Masayoshi Son
Information Revolution
Masayoshi Son founded SoftBank in 1981 and has run it as a vehicle for what he calls the Information Revolution — the thesis that computing, connectivity and now artificial intelligence reorganise every industry they touch, and that the returns accrue to whoever reaches scale first. His method is venture capital applied at unusual size: the Vision Fund underwrote a distribution of outcomes in which a small number of extreme winners were expected to pay for many failures, rather than seeking a margin of safety on each position. Son will accept losses, and rich prices, where he believes the loss is buying a durable position and the downside is bounded. He is explicit that he judges founders over a very long horizon — he has described a 300-year vision for SoftBank — which makes his framework the opposite of the value lenses elsewhere in this app rather than a variation on them.
Methodology fidelity
Thematic proxySources reviewed 2026-08-24SoftBank publicly documents its Information Revolution/AI thesis, backing leading technology companies and ambitious entrepreneurs at scale. It does not publish this eight-rule public-equity checklist. Every numerical bar—including 25% growth, $500M revenue, two years of runway, 40% gross margin, 30% balance-sheet backing, and the 5-of-8 verdict—is an app invention. FF48 industry membership is especially coarse and cannot establish that a company fits Son’s thesis.
Valuation Method
No intrinsic value computed
No investor formulaSon underwrites a distribution of venture outcomes dominated by a few extreme winners, not a per-share value to buy below. Attaching a discounted cash flow here would attribute to him a method he does not use, so this framework scores criteria only.
Screening implementation (8 criteria)
| # | Criterion | Metric | Threshold | Data Source |
|---|---|---|---|---|
| 1 | Hypergrowth Revenue | 3-year Revenue CAGR | CAGR > 25% | Income Statement: Revenue, 3-year history |
| 2 | Scale Achieved | Latest annual revenue | Above $500M — a real business, not a concept | Income Statement: Revenue |
| 3 | Growth-Justified Losses | Net income sign vs. revenue growth | Profitable, or loss-making while revenue grows above 20% | Income Statement: Net Income and Revenue, 3-year history |
| 4 | Balance-Sheet Runway | Cash / annual operating burn | At least 2 years of runway, or positive operating cash flow | Balance Sheet: Cash; Cash Flow Statement: Operating Cash Flow |
| 5 | Margin Scalability | Gross margin | Above 40% — headroom for the model to scale into | Income Statement: Gross Profit / Revenue. Falls back to the operating-margin trend when gross profit is untagged, at reduced confidence. |
| 6 | Information Revolution Alignment | Fama-French 48 industry | Operates in an information-driven industry | companies.ff48_industry, derived from the SIC code |
| 7 | Asymmetric Payoff | (Cash + Equity) / Market Cap, alongside revenue growth | Balance-sheet backing at least 30% of market value AND revenue CAGR above 25% | Balance Sheet: Cash, Shareholders Equity; Market Price × Shares Outstanding |
| 8 | Founder-Led Conviction | Founder involvement and vision horizon | Is a founder still setting a decades-long direction? | Qualitative: SoftBank Vision Fund presentations, founder track record |
Verdict Logic
PASS when at least 5 of 8 criteria hold, applied as a proportion of the criteria actually evaluated. Note that criterion 3 inverts the usual convention: a loss is not disqualifying here, an unproductive loss is.
Sources
SoftBank Group Annual Reports and Investor Presentations
primary2017–presentSoftBank Group Corp
The primary source for the Information Revolution thesis, the Vision Fund portfolio construction logic, and the stated expectation that a small number of positions carry the whole return.
View source →SoftBank Vision Fund Strategy Materials
primary2017–presentSoftBank Investment Advisers
Sets out the fund-level approach: concentration in category leaders, willingness to fund losses that buy scale, and judging founders over a very long horizon.
View source →