy distributor riding the PC boom, then an internet investor in the 1990s, when Son made early, enormous bets on Yahoo and other web properties that briefly made him the richest man in the world in 2000, before the dot-com crash wiped away most of that paper wealth.

That crash is essential to understanding Son. Rather than retreat, he doubled down on the thesis that the internet would reorganize civilization. He invested in broadband in Japan when competitors thought it was madness. He bought Vodafone Japan and turned it into a mobile carrier. Each time the market told him he was wrong, he treated the verdict as premature.

The pattern is worth naming, because it recurs at every stage of his career: Son treats market skepticism as a lagging indicator rather than a verdict. In the late 1990s, when Japanese broadband was dominated by incumbents who controlled the last mile, he launched Yahoo BB and priced it aggressively, forcing a price war that reshaped the country’s internet access market. The bet looked reckless on paper; it worked because Son was willing to absorb losses long enough for scale to arrive. That is the same logic he later applied to ride-hailing, to co-working, and to chip design — patience financed by cash flow from elsewhere in the empire.

The Vision Fund: Institutionalizing Conviction

In 2017, Son did something no one had done at scale: he turned his personal investment style into a product. The SoftBank Vision Fund, seeded with $60 billion from Saudi Arabia’s Public Investment Fund and $15 billion from Abu Dhabi’s Mubadala, alongside SoftBank’s own capital, became the largest technology investment vehicle ever assembled. A second fund followed, backed heavily by SoftBank itself after the first fund’s outside investors grew wary.

The Vision Fund’s logic was Son’s logic, industrialized. Find founders with world-conquering ambition, give them more money than they know how to spend, and let them crush competitors through sheer capital advantage. The fund poured tens of billions into companies like WeWork, Uber, DoorDash, Slack, and a long tail of startups across ride-hailing, fintech, and real estate.

The mechanism deserves scrutiny, because it was not merely a matter of writing bigger checks. The Vision Fund’s structure concentrated decision rights in Son and a small circle of lieutenants, with an investment committee that functioned more as a ratification body than a check on his judgment. Deals could move from first meeting to term sheet in weeks. Founders were offered something no traditional venture firm could match: not just capital but the implicit promise of follow-on support, access to SoftBank’s network of portfolio companies, and the threat of that capital being deployed against them if they declined. That dynamic — the fund as both partner and potential competitor — changed the bargaining position of founders across Silicon Valley.

The results were spectacular in both directions. Some bets — Alibaba above all — became legendary. Son’s early stake in Jack Ma’s company, reportedly acquired for around $20 million, grew to be worth well over $100 billion at its peak, the single greatest venture return in history and the financial backbone of Son’s mystique. Other bets became cautionary tales. WeWork’s collapse ahead of its 2019 IPO, in which SoftBank had invested more than $10 billion, became a case study in what happens when conviction outruns diligence. Son has been candid about it: he called the WeWork episode a personal failure of judgment, a rare admission from a man who rarely concedes error.

The Alibaba stake is worth dwelling on, because it explains why the Vision Fund could exist at all. That single position generated returns that dwarfed the fund’s losses elsewhere, giving Son both the capital and the credibility to keep raising and deploying. It is the financial equivalent of a get-out-of-jail card — one that he has repeatedly played, including by selling down large portions of the stake to fund buybacks and new bets. The concentration that makes Son vulnerable is the same concentration that made him rich.

The Mechanics of the Conviction Machine

What distinguishes Son is not simply risk tolerance but the architecture he built to sustain it. SoftBank’s holding structure, its telecom cash flows, its stakes in Arm and Alibaba, and its willingness to leverage — including through complex instruments and margin loans against its own shares — create a perpetual reservoir of capital that can be redeployed toward whatever Son believes is next.

This is the machine: cash-generating assets fund speculative ones; speculative wins fund bigger speculative ones; and the whole apparatus is held together by Son’s credibility with lenders and sovereign funds. When that credibility wobbles — as it did in 2020 and again in 2022, when SoftBank posted record losses and Son announced a defensive posture — the machine slows. But it does not stop. Son has repeatedly sold down assets, including large portions of the Alibaba stake, to reload.

