Part BOutcome & teaching note

Operating Quality · 2015–2017

Volkswagen emissions-control response and governance

After the public EPA notice, Volkswagen halted some U.S. diesel sales and announced an investigation; later issuer and court records document formal Supervisory Board action, an expanded disclosed scope, forensic recovery and compliance reforms, a 2.0-liter partial civil settlement, and a corporate guilty judgment, while also recording incomplete remediation.

Outcome-reveal packet. Read only after freezing Part A. This report is bound to canonical Part A digest 902c6e21d669d13d79be2cc7eaaa79104464b360dfe79fc222754a872a94821f. The Part A decision is the post-notice response at September 21, 2015—not whether to commit or continue the underlying conduct.

Executive Summary

This is a mixed post-NOV response inside an instructor-classified realized failure. The failure label describes the episode's realized legal, remedy, and accounting consequences. It is not a causal grade on the September 21 response decision.

The later record supports four bounded conclusions:

  1. Volkswagen's annual report records Supervisory Board discussion of the diesel issue and a special committee on September 25, leadership changes, and formal establishment of the committee on October 7. Those actions occurred after the Part A decision time; the selected record does not show that equivalent formal authority existed by September 21. source · volkswagen.annual-report-2015-text
  2. The signed Rule 11 plea agreement records outside-counsel disclosure, forensic recovery of large portions of deleted documents, an investigation, and compliance reforms. The same agreement says remediation remained incomplete. This supports a mixed process assessment, not a claim of either complete failure or complete control. source · doj.volkswagen-plea-2017-text
  3. The April 21, 2017 judgment records a guilty plea and adjudication of guilt on three corporate counts, three years of probation, and a $2.8 billion criminal fine. The June 28, 2016 filed proposed partial consent decree contains limited 2.0-liter admissions and separately stated civil-remedy components; EPA later reported that the court entered a rounded $14.7 billion partial settlement on October 25, 2016. source · court.volkswagen-judgment-2017-text source · court.volkswagen-2l-consent-decree-2016-text source · epa.volkswagen-generation3-remedy-2017-text
  4. Volkswagen reported substantial operating scale and Automotive Division net liquidity, but also material diesel-related expenses in operating result. Accounting expenses, settlement terms, the criminal fine, cash payments, and currencies are not interchangeable. This bundle therefore abstains from a total-loss bridge, response attribution, or target price. source · volkswagen.annual-report-2016-text

The most transferable learning is procedural: when a regulator publicly alleges a test-versus-operation control divergence and says scope remains open, preserve evidence, reconcile product scope, govern disclosure, and map liquidity under independent board authority before narrowing the issue or making material commitments. The rule is a single-case candidate. Its 24-hour complete-hold threshold is a precautionary, analyst-designed emergency gate, not an empirically calibrated optimum.

Decision and outcome boundary

Part A froze evidence available by September 20, 2015 at 23:59:59 UTC and posed a September 21 response-governance decision. By then, the EPA had publicly issued its notice, Volkswagen had announced a partial U.S. diesel sales halt and investigation, and the alleged control behavior was no longer a decision the board could avoid having occurred. The decision still open was how to govern evidence, technical scope, disclosure, remediation, and liquidity. source · epa.volkswagen-nov-2015-ocr source · reuters.volkswagen-stop-sale-2015-reader

This distinction controls the causal reading. The later consent-decree admissions, corporate conviction, remedy obligations, and accounting charges principally concern pre-cutoff conduct and its consequences. They show that the episode was severe; they do not identify how much the September 21 response improved or worsened the outcome.

