The SoftBank Vision Fund is a $98.6 billion global technology investment vehicle launched in 2017 by SoftBank Group Corp., deploying capital across late-stage venture and growth equity in AI, cloud computing, and enterprise software. It is structured as a limited partnership with institutional and sovereign investors, managed by SoftBank Investment Advisers.
The SoftBank Vision Fund is a $98.6 billion global technology investment vehicle launched in 2017 by SoftBank Group Corp., deploying capital across late-stage venture and growth equity in AI, cloud computing, and enterprise software. It is structured as a limited partnership with institutional and sovereign investors, managed by SoftBank Investment Advisers.
The Vision Fund model emerged from Masayoshi Son's conviction that technology-driven transformation would drive long-term value creation across all sectors. With capital commitments from the Saudi Public Investment Fund, Abu Dhabi's Mubadala, Temasek, and other institutional investors, SVF represented a departure from traditional venture capital structures: instead of managing $1–3 billion vehicles with dispersed ticket sizes, SVF deployed multibillion-dollar checks into late-stage growth companies with clear paths to unicorn status and eventual public markets.
For institutional allocators—sovereign wealth funds, large pension plans, and endowments—understanding the Vision Fund's structure, performance track record, and strategic implications is essential to assessing venture capital allocation within diversified portfolios.
What is the SoftBank Vision Fund's core structure and capitalization?
The SoftBank Vision Fund launched in May 2017 with final commitments of $98.6 billion. The fund was anchored by a $45 billion commitment from Saudi Arabia's Public Investment Fund (PIF), the kingdom's sovereign wealth manager. Additional major commitments came from Abu Dhabi's Mubadala Investment Company, Temasek Holdings (Singapore's sovereign vehicle), SoftBank Group itself ($28.3 billion), and U.S. corporations including Raytheon Technologies and Qualcomm.
This investor composition reflected a strategic shift in venture capital: institutional capital—historically allocated to buyout and infrastructure funds—began flowing into late-stage technology ventures. The fund's limited partnership agreement followed pension fund governance standards, with quarterly reporting, independent oversight committees, and transparent fee structures common to institutional asset management.
A second vehicle, the SoftBank Vision Fund 2, launched in 2019 and closed at $49.9 billion in 2021. Fund 2 followed a similar thesis but with modified investor composition and a narrower geographic mandate. By 2024, SoftBank had also established Vision Fund 3 with approximately $20 billion in commitments, signaling management's confidence in the model despite Fund 1's underperformance.
How did the Vision Fund differ from traditional venture capital models?
Traditional venture capital partnerships typically manage $500 million to $2 billion in assets, deploy over 7–10 years, and focus on early-stage companies (seed through Series B) with 20–40 portfolio companies. General partners—often founders or operating executives—take active board seats and work closely with company founders on product-market fit and team building.
SoftBank Vision Fund operated at a fundamentally different scale. With $98.6 billion under management, it could deploy $500 million to $2 billion per investment, often as solo or lead investor in growth-stage rounds. Rather than backing 100+ companies with modest checks, SVF concentrated capital into 80–90 carefully selected companies where the fund took substantial minority stakes (10–50% ownership in some cases).
This megafund model required different operational practices. SoftBank established a 70-person investment team across Tokyo, London, and San Francisco, supported by dedicated operational partners for post-investment value creation. The fund created a secondary portfolio company advisory network: executive advisors from technology, logistics, and telecommunications sectors were deployed to SVF portfolio companies for strategic guidance, capital expenditure optimization, and cross-portfolio synergy development.
The megafund approach also shifted return expectations. Traditional venture funds targeted 25–30% internal rates of return (IRRs) and 3–4x multiple on invested capital (MOIC) to justify the illiquidity and operational intensity of early-stage investing. SoftBank Vision Fund, with its institutional LP base and long capital duration, modeled for 15–20% IRRs and 2–3x MOICs, acknowledging that late-stage, lower-risk companies offer lower return multiples than seed-stage bets.
What were the Vision Fund's investment criteria and sector concentration?
