The WeWork S-1: How Six Weeks of Disclosure Erased $37 Billion

On 14 August 2019, The We Company filed its S-1 with the SEC, carrying an implied valuation of roughly $47 billion — the mark set by SoftBank in a private round that January, and among the highest ever assigned to a US venture-backed company.

On 30 September 2019, the company withdrew the filing. In the six weeks between, the marketed valuation fell below $10 billion, the chief executive was removed, and the institutional buyer base declined to take the deal at any price near the private mark.

Here is the part worth sitting with: the business did not deteriorate during those six weeks. Revenue did not fall. No customer left. No contract was cancelled. The only thing that changed was that the numbers became public and a class of buyer with no prior position, no relationship, and a professional obligation to be sceptical read them for the first time.

That makes this the cleanest natural experiment the market has produced on the difference between a private mark and a price.

1. The Timeline, Compressed

• January 2019 — SoftBank leads a round implying roughly $47 billion.
• 14 August 2019 — the S-1 becomes public. Financials and governance disclosed for the first time.
• Late August — press and analyst coverage turns hostile. The gap between the growth narrative and the unit economics is dissected publicly.
• Early September — the marketed range is cut repeatedly. Figures in the high teens circulate, then low teens, then below $10 billion.
• 24 September — Adam Neumann resigns as chief executive under board pressure.
• 30 September — the S-1 is withdrawn.
• October — SoftBank provides a rescue package reported at roughly $9.5 billion, at a valuation around $8 billion.

Six weeks, from $47 billion to $8 billion, on no new operating information.

2. What the S-1 Actually Revealed

The financials were not hidden. They were simply seen.

For 2018 the company reported revenue of roughly $1.8 billion against a net loss of roughly $1.6 billion — close to a dollar of loss for every dollar of revenue, with losses growing broadly in line with growth rather than tapering. That is not a business approaching scale economics. It is a business whose cost structure moves with its revenue.

The genuinely decisive number was the balance sheet commitment: approximately $47 billion in future lease obligations against roughly $4 billion in committed future revenue.

That single comparison is the whole thesis. The company had signed long-dated, non-cancellable obligations to landlords, and sold short-dated, cancellable memberships to tenants. In a strong market that spread is a business. In a downturn, the obligations remain and the revenue leaves.

This is a duration mismatch, and it is the same structural risk that has periodically destroyed banks and insurers. Presented as a technology company's growth story, it was novel enough to survive private diligence. Presented in a registration statement to public credit and equity analysts, it was recognised immediately — because those people spend their careers looking for exactly this.

3. "Community Adjusted EBITDA"

The filing introduced a metric that excluded not only interest, tax, depreciation and amortisation, but also basic operating costs of running the locations — marketing, administration, and development expenses among them.

It became a punchline within days, and deservedly. But the analytical point is more useful than the mockery.

A bespoke metric is a disclosure in itself. When a company defines a measure that exists nowhere else, it is telling you which costs it wants excluded from your assessment. Sophisticated readers do not argue with the metric; they simply reconstruct the standard one and note the size of the gap. The larger the adjustment required to reach profitability, the clearer the signal about what management believes it cannot defend.

Adjusted metrics are legitimate and common — the question is always whether the adjustments remove genuinely non-recurring items or recurring costs of doing business. Excluding the cost of operating the buildings, in a building-operating company, answered that question.

4. The Governance Findings

Institutions can price a loss-making business. What they discount permanently is a structure that removes their ability to do anything about it.

The S-1 disclosed, among other things:

• Super-voting stock giving the founder multiple votes per share — initially twenty, reduced to ten in a late concession — placing control beyond the reach of any public shareholder vote.
• Related-party property transactions, in which the founder held interests in buildings leased by the company — putting him on both sides of the rent.
• A payment of roughly $5.9 million to an entity he controlled, for the rights to the "We" trademark. It was returned after the disclosure drew attention, which underlined rather than resolved the problem.
• A succession provision under which, in defined circumstances, the founder's wife would be among those selecting a successor chief executive.
• Reported personal liquidity of more than $700 million extracted ahead of the listing through share sales and borrowings against stock.

Any one of these would be an argument. Together they described a company where the founder's interests and the shareholders' were structurally separable, and where no mechanism existed to correct it.

The market's response was not to demand a lower price. It was to decline. That distinction matters: governance defects of this kind are not priced as a discount to be negotiated, they are priced as a reason not to participate — which is why the questions in our guide to boards and governance belong on the table years before a filing, not in the weeks before one.

5. Why the Private Mark Was Never a Price

This is the most transferable lesson in the episode, and it is still routinely misunderstood.

A late-stage private round is not a valuation of the company. It is one buyer's price for a specific, structured instrument. That instrument typically carries a liquidation preference, anti-dilution protection, information rights, board representation and often a ratchet tied to a future listing price.

Strip those protections away and you have common stock — which is precisely what an IPO sells.