The signed plea's Statement of Facts describes document destruction as Volkswagen employees prepared to admit the defeat device to U.S. regulators. The court judgment dates the obstruction offense to September 2, 2015. That record makes evidence preservation urgent at the Part A cutoff, but it cannot be used as a negative grade on post-NOV conduct that had not yet occurred. source · doj.volkswagen-plea-2017-text source · court.volkswagen-judgment-2017-text

What happened after the cutoff

  • September 22, 2015: Volkswagen's 2016 annual report later recounted an issuer announcement of noticeable test-versus-road discrepancies in around eleven million vehicles worldwide. It also recounted a November 2 EPA notice involving V6 3.0-liter engines. This is a later issuer account and not an independently reconciled same-day denominator. source · volkswagen.annual-report-2016-text
  • September 25 and October 7, 2015: Volkswagen's 2015 annual report records Supervisory Board consideration and then formal establishment of the Special Committee on Diesel Engines, alongside leadership changes. source · volkswagen.annual-report-2015-text
  • June 28 and October 25, 2016: A proposed 2.0-liter partial consent decree was filed on June 28. EPA later stated that the court entered a rounded $14.7 billion partial settlement on October 25. The procedural stages are kept separate. source · court.volkswagen-2l-consent-decree-2016-text source · epa.volkswagen-generation3-remedy-2017-text
  • January 11, 2017: Volkswagen signed a Rule 11 plea agreement, agreed that the attached Statement of Facts was true and correct, and accepted responsibility under U.S. law for the described employee acts. The agreement alone is not labeled a court conviction here. source · doj.volkswagen-plea-2017-text
  • April 21, 2017: The court judgment separately establishes the accepted guilty plea, adjudication on three corporate counts, probation, and fine. source · court.volkswagen-judgment-2017-text

Financial reconstruction

All amounts below preserve reported scopes and source vintages. Parentheses indicate negative values.

Reported scale

EUR millionsFY2015FY2016
Sales revenue213,292217,267
Operating result before special items12,82414,623
Operating result(4,069)7,103
Earnings after tax(1,361)5,379
Automotive Division net liquidity at year-end24,52227,180

The FY2015 figures come from the 2015 annual report; the FY2016 figures come from the 2016 annual report. Automotive Division net liquidity is an issuer-defined divisional measure, not consolidated cash, unrestricted cash, or proof of legal-entity access. source · volkswagen.annual-report-2015-text source · volkswagen.annual-report-2016-text

Diesel-related accounting expense

EUR millionsFY2015FY2016
Diesel-related expenses recognized in operating result16,2006,400

Volkswagen's 2016 annual report said the €16.2 billion 2015 amount primarily included €7.8 billion of provisions for field activities and repurchases and €7.0 billion for legal risks. It described the additional €6.4 billion of 2016 expenses through several components. These are accounting expenses recognized in operating result. They are not a schedule of cash paid and must not be added to settlement face values without testing overlap, timing, currency, and legal entity. source · volkswagen.annual-report-2016-text

U.S. 2.0-liter partial settlement—separately stated amounts

USD millionsFiled proposal—June 28, 2016Court entry—October 25, 2016; reported January 6, 2017
Court-entered partial settlement headline, rounded14,700
Consumer program, estimated up to10,033Not separately verified in the selected court-entry source
Environmental mitigation obligation2,700Not separately verified in the selected court-entry source
Zero-emission-vehicle investments over ten years2,000Not separately verified in the selected court-entry source

The $14.7 billion figure is EPA's rounded headline for the court-entered partial settlement. The other values are distinct components stated in the June 28 filed proposal; the consumer value is an estimate and ceiling, and the ZEV commitment spans ten years. The headline is not an additional amount to add to the components. The scope is U.S. 2.0-liter vehicles, not the issuer's later worldwide vehicle count. source · court.volkswagen-2l-consent-decree-2016-text source · epa.volkswagen-generation3-remedy-2017-text

U.S. criminal judgment—separate penalty

USD millionsApril 21, 2017 judgment
Court-imposed criminal fine2,800

The criminal fine is in a separate judgment and scope. It is not added here to the civil headline, civil components, or euro-denominated accounting expenses. source · court.volkswagen-judgment-2017-text

