← Return to Datacendia

These are not hypotheticals. Each simulation uses publicly available data — SEC filings, earnings calls, news reports, and regulatory documents — that existed before the disaster. We feed this data to The Council and ask: "Would you have approved this decision?"

Banking / Treasury Risk

Silicon Valley Bank Collapse

March 2023 · 48-hour bank run · FDIC receivership

Total Loss
$209B

The Decision Under Analysis

Q4 2021: SVB Treasury team proposes extending duration on the Held-to-Maturity (HTM) securities portfolio to capture yield in a low-rate environment. The portfolio grows from $49B to $91B in long-dated bonds. No interest rate hedges are purchased.

The Council's Analysis

CFO Agent
HTM classification masks $15B+ unrealized losses from balance sheet. If forced to sell, capital ratios collapse below regulatory minimums.
Risk Agent
Deposit base is 97% uninsured. Concentration in tech/VC sector creates correlated withdrawal risk. Duration mismatch: 3.6yr assets vs. overnight liabilities.
RedTeam Agent
Simulated scenario: Fed raises rates 400bps over 18 months. Result: $17B unrealized loss, negative tangible equity, bank failure within 72 hours of deposit acceleration.
Compliance Agent
No regulatory breach at time of decision. However, 10-K risk disclosures understate interest rate sensitivity. Potential securities law exposure if losses crystallize.

CendiaVeto Triggered

Rationale: Unhedged duration risk exceeds policy limits. Concentration of uninsured deposits creates liquidity risk that cannot be mitigated by Fed discount window (stigma effect). Recommendation: Cap HTM portfolio at $60B, purchase interest rate swaps, diversify deposit base.

Sovereign Alternate History

With hedges in place and diversified funding, SVB survives the 2022-2023 rate cycle with ~$800M in hedge gains offsetting HTM losses. No bank run. No FDIC receivership. Estimated value preserved: $16B+ in equity value.

Public Data Sources Used

SVB 10-K (2021, 2022), FDIC Call Reports, Federal Reserve H.15 yield curve data, Pitchbook VC funding data, Twitter/X sentiment analysis (March 9-10, 2023).

Aerospace / Safety Engineering

Boeing 737 MAX MCAS Failures

2018-2019 · Lion Air 610 & Ethiopian 302 · 346 fatalities

Total Cost
$20B+

The Decision Under Analysis

2015-2016: Boeing engineering team proposes MCAS (Maneuvering Characteristics Augmentation System) to compensate for aerodynamic changes from larger engines. System relies on single Angle of Attack (AOA) sensor. Pilot training classified as "Level B" (iPad course, no simulator time) to maintain 737 type rating compatibility.

The Council's Analysis

Risk Agent
Single-point-of-failure architecture. AOA sensor failure rate: ~1 per 100,000 flight hours. With 4,000+ aircraft planned, statistical certainty of multiple sensor failures over fleet life.
RedTeam Agent
Simulated: AOA sensor failure at V2+10 on takeoff. MCAS activates, pushes nose down. Pilot has 10 seconds to diagnose and disable. Failure rate in simulator: 67% of pilots could not recover.
Compliance Agent
FAA delegation allows Boeing self-certification. MCAS authority increased from 0.6° to 2.5° without updated safety analysis. Potential violation of 14 CFR 25.1309 (catastrophic failure probability).
Mirror Agent (Pilot)
Pilots not informed of MCAS existence. Runaway stabilizer checklist does not address MCAS-specific behavior. Training gap creates "automation surprise" risk.

CendiaVeto Triggered

Rationale: Single-sensor architecture for flight-critical system violates fundamental safety engineering principles. Pilot training inadequate for system complexity. Recommendation: Dual AOA sensor requirement, disagree light standard, Level D simulator training mandatory.

Sovereign Alternate History

With dual-sensor architecture and proper training, MCAS performs as designed safety system. No crashes. No 20-month grounding. No $20B in settlements, lost orders, and production delays. 346 lives preserved.

