Four decades of technology-empire compounding, on one filterable page — anchored by the greatest venture investment in history.
SoftBank was founded by Masayoshi Son in Tokyo in September 1981 as a personal-computer software distributor. Through the 1990s the company transformed into a technology investor — anchored by the 1996 Yahoo! Japan joint venture and, in January 2000, an approximately $20M investment in Alibaba for ~34% of the company, a decision now routinely cited as the single greatest venture investment in history in absolute-dollar terms. The 2006 Vodafone Japan acquisition ($15B) built the domestic telecom platform; the 2013 Sprint Corporation deal ($22B for 80%) landed SoftBank in the US carrier market; the 2016 ARM Holdings acquisition ($32B) added the world's dominant chip-designer platform. The 2017 launch of the $100B Vision Fund — anchored by Saudi Arabia's PIF and Abu Dhabi's Mubadala — scaled SoftBank into a technology capital allocator of unprecedented size, with concentrated late-stage bets on Uber, DoorDash, Coupang, DiDi, and the catastrophic WeWork commitment. The 2023 ARM Nasdaq IPO validated the whole-company thesis; the 2024-2026 pivot into OpenAI, Perplexity, and AI infrastructure defines the current chapter. This page catalogs the material record from the 1981 founding through today. It is intentionally a living reference — as new positions are entered or divestitures announced, the row is added, the roll-ups reflow, and the sitemap timestamp bumps. Nothing here is investment advice. Everything here is a fact-checkable practitioner reference for a very specific question — what does four decades of concentrated technology-empire investing actually look like in list form?
Nine columns. Year of announcement or close. Target name. Region at time of investment (Japan, US, UK, China, India, Southeast Asia, Middle East, Global for fund-level events, or Corporate for structural). Approximate consideration in USD — several venture rounds and private follow-ons were never publicly disclosed; where the individual round size is undisclosed the row is flagged "n/d" or "approx" rather than fabricated. Deal structure (whole-company, majority stake, minority stake, Vision Fund position, merger, spinoff, IPO, divestiture, or fund launch). Counterparty type. The SoftBank / Vision Fund flag — SoftBank's signature is concentrated late-stage technology positions retained for long-duration compounding through market cycles. Distinctive notes. Current status — held, divested, partially divested, IPO'd, or merged.
Sort and filter. Click any column header to sort. Use the decade, region, structure, counterparty, and SoftBank / Vision Fund filters to isolate a slice. The search box matches target names and notes.
Two-structure framework. SoftBank Group Corp. (TSE: 9984) is the Tokyo-listed parent holding company — it books the ARM Holdings majority stake, the residual Alibaba position, the SoftBank Corp. Japanese telecom stake, the T-Mobile equity received in the 2020 Sprint merger, and the direct-investment book. Vision Fund 1 (VF1, $100B committed 2017, anchor LPs Saudi PIF ~$45B and Abu Dhabi Mubadala ~$15B) and Vision Fund 2 (VF2, ~$30-40B raised 2019, primarily SoftBank Group capital) are the fund vehicles that hold most of the modern late-stage venture portfolio. Vision Fund attribution is sometimes ambiguous where SoftBank Group participated alongside or ahead of the fund; where possible individual rows note which vehicle led. A prospective Vision Fund 3 discussed in 2024-2026 has not yet closed at the time of publication.
SoftBank / Vision Fund marker. A YES flag means the transaction was a concentrated technology position — typically a large minority-stake or Vision Fund late-stage round — retained with long-duration compounding intent. This is SoftBank's signature move: Alibaba is the archetype; Uber, DoorDash, Coupang, ARM, DiDi, and (with catastrophic outcome) WeWork are variations. A NO flag means a whole-company control acquisition (ARM, Sprint, Vodafone Japan, Ziff-Davis), a structural corporate event (SoftBank IPO, Sprint/T-Mobile merger, Vision Fund launches, ARM IPO), or a divestiture. Roughly 55-65% of the cataloged events are concentrated-position bets; the balance are control acquisitions, structural events, and portfolio divestitures.
