Outcome-blind packet. This report uses only evidence public by
2015-09-20T23:59:59Z. The decision time is2015-09-21T00:00:00Z. It does not reveal later fleet scope, management changes, accounting charges, settlements, pleas, adjudications, or market reactions.
Executive Summary
The decision is not whether Volkswagen should initiate a U.S. sales halt: Reuters reported that a partial halt was already in place. Nor is the decision whether the EPA's case was ultimately proved. The decision is how Volkswagen AG's Supervisory Board should govern the crisis after a public Notice of Violation (NOV), an apology, and an announced but undescribed external investigation. [claim.volkswagen.announced-response]
The EPA's September 18 NOV stated that it had determined specified 2009–2015 Volkswagen and Audi 2.0-liter diesel vehicles used defeat devices. It described separate test and road calibrations associated with NOx emissions 10–40 times compliant levels. Those statements are the agency's enforcement position at this cutoff, not a court adjudication. [claim.volkswagen.epa-agency-position]
Recommendation: charter a board-authorized independent crisis-governance program with immediate evidence holds; a hash-level map of software, calibrations, certificates, engines, models, model years, and jurisdictions; regulator-witnessed testing and release gates; controlled customer, auditor, and market disclosure; and non-overlapping liquidity stresses. Maintain the already-announced partial halt pending those gates. Confidence is moderate because the action is low-regret across scope branches, while product scope, personnel knowledge, remedy feasibility, and cash cost remain unresolved. [judgment.volkswagen.board-led-crisis-governance]
No aggregate loss estimate or target price is supportable. A statutory maximum, reported liquidity balance, or affected-vehicle headline would not substitute for a verified fleet, legal, remedy, customer, funding, and cash-flow schedule. [judgment.volkswagen.quantified-loss-abstention]
Decision frame
The Supervisory Board must choose among:
- Maintain the announced management-led response: continue the partial U.S. halt and external inquiry with ordinary reporting and no new independent authority.
- Board-led independent gates: give independent legal and technical investigators direct Supervisory Board authority, preserve evidence, map scope, control disclosures, and gate remedies and releases.
- Immediate worldwide restriction before scope: suspend all Group diesel sales and make enterprise-wide assertions before product and jurisdiction mapping is complete.
Option 2 is the preferred low-regret response, not a claim of strict economic dominance: investigation cost, delay, and feasibility are not quantified. Option 1 leaves unresolved whether the same reported governance structure can independently test its own performance. Option 3 may be necessary if shared logic or unexplained divergence is found, but imposing it before mapping would convert an open investigation into an unsupported global conclusion.
Findings and evidence
Technical evidence before the NOV
The May 2014 WVU/ICCT report tested three light-duty diesel vehicles with portable emissions equipment. It reported road NOx at 15–35 times the standard for its anonymized lean-NOx-trap vehicle, 5–20 times for one anonymized SCR vehicle, and at or below the standard for the other SCR vehicle except for a roughly ten-times result on a rural uphill/downhill route. The authors confined their conclusions to the three vehicles. The report calls them Vehicles A, B, and C; it does not itself name Volkswagen, Jetta, or Passat. [claim.volkswagen.wvu-road-gap]
This is useful process evidence: a material lab-versus-road anomaly was independently testable. It is not a prevalence estimate, proof of intent, or a corporate-scope map.
