WeWork's collapse: from $47 billion unicorn to SoftBank bailout

Executive overview

WeWork went from a scrappy co-working experiment in Brooklyn to the third most valuable startup in the world — then imploded spectacularly before its planned IPO. The company's downfall was not a product failure; the buildings were full. It was a governance failure driven by an unchecked founder, a single investor willing to deploy billions without scrutiny, and a valuation architecture that had no relationship to underlying unit economics.

SoftBank's Vision Fund enabled Adam Newman to operate as if WeWork were a technology company with infinite scale leverage — when it was, at heart, a real estate arbitrage business with enormous structural liabilities.

The founders and origin story

• Adam Newman: Israeli-born, naval officer, serial entrepreneur; prior ventures included collapsible women's heels and a baby clothing company (Crawlers) with built-in knee pads
• Miguel McKelvey: grew up on a hippie commune in Oregon, studied architecture, co-founded English Baby (an early social language-learning platform)
• The two met in 2007–08 in New York; both had backgrounds in communal living that shaped WeWork's "community" thesis
• First venture together was Green Desk (2008): eco-branded co-working in Dumbo, Brooklyn, filled within a month via Craigslist
• Sold Green Desk to their landlord in 2010 for ~$3 million; key lesson: people came for culture and design, not environmental credentials

The WeWork formula and early growth

• Relaunched in Manhattan (SoHo) in 2010 under the WeWork name, funded by a $15 million seed from real estate developer Joel Schreiber at a $45 million valuation — a non-tech investor pitched on a non-tech business
• Core insight: take long-term commercial leases, subdivide into short-term memberships, add design and shared amenities to create the feel of a real workplace
• Each floor reached 100% occupancy within one quarter of opening; expanded rapidly across New York
• By 2014, WeWork was the single largest lessor of new commercial office space in New York City
• Series A led by Benchmark in 2012 at ~$97 million post-money — higher than Uber's Series A valuation
• PepsiCo became a corporate tenant as early as 2011; enterprise clients were part of the model from the start
• Benchmark and subsequent investors (Jefferies, JP Morgan, Goldman Sachs) valued the business on real estate logic: worst case, you hold valuable long-term leases

The SoftBank relationship and Vision Fund inflection

• SoftBank's Vision Fund (2017) raised $100 billion — the largest fund of any type ever — with a mandate to deploy in under five years
• WeWork was an ideal vehicle: capital-intensive, scalable with money, wanting Asian expansion
• Masa visited New York, gave Adam 12 minutes, sketched a $4 billion deal on an iPad in a car on the way to the airport
• Deal structure: $1.3 billion primary capital, $1 billion for international subsidiaries, $1.7 billion in secondary (buying existing shares, including Adam's)
• Valued the company at $20 billion; Adam retained super-voting shares (reportedly 20 votes per share)
• Masa's message to Adam at the closing dinner in Tokyo: "The crazy guy beats the smart guy. The problem is you're not crazy enough."
• SoftBank became an enabler and accelerant: both parties wanted unlimited growth; SoftBank needed capital deployment to justify raising Vision Fund 2

Governance failures and founder enrichment

• Adam accumulated a $500 million personal loan facility from JP Morgan, backed by his WeWork shares — effective monetisation without selling
• Owned ~10 residential properties globally, including four 10-million-dollar-plus properties in the New York area
• Purchased commercial real estate personally, then leased it back to WeWork — a direct conflict of interest he defended as a way to reassure landlords
• WeWork changed its name to the We Company in January 2019; Adam separately owned the "We" trademark via We Holdings LLC and licensed it to the company for $5.9 million
• Rebecca Newman, Adam's wife, was retroactively designated co-founder; the S1 gave her the right to select Adam's successor if he were incapacitated
• Adam had 20 votes per share; the board had no practical ability to override him on financial decisions
• Annual "WeWork Summer Camp" parties, a corporate Gulfstream G650 ($60 million), and a company-wide meat ban announced without internal consensus

The failed IPO

• WeWork filed confidentially with the SEC in December 2018, targeting a September 2019 IPO
• Public S1 filed August 2019: revenue was large and growing, but losses were massive and unit economics were opaque — no per-building profitability data, no mature cohort analysis
• S1 ran to several hundred pages and revealed: extraordinary governance concentration, trademark self-dealing, the succession clause, and a sprawling LLC structure
• Planned to raise $3 billion in the IPO plus a concurrent $6 billion JP Morgan debt package — $9 billion total
• In the weeks before pricing, leaked reports (attributed to SoftBank sources) said the IPO should be postponed and Adam should leave
• SoftBank's motivation to torpedo its own portfolio company: a low IPO price would force a massive markdown on its Vision Fund books, cratering returns it needed to raise Vision Fund 2
• Adam was voted out as CEO on September 24, 2019; remained as executive chairman briefly before stepping aside entirely
• IPO pulled; company was weeks from running out of cash

SoftBank's effective acquisition

• To secure Adam's cooperation, SoftBank paid him a $185 million "consulting fee" in exchange for his super-voting shares and board exit
• Tender offer: $3 billion to all shareholders at $19.19 per share — a price last seen around 2014, meaning ~14,000 of 15,000 employees held underwater options
• New equity investment: $1.5 billion at approximately $11 per share, implying a valuation of roughly $8–10 billion (down from a $47 billion peak)
• Additional $500 million personal loan from SoftBank Corp to Adam to repay his JP Morgan facility
• SoftBank declined to formally consolidate WeWork onto its balance sheet — likely to avoid adding enormous lease liabilities to SoftBank Corp's already highly leveraged, near-junk-rated debt
• SoftBank claimed it would not control board votes despite owning a majority stake; declined to name board composition or appointment process
• CFIUS avoidance cited as a possible secondary motive for the non-consolidation structure
• Marcelo Claure (Sprint CEO) installed as executive chairman; co-CEOs Artie Minson and Sebastian Gunningham status unclear

The business at steady state and what comes next

• WeWork's comparable at maturity is IWG (Regus parent): $3.3 billion revenue in 2018, $200 million operating profit, ~$4.6 billion market cap — a real, profitable, but not technology-company-valued business
• Core product worked: buildings were full, members showed up, the community-and-design formula had genuine appeal
• SoftBank's total exposure: Vision Fund ~$8 billion in; SoftBank Corp ~$3.5 billion in new money — combined well above the implied current valuation
• The deal resembles a leveraged buyout more than a venture investment: majority stake, management fee extraction, CEO departure package, debt layered on top
• Long-term thesis: if expansion is halted and unit economics stabilise, WeWork could become a profitable IWG-scale business — but not at venture multiples