Penny × MadisonGoverned Financial IntelligenceDETERMINISTIC SAMPLE

PENNY × MADISON · GOVERNED FINANCIAL INTELLIGENCE

See the thesis.
See what could break it.

Point-in-time company research that keeps evidence, uncertainty, and every change visible—so a human can make the final call.

Explore the sample thesis Research only · No trading or execution
01Point-in-time truthNo hindsight silently added
02Visible uncertaintyUnknown stays unknown
03Contradiction firstDisconfirming evidence stays visible
04Human authorityResearch supports, people decide

MADISON RESEARCH WORKSPACE

A thesis built to be challenged.

Every conclusion connects to dated evidence, explicit assumptions, and a condition that would invalidate it.

01 · INVESTMENT THESISMODERATE

Grid modernization can support durable growth—if execution quality catches up.

The sample evidence suggests a growing installed base and improving recurring mix. The thesis is constrained by customer concentration, uneven margins, and incomplete supplier data.

INVALIDATION CONDITIONTwo consecutive quarters of backlog contraction, service retention below 90%, or gross margin below 24%.
SUPPORTING EVIDENCE3 CURRENT
Contracted backlog expanded

Sample backlog reached $4.8B, up 18% from the prior sample period.

FY26 Q2 shareholder letterPublished Jul 28, 2026
Recurring service mix improved

Illustrative recurring revenue rose to 38% of sales from 31% one year earlier.

FY26 Q2 sample filingCaptured Jul 29, 2026
Free-cash-flow conversion recovered

Sample trailing conversion returned to 82% after two integration-heavy quarters.

Penny sample calculationInputs frozen Jun 30, 2026
CONTRADICTORY EVIDENCE2 ACTIVE
Customer concentration remains elevated

The largest three sample customers represent 41% of revenue.

Revenue concentration scheduleCurrent · active risk
Gross margin missed the prior thesis

Illustrative margin was 120 bps below the assumption recorded in March.

Thesis variance calculationCurrent · thesis pressure
MISSING / STALE EVIDENCE1 MISSING · 2 STALE
Supplier renewal schedule

Not present in the sample evidence set. The related thesis claim remains unresolved.

Required source unavailableMISSING · confidence UNKNOWN
Regional order mix

Last observation predates the current reporting period.

FY25 segment scheduleSTALE · review required

ASSUMPTIONS, NOT ANSWERS

Valuation with the inputs exposed.

SAMPLE MODEL · AS OF 2026-06-30
SAMPLE PRICE$42.80
$42.80 current$46$58
MODELED RANGE$46–$58
BEAR CASE$31
Growth
2%
Margin
21%
Demand slips; concentration risk realizes
BASE CASE · REFERENCE$51
Growth
9%
Margin
25.5%
Backlog converts near historical rates
BULL CASE$67
Growth
14%
Margin
29%
Service mix compounds faster than expected

Illustrative discounted cash-flow outputs. Not a price target, recommendation, or live market value.

POINT-IN-TIME HISTORY

Memory without revisionism.

Penny keeps each prior view intact so a reviewer can see what changed, when it changed, and which evidence caused the change.

CATALYSTService renewal disclosuresExpected sample window · Q4 2026

PRIMARY RISKMargin recovery stallsReview threshold · below 24%

  1. VERSION 01

    Initial thesis recorded

    Moderate confidence · evidence set v1

    Baseline formed from the frozen FY25 evidence set.

  2. VERSION 02

    Risk flag added

    Supplier concentration disclosed

    A new filing introduced a previously unobserved dependency.

  3. CURRENT VERSION

    Thesis revised—not replaced

    Moderate confidence · range narrowed

    Margin pressure offset stronger backlog quality.

PROVENANCE / SOURCE STATE

Every claim should have a receipt.

Source dates and observation states travel with the research. Stale or absent evidence remains visible.

12
Sample source recordsFilings · transcripts · calculations
2
Stale evidence flagsVisible, never silently discarded
1
Missing evidence itemConfidence remains unknown
100%
Dated observationsReplayable point-in-time snapshot

FIXED AUTHORITY BOUNDARY

Intelligence for the decision.
Authority stays with the human.

Penny and Madison support a governed research process. Neither system can move capital or make the final decision.

01
P
SYSTEM OF RECORD

Penny Core

Preserves dated observations, prior thesis versions, and governed provenance.

02
M
RESEARCH LAYER

Madison

Organizes evidence, surfaces contradictions, and monitors invalidation conditions.

03
H
FINAL AUTHORITY

Human reviewer

Interprets the research, approves conclusions, and retains every consequential decision.

Research and evidence organizationChange and invalidation monitoringNo brokerage or order routingNo autonomous capital allocation