Improved Discounted Cash Flow (DCF) Valuation Model

A 5-year DCF model to estimate the intrinsic value of publicly traded companies. Built in Python with automated financial data retrieval, WACC calculation via CAPM, sensitivity analysis, and visualization. Pls try it ;)

Features

Feature	Description
5-Year FCF Projections	Revenue, EBIT, EBIAT, D&A, CapEx, and NWC projected using weighted historical margins and analyst consensus estimates
WACC via CAPM	Automated calculation of Cost of Equity (CAPM), Cost of Debt, and capital structure weights
Terminal Value	Perpetual Growth (Gordon Growth) and EV/EBITDA Exit Multiple methods, with sector-aware default multiples
Monte Carlo Simulation	10,000-run probabilistic valuation randomizing growth, margins, WACC, and TGR — outputs a price distribution with percentiles
Scenario Analysis	Bull / Base / Bear cases with different growth and margin assumptions
Sensitivity Analysis	Matrix of implied prices across WACC and Terminal Growth Rate scenarios
Comparable Company Analysis	Auto-fetches P/E, EV/EBITDA, EV/Revenue, and P/S for peer companies
Implied Growth Rate	Bisection algorithm to reverse-engineer the growth rate priced into the current stock price
Multi-Stock Comparison	Run the model across multiple tickers and compare valuations side by side
Visualization	6-chart report: FCF bars, sensitivity heatmap, valuation bridge, Monte Carlo histogram, scenario analysis, and key metrics
JSON Export	Export model outputs for integration with other tools

Quick Start

# Clone the repo
git clone https://github.com/EmanueleSturzo/DCF-Valuation-Model.git
(download zip file and extract)

for Mac:
cd ~/Downloads/DCF-Valuation-Model-main

for windows:
cd C:\Users\UserName\Downloads\DCF-Valuation-Model-main

# Install dependencies
pip install -r requirements.txt

# Run the model
python dcf\_model.py --ticker AAPL

Usage

Single Stock Valuation

python dcf\_model.py --ticker AAPL

Output:

============================================================
  DCF VALUATION SUMMARY — Apple Inc. (AAPL)
============================================================
  Sector:            Technology
  Industry:          Consumer Electronics
  Current Price:     $217.90
  Shares Outstanding: 15.115B
============================================================

  KEY ASSUMPTIONS
  ────────────────────────────────────────
  Revenue Growth (Yr1): 7.32%
  EBIT Margin:          31.52%
  Effective Tax Rate:   16.02%
  ...

  VALUATION OUTPUT
  ────────────────────────────────────────
  Implied Share Price:  $243.18
  Current Market Price: $217.90
  Margin of Safety:     11.60% (UNDERVALUED ↑)
  Implied Growth Rate:  5.14%

Custom Parameters

python dcf\_model.py --ticker NVDA --tgr 0.04 --riskfree ^FVX --market VT

Argument	Description	Default
--ticker	Stock ticker symbol (required)	—
--tgr	Terminal Growth Rate	0.03
--riskfree	Risk-free rate proxy ticker	^TNX (US 10Y)
--market	Market return proxy ticker	VTI
--json	Export results to JSON file	None

Compare Multiple Stocks

python compare.py --tickers AAPL MSFT GOOGL META AMZN

Outputs a side-by-side comparison table with implied prices, margins of safety, and buy/hold/sell signals.

Generate Visual Report

python visualize.py --ticker AAPL --save

Produces a full-page report with 4 charts saved to output/AAPL\_dcf\_report.png.

Use as a Python Module

from dcf\_model import DCFModel

model = DCFModel("MSFT")

# Core valuation
print(f"Implied Price: ${model.implied\_share\_price:.2f}")
print(f"WACC: {model.wacc:.2%}")
print(f"Margin of Safety: {model.margin\_of\_safety:.2%}")

# Monte Carlo simulation
mc = model.monte\_carlo()
print(f"Monte Carlo Median: ${mc\['median']:.2f}")
print(f"P10–P90 Range: ${mc\['p10']:.0f}–${mc\['p90']:.0f}")
print(f"Probability Undervalued: {mc\['prob\_undervalued']:.1f}%")

# Scenario analysis
print(model.scenario\_analysis())

# Exit multiple valuation
em = model.exit\_multiple\_valuation()
print(f"Exit Multiple Price: ${em\['implied\_price']:.2f} ({em\['multiple']:.0f}x EV/EBITDA)")

# Sensitivity table
print(model.sensitivity\_analysis())

# Comparable company analysis (requires internet)
print(model.comparable\_analysis(\["GOOGL", "META", "AMZN"]))

# Export
model.to\_json("output/msft\_valuation.json")

Comparable Company Analysis

# Auto-detect peers by sector
python dcf\_model.py --ticker AAPL --comps

# Specify custom peers
python dcf\_model.py --ticker AAPL --comps MSFT GOOGL AMZN META

---

Methodology

1. Free Cash Flow (FCF)

Free Cash Flow represents the cash available to all capital providers after operating expenses and reinvestment:

$$\text{FCF} = \text{EBIT} \times (1 - \text{Tax Rate}) + \text{D and A} - \text{CapEx} - \Delta\text{NWC}$$

Where:

• EBIT = Earnings Before Interest and Taxes
• D & A = Depreciation & Amortization (non-cash add-back)
• CapEx = Capital Expenditures (reinvestment in fixed assets)
• ΔNWC = Change in Net Working Capital

Each component is projected as a weighted-average margin of revenue, with more recent years weighted more heavily (40/30/20/10).

