dcf-valuation-workflow

name	dcf-valuation-workflow
description	DCF-only valuation workflow orchestrator for equity research. Use when the user asks for DCF估值, DCF闭环, intrinsic value, Formal DCF, Scenario DCF, Reverse DCF, or wants a one-pass workflow that first prepares financial-modeling DCF-ready inputs and then runs dcf-model outputs. This skill coordinates financial-modeling to dcf-model for DCF deliverables only.

DCF Valuation Workflow

Overview

This skill is the DCF control layer. It delegates three-statement preparation to financial-modeling, delegates discounting and validation to dcf-model, and leaves PEG comparison to the report layer.

It orchestrates two independent skills:

financial-modeling -> DCF-ready package / UFCF bridge / assumption ledger / data gaps
dcf-model -> Formal, Scenario, or Reverse DCF model / summary / validation

Core Rules

• Always use $financial-modeling before any forward DCF path.
• Use $dcf-model only after a DCF-ready handoff, assumption ledger, or explicit data-gap decision exists.
• DCF model inputs are limited to operating forecasts, UFCF bridge, WACC, terminal assumptions, and EV-to-equity bridge; PEG target market cap, PEG bands, and PEG scorecards are outside the DCF input set.
• Output mode, valuation files, validation status, and data gaps; omit buy/sell advice, target position size, and certainty language.
• If the DCF inputs are not strong enough for Formal DCF, downgrade to Scenario DCF, Reverse DCF, or preparation-only instead of inventing precision.
• If the user asks to compare PEG and DCF, complete this DCF workflow first and leave cross-model comparison to the report layer.

Progressive Disclosure

Read these references as needed:

• Always read workflow.md before acting.
• Read input-contract.md when preparing, checking, or repairing DCF-ready inputs.
• Read output-checklist.md before final delivery or when deciding whether the DCF workflow is complete.

Also load and follow the upstream skill files at the point of use:

/Users/a/.codex/skills/financial-modeling/SKILL.md
/Users/a/.codex/skills/dcf-model/SKILL.md

Workflow

1. Identify the company, ticker, currency, artifact directory, and available research files.
2. Run the financial-modeling stage to create or verify the DCF-ready package.
3. Decide the DCF mode using the workflow reference:
   • Formal DCF
   • Scenario DCF
   • Reverse DCF
   • Preparation-only
4. Run the dcf-model stage using only admitted DCF inputs.
5. Validate formulas, source comments, assumption ledger completeness, terminal constraints, UFCF tie-out, and EV-to-equity bridge.
6. Return the generated files, mode, validation status, and remaining data gaps.

Required Artifact Boundary

Default research directory:

research_artifacts/<target>/

Financial-modeling stage should produce or confirm:

<target>_dcf_ready_package.md
<target>_dcf_assumption_ledger.md
<target>_dcf_data_gaps.md

DCF-model stage should produce one mode-specific set:

Formal:
  <target>_dcf_model.xlsx
  <target>_dcf_summary.md
  <target>_dcf_validation.json

Scenario:
  <target>_scenario_dcf_model.xlsx
  <target>_scenario_dcf_summary.md
  <target>_scenario_dcf_validation.json
  <target>_dcf_assumption_ledger.md

Reverse:
  <target>_reverse_dcf_summary.md
  <target>_dcf_assumption_ledger.md

Preparation-only:
  <target>_dcf_data_gaps.md

Non-Negotiables

1. No forward DCF without a UFCF bridge.
2. No Formal DCF without sourced or auditable Revenue, EBIT, Tax, D&A, Capex, Delta NWC, UFCF, cash, debt, shares, WACC inputs, and terminal assumptions.
3. No non-factual input enters the model without source type, confidence, range, and sensitivity treatment.
4. No D-confidence clue can drive Base or Bull valuation.
5. No net income, adjusted profit, EBITDA, or PEG profit anchor may substitute for UFCF.
6. No DCF delivery is complete without mode label, source/date notes, data gaps, and validation status.