SoftBank's investment thesis centered on "information revolution"—the idea that artificial intelligence, cloud computing, and data analytics would transform every major industry. Within this framework, the fund prioritized companies meeting specific criteria:
Company stage and size: Series C and later rounds, with $1 billion+ valuation targets. The fund sought companies with proven product-market fit, defined business models, and clear paths to $10 billion+ valuations within 5–7 years.
Geography: Predominantly North America (approximately 70% of capital), with secondary exposure to Europe (15%) and Asia-Pacific (15%). Japan and India received particular attention as growth markets with favorable demographics and technology adoption curves.
Sector concentration: Enterprise software and cloud infrastructure (40% of fund), mobility and transportation (20%, including Uber, Grab, and Didi), e-commerce and fintech (20%), and emerging technologies (20%, including robotics and AI chips).
Size of investment: Typical checks ranged from $500 million to $2 billion, occasionally exceeding $3 billion (as in Uber's series valuations). This ticket size gave SoftBank significant negotiating power over board representation, governance rights, and information access.
By 2024, notable SVF portfolio companies included Arm Holdings (valued at $61 billion in its ARM Holdings PLC IPO in September 2023), Grab Holdings (valued at $40 billion at its 2021 SPAC transaction), Raytheon-acquired manufacturer interests, and earlier-stage cloud and AI infrastructure startups. The fund also maintained significant positions in publicly traded companies, including its stake in Alibaba Group (originally acquired for $4.4 billion in 2017, valued at $23 billion+ by 2024 depending on market conditions).
How did the Vision Fund's performance unfold through 2024?
The SoftBank Vision Fund's investment returns have been substantially negative, reflecting broader venture capital headwinds and manager-specific execution challenges.
As of the second quarter of 2024, Vision Fund 1 reported the following metrics (per SoftBank Group's investor presentation and annual earnings reports):
- Cumulative net loss: Approximately $9.3 billion (as of Q2 2024)
- Gross return: Approximately 5.1x invested capital gross of fees
- Net return: 0.82x invested capital net of fees and expenses
- Internal rate of return (IRR): Approximately -5% net of fees
- Proportion of portfolio in loss: Approximately 42% of SVF 1 holdings were marked below cost basis
These results reflected several overlapping challenges:
Valuation correction (2021–2023): The fund's largest positions—WeWork, Slack, Grab, Didi—faced public market valuations substantially below their private round pricing. WeWork's valuation fell from $47 billion (2019, private) to near-zero (following failed IPO and subsequent restructuring). Grab, acquired by SoftBank at $40 billion (2021 SPAC valuation), traded significantly below that level by 2023. These portfolio stumbles directly reduced fund returns.
Concentration and execution risk: Unlike diversified venture funds that accept high failure rates through large portfolio size, SVF's 80–90 company concentration required near-perfect execution from management. Several exits—including the failed WeWork IPO and underperformance of mobility bets—significantly underweighted returns.
Macro environment: The 2022–2023 interest rate cycle compressed venture valuations and extended exit timelines. Companies that had expected 3–5 year paths to IPO instead faced extended private rounds, refinancing challenges, and dilutive downrounds. SVF's illiquid positions were marked down accordingly.
Vision Fund 2, with a later vintage (2019–2021), had deployed approximately $32 billion by 2024 but reported similarly poor returns. Early MOIC stood near 0.9–1.0x, reflecting similar valuation pressures in 2022–2023.
How is the Vision Fund governed and managed?
SoftBank Investment Advisers (SBIA), a wholly-owned subsidiary of SoftBank Group Corp., serves as the general partner and investment manager. The organizational structure consists of:
Executive leadership: Masayoshi Son, SoftBank Group CEO, holds the title of Vision Fund chairman and retains final investment approval authority. Rajeev Misra, a former Deutsche Bank and Masayoshi Son advisor, leads SBIA as chief operating officer and principal investment authority for day-to-day decision-making.