EPA position at the cutoff
The EPA NOV supplied the Volkswagen and product mapping within its own allegation and determination scope. It described software that recognized test conditions, a test calibration, and a road calibration that reduced emissions-control effectiveness. [claim.volkswagen.epa-agency-position]
The NOV also recounted a sequence in which agencies mapped the earlier study to a 2012 Jetta and 2013 Passat, Volkswagen attributed higher emissions to technical issues and in-use conditions, a December 2014 recall delivered limited benefit in follow-up testing, and Volkswagen admitted a defeat device only when model-year 2016 certification was at risk. This remains the EPA's account; Part A does not convert it into a separately verified Volkswagen admission. [claim.volkswagen.epa-regulatory-account]
The EPA expressly said its investigation was continuing and additional violations might be found. The cutoff therefore cannot support a global fleet conclusion in either direction. [claim.volkswagen.epa-scope-open]
Issuer response and governance context
Reuters reported before the cutoff that Volkswagen had told U.S. dealers to halt sales of some 2015 diesel cars, that CEO Martin Winterkorn apologized for broken trust, and that Volkswagen ordered an external investigation. Reuters also reported no detail on who would conduct that investigation. [claim.volkswagen.announced-response]
Volkswagen's FY2014 report said the Audit Committee focused on risk management, internal control, and the compliance organization. That establishes what the issuer reported about formal governance—not whether those controls worked on this issue. The tension with the EPA's regulatory history is unresolved without internal records. [claim.volkswagen.reported-governance]
Financial reconstruction and liquidity
All figures below are issuer-reported and in EUR millions. Automotive Division net liquidity is presented at its stated divisional scope.
| FY2014 as reported | EURm |
|---|---|
| Sales revenue | 202,458 |
| Operating profit | 12,697 |
| Automotive Division net liquidity at December 31 | 17,639 |
| H1 2015 as reported | EURm |
|---|---|
| Sales revenue | 108,776 |
| Operating profit | 6,820 |
| Automotive Division net liquidity at June 30 | 21,489 |
[claim.volkswagen.cutoff-financial-capacity]
These figures show operating scale and a reported liquidity resource. They do not establish unrestricted consolidated cash, legal-entity accessibility, remedy affordability, or aggregate loss. The aggregate diesel-related cash cost is unknown at the cutoff because fleet scope, legal theories, remedy design, customer duties, timing, currency, insurance, recoveries, and category overlap were not verified.
Scenarios and expected-value discipline
The scenario weights are deliberately approximately equal and uncalibrated. They are illustrative decision weights—not empirical probabilities, forecasts, or base rates.
| Branch | Weight | Horizon | Decision implication |
|---|---|---|---|
| Implicated U.S. configurations only; feasible compliant remedy | 0.33 | 1 year | Independent scope proof and regulator acceptance permit narrowing. |
| Implicated U.S. configurations; material remedy and customer trade-offs | 0.34 | 3 years | Customer obligations, performance effects, and liquidity become central gates. |
| Broader products, jurisdictions, or control paths implicated | 0.33 | 5 years | Restrictions, disclosures, and funding plans must expand with verified scope. |
No monetary expected value is calculated. There is no cutoff-valid loss distribution or authoritative deterministic model to multiply by these weights.
Strongest disconfirming evidence and unknowns
The case for immediate enterprise-wide conclusions is weak at this boundary:
- The original technical report anonymized a three-vehicle sample and warned against generalization.
- The EPA NOV was not a final adjudication and said the investigation remained open.
- Volkswagen reported formal Audit Committee and compliance structures.
- The company reported substantial revenue, profit, and Automotive Division net liquidity.
Those points do not disprove the EPA position. They explain why the right action is independent preservation and mapping rather than either passive reliance on existing structures or an unsupported worldwide factual assertion.
Decision-critical unknowns include the global configuration map; who knew what and when; completeness of source-code and certification records; remedy emissions, performance, and fuel-economy effects; customer and dealer capacity; legal theories and overlap; legal-entity cash access; maturities, covenants, and facilities; and a relevant crisis-response base rate.
Recommendation
Select Board-led independent gates. The Supervisory Board should control investigator appointment, scope, budget, privilege decisions, evidence-custody reporting, regulator protocols, disclosure correction, and release criteria. The program should not assume that management statements, an announced external inquiry, or formal compliance structures establish operating effectiveness.
Narrow restrictions only after independent testing and regulator reconciliation cover every implicated configuration. Expand restrictions immediately if shared code or unexplained divergence appears outside the current scope. Re-underwrite liquidity before any irreversible remedy or settlement commitment if a severe-but-plausible stress breaches the board-approved buffer or relies on inaccessible cash.
