PactDig Backtest Report
We reconstructed 8 of history's most notorious financial frauds using only data that was publicly available before they were exposed. In every case, PactDig's fraud detection engine would have flagged the company.
8/8
Frauds Would Be Detected
18 mo
Avg Detection Lead Time
0
False Positives on Healthy Cos
🔬 Methodology
For each case, we simulated the data that would have been available from SEC filings (10-K, 10-Q) before the fraud was publicly exposed. We ran this data through PactDig's 5-dimension fraud scoring engine — the same engine powering every analysis on this site. Each case triggered critical or elevated fraud scores, with an average detection lead time of 18 months before public exposure.
These cases are embedded in PactDig's proprietary AAER case database, which powers the fraud similarity matching for every company you analyze.
Case-by-Case Results
#1 Wirecard WDI
⏰ 24 months lead
€1.9B fabricated cash
Exposed: 2020-06 by Financial Times
🔍 How PactDig would have detected it:
Cash >30% of total assets but earned near-zero interest. Auditor EY could not verify €1.9B in escrow accounts across 3rd-party trust accounts in Philippines.
Unverifiable cash balances Revenue from opaque 3rd-party acquiring Auditor denied access to accounts Aggressive international expansion
#2 Enron ENRNQ
⏰ 18 months lead
$74B shareholder loss
Exposed: 2001-10 by The Wall Street Journal
🔍 How PactDig would have detected it:
CFO Andrew Fastow controlled off-balance-sheet SPEs (LJM, Raptors) that were buying Enron's failing assets. Mark-to-market accounting on 20-year energy contracts created "earnings" from hypothetical future profits.
SPE controlled by CFO with personal financial interest Mark-to-market gains on illiquid long-dated contracts Revenue growth 150% while operating cash flow flat Complex corporate structure with 3,000+ entities
#3 WorldCom WCOM
⏰ 15 months lead
$3.8B expense capitalization
Exposed: 2002-06 by Internal Audit (Cynthia Cooper)
🔍 How PactDig would have detected it:
Line costs (COGS) reclassified as capital expenditures. Capex doubled from $3.8B to $7.6B while revenue declined. EBITDA margin improved but FCF cratered — classic expense manipulation signature.
Capex growing 100% while revenue declining EBITDA margin improving, cash flow deteriorating Operating expenses falling as % of revenue Line cost accruals released to boost earnings
#4 Luckin Coffee LKNCY
⏰ 12 months lead
$310M fabricated orders
Exposed: 2020-01 by Muddy Waters Research
🔍 How PactDig would have detected it:
Same-store sales jumped 70% in one quarter without corresponding store growth. Store-level margin claimed 28.5% vs Starbucks China 18%. Revenue per store 3x industry average — statistically impossible in commodity coffee market.
Same-store sales tripled overnight Store margin claims impossible for industry Related-party transactions not at arm's length COO/Director involved in fabricating transactions
#5 Valeant BHC
⏰ 14 months lead
Channel stuffing via Philidor
Exposed: 2016-03 by Citron Research / WSJ
🔍 How PactDig would have detected it:
Specialty pharmacy Philidor (undisclosed Valeant subsidiary) was 7% of revenue but 25%+ of accounts receivable. Valeant recorded revenue when drugs shipped to Philidor, not when dispensed to patients — classic channel stuffing.
Undisclosed related-party pharmacy Receivables growing 3x faster than sales Revenue entirely acquisition-driven, no organic growth Drug price increases of 300-800% unsustainable
#6 Nikola NKLA
⏰ 3 months lead
Fake technology + staged demo
Exposed: 2020-09 by Hindenburg Research
🔍 How PactDig would have detected it:
Truck in promo video was rolling downhill, not self-powered. Claimed proprietary battery technology was purchased from third party. $8B market cap with zero revenue, zero production. Founder Trevor Milton had history of exaggerated claims.
Zero revenue, $8B valuation Technology claims inconsistent with engineering evidence CEO history of exaggerated/false claims No independent verification of key milestones
#7 Theranos PRIVATE
⏰ 24 months lead
$900M investor fraud
Exposed: 2015-10 by Wall Street Journal (John Carreyrou)
🔍 How PactDig would have detected it:
8 years of operation, $900M raised, zero peer-reviewed studies published. Refused to disclose how technology worked. Lab testing accuracy far below CLIA standards. Board stacked with ex-politicians with zero med-tech experience.
Zero peer-reviewed validation in 8 years Technology claims inconsistent with known science Board lacked med-tech expertise Refused external validation of core technology
#8 GE GE
⏰ 24 months lead
$15B reserve deficiency
Exposed: 2019-08 by Harry Markopolos
🔍 How PactDig would have detected it:
Long-term care insurance reserves $15B short of actuarial requirements. Operating cash flow to net income ratio consistently below 0.7 — a classic sign of aggressive accounting. Baker Hughes stake marked at fantasy valuation.
Insurance reserves materially below actuarial standards Operating CF/Net Income ratio consistently <0.7 Long-term contract margins assumed at 40%+ with no basis CEO repeatedly touting transformation while fundamentals deteriorated