Public Data Sources Used

House Transportation Committee Report (2020), JATR Report, Lion Air KNKT Final Report, Ethiopian AIB Report, Boeing internal emails (Congressional exhibits), FAA TARAM safety assessments.

Financial Services / Fraud Detection

Wirecard Accounting Fraud

2015-2020 · €1.9B "missing" · CEO arrested

Investor Loss
€24B

The Decision Under Analysis

Q2 2019: Institutional investor (pension fund) considers €50M position in Wirecard AG following strong earnings. FT has published allegations of accounting irregularities in Asia operations, but Wirecard denies and threatens legal action. EY has issued clean audits for 10+ years.

The Council's Analysis

Analyst Agent
Third-party acquiring (TPA) revenue in Asia represents 50%+ of EBITDA but only 5% of transaction volume. Margin profile (30%+) inconsistent with industry benchmarks (3-5%).
Risk Agent
TPA partners are shell companies in Philippines and Singapore. €1.9B in "escrow" cannot be independently verified. Auditor has not physically confirmed cash balances.
RedTeam Agent
Scenario: TPA revenue is fabricated. If removed, Wirecard is unprofitable and insolvent. Probability assessment: 73% likelihood of material fraud based on forensic accounting indicators.
Compliance Agent
BaFin (German regulator) has banned short-selling of Wirecard stock — regulatory capture risk. Legal threats against journalists indicate potential obstruction pattern.

CendiaVeto Triggered

Rationale: Unverifiable cash balances in opaque jurisdictions. Margin profile statistically inconsistent with disclosed business model. Aggressive legal posture toward journalists is fraud indicator. Recommendation: Do not invest. If existing position, exit immediately.

Sovereign Alternate History

Pension fund avoids €50M loss. Portfolio reallocated to diversified fintech ETF. 18 months later, Wirecard stock goes to zero. €50M in beneficiary assets preserved.

Public Data Sources Used

Financial Times investigative series (2015-2019), Wirecard AG Annual Reports, EY audit opinions, BaFin regulatory filings, Philippine SEC corporate records, Singapore ACRA filings.

Healthcare / Due Diligence

Theranos Investment Due Diligence

2013-2015 · $700M raised · Technology never worked

Investor Loss
$600M+

The Decision Under Analysis

Q3 2014: Family office considers $25M investment in Theranos Series C at $9B valuation. Company claims revolutionary blood testing from finger prick. Board includes Henry Kissinger, George Shultz, James Mattis. Walgreens partnership announced.

The Council's Analysis

Analyst Agent
No peer-reviewed publications validating technology. No FDA 510(k) clearances for proprietary device. Lab operations run on modified Siemens machines, not "Edison" device.
Risk Agent
Board composition: Zero healthcare/diagnostics expertise. Red flag: experienced operators avoid, celebrity/political figures accept. Classic "affinity fraud" pattern.
RedTeam Agent
Technical due diligence blocked. "Trade secret" justification used to prevent independent validation. Scenario: Technology doesn't work. Probability: 89% based on missing technical validation.
Compliance Agent
CLIA lab certification obtained, but proficiency testing results not disclosed. CMS inspection records requested but not provided. Potential regulatory violations if results are inaccurate.

CendiaVeto Triggered

Rationale: No independent technical validation. Refusal of due diligence access. Board lacks domain expertise. Recommendation: Pass on investment. Require peer-reviewed validation and FDA clearance before any future consideration.

Sovereign Alternate History

Family office declines investment. Capital redeployed to validated diagnostics companies (Exact Sciences, Guardant Health). 5 years later, Theranos dissolved, founder convicted. $25M preserved, opportunity cost avoided.

Public Data Sources Used

Theranos investor presentations (leaked), CMS inspection reports (2015-2016), FDA warning letters, John Carreyrou WSJ articles (2015), Patent filings (USPTO), Board member disclosures.

Run Your Own War Game

Bring us your strategic decision. We'll stress-test it with The Council using your real data — before it becomes a case study for someone else.

Request a War Game Session

SYSTEM STATUS: AIR-GAP READY · NO TRACKER PIXELS DETECTED