One Hundred Billion Dollars and What It Did to the Market

When SoftBank's Vision Fund held its final close in May 2017 at $98.6 billion, it was not simply the largest venture fund ever raised. According to PitchBook data from that period, it was roughly twice the size of all other active venture funds combined — a single pool of capital larger than the entire asset class it was entering. Masayoshi Son, SoftBank's founder and chief executive, had spent less than a year assembling the commitments after an early conversation with Saudi Arabia's Crown Prince Mohammed bin Salman in 2016. The speed and scale of that fundraise had no precedent in the history of institutional venture.

The LP roster reflected Son's conviction that sovereign wealth, not Sand Hill Road partnerships, would define the next era of technology investment. The Public Investment Fund of Saudi Arabia committed $45 billion. The Abu Dhabi state fund Mubadala committed $15 billion. SoftBank itself put in $28 billion, a figure that required the Japanese conglomerate to lever its own balance sheet. Apple, Qualcomm, Foxconn, and Sharp contributed smaller tranches, each bringing strategic as well as financial motivation. The structure was unusual in another respect: roughly $40 billion of the capital came in as preferred equity carrying a fixed coupon of 7 percent annually — debt-like obligations embedded inside what was nominally a venture vehicle, creating repayment pressure that pure equity funds do not carry.

The Vision Fund did not deploy capital the way conventional venture funds did. Where a top-tier Series A firm might write a $20 million cheque and consider itself a meaningful investor, the Vision Fund's target ticket was $100 million at minimum. Son had articulated a thesis — sometimes called "singularity investing" — that the most important technology companies of the next thirty years would be structurally dominant and that the correct strategy was to identify them early and fund them so heavily that competition became impossible. In practice this meant arriving at companies that had already established some traction and writing amounts that allowed, or required, aggressive expansion.

The downstream effect was immediate and measurable. Late-stage round sizes across the industry reset upward. Crunchbase, PitchBook, and CB Insights all recorded sharp increases in median and mean round sizes for growth-stage deals between 2017 and 2019, with the Vision Fund's activity a primary driver. Founders who had been told a $30 million Series B was a good outcome found themselves fielding term sheets from SoftBank for $300 million or more. Some declined. Many accepted, because the alternative — competing against a well-capitalized Vision Fund portfolio company — seemed worse.

The valuation effect was equally direct. When a fund writes a $300 million cheque into a company, it does so at a negotiated post-money valuation, and that valuation becomes the headline number that subsequent investors, employees, and journalists use to assess the company's worth. Vision Fund investments arrived at valuations that often exceeded what conventional discounted-cash-flow or comparable-company analysis would have supported. WeWork received Vision Fund investment at a valuation that eventually reached $47 billion before its S-1 filing in 2019 exposed the underlying economics; the S-1 disclosure triggered a collapse in investor confidence that killed the IPO. Uber, Grab, Didi, OYO, Katerra, Greensill, and Zume Pizza each received Vision Fund capital; several would collapse, retrench, or write down to fractions of their funded valuations within four years.

The broader industry absorbed the pricing signals even when SoftBank was not in the room. If the Vision Fund was willing to value a food-delivery company at $10 billion, rival funds felt pressure to mark their own food-delivery holdings at comparable levels or risk appearing to underperform. Benchmark, Sequoia, and others raised their own growth funds in part to compete at sizes that would have been unthinkable before 2017. The term unicorn — coined by Aileen Lee in November 2013 to describe the then-rare privately held company valued at $1 billion or more — became inadequate; "decacorn" entered usage for $10 billion companies, and the Vision Fund manufactured them at speed.

The fund's fiscal year 2019 results, reported by SoftBank, recorded a loss of approximately $17.7 billion on Vision Fund investments — the largest annual loss in Japanese corporate history at that point. WeWork's failed IPO accounted for a substantial portion, but write-downs on Katerra, Greensill Capital, and OYO contributed materially. The collapse of Silicon Valley Bank in March 2023, which held deposits for a large share of the Vision Fund's portfolio companies, added further strain to an ecosystem that the fund itself had helped inflate.

Masayoshi Son announced a second fund, Vision Fund 2, in 2019. It raised substantially less than its predecessor — approximately $56 billion, and largely from SoftBank itself after the Vision Fund 1 results discouraged external LPs. The sovereign wealth funds that had anchored the first vehicle did not return at comparable scale. Saudi Arabia's Public Investment Fund shifted its attention toward direct co-investments and its own domestic technology initiatives.

The structural lesson the Vision Fund demonstrated was one the venture industry had known theoretically: capital is not neutral. When a single fund writes cheques large enough to distort competitive dynamics, it also distorts the signals that founders, co-investors, and employees use to assess whether a business is working. A company that raises $500 million can remain solvent long past the point at which organic demand would have forced a reckoning. This is sometimes called "blitzscaling" — the deliberate prioritisation of speed over efficiency — but the Vision Fund pushed the strategy into territory where Reid Hoffman's 2018 formulation did not quite reach: not a calculated risk on winner-take-all markets, but a bet large enough to create, and then hide, the conditions for catastrophic failure.

The fund did produce genuine returns in some holdings. Its early position in Coupang, the South Korean e-commerce company, generated significant gains when Coupang listed on the New York Stock Exchange in March 2021. DoorDash and Guardant Health also returned capital. But the aggregate picture, documented in SoftBank's own filings with Japanese regulators, showed that the headline number — $98.6 billion deployed with ambition — produced returns far below what the venture industry's historical J-curve would have suggested, and at a cost to market discipline that the industry spent the correction years of 2022 and 2023 slowly absorbing.