원문 보기
This week, SoftBank CEO Masayoshi Son was one of three tech executives invited to the White House to announce the quote Stargate joint venture that would invest billions of dollars to boost AI infrastructure in the US. In 2000, Masayoshi Son wrote a $20 million check to a small Chinese e-commerce startup.
That single bet would eventually be worth over $100 billion, one of the greatest venture capital returns in history. In 2019, he poured billions into a company that rented office space, pushed its valuation to $47 billion, and watched it collapse into bankruptcy, one of the most public humiliations in venture investing.
In 2025, he sold his entire Nvidia stake, fully exited Alibaba after 25 years, and put $64.6 billion into a single company, OpenAI. His net worth hit $80 billion this year, making him the richest person in Asia. One man, three decades, the same pattern every time. Bet everything on a future he claims to see before anyone else.
Sometimes catastrophically wrong, sometimes the greatest investor alive. This is Masayoshi Son and the company he built to make civilization-scale bets, SoftBank. Son was born on August 11th, 1957, on the edge of a shantytown in Tosu, Kyushu. His grandfather had migrated from Daegu, Korea, during Japanese colonial rule and worked as a miner.
His father raised pigs and chickens on land the family did not own, illegally built their home on railway property, and ran an illegal sake brewing operation that eventually funded a chain of pachinko parlors. Zainichi Koreans faced open discrimination in post-war Japan. Son has described collecting leftover food from neighbors to feed the family's pigs.
His biographer, Lionel Barber, has said he was raised effectively pretending to be Japanese undercover in the only country he had ever lived in. At 16, he made 60 long-distance phone calls until a Japanese business icon finally agreed to meet him. That same year, he left for California. At UC Berkeley, he became fixated on the idea that the microchip was about to trigger a transformation as significant as the Industrial Revolution.
He also invented an electronic pocket translator and sold the concept to Sharp for roughly $1 million, his first real money. He mapped out a 300-year vision for his life's work and returned to Japan in 1981 to found SoftBank. Three years later at 24, he was diagnosed with chronic hepatitis and given roughly 5 years to live.
He ran the company from a hospital bed, crying alone during treatment, refusing to let go of what he had built. An experimental treatment saved him. The doctors were wrong. SoftBank started as a software distributor. Sun literally stood on a soapbox in electronic stores pitching software to confused shoppers, but he understood the value was not in the box.
It was in the network the box connected to. When the internet arrived in the 1990s, he pivoted SoftBank from distributor to investor, pouring money into early internet companies including a major stake in Yahoo. Most of it got wiped out when the dot-com bubble burst. Sun's personal net worth reportedly fell by tens of billions in months, one of the largest paper losses any individual has recorded.
He kept going. The Alibaba bet in 2000 came after a conversation with Jack Ma lasting only a few minutes. It gave Sun a reputation as a kingmaker and an ocean of capital to deploy at the scale he had always imagined. He used it to build a hybrid empire. In 2006, he executed a leveraged buyout of Vodafone's Japanese mobile business, turning SoftBank overnight into a telecom operator with predictable cash flow.
In 2013, he acquired Sprint, attempting to break the AT&T Verizon duopoly. Sprint struggled for years before merging with T-Mobile in 2020. SoftBank fully exited that by 2024 for a substantial profit, one of Sun's messiest bets and eventually paying off on a long enough timeline. In 2017, Sun launched the SoftBank Vision Fund, roughly 98.6 billion dollars.
At the time, the largest technology-focused investment vehicle ever assembled. The strategy was blitzscaling. Pour billions into late-stage startups so they could outspend every competitor into irrelevance across a portfolio that eventually spanned 400 countries. Then came WeWork. Sun saw a kindred spirit in founder Adam Neumann and pushed the private valuation to 47 billion dollars for a company that rented office space.
When WeWork tried to go public in 2019, its actual financials were exposed. Staggering losses, bizarre governance. The IPO collapsed. WeWork eventually filed for Chapter 11 bankruptcy in 2023. Total losses tied to the investment exceeded 11.5 billion dollars. Sun later admitted the mistake directly, telling shareholders his own judgment had not been good.
The money behind the fund carried its own complications. The largest outside investor was Saudi Arabia's Public Investment Fund run by Crown Prince Mohammed bin Salman, who committed 45 billion dollars after a 45-minute meeting. Sun later bragged about the pace, calling it a billion dollars a minute.
When journalist Jamal Khashoggi was murdered inside the Saudi consulate in Istanbul in 2018, Son publicly condemned the killing while declining to walk away from the money already committed. SoftBank's market value dropped by billions as investors questioned the relationship. Son has since acknowledged the partnership has not delivered the returns he promised MBS.
While Vision Fund headlines were dominated by WeWork, a quiet bet that was already in motion. In 2016, SoftBank paid $32 billion for ARM Holdings, a company that designed the architecture used in most of the world's smartphones. Analysts were confused. Son said he was betting on what mobile would become.
ARM went public again in 2023. SoftBank retains roughly 90% ownership, and ARM now predicts annual revenue could reach $25 billion within 5 years. Then, the largest bet of his career. Throughout 2025, SoftBank sold its entire Nvidia stake, fully divested from Alibaba after 25 years, and liquidated large portions of the Vision Fund portfolio.
All of it funneled into OpenAI. By March 2026, SoftBank had committed $64.6 billion for roughly a 13% stake. Following OpenAI's conversion to a for-profit structure and the funding round valuing the company at $852 billion. On paper, SoftBank is already sitting on approximately $45 billion in gains, a 129% return.
SoftBank is also a core partner in the Stargate project alongside OpenAI and Oracle. A proposed $500 billion AI infrastructure build-out announced from the White House alongside President Trump in January 2025. Son has positioned SoftBank as essential to reassuring American chip manufacturing, giving the company direct access to US policy on chips, AI, and trade at the exact moment those rules are being written.
For the fiscal year ending March 2026, SoftBank reported net profit of approximately $31 billion, more than quadruple the prior year. As the stock more than tripled, Son's personal net worth reached $80 billion, an 80% jump from his previous peak, making him the richest person in Japan and briefly all of Asia.
Not every analyst is convinced the run continues cleanly. Anthropic surpassed OpenAI in valuation following a May 2026 round pricing it at $965 billion, and cheaper open-source models could pressure OpenAI's growth from below. SoftBank has also taken on a $40 billion bridge loan to keep funding the buildout, meaning the current strategy runs partially on borrowed money layered into a stock price some analysts describe as showing signs of mania.
Every major decision in this story traces back to two moments most people never hear about. A teenager who made 60 phone calls until someone finally said yes, and a 24-year-old running a company from a hospital bed after being told he had 5 years to live. That is not really a story about patience, it is a story about someone who has never treated a no as final.
Not from a skeptical father, not from a market crashing around his bets, not from a body that was supposed to fail him decades ago. The uncomfortable truth is that Son's genius and his recklessness are not two different traits, they are the same trait pointed at different outcomes. The conviction that produced Alibaba is identical to the conviction that produced WeWork.
You cannot have one without risking the other. Right now that conviction is pointed at artificial super intelligence, backed by tens of billions of dollars and a growing pile of debt. Drop your take in the comments. Is the OpenAI bet the next Alibaba or is SoftBank building the next WeWork at 10 times the scale?
Son himself has shown us it could genuinely be either. Subscribe to Stig Research for more breakdowns like this. The investors, the empires, and the bets reshaping the global economy.
Comments 0