dcf-thinking
Skill2 starsChanged 5 months agoMITView on GitHub
Discounted Cash Flow valuation with step-by-step auditable reasoning. Claude reasons through revenue projections, EBITDA margins, CapEx, working capital, WACC components, and terminal value — writing every assumption, its justification, and the sensitivity analysis into the visible output. The written reasoning is the deliverable: every assumption is auditable.
In this repository22 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- chart-readerSkill
- credit-analyzerSkill
- drip-optimizerSkill
- earnings-reviewerSkill
- finops-agentSkill
- gl-reconcilerSkill
- kyc-screenerSkill
- macro-calendarSkill
- macro-stressSkill
- market-researcherSkill
- ma-screenerSkill
- meeting-preparerSkill
- model-builderSkill
- monte-carloSkill
- month-end-closerSkill
- options-pricerSkill
- pdf-analyzerSkill
- pitch-builderSkill
- statement-auditorSkill
- tax-harvesterSkill
- valuation-reviewerSkill
The file
Files
DCF Valuation (Auditable Reasoning)
What Claude does here: Builds a full DCF model step-by-step, writing every assumption and its justification into the response and the output artifact — not leaving it in internal thinking, which is never visible to the reader. Unlike a spreadsheet, each cell is explained — why this growth rate, why this margin, why this WACC. The written reasoning is as valuable as the number.
Trigger phrases
- "NVDA DCF değerlemesi yap"
- "Intrinsic value hesapla: [ticker]"
- "DCF modeli kur — [şirket]"
- "Fair value analizi: [ticker]"
- "Build a DCF for [company]"
- "Step-by-step valuation: [ticker]"
Phases
Phase 1 — Company profile & data collection
Fetch via WebSearch + WebFetch (Yahoo Finance quoteSummary):
- Current revenue, EBITDA, net income, EPS (last 12 months)
- Revenue growth rate (3-year historical CAGR)
- EBITDA margin (current and 3-year trend)
- CapEx / Revenue ratio
- Net working capital / Revenue ratio
- Current debt, cash, shares outstanding
- Beta (from Yahoo or compute vs benchmark)
- Sector median multiples for sanity check
Phase 2 — Revenue projection (5 years)
Reasoning step (write it out): Reason through growth rate assumptions:
- Historical growth rate — starting point
- Analyst consensus (WebSearch: "[company] revenue forecast 2025 2026")
- Market growth rate — industry TAM expansion
- Competitive position adjustment — gaining/losing share?
- Macro headwinds/tailwinds — FX, rates, demand cycle
- Management guidance — recent earnings call commentary
Output: year 1-5 revenue with explicit rationale for each year's growth rate.
Phase 3 — EBITDA margin projection
Reasoning step (write it out):
- Operating leverage: how does margin behave as revenue grows?
- Investment cycle: is the company in heavy CapEx mode (margin compression) or harvesting?
- Sector comparison: is current margin above/at/below peers?
- One-time items to strip out
Output: year 1-5 EBITDA margin with reasoning.
Phase 4 — Free cash flow build
For each projection year:
Revenue
× EBITDA margin = EBITDA
- D&A (estimate from CapEx trend)
= EBIT
× (1 - tax rate) = NOPAT
+ D&A
- CapEx (% of revenue, declining from heavy investment phase)
- ΔNWC (% of revenue growth × NWC/Revenue ratio)
= Free Cash Flow to Firm (FCFF)Tax rate: use effective rate from financials (typically 20-28% for US, 22% for Turkey).
Phase 5 — WACC computation
Reasoning step — each component justified in the output:
Cost of equity (CAPM):
Ke = Rf + β × (Rm - Rf) + country_risk_premium- Rf: current 10-year US Treasury (WebFetch or state assumption)
- β: from Yahoo or estimated
- Equity risk premium: 5.5% (US), 6.5% (Turkey/EM)
- Country risk premium: 0% (US), 3-5% (Turkey)
Cost of debt:
Kd = (Interest expense / Total debt) × (1 - tax rate)WACC:
WACC = (E/V) × Ke + (D/V) × KdWhere E = market cap, D = net debt, V = E + D.
Sanity check (state it in the output): Is this WACC reasonable for this sector? Compare to sector WACC range. Flag if >15% or <6%.
Phase 6 — Terminal value
Gordon Growth Model:
TV = FCF_year5 × (1 + g) / (WACC - g)Terminal growth rate g: typically 2-3% (GDP growth, conservative).
Exit multiple cross-check:
TV_check = EBITDA_year5 × EV/EBITDA_exit_multipleExit multiple = current sector median. If TV_GGM vs TV_check differ > 30%, explain why.
Phase 7 — Enterprise value → equity value
Enterprise Value = PV(FCF year 1-5) + PV(Terminal Value)
[discount each year's FCF at WACC]
Equity Value = EV - Net Debt
Intrinsic value per share = Equity Value / Diluted shares outstanding
Upside/downside vs current price = (Intrinsic - Current) / Current × 100%Phase 8 — Sensitivity analysis
Build a 3×3 sensitivity table:
| WACC - 1% | WACC | WACC + 1% | |
| g + 0.5% | |||
| g | |||
| g - 0.5% |
And a revenue growth sensitivity:
| Growth - 3% | Base | Growth + 3% | |
| Margin - 2% | |||
| Base margin | |||
| Margin + 2% |
Phase 9 — Bull / base / bear valuation
| Scenario | Key assumption | Intrinsic value | Upside |
| Bear | Low growth, margin compression, high WACC | ||
| Base | Consensus assumptions | ||
| Bull | High growth, margin expansion, low WACC |
Phase 10 — Output
Write output/dcf-<ticker>-<YYYYMMDD>.md:
# DCF Valuation: <Company> (<Ticker>)
## Summary
Intrinsic value: $X | Current price: $Y | Upside: Z%
Verdict: UNDERVALUED / FAIRLY VALUED / OVERVALUED
## Key Assumptions (with reasoning)
## 5-Year Projections
## WACC Derivation
## DCF Computation
## Terminal Value
## Sensitivity Analysis
## Bull / Base / Bear
## Comparison to Market Multiples
## Risks to the thesisDisclaimer: DCF is highly sensitive to assumptions. Small changes in WACC or terminal growth rate produce large value swings. Not investment advice.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