Process quality versus outcome quality

The response process is best classified as mixed:

  • The public cutoff record confirmed a sales halt, an investigation, and cooperation, but did not disclose who would conduct the external investigation. source · reuters.volkswagen-stop-sale-2015-reader
  • Later issuer records document formal board attention and a special committee.
  • The signed plea records evidence recovery, outside-counsel disclosure, investigation, and specific compliance reforms.
  • The same plea says remediation remained incomplete.
  • No captured source verifies the Part A indicator's independently defined denominator, 100% preservation coverage within 24 hours, complete configuration reconciliation, or non-overlapping funding map.

The realized outcome is clearly adverse in the selected record: admissions, a large civil remedy, corporate conviction, probation, criminal fine, and material accounting expense. Yet adverse outcomes do not prove that every post-notice response action was poor. Likewise, Volkswagen's continuing scale and liquidity do not prove good response governance. Evidence confidence, response-process quality, realized outcome, and financial capacity remain separate scores.

Causal assessment

The primary hypothesis is deliberately narrow: the documented post-NOV response implemented elements of the Part A board-governance, evidence-recovery, and compliance logic, but the committee's independence and authority remain unverified and the public record does not identify the response's incremental effect.

The sequence supports that bounded reading:

  1. At cutoff, regulator scope remained open and public investigation authority was unspecified.
  2. Formal Supervisory Board actions were later documented.
  3. The signed plea later recorded forensic recovery and reforms, with incomplete remediation.
  4. Civil admissions and the corporate judgment concerned conduct that principally predated the response decision.

Confidence is moderate on the description of the documented process and low on any causal effect estimate. The strongest rival hypothesis is that the enforcement path was already so determined by pre-cutoff misconduct that post-NOV response quality was economically immaterial. The later consequences are consistent with that rival, but the recorded investigation, recovery, reforms, and regulator-approved remedy prevent the bundle from treating immateriality as established. A matched or within-case study of timing, evidence integrity, disclosure correction, remedy approval, customer cost, and financing outcomes would be needed to discriminate the hypotheses.

Counterfactual

The nearest feasible response-only counterfactual is the Part A board-led crisis gate: on September 21, establish an independently reporting Supervisory Board authority; preserve the population independently derived from HR, IT, engineering, legal, testing, and certification systems; reconcile every possibly shared configuration to regulator evidence; control quantitative disclosures; and build a non-overlapping funding map before material commitments.

The sales halt and investigation already underway, followed within days by documented board actions, support operational plausibility. They do not prove that the full control system could have been completed on the proposed timetable. source · reuters.volkswagen-stop-sale-2015-reader source · volkswagen.annual-report-2015-text

The counterfactual cannot undo pre-cutoff misconduct or destruction. No deterministic model in this bundle estimates reduced penalties, faster remedies, lower customer cost, less operating disruption, or a valuation effect.

Transferable candidate rule

Regulator-notice crisis gate. If a public regulator notice alleges software or another control behaves differently in certification testing and ordinary operation, scope remains open, and the issuer has not independently reconciled authority, custody, configurations, disclosure, and liquidity, then establish an independently reporting board crisis authority and complete those reconciliations before narrowing scope, releasing affected products, or making material commitments. A narrowly tailored emergency safety action that cannot await the gate may proceed while preservation and reconciliation run in parallel. [rule.volkswagen.regulator-notice-crisis-gate]

Three indicators govern the candidate rule:

  • Preservation coverage: verified held population divided by the independently defined population. The Part A 100%-within-24-hours threshold is a precautionary emergency gate, not an evidence-calibrated optimum. Any deletion, missing repository, or custody break triggers board and regulator escalation; human counsel may replace the timing with a documented safer executable standard.
  • Scope reconciliation: no unresolved material configuration-to-regulator difference before a quantitative public scope statement, certification request, product release, or narrowing decision.
  • Funding bridge: no material unsupported overlap, double count, currency conversion, or inaccessible-cash assumption before an irreversible commitment.