2. Weighted Average Cost of Capital (WACC)

WACC is the blended cost of financing used as the discount rate:

$$\text{WACC} = W_D \cdot R_D \cdot (1 - T) + W_E \cdot R_E$$

Cost of Equity is derived from the Capital Asset Pricing Model (CAPM):

$$R_E = R_f + \beta \cdot (R_m - R_f)$$

Variable	Meaning	Source
$R_f$	Risk-Free Rate	US 10-Year Treasury Yield (^TNX)
$\beta$	Systematic risk relative to market	Yahoo Finance
$R_m$	Expected Market Return	VTI 3-year average return
$R_D$	Cost of Debt = Interest Expense / Total Debt	Income Statement
$W_D, W_E$	Debt and Equity weights in capital structure	Balance Sheet + Market Cap

3. Terminal Value — Perpetual Growth Method

Assumes the company generates free cash flow at a constant growth rate in perpetuity after the projection period:

$$\text{TV} = \frac{\text{FCF}_n \times (1 + g)}{(\text{WACC} - g)}$$

Where $g$ is the terminal growth rate (default: 3%, approximating long-run nominal GDP growth).

4. Discounted Cash Flow (Enterprise Value)

$$\text{EV} = \sum_{t=1}^{n} \frac{\text{FCF}_t}{(1 + \text{WACC})^{t - 0.5}} + \frac{\text{TV}}{(1 + \text{WACC})^n}$$

Uses mid-year convention ($t - 0.5$) to reflect that cash flows are received throughout the year, not only at year-end.

5. Equity Value → Implied Share Price

$$\text{Equity Value} = \text{EV} - \text{Debt} + \text{Cash}$$

$$\text{Implied Price} = \frac{\text{Equity Value}}{\text{Shares Outstanding}}$$

6. Implied Growth Rate (Bisection Method)

Reverse-engineers the revenue growth rate the market is currently pricing in by finding $g^*$ such that:

$$\text{ImpliedPrice}(g^*) = \text{CurrentPrice}$$

Solved iteratively using the bisection algorithm with convergence tolerance of $0.01.

7. Sensitivity Analysis

Generates a matrix of implied share prices across a range of WACC and Terminal Growth Rate assumptions, enabling the analyst to assess how sensitive the valuation is to key inputs.

8. Monte Carlo Simulation

Randomizes key model inputs across distributions to produce a probability distribution of implied prices:

Input	Distribution
Revenue Growth	Normal (μ = base estimate, σ = 3%)
EBIT Margin	Normal (μ = historical avg, σ = 2%)
WACC	Normal (μ = calculated, σ = 1.5%)
Terminal Growth Rate	Uniform (1.5% – 4.0%)

10,000 simulations produce percentiles (P10, P25, P50, P75, P90) and the probability that the stock is undervalued.

9. Scenario Analysis

Three cases with distinct assumptions:

Scenario	Growth Adj	Margin Adj
Bull	Base + 2%	Base + 1%
Base	Model default	Model default
Bear	Base − 3%	Base − 2%

10. Exit Multiple Terminal Value

An alternative to the Gordon Growth method:

$$\text{TV}_{exit} = \text{EBITDA}_n \times \text{EV/EBITDA Multiple}$$

The model auto-selects a sector-appropriate multiple (e.g., 20x for Technology, 6x for Energy) or accepts a custom input.

11. Comparable Company Analysis

Automatically retrieves valuation multiples (P/E, EV/EBITDA, EV/Revenue, P/S) for peer companies and displays a comparison table. Peers are either specified manually or auto-selected by sector.

---

Project Structure

DCF-Valuation-Model/
├── dcf\_model.py          # Core DCF model (CLI + importable class)
├── compare.py            # Multi-stock comparison tool
├── visualize.py          # Chart generation and visual reports
├── requirements.txt      # Python dependencies
├── LICENSE               # MIT License
├── .gitignore
├── output/               # Generated reports (gitignored)
└── README.md

Data Sources

Data	Source
Financial Statements	Yahoo Finance via yfinance
Analyst Revenue Estimates	Yahoo Finance consensus
Risk-Free Rate	US Treasury Yields (^TNX, ^FVX)
Market Return	Vanguard Total Stock Market ETF (VTI)
Beta	Yahoo Finance

Limitations

• Analyst estimates: Revenue projections rely on Yahoo Finance consensus; these may be stale or unavailable for smaller companies. The model falls back to historical CAGR when estimates are not available.
• Margin stability: The model assumes historical margins (EBIT, D&A, CapEx) are representative of the future. This may not hold for companies undergoing transformation.
• Terminal Value dominance: In most DCFs, the Terminal Value accounts for 60–80% of the Enterprise Value. Small changes in WACC or TGR can dramatically change the output — hence the sensitivity table.
• Single-scenario: This is a deterministic model. A Monte Carlo extension would provide a probability distribution of outcomes.

Potential Extensions

• Revenue build-up by business segment (e.g., iPhone, Services, Mac)
• Regression-based beta calculation from historical returns
• ROIC and reinvestment rate consistency check
• Streamlit web dashboard for interactive analysis
• Historical DCF accuracy backtesting
• LBO model add-on

References

• Damodaran, A. — NYU Stern Valuation Resources
• Investopedia — DCF Analysis
• Wall Street Oasis — Terminal Value
• