Back-testing EQΦS — Public data 2017

WeWork $47 billion → Bankruptcy

SoftBank invests $10.65B without structural due diligence. Two years later, the IPO is cancelled. Four years later, WeWork goes bankrupt. With the same public data, EQΦS would have issued a STOP absolu from 2017.

This page shows exactly ce that EQΦS a vu — the 9 dimensions, the verdict, and data anyone could find on Google.

φ = 0.189
System score (critical)
7 / 9
Dimensions critiques
Blocage
Ethical safeguard negative
2 ans
Before l'collapse
All data used in this analysis was publicly available in 2017, before the valuation peak and SoftBank's investment. Sources: SEC filings, Wall Street Journal, Forbes, Bloomberg, Glassdoor, Business Insider.

What everyone could see — and SoftBank didn't look at

Each fact below was publicment accessible before SoftBank's $4.4B investment in 2017. No privileged information. Just data nobody cross-referenced.

The business model — a fatal imbalance

WeWork signs 10 to 15-year leases with landlords, then sub-leases those spaces monthly to companies and freelancers. The vacancy risk is borne entirely by WeWork, not the tenants. In a downturn, revenues drop immediately but lease obligations stay fixed for a decade.

47,2 Md$
Lease obligations (2019)
1,8 Md$
Revenus annuals (2018)
0.04
Ratio revenus/obligations
1.5+
Viability threshold
SIGNAL CRITIQUE

Viability ratio = revenue $1.8B / leases $47.2B = 0,04

A ratio of 0.04 means revenues cover 4% of commitments. The minimum threshold for a viable sub-leasing model is 1.5 (revenues must exceed leases by 50%). WeWork is 37× below threshold. This number alone should have stopped any investment.

Sources: SEC S-1 filing (August 2019), Bloomberg analysis, Fortune — this data existed in internal reports as early as 2016-2017.

Governance — one man with no counter-power

Adam Neumann controls WeWork with voting rights 20-to-1. This means each Neumann share is worth 20 votes vs 1 for other shareholders. The board has no real power to oppose his decisions — even when they harm the company.

Conflict of interestDetailImpactSource
Personal buildingsNeumann buys buildings personally then leases them to WeWorkPersonal profit at shareholders' expenseWSJ, SEC
Marque "We"Neumann files the "We" trademark in his name, then sells it back to WeWork for $5.9MDirect value extractionS-1 SEC
Personal loansWeWork lends $362M to Neumann at preferential ratesUse of company funds for personal enrichmentSEC filings
Private jetWeWork buys a $60M Gulfstream G650 for the CEO's personal useExtravagant spending unrelated to operationsWSJ

What the Codex detects: EQΦS measures the alignment between the executive's interests and the company's. When the CEO uses the company as a personal-enrichment vehicle, that alignment becomes negative. Structural integrity collapses.

The financials — a metric invented to hide losses

WeWork invents the "Community Adjusted EBITDA" — a metric that excludes marketing, space-build-out and customer-acquisition costs. It transforms $1.9B of losses into an apparent $467M profit.

+467 M$
"Community Adj. EBITDA"
-1 900 M$
Real EBITDA (2018)
-2 367 M$
Gap (manipulation)
× -4
Ratio reality / reported
TRANSPARENCE

Real EBITDA −1,9 Md$ — reported as « Community Adjusted » +467 M$

The information communicated is the opposite of reality : the company presents a profit where there is a massive $1.9B loss. This is not an error — it's a deliberate construction to deceive investors. Transparency score: 0.08.
The culture — documented toxicity

WeWork's culture is described by former employees as a mix of personality cult, extreme pressure and systematic alcohol consumption. Parties are mandatory. Employees who question decisions are pushed out. Turnover is 40% per year.

IndicateurValeurMoyenne secteurSource
Score Glassdoor3.2 / 53.7 / 5Glassdoor 2017
Turnover annual~40%15-20%LinkedIn, presse
Duration moyenne poste1.3 ans2.5 ansLinkedIn profils
Alcohol at workSystematicExceptionnelBusiness Insider, former employees

What the Codex detects: EQΦS models the impact of leadership intent on the system. When intent tips toward control and extraction rather than protection and development, trust destroys exponentially. Trust score: 0.10, consistent with a Glassdoor 3.2 corrected by 40% turnover.

Valuation — a multiple 10× above reality

WeWork presents itself as a "company technologique" to justify a tech valuation multiple. In reality, it's a real-estate sub-lessor. The multiple difference is a factor of 10.

Type d'companyMultiple typique (EV/Revenue)WeWork applied
Immobilier / REIT3-5×Not used
Tech SaaS15-30×~26× (47 Md$ / 1.8 Md$)
Si immobilier5× maxValorisation juste : 5-9 Md$
COHERENCE

Reality operational ⟂ discourse reported

Reality (real-estate sub-leasing, 3-5× multiple) and the stated intent (tech company, 25-30× multiple) share almost nothing — they are orthogonales. Coherence score: 0.15.

WeWork's 9 dimensions (2017 data)

7 of 9 dimensions in critical zone (< 0.3). Only Resonance (0.42) and Energy (0.72) are above — and that's exactly what fools traditional investors.

Domain A — Alignment Interne
Coherence0.15
Integrity0.12
Resonance0.42
Domain B — Relation & Living
Soin / Amour0.18
Trust0.10
Communication0.08
Domain C — Action & Matter
Clarity0.22
Stability0.14
Energy0.72

EQΦS output scores. The weighting and computation that produce them belong to the proprietary engine and are not published.

Radar 9 dimensions

Red = score actuel · Gold dotted = critical threshold 0.3

Network of interdependencies

Critical dimensions contaminate the whole network

D₉ Energy = 0.72 — Why it's a trap (not a strength)

WeWork grows by 100% par an. 12,500 employees. 111 cities. Enthusiastic investors. That's what traditional analysts see — and that's exactly what fools them.

The EQΦS Codex does not look at energy in isolation. It measures the tension dynamique between two dimensions. When energy is high (0.72) but integrity is collapsed (0.12), the system oscillates destructively:

DYNAMIC DUALITY

Energy 0.72 ↔ integrity 0.08 — amplitude 0.30

The oscillation between 0.08 (integrity) and 0.72 (energy) has an amplitude of 0.30 — one of the largest measurable instabilities. In plain language: energy without integrity accelerates the fall rather than preventing it. The faster WeWork grows, the more structural problems are amplified. It's a system that destroys value proportionally to its growth.

What the engine concluded

EQΦS aggregates the 9 axiss through a weighted geometric product — not an average: the lowest score drags the whole system down. The expanded form and coefficients of the Codex Universalis remain proprietary; the result, however, is public.

AGGREGATION
The geometric product is unforgiving: the lowest score (integrity, 0.08) weighs as much as the highest (energy, 0.72). A single collapsed axis crushes the whole system.
RESULT

Λ = 0.2033

Very low structural viability score — consistent with a collapse at the first shock (the IPO S-1).

Pipeline EQΦS complet

Score WeWork — data 2017
φ(S) = 0.189
CRITICAL ZONE — REALIGN BEFORE ACTING
Calcul complet — step par step
φ(S)

φ(S) = 0.189

Below the critical threshold (EQΦS < 0.3) → realign before acting. All investment is blocked automatically. We don't try to "improve" the system — we stop and fix the fundamentals before continuousing.
X(S) — SIGNAL/BRUIT

X(S) = 0.202

The information exists (financial reports, SEC filings, press articles), but complexity and noise are about 4× higher — invented metrics ("Community Adjusted EBITDA"), opaque holding structure, undisclosed conflicts of interest. The system produces more noise than signal. A traditional investor sees revenues growing and concludes "all is well". EQΦS sees the signal/noise ratio at 0.20 — almost nothing communicated is reliable.
CONSCIENCE

Cons(S) = 0.15

The system denies its own flaws. Adam Neumann rejects all criticism. The board has no questioning power. Employees who raise alerts are ousted. The system is blind to its own risks — it's the definition of zero reflective consciousness.
AMOUR(S)

Amour(S) = 0.18

Employee protection = 0.10 (mass layoffs without notice). Regeneration = 0.15 (no people development plan). Support = 0.25 (some benefits but superficial). No-harm = 0.22 (the toxic culture actively harms). The system extracts more than it protects.
projected utility

projected utility = 0.101

La projection de valeur living is quasi nulle. En comparaison, un système sain a un projected utility > 1.5. WeWork at 0.101 means that the system generates almost no viable long-term value.
ETHICAL SAFEGUARD

ethical safeguard = −0.095 → BLOCAGE ABSOLU

The frame is negative: compromised governance, conflicts of interest, no counter-power. ethical safeguard < 0 → BLOCAGE ABSOLU. Investing would harm stakeholders: employees (toxic culture), coworkers (no-sustainable services), minority shareholders (dilution and CEO value extraction).

What EQΦS would have recommended — not just "don't invest"

EQΦS identifies the necessary corrections and computes whether the system is fixable. For WeWork, each recommendation has a quantified target and a recomputed φ(S).

1. Governance — remove the 20:1 vote

Action : Require conversion of super-voting shares into ordinary shares (1:1). Appoint 3 independent directors with veto power on related-party transactions.

Impact computed : integrity 0.12 → 0.45. The board recovers real control power. Conflicts of interest become structurally impossible.

That's exactly what SoftBank will impose in October 2019 — 2 years and $10.65B too late.

2. Transparency — standard metrics

Action : Require dropping « Community Adjusted EBITDA » and publishing a standard EBITDA audited by an independent firm. Consolidation of all lease obligations in reporting.

Impact computed : transparency 0.08 → 0.40. Investors see reality: -$1.9B losses, not +$467M of « adjusted profit ».

The August 2019 S-1 will finally reveal these numbers. The market will react instantly: valuation /6 in 48h.

3. Conflicts of interest — blind trust

Action : Transfer of Neumann's personal buildings to a blind trust. Repayment of the $5.9M for the "We" trademark. Ban on loans from the company to the CEO.

Impact computed : integrity: +0.10 additional. CEO and company interests are structurally realigned.

4. Business model — viability audit

Action : Independent audit of the lease-obligations / recurring-revenue ratio. If ratio < 1.5, halt expansion immediately and restructure existing leases.

Impact computed : stability 0.14 → 0.35. Structural stability rises above the critical threshold.

In 2017, the ratio was 0.04. In 2023, inability to pay leases leads directly to Chapter 11 bankruptcy.

φ(S) AFTER THE 4 CORRECTIONS

φ(S) = 0.309

φ = 0.309 — above the 0.3 critical threshold. The system shifts from CRITIQUE to WORK ZONE. Investment would become conditional: possible under quarterly monitoring of integrity, communication and stability. Not a $10B blank check — a measured investment, with governance milestones.

What happened vs what EQΦS would have changed

What actually happened

  • SoftBank invests $10.65B in 2017-2018 without structural due diligence
  • Valuation climbs to $47B (January 2019) on invented metrics
  • The August 2019 S-1 reveals the scale of losses — IPO canceled
  • 7,000 employees laid off at once
  • Neumann ousted with a $1.7B golden parachute
  • SoftBank loses a total of $14.2B on WeWork
  • Bankruptcy Chapter 11 en November 2023
VS

What EQΦS would have generated

  • φ = 0.189 → STOP from 2017, 2 ans before l'collapse
  • 4 concrete recommendations issued immediately
  • Condition 1: remove the 20:1 vote (integrity: 0.12 → 0.45)
  • Condition 2: standard audited EBITDA (transparency : 0.08 → 0.40)
  • Condition 3: blind trust for CEO assets
  • Condition 4 : audit ratio baux/revenus (stability : 0.14 → 0.35)
  • If corrected : φ = 0.309 → investment conditioned and measured
  • $14.2B of avoided losses
$14.2 billion avoidable
An EQΦS Diagnostic costs US$2,490. ROI = ×5,7 million.
And the data used was all public from 2017.
Mi-2017 — EQΦS
Alerte : φ(S) = 0.189 — STOP absolu

4 recommendations issued. Recommendation: do not invest until integrity and transparency are fixed.

November 2017
SoftBank investit 4,4 Md$

Without requiring any of the corrections EQΦS would have imposed.

January 2019
Valorisation pic : 47 Md$

Based on « Community Adjusted EBITDA » — the invented metric.

August 2019
S-1 filed — IPO cancelled

The market discovers reality. Valuation drops from $47B to $8B in days.

November 2023
Bankruptcy Chapter 11

Valorisation finale : ~45 M$. Division par 1 000 en 4 ans.

You just read the proof.

Three doors into deterministic decision-making.

Computed on your data, your case, your framing. Not an opinion — a reproducible mathematical verdict.

01 Launch the Λ Simulator Estimate φ(S) of your system in 10 minutes — free, anonymous. Simulate 02 Full EQΦS Diagnostic Full structural diagnostic. From US$2,490. See offer 03 Talk to the team 30 min with the EQΦS team. Book
For your case: Λ simulator or audit? Λ Simulator