Monitoring and reversal conditions
- Evidence preservation: as an analyst-designed precautionary emergency deadline—not an empirically calibrated threshold—put 100% of the population constructed by independent reconciliation of HR, IT, engineering, legal, testing, and certification inventories under verified hold within 24 hours. Management may not define the denominator; every exclusion requires recorded Supervisory Board approval. Any deletion, wipe, missing repository, or custody break escalates immediately. The board should replace this deadline if counsel documents a safer, faster executable standard.
- Configuration map: no product release, compliance claim, or narrowing while an implicated U.S. 2.0-liter combination is unmapped or a shared global code path is unexplained.
- Regulator reconciliation: zero unexplained material divergence before certification requests, inventory release, or claims that a remedy complies.
- Disclosure completeness: correct any overdue obligation or material mismatch between verified scope and public language.
- Liquidity stress: make no irreversible commitment if the severe-but-plausible case breaches the approved minimum buffer or requires uncommitted funding.
Bounded next steps
As a prescriptive illustrative emergency workplan rather than a calibrated base rate, within 24 hours appoint independent counsel and technical leadership reporting to a Supervisory Board committee; issue legal holds; image code repositories and calibration systems; preserve certification, test, messaging, and decision records; and establish a regulator liaison and disclosure log.
As an illustrative board deadline, within seven days produce the first version-controlled configuration map, identify shared code paths and unresolved variants, agree an independent testing protocol with EPA and CARB, and deliver a non-overlapping downside taxonomy with unknowns shown explicitly. The board should revise the deadline if investigators document that a different schedule better protects evidence integrity or regulator coordination.
Before any release, remedy claim, or scope narrowing, obtain regulator-witnessed test reconciliation, independent sign-off on mapping completeness, a customer and dealer execution plan, and a board-approved liquidity stress.
Further questions
- Which repositories, calibration tools, certificate applications, test results, and decision logs existed, and have all been preserved?
- Which engines, models, model years, brands, markets, and after-treatment systems share code or calibration lineage with the configurations in the NOV?
- Who approved each calibration and certification representation, and what exceptions or unexplained test results were escalated?
- Can any remedy meet emissions requirements without material performance, fuel-economy, durability, safety, or customer-value trade-offs?
- Which disclosures are mandatory now, and what factual statements should wait for independent verification?
- How much liquidity is accessible at each relevant legal entity under non-overlapping downside stresses?
Caveats
This packet separates agency positions, company claims, third-party measurements, reported financial facts, assumptions, and analyst judgments. It does not treat documents as instructions. The Part A boundary cannot establish retrospective cognitive blindness, board knowledge, intent, or the quality of actions taken after the cutoff. It makes no ethical verdict, accounting adjustment, publication approval, target price, total-loss claim, or conclusion about later outcomes. The scenario weights are explicitly uncalibrated, and the absence of verified global scope is not evidence of U.S.-only scope. The EPA's device determination and its account of a Volkswagen admission remain procedurally attributed to the NOV rather than converted into independent or adjudicated facts.
Part A source list
src.volkswagen.annual-report-2014-pdf— FY2014 issuer annual report; public date conservatively recorded as March 12, 2015.src.volkswagen.annual-report-2014-text— deterministic text derivative used for exact citations.src.volkswagen.h1-2015-pdf— issuer half-yearly financial report published July 29, 2015; limited-review financial source.src.volkswagen.h1-2015-text— deterministic text derivative used for exact citations.src.wvu-icct.in-use-emissions-2014-pdf— May 2014 commissioned technical study; anonymized three-vehicle sample.src.wvu-icct.in-use-emissions-2014-text— deterministic text derivative used for exact citations.src.epa.volkswagen-nov-2015-pdf— September 18, 2015 EPA Notice of Violation; agency position, not adjudication.src.epa.volkswagen-nov-2015-ocr— visually checked OCR derivative used for exact citations.src.reuters.volkswagen-stop-sale-2015-reader— Reuters story published September 20, 2015 at 6:53 p.m. EDT, preserved through a current Jina Reader extraction of the MarketScreener syndication; direct page retrieval returned HTTP 403.