The rule remains candidate, has one supporting case, no tested counterexample, and low confidence. It governs response after notice; it does not detect or explain the underlying misconduct. Accounting adjustments, legal conclusions, publication, and external commitments require human approval.

Evidence gaps and abstentions

The record is missing complete board minutes, investigation mandates, custody and deletion logs, configuration ledgers, regulator communications, disclosure drafts, customer-remedy execution data, cash-payment schedules, insurance and recoveries, legal-entity liquidity, covenants, rejected alternatives, and a reference class for post-notice responses.

Accordingly, this case abstains from:

  • assigning intent or liability beyond the captured corporate admissions and judgment;
  • treating individual charging allegations as adjudicated facts;
  • determining whether every Part A control was timely implemented;
  • allocating total consequences between pre-cutoff conduct and post-cutoff response;
  • adding provisions, expenses, settlement amounts, penalties, or currencies;
  • estimating total cash loss, response benefit, enterprise value, or target price.

Bounded next steps

  1. Obtain the entered October 25, 2016 consent decree itself and compare it line by line with the June 28 filed proposal before promoting any term beyond the separately corroborated headline and components.
  2. Build a source-lined cash schedule by obligation, currency, legal entity, payment date, provision utilization, insurance, recovery, and overlap; keep it non-authoritative until a human approves material accounting adjustments.
  3. Seek contemporaneous board, investigation-authority, custody, regulator-reconciliation, and disclosure-control records to test the September 21 response rather than inferring it from later outcomes.
  4. Test the candidate rule against other regulator-notice cases, including at least one counterexample, before considering corpus validation or calibrating the emergency threshold.
  5. Keep publication approval pending. This teaching note is evidence-bounded research, not legal advice, an accounting determination, or an investment recommendation.

Further questions

  • When did independent investigators receive authority, budget, system access, and a direct reporting line, and how did those dates compare with September 21?
  • What was the independently derived preservation denominator, what exceptions remained after 24 hours, and which deleted records were not recovered?
  • When did the global engine/model/software/certificate denominator become reconciled to EPA, CARB, and other regulator evidence?
  • Which disclosure corrections were delayed or accelerated by governance choices after the notice?
  • How did actual cash payments map to the €16.2 billion and €6.4 billion accounting expenses, civil components, criminal fine, and later obligations without overlap?
  • Did earlier independently reporting board governance measurably change remedy approval time, customer participation, dealer disruption, financing costs, or legal outcomes?

Caveats

  • Annual reports are primary issuer disclosures but remain interested evidence for governance, investigation, remediation, and causality.
  • The signed Rule 11 agreement establishes Volkswagen's agreement to its factual basis; the April 21 judgment separately establishes court acceptance, adjudication, probation, and fine.
  • The June 28 consent-decree artifact was filed as a proposed partial decree. EPA's January 6 statement separately establishes the October 25 court entry and rounded $14.7 billion headline.
  • Consent-decree admissions are limited by the decree language and the captured scope is U.S. 2.0-liter vehicles. The issuer's around-eleven-million worldwide announcement is a different scope.
  • The obstruction offense ended before the Part A cutoff. It supports preservation urgency, not a finding that the later response caused the destruction.
  • Financial figures are as reported; no model run, currency translation, cash bridge, overlap adjustment, or valuation is present.
  • This case distinguishes facts, issuer claims, analyst judgments, assumptions, and unknowns. Missing decision-critical evidence triggers abstention.

