The 300-Year Vision Anchor
Lock onto a multi-decade technology trend, then hold through every crash
Masayoshi Son does not invest in companies — he invests in technological eras. His mental model starts with identifying which technology wave is next (microchip, internet, mobile, AI) and then backing that wave through downturns, crashes, and personal humiliation rather than exiting at the first sign of trouble. Masa held his Alibaba stake for fifteen years before beginning to sell, and stayed in broadband — losing a billion a year — because he was convinced the trend was inevitable, just early.
The practical implication is that the investment thesis must be timeframe-matched to the technology cycle, not to quarterly earnings or even annual returns. When the dot-com bubble burst in 2000, Masa was saying within months that markets had fallen so far that now was the time to invest in broadband. That contrarian re-entry, grounded in an unchanged long-run thesis, is the engine of his biggest wins.
Barber frames this as the defining trait that separates Masa from a reckless speculator: not the absence of losses, but the constancy of the underlying directional conviction. The question to ask is not 'is this investment profitable right now' but 'is the technology wave I identified still coming?'
- Identify the technology era first; pick companies second — the wave matters more than any individual bet
- Measure holding periods in technology cycles (decades), not market cycles (years)
- A crash in price is not a crash in thesis — re-enter or hold if the underlying trend is intact
- Accept losing a billion a year as tuition fees if the trend is confirmed — short-term losses are the cost of being early
- Sell only when the technology era is over, not when sentiment turns negative
- Map the technology S-curveIdentify which technology is transitioning from early adoption to mass market. Son mapped microchip → internet → mobile → AI as sequential eras, each predictable by looking at the previous curve's trajectory.Pro tipLook for the infrastructure bets — ARM for chips, Yahoo for internet, Sprint for mobile — rather than just application-layer companies. Infrastructure plays compound across the whole era.WarningDo not confuse 'a company in this space' with 'the wave itself.' Son lost billions on WeWork by funding a real-estate business dressed in tech language.
- Size the bet to survive the waitCommit enough capital to matter but ensure you can survive the inevitable trough between early bet and mass adoption. Son used leverage against Japanese low interest rates to magnify position size, but also securitized assets to service debt during down periods.Pro tipPair the long-horizon thesis with a liquidity buffer — Son's mistake in the Vision Fund era was deploying capital faster than the companies could absorb it, which forced poor allocation decisions.WarningLeverage amplifies both the patience advantage and the solvency risk. Over-leverage forced asset sales at the worst times (SoftBank stock crashes of 2000 and 2020).
- Reframe crashes as confirmation or evidenceWhen a market crash hits, return to the original thesis question: 'Is the technology wave I identified still coming?' If yes, treat the crash as a buying opportunity or hold signal. If no, exit cleanly. Son treated the dot-com crash as an interruption, not a refutation.Pro tipBarber notes Son's post-crash public posture shifted within months from loss acknowledgement to identifying the next entry point — the emotional reset was deliberate, not denial.
- Hold through social pressureInstitutional investors, board members, and analysts will pressure an exit during drawdowns. Son held Alibaba for fifteen years despite SoftBank's own financial pressures because selling would have been thesis-betrayal. Develop a written thesis document that can be re-read during pressure periods.Pro tipSon used public declarations ('AI is the next great thing') as self-commitment devices — broadcasting the thesis made reversal more costly than holding.WarningPublic commitment is a double-edged tool: it locks you into holding even when the thesis has genuinely broken. Build in explicit thesis-exit criteria before you commit publicly.
- Exit only at era transition, not at valuation peakSon's biggest regret was selling Nvidia near a 5% stake in 2019. He needed liquidity, but his thesis — that AI compute would be the defining infrastructure of the next era — was intact. Valuation peaks inside a still-running era are not exit signals.WarningFailing to distinguish 'I need liquidity' from 'my thesis is over' is how multi-generational returns get cut to merely good ones. Separate liquidity management from thesis management.
Son met Jack Ma in Beijing in 1999 and committed capital to Alibaba after a twenty-minute meeting, famously saying he could 'smell the horse flesh' — he sensed the founder quality and matched it to his Chinese internet era thesis. He held the stake for fifteen years through multiple SoftBank financial crises.
After the dot-com crash wiped 98% of Son's paper wealth, he pivoted within months to broadband infrastructure in Japan. He lost approximately a billion dollars a year during the build-out phase, which analysts and investors treated as evidence of recklessness.
Son held a near-5% stake in Nvidia — a perfect fit for his AI era thesis. In 2019, under Vision Fund liquidity pressure, he sold most of it. His AI thesis remained unchanged; his treasury management forced the exit.
This framework is Masayoshi Son's own, as reconstructed by Barber through six research trips to Japan over fourteen months and direct interviews with Son. Barber describes how Son articulated his investing logic across decades — from Yahoo in 1995 to Alibaba in 1999 to broadband in 2001 to AI from 2023 — as sequential bets on the same underlying thesis: that information technology would restructure every industry. Barber's contribution is the journalistic framing: he positions this not as recklessness but as a coherent, if extreme, application of conviction investing.