Every material SoftBank Group acquisition and Vision Fund investment from Masayoshi Son's 1981 founding through today, anchored by the transformative 2000 Alibaba stake, the 2016 ARM Holdings acquisition, the 2017 launch of the $100B Vision Fund, and the 2023 ARM Nasdaq IPO. Sortable by year, region, deal size, structure, and counterparty type — with the SoftBank / Vision Fund concentrated-position pattern flagged across four decades of activity. Search by target name (Alibaba, ARM Holdings, Sprint, Vodafone Japan, Yahoo Japan, WeWork, Uber, DoorDash, Coupang, DiDi, Nvidia, Slack, Flipkart, OpenAI, Perplexity, Supercell, GungHo, Ziff-Davis, Kingston Technology, T-Mobile), by region (Japan, United States, United Kingdom, China, India, Southeast Asia, Middle East), or by structural term (Vision Fund, minority stake, whole-company, IPO, divestiture, fund launch). Every row is a fact-checkable reference. This is a living dataset — updated whenever SoftBank closes a new material deal or announces a portfolio divestiture.
| Year | Target | Region | Consideration | Structure | Counterparty | SoftBank / Vision Fund | Notes | Status |
|---|
Roll-ups reflect the investments cataloged in the table above. Dollar totals are directional at best and reflect only the subset of transactions where consideration was publicly disclosed — several venture-round investments and follow-ons were not individually disclosed. The 2017 Vision Fund launch is treated as a fund-launch structural event rather than a cash-deployment event (individual Vision Fund investments are booked as separate rows). The 2000 Alibaba investment ($20M) is by dollar count trivial next to Vision Fund late-stage rounds; by outcome it is the entire record.
Includes only the subset of SoftBank / Vision Fund investments where consideration was individually disclosed. The 2000 Alibaba investment ($20M) is dwarfed on a dollar basis by 2013-2019 Sprint, ARM, and Vision Fund rounds — a fact that is the entire investing lesson of this record. Bar length is proportional within this table only. Directional; not audit-grade.
Minority stakes and Vision Fund positions are the near-total backbone of the SoftBank record. Whole-company acquisitions cluster around control platforms (ARM, Sprint, Vodafone Japan, Ziff-Davis, Supercell). Fund launches, IPOs, and divestitures sit in their own buckets.
Region is captured at time of the transaction. Japan concentrates the early telecom and internet-portal buildout years. The United States dominates the Vision Fund era. China is anchored by the single Alibaba line plus DiDi. India, Southeast Asia, and the United Kingdom (ARM) carry meaningful weight of the modern portfolio.
Roughly 55-65% of the cataloged events are concentrated-position bets — SoftBank's signature move, of which the 2000 Alibaba stake is the archetype. The balance are control acquisitions (ARM, Sprint, Vodafone Japan), structural events (SoftBank IPO, Vision Fund launches, ARM IPO, Sprint/T-Mobile merger), and divestitures (the ongoing Alibaba and T-Mobile trims).
An acquisition record is a lagging indicator. The leading indicator, for SoftBank specifically, is the capital-allocation posture between four ongoing programs: (1) the systematic partial-divestiture of the Alibaba and T-Mobile stakes to fund share buybacks and de-lever the balance sheet — a discount-arbitrage move that assumes the SoftBank Group NAV discount is worth attacking directly rather than waiting for the market to close it; (2) the reinvestment of proceeds and cash flow into the AI-infrastructure thesis, including outsized positions in OpenAI, Perplexity, and the ARM-adjacent chip stack; (3) the continued operation and periodic monetization of ARM Holdings, now Nasdaq-listed but majority-controlled; and (4) selective Vision Fund 2 late-stage positions and prospective Vision Fund 3 discussions.
