canonical/modules/M002-the-research-audit-trail.md

<!-- Generated loss-aware reference mirror from God_Level_Public_Company_Financial_Analyst_Job_Guide_V6_99_ALL_SUB70_FIXED.docx. Canonical source remains the bundled DOCX. -->

<!-- Module: 002 | Title: The Research Audit Trail -->

## PART I - ANALYST OPERATING SYSTEM | MODULE 002

# The Research Audit Trail

> Mission. Create a source log, assumption register, model-change log, and decision journal that another analyst can reproduce.

## Decision output

Objective: Create a source log, assumption register, model-change log, and decision journal that another analyst can reproduce. The completed work product must be reproducible from evidence, show the downstream financial or decision effect when material, state the strongest contrary case, and define a dated update rule.

## Explicit operating procedure

1. Create a source log containing issuer, document, form/type, filing or publication date, period covered, accession/URL, exact section/table, retrieved timestamp, and version.

1. Create an assumption register with assumption name, historical basis, source, range, confidence, owner, valuation sensitivity, and next review date.

1. Maintain a model-change log that records prior value/formula, new value/formula, reason, source, affected outputs, reviewer, and timestamp.

1. Maintain a decision journal that freezes the conclusion, expected path, key uncertainties, disconfirming evidence, and thesis-break conditions before later outcomes are known.

1. Version raw sources, extracted data, model files, and memos so a reviewer can reproduce any past conclusion. Never silently replace an amended filing or revised dataset.

1. At each review, test whether the audit trail is sufficient for another analyst to rebuild the material conclusion without asking the original author.

## Required evidence and model bridge

- Primary-source set: source log, assumption register, decision journal, model-change log. Preserve exact document/version, date, period, and source location for every material factual input used in the research audit trail.

- For each key concept - source IDs, accession numbers, raw versus transformed data, assumption taxonomy, model lineage, decision journal - state whether it is a reported fact, analyst calculation, management claim, external estimate, or judgment. Quantitative concepts must retain raw components and units; qualitative concepts must retain the specific evidence and counterevidence.

- Map only economically relevant findings into the model or decision record. Process-control modules such as the research audit trail may have no direct valuation line; in that case document the downstream error or governance risk the control prevents.

## Metrics and calculation controls

| Metric / concept | Construction | Required validation |
| --- | --- | --- |
| claim traceability | % of material factual claims that link to a dated source, exact page/section, and model or memo location. | claim traceability: Audit a sample back to dated evidence and decision records; verify the stated threshold/score is reproducible by an independent reviewer and tied to a defined decision consequence. |
| model-change coverage | % of material model changes with a documented source, rationale, date, author, and affected output/sensitivity. | model-change coverage: Recalculate from same-scope numerator and denominator; confirm period, units, cohort/geography, and issuer definition; reconcile material differences to filings or operating data. |
| reproduction time | Elapsed analyst time for an independent reviewer to reproduce the key conclusion from the archived sources, assumptions, and model. | reproduction time: Verify start/end timestamps or periods from source records, use a consistent calendar/business-day convention, and test outliers rather than averaging them away. |



## Worked application

> Case: management changes a KPI definition and recasts only part of history.

- Reconstruct the relevant reported fact from primary evidence before interpreting the case. For the research audit trail, show the raw components rather than only the resulting ratio or narrative.

- Build the causal chain through source IDs, accession numbers, raw versus transformed data, assumption taxonomy, then identify which link is directly observed and which link remains an assumption.

- Calculate claim traceability, model-change coverage, reproduction time from sourced components under the reported/base interpretation and at least one skeptical alternative interpretation.

- Translate the difference between cases into the variable that matters for the research audit trail: evidence quality, revenue, operating profit/NOPAT, free cash flow, invested capital, financing/dilution, risk, or valuation. Mark non-applicable links instead of inventing them.

- Expert consistency test: preserve rejected hypotheses and prior model versions so hindsight cannot rewrite the record.

- Precommit the specific future filing, KPI, customer/supplier observation, regulator action, or market input that would materially invalidate the the research audit trail conclusion.

## Failure tests

- FAIL if source IDs cannot be defined and reproduced from the source pack.

- FAIL if a reviewer cannot reproduce a material claim, assumption, model change, and decision from dated source/version records without asking the original analyst.

- FAIL if the the research audit trail conclusion depends on an unstated assumption, unreconciled definition, or evidence that cannot be traced to its source/version.

- FAIL if evidence materially inconsistent with the the research audit trail conclusion is omitted, reclassified, or dismissed without a documented definition, materiality, causal, timing, and source-quality analysis.

## Completion test

A senior reviewer must be able to reproduce the the research audit trail conclusion, vary the most sensitive assumption independently, trace the change through the model, understand the strongest opposing case, and identify the next evidence that would force an update. If any link is missing, the module remains open.