So the headline number attached to a late private round systematically overstates the value of the common. An investor with downside protection can rationally accept a high nominal price, because the structure limits what they lose if they are wrong. A public buyer has no such protection and must therefore underwrite the whole risk at the headline.

Two further distortions compound it:

• A single marginal buyer sets the mark. One investor willing to pay $47 billion establishes a $47 billion valuation for reporting purposes, even if the second-highest bid was half that. Public markets require thousands of buyers to clear a large float — which means the price is set closer to the median view than the maximum one.
• Private diligence is bounded by relationship. A lead investor negotiating a competitive round has limited time, limited access and a commercial incentive to close. A public analyst has the complete filing, no relationship to protect, and a professional reputation that improves when they are early to a problem.

The same structural gap is live in the current cohort of very large private companies, and it is why our analysis of the AI listing pipeline focuses on what those companies will have to disclose rather than on what they are currently marked at.

6. Was the Market Right?

It is worth testing the verdict rather than assuming it.

The company eventually went public in October 2021 through a SPAC merger — a route that requires far less scrutiny than a traditional registered offering, and one covered in our comparison of listing structures. Shares rose more than 13% on debut at a valuation a fraction of the 2019 mark.

It filed for Chapter 11 bankruptcy protection in November 2023.

The lease obligations that public analysts identified in August 2019 as the central risk were, in the end, the thing that took the company down — after a pandemic did to office demand roughly what the sceptics had said a downturn would do. The market's reading was not merely fashionable scepticism. It was correct, on the specific mechanism it named, four years early.

7. The Lessons That Transfer

• Disclosure is a valuation event. If your economics only survive limited scrutiny, you do not have a private-versus-public problem. You have a business problem with a delayed reveal.
• Model the mismatch, not the growth rate. Whenever obligations are long and cancellable revenue is short, that spread is the business — and it is the first thing a credit analyst will find.
• Custom metrics invite reconstruction. Define one and sophisticated readers will immediately rebuild the standard figure and measure the gap. The gap becomes the story.
• Governance is priced as a refusal, not a discount. Control structures that leave shareholders no remedy remove buyers from the book entirely.
• Founder liquidity ahead of a listing is read as a signal. Structured secondaries and tender offers are legitimate and increasingly normal — large, informal, pre-IPO extraction is not the same thing.
• Your last round is not your value. Founders who anchor on the headline of a structured round are anchoring on a number that includes protections public buyers will never receive.

Frequently Asked Questions

Could better bankers have got this deal done?

At a lower price, possibly. At anything close to $47 billion, no. Underwriters can shape a narrative, sequence a roadshow and build a book — they cannot conceal a balance sheet in a registered filing, and they cannot manufacture buyers for a governance structure institutions have decided against. Our guide to what investment banks actually do is useful on where their influence ends.

Was the problem the business model or the execution?

Both, but they are separable. Flexible workspace is a real and durable business, operated profitably by others — usually with management agreements or revenue shares rather than long fixed leases, which is precisely the mismatch WeWork took onto its own balance sheet. The model was viable; the specific structure was fragile, and the governance made it unfixable from outside.

What should a founder take from this before filing?

Read your own S-1 as an adversary would, twelve months early. Every related-party transaction, every control provision, every non-standard metric will be read by people whose job is finding exactly those things. If any of them cannot be defended in a sentence, fix it while you still have the option — the preparation discipline in our guide to investor due diligence is the same exercise at smaller scale.

Does a withdrawn IPO end a company?

No. Withdrawals happen for many reasons, including market conditions that have nothing to do with the issuer, and companies routinely return later. What is genuinely damaging is a withdrawal driven by disclosure, because the disclosure does not go away — the filing remains public, and the next attempt begins from the sceptic's version of the story.

How is this different from a company that simply prices low?

Pricing low is a negotiation about value. What happened here was a refusal to transact on governance and structure. A company can recover quickly from the first — many strong businesses have listed below their last private round and performed well since. Recovering from the second requires changing the company, not the price.

Is the private-to-public valuation gap still a live risk?

Very much so, and arguably larger now. Companies stay private far longer, raise far more, and accumulate more structural protections in later rounds — all of which widens the distance between the headline mark and what common stock is worth. The current generation of large listings is testing exactly this, as our coverage of the SpaceX offering and the 2012 Facebook listing both illustrate from different directions.

The Bottom Line

WeWork did not fail because the market turned or the story got stale. It failed because a document was published, and the people who read it were not the people who had set the price.

A private mark is one buyer's price for a protected instrument. A public price is thousands of buyers pricing unprotected common stock with the full filing in front of them. When those two numbers are far apart, the filing is what closes the gap — and it closes it in one direction.

Global Capital Network connects founders, investors and the bankers, counsel and advisors who prepare companies for this scrutiny. See upcoming events or get in touch.

Accurate as at 2 August 2026. Figures are as reported contemporaneously. Primary sources: CNBC on the withdrawal, CNBC on the 2021 SPAC debut, and the S-1 and subsequent filings via SEC EDGAR. This article is general information and market history, not investment advice.