Under founder-CEO Masayoshi Son, publicly framing SoftBank's next chapter as a "10x on AI" positioning, the emphasis has shifted decisively toward AI-infrastructure and away from the late-cycle consumer-technology venture rounds that defined the 2018-2021 Vision Fund vintage. Whether ARM can continue to appreciate on AI demand, whether the OpenAI and Perplexity positions can compound like Alibaba, whether WeWork-style catastrophic write-downs can be avoided in the AI wave, and whether the group's leverage can be responsibly managed alongside these commitments are the four defining questions of the next chapter.
ARM Holdings (Nasdaq: ARM) is the SoftBank-controlled UK-based chip-designer platform, IPO'd on Nasdaq in September 2023 at approximately $54B valuation. SoftBank retained approximately 90% of the post-IPO shares. Through 2024-2026 ARM's share price appreciated substantially on the AI-infrastructure thesis, making the residual stake one of the most valuable single lines on the SoftBank Group balance sheet.
SoftBank continues to reduce the Alibaba (NYSE: BABA) position that anchored the 2000-era record. Cumulative Alibaba divestiture proceeds since 2021-2022 run into the tens of billions of dollars. The residual position is materially smaller than the peak but remains a meaningful line, and further reductions continue to be executed opportunistically.
SoftBank participation in OpenAI — via Vision Fund and direct SoftBank Group capital — positions the group as one of the largest single AI infrastructure investors. Individual round sizes are not fully disclosed and the position is expected to compound significantly if the OpenAI thesis plays out. The commitment posture is characteristic of the Masayoshi Son concentrated-position playbook.
Discussions of a prospective Vision Fund 3 focused on AI infrastructure have been ongoing in 2024-2026. No formal close has been announced at the time of publication. Whether external LPs return at the scale of the 2017 Saudi PIF and Abu Dhabi Mubadala commitments remains the key open question.
SoftBank belongs to the small handful of records that define the modern compounding era — and it is the only one anchored by a single 2000-vintage venture investment ($20M in Alibaba for approximately 34%) that outperformed nearly every other deal on the record around it. These are the companion practitioner references and case memos already published on the Institute site.
Educational reference. Not investment advice. Not a solicitation. Not affiliated with SoftBank Group Corp., SoftBank Corp., SoftBank Vision Fund, ARM Holdings plc, Alibaba Group Holding Limited, Sprint Corporation, T-Mobile US, Vodafone Group, Yahoo! Inc., LY Corporation, WeWork Inc., Uber Technologies, DoorDash, Coupang, DiDi Global, Nvidia, OpenAI, Perplexity, or any of their subsidiaries or affiliates, nor with Masayoshi Son, Jack Ma, Pony Ma, Rene Haas, Marcelo Claure, Rajeev Misra, Ronaldo Fiorini, or any past or present SoftBank, Vision Fund, or portfolio-company executive. The Baratelli Institute publishes under the Lowe v. SEC publisher exception; neutral positioning maintained throughout. Deal figures cited in this catalog are sourced primarily to SoftBank Group Corp. annual reports and integrated reports, Tokyo Stock Exchange filings, Vision Fund periodic disclosures, contemporaneous press coverage (Reuters, Bloomberg, Financial Times, Nikkei, The Wall Street Journal, TechCrunch, The Information), and standard practitioner references. Dollar amounts are approximate; where original consideration was denominated in JPY, GBP, HKD, INR, EUR, or other non-USD currencies the USD equivalent is directional and reflects contemporaneous FX rates. Several venture-round investments and follow-on rounds are individually undisclosed and are flagged with "approx" or "n/d" (not disclosed) rather than fabricating precision. The Alibaba stake value is a point-in-time reference that fluctuates daily with the New York and Hong Kong share prices; the historical arc from $20M in 2000 to a $150B+ peak in 2020 is the fact-checkable through-line of the entire record. Vision Fund attribution can be ambiguous where SoftBank Group participated alongside the fund; where possible individual rows note vehicle. Corrections welcome via the link in the footer.
“When I met Jack Ma for the first time, he did not have any business plan and zero revenue, employees only 35 or 36. But his eyes were shining. I decided in five minutes to invest.”