templates/plan.md

# Research plan and procedure

Status: not started. Populate the mandate and this plan before substantive screening.

## Decision contract

Record industry boundaries, as-of cutoff, universe, horizon, currency, benchmark, return convention, success event, policy thresholds and user overrides. Identify tools available and actual delegation mode.

## Procedure checklist

- [ ] Publish mandate, assumptions, sources, expert assignments and this plan.
- [ ] Build industry/value-chain map, candidate universe and exclusion ledger.
- [ ] Investigate business quality, financials, stewardship, risk and expectations.
- [ ] Resolve source conflicts and normalize periods, units and per-share inputs.
- [ ] Select finalists and closest rejected alternatives; record coverage limits.
- [ ] Model current-price expectations, central scenarios and adverse sensitivity.
- [ ] Rebuild 5-/10-year forecasts and check rank-reversing assumptions.
- [ ] Obtain independent countercases and neutral source/calculation checks.
- [ ] Correct material errors and recompute affected models and ranks.
- [ ] Apply declared gates; show primary, probability-first and upside comparisons.
- [ ] Complete report, ledgers, forecast events and final coverage statement.

## Assignments and dependencies

For each assignment record owner/skill, bounded question, inputs, sources, output path, dependencies, acceptance check and status. Preserve one owner for each canonical model.

## Decision-relevant open questions

List uncertainties, plausible impact, evidence needed and next investigation. Track material tangents in tangents.csv.

## Acceptance and stopping

Decisive sources and prices checked; model conventions consistent; calculations reproduced; conflicts resolved or reflected in conditional status; coverage and residual uncertainty visible. Continue if an accessible unanswered question could plausibly reverse the decision.

## Changes to plan

Record new evidence, procedure change, reason and impact. Do not silently change the ranking objective.