Primary-source map

  • src.volkswagen.annual-report-2015-pdf and its deterministic text derivative — FY2015 financial scale and Volkswagen's account of Supervisory Board actions.
  • src.volkswagen.annual-report-2016-pdf and its deterministic text derivative — FY2016 financial scale, issuer-reported scope, and diesel-related accounting expenses.
  • src.court.volkswagen-2l-consent-decree-2016-pdf and its deterministic text derivative — June 28 filed proposed 2.0-liter decree, limited admissions, and separately stated components.
  • src.doj.volkswagen-plea-2017-pdf and its deterministic text derivative — signed Rule 11 agreement, Statement of Facts, recovery, and remediation record.
  • src.court.volkswagen-judgment-2017-pdf and its deterministic text derivative — April 21 corporate judgment, offenses, probation, and fine.
  • src.epa.volkswagen-generation3-remedy-2017-html and its deterministic text derivative — EPA statement on court entry and one approved Generation 3 remedy.
  • Frozen Part A primary sources — EPA's September 18 notice, WVU/ICCT in-use research, Volkswagen financial disclosures, and contemporaneous Reuters reporting. No later outcome evidence was introduced into Part A.

Observed after the cutoff

Outcome financials

4 tables

Later values do not backfill Part A. Definition changes, unknowns, and derived endpoints remain labeled.

Reported financial scale around the outcome periodAs Reported At Horizon · EURm
MeasureFY2015FY2016
Sales revenue213,2921217,2671
Operating result before special items12,824114,6231
Operating result-4,06917,1031
Earnings after tax-1,36115,3791
Automotive Division net liquidity at year-end24,522127,1801
EUR · EURmReported values remain strings; no browser-side recalculation.
Issuer-reported diesel-related expenses recognized in operating resultAs Reported At Horizon · EURm
MeasureFY2015FY2016
Diesel-related operating-result expenses16,20016,4001
EUR · EURmReported values remain strings; no browser-side recalculation.
U.S. 2.0-liter settlement amounts by procedural dateAs Reported At Horizon · USDm
MeasureFiled proposal—June 28, 2016Court entry—October 25, 2016; reported January 6, 2017
Court-entered partial settlement headline, roundedNot established14,7001
Consumer program, estimated up to10,0331Not established
Environmental mitigation obligation2,7001Not established
Zero-emission vehicle investments over ten years2,0001Not established
USD · USDmReported values remain strings; no browser-side recalculation.
U.S. criminal judgment—separate penalty scopeAs Reported At Horizon · USDm
MeasureApril 21, 2017 judgment
Court-imposed criminal fine2,8001
USD · USDmReported values remain strings; no browser-side recalculation.

Transferable—but not universal

Candidate decision rules

1 hypotheses

These rules are case-derived hypotheses. Each retains “unless” conditions, kill criteria, counterexamples, and promotion gaps.

Candidatelow confidence

rule.volkswagen.regulator-notice-crisis-gate

Establish an independently reporting board crisis authority, preserve the independently defined evidence population, reconcile every potentially shared configuration to regulator evidence, control quantitative disclosures, and map non-overlapping liquidity exposures before material remedy or settlement commitments.

Open technical and legal scope creates asymmetric risk from lost evidence, premature narrowing, inconsistent disclosures, and overlapping commitments; a board-controlled gate preserves information and financing options while facts are reconciled.

Use when

  • A public regulator notice alleges that software or another control behaves differently in certification testing and ordinary operation.
  • The regulator says its investigation or affected-product scope remains open.
  • The issuer has not independently reconciled investigation authority, evidence custody, affected configurations, disclosure obligations, and liquidity exposure.

Do not transfer when

  • An independent regulator-witnessed review already verifies complete evidence custody, product scope, disclosure controls, and downside funding.
  • A narrowly tailored emergency safety action cannot await the gate; the action should proceed while preservation and reconciliation run in parallel.