Investment team: Approximately 70 investment professionals across technology, operations, and sector specialization, concentrated in Tokyo, San Francisco, and London. Deal sourcing comes through a combination of founder networks, secondary market brokers, and proactive outreach to late-stage companies.
Limited partner governance: SVF's limited partnership agreement includes an independent advisory committee, typically composed of 5–7 institutional LP representatives, which meets quarterly to review performance, investment decisions, and fee structures. Major LPs including Saudi Arabia's PIF and Abu Dhabi's Mubadala hold board observation rights and information access consistent with global pension fund governance standards.
Fee structure: SoftBank charges a 1.5% management fee on committed capital and 20% carried interest (profit share) on net returns exceeding a specified benchmark. This fee structure is standard for megafunds but represents substantial annual expenses given the $98.6 billion base (approximately $1.5 billion in annual management fees during the deployment period).
This governance model concentrated significant decision-making power within SoftBank and Masayoshi Son, which both enabled rapid large-scale investments (critical in competitive growth-stage rounds) but also concentrated execution risk. Major investment decisions—including the Grab $40 billion SPAC valuation and the WeWork $47 billion private valuation—reflected Masayoshi Son's personal conviction rather than consensus-driven committee review, exposing LPs to individual decision-maker risk.
What are the strategic implications for sovereign wealth funds and pension allocators?
The SoftBank Vision Fund's performance and governance model offer several lessons for long-term institutional allocators.
Scale and illiquidity trade-offs: While megafund scale provided SVF access to large growth-stage rounds unavailable to smaller venture funds, it also created illiquidity and concentration risk. Unlike diversified venture funds with 40–60 portfolio companies, SVF's 80–90 investments meant that a handful of underperformers (WeWork, certain mobility bets) significantly depressed overall returns. Allocators must determine whether megafund access justifies these concentration risks versus allocating to diversified fund-of-funds structures like those offered by CalSTRS, the world's largest educator pension fund.
Sovereign capital and private markets correlation: SVF demonstrated that even sovereign wealth capital can experience significant drawdowns in venture equity. The PIF's $45 billion commitment, while part of a diversified portfolio including SOFAZ (Azerbaijan's State Oil Fund) model sovereign allocations, still resulted in substantial unrealized losses during the 2022–2023 correction. Long-term allocators should assume venture equity correlates with broad equity markets during risk-off cycles rather than treating it as a diversifying asset class.
Manager concentration and operational oversight: SVF's reliance on Masayoshi Son's personal conviction and SoftBank's operational team created both upside (rapid capital deployment, cross-portfolio synergies) and downside (centralized decision-making, limited external governance friction). Allocators like GIC (Singapore's Sovereign Wealth Fund), which maintains deep internal venture investment teams, can exercise more granular oversight than LPs in manager-led structures. Assessing whether SVF's historical LP governance committees provided meaningful friction on investment decisions is critical to evaluating future capital commitments.
Venture capital fee models under pressure: SoftBank's 1.5% management fee on $98.6 billion generates approximately $1.5 billion in annual fees, a scale unprecedented in venture capital. Net returns of -5% (Vision Fund 1 through 2024) mean that LPs paid approximately $1.5 billion annually for negative returns. This fee drag raises questions about whether future megafund ventures can justify equivalent fee structures or whether institutional pressure will push management companies toward performance-based fee models more aligned with traditional asset management.
Alternative institutional venture models: Some allocators have shifted toward hybrid models combining direct investment, secondaries, and diversified fund commitments. Mumtalakat (Bahrain's Sovereign Wealth Fund) and similar entities have built internal venture teams rather than relying solely on manager-led vehicles. This approach requires significant internal capability but offers better alignment with specific return targets and portfolio construction goals.
Conclusion: Implications for long-term allocators
The SoftBank Vision Fund represents a significant experiment in scaling institutional venture capital. With $98.6 billion under management and leading institutional and sovereign investors, SVF demonstrated both the promise (access to large technology rounds, meaningful portfolio influence) and perils (concentration risk, valuation sensitivity, management discretion) of megafund venture investing.