The telecom business is the quiet foundation of this structure. SoftBank’s mobile carrier in Japan generates predictable cash flow, and that cash flow has historically served as collateral and as a source of dividends that can be redirected. The 2016 acquisition of Arm, the British chip designer whose architecture underpins most of the world’s smartphones, added a different kind of asset: a royalty stream tied to the entire mobile ecosystem, and a strategic position in the semiconductor supply chain that Son has since framed as central to the AI era. Taking Arm public in 2023 gave SoftBank a liquid, highly valued asset that could be used to raise capital without surrendering control.

The leverage is the part that makes outsiders nervous, and for good reason. Borrowing against your own shares works beautifully when the shares rise and painfully when they fall, because margin calls can force sales at the worst possible moment. Son has navigated several such moments by pre-emptively selling assets and by leaning on his relationships with sovereign wealth funds, which have tolerated volatility in exchange for access to deal flow. That relationship is itself a form of capital — reputational collateral that can be drawn upon when financial collateral tightens.

The conviction is also thematic. Son has cycled through eras of belief: the PC revolution, the internet, mobile, and now artificial intelligence. In recent years he has positioned Arm — the chip-design company SoftBank acquired in 2016 and took public in 2023 — as the keystone of an AI future, and he has spoken of a coming “singularity” in which artificial intelligence surpasses human capability. Whether one shares that belief is almost beside the point; the belief is what organizes the capital.

The Three-Hundred-Year Horizon

Son’s most revealing habit is his time horizon. He speaks of SoftBank as a company meant to endure for three centuries, and he has structured his succession planning around that idea. He has groomed and lost lieutenants, most notably former COO Marcelo Claure, and has repeatedly returned to the operational front himself. At an age when most founders retire, Son remains the central decision-maker, the single point of conviction upon which the entire structure depends.

That concentration is both the empire’s strength and its fragility. SoftBank’s fortunes are not diversified across a committee of cautious managers; they are an expression of one man’s worldview. When Son is right, the returns are historic. When he is wrong, the losses are correspondingly enormous.

The three-hundred-year framing is easy to dismiss as rhetoric, but it has practical consequences. It justifies holding positions through drawdowns that would force a quarterly-results-driven fund to sell. It gives Son a narrative to offer lenders and sovereign partners who might otherwise balk at volatility. And it creates a succession problem: a plan measured in centuries requires an institution that can outlive its founder, yet SoftBank’s decision-making has remained stubbornly personal. The lieutenants who might have inherited the machine — Claure most visibly — departed, and Son returned to the front each time. The horizon is long; the bench, so far, has been short.

What the Machine Teaches

Masayoshi Son’s career is a rebuttal to the idea that investing is primarily about avoiding mistakes. His record is full of them — dot-com blowups, WeWork, and a string of Vision Fund write-downs. But the handful of bets he got spectacularly right, held with a conviction that most investors could not stomach, outweighed everything else.

The lesson is not that conviction always wins. It is that conviction, backed by capital and a structure designed to survive its own errors, can compound in ways that careful diversification cannot. Son built a machine that turns belief into leverage and leverage into scale. Whether the next era — AI, chips, and whatever Son sees beyond them — vindicates that machine is the open question of his career. But the machine itself, and the man who runs it, have already reshaped how the world’s most ambitious technology companies get funded.

For the reader, the practical takeaway is narrower than the mythology. Son’s approach is not a template most investors can copy, because most investors lack the cash-generating assets, the sovereign backers, and the stomach for drawdowns that the model requires. What is transferable is the discipline of asking which bets are worth holding through public skepticism — and which are simply conviction unmoored from evidence. The difference between the Alibaba stake and the WeWork episode is not the size of the bet or the strength of the belief. It is whether the underlying business could eventually justify the price. Son’s machine works when his conviction is paired with a business that compounds; it fails when the story outruns the economics. That distinction, more than any three-hundred-year plan, is the real lesson of his career.

For better and worse, Masayoshi Son is the proof that in modern capitalism, the size of the bet can be the strategy.