Reverse or kill if

  • Withdraw or narrow the rule if cross-case testing shows that the trigger produces frequent costly escalations without improving evidence integrity, scope accuracy, disclosure quality, or financing option value.
  • Reopen every scope or release decision when independent testing finds an unexplained configuration or regulator mismatch.
  • Halt nonessential commitments when the non-overlapping downside bridge breaches a human-approved minimum liquidity buffer or relies on uncommitted funding.
Limitations and promotion gaps
  • This is a candidate from one retrospective case and is not corpus-validated.
  • The 24-hour and 100% preservation threshold is a precautionary analyst-designed gate, not a threshold calibrated from a reference class.
  • The rule governs response after a public regulator notice; it does not prevent or explain the underlying misconduct.
  • Investigation authority, accounting adjustments, legal conclusions, publication, and external commitments still require appropriate human approval.

Lineage

Complete case source ledger

21 records

This list combines decision-cutoff and outcome evidence. Each report citation resolves to a source ID below. Third-party documents remain with their original publishers.

T2

src.volkswagen.annual-report-2014-pdf

Volkswagen Group Annual Report 2014

Volkswagen AG · Mar 12, 2015

Issuer DisclosurePrimaryContemporaneous

Used for: Cutoff-valid financial scale and Automotive Division liquidity · Issuer-described supervisory and compliance governance · Parent object for deterministic text extraction

T3

src.wvu-icct.in-use-emissions-2014-pdf

In-Use Emissions Testing of Light-Duty Diesel Vehicles in the United States

West Virginia University Center for Alternative Fuels, Engines and Emissions · May 30, 2014

Academic ResearchSecondaryContemporaneous

Used for: Pre-NOV technical evidence about test-versus-road emissions divergence · Method and sample-size limitations · Parent object for deterministic text extraction

T1

src.epa.volkswagen-nov-2015-pdf

Notice of Violation to Volkswagen AG, Audi AG, and Volkswagen Group of America

U.S. Environmental Protection Agency · Sep 18, 2015

Court Or Government RecordPrimaryContemporaneous

Used for: Cutoff-valid agency determinations and allegations · Volkswagen and model mapping within the NOV's scope · Evidence that the investigation and violation scope remained open · Parent object for OCR extraction

T3

src.reuters.volkswagen-stop-sale-2015-reader

Volkswagen to halt U.S. sales of some 2015 diesel cars

Reuters via MarketScreener and Jina Reader · Sep 20, 2015

Reputable NewsSecondaryContemporaneous

Used for: Confirmation that a partial U.S. sales halt was already in place before the cutoff · Contemporaneous issuer response and announced external investigation

T2

src.volkswagen.annual-report-2015-pdf

Volkswagen Group Annual Report 2015

Volkswagen AG · Apr 28, 2016

Issuer DisclosurePrimaryContemporaneous

Used for: Reported 2015 financial scale and diesel-related provisions · Issuer account of post-cutoff scope, board action, investigation, and evidence preservation · Parent object for deterministic text extraction

T2

src.volkswagen.annual-report-2016-pdf

Volkswagen Group Annual Report 2016

Volkswagen AG · Mar 14, 2017

Issuer DisclosurePrimaryContemporaneous

Used for: Reported 2016 financial scale and additional diesel-related special items · Issuer decomposition of 2015 and 2016 diesel-related expenses · Parent object for deterministic text extraction

T1

src.court.volkswagen-judgment-2017-pdf

United States v. Volkswagen AG — Judgment in a Criminal Case

U.S. District Court for the Eastern District of Michigan · Apr 21, 2017

Court Or Government RecordPrimaryContemporaneous

Used for: Court-entered guilty judgment, offense counts, probation, and criminal fine · Parent object for deterministic text extraction

T1

src.epa.volkswagen-generation3-remedy-2017-html

EPA and CARB Approve Remedy for Generation 3 2.0-Liter Diesel Vehicles

U.S. Environmental Protection Agency · Jan 6, 2017

Court Or Government RecordPrimaryContemporaneous

Used for: Contemporaneous agency confirmation that the court entered the 2.0-liter partial settlement · Approved Generation 3 remedy and its limited vehicle scope · Parent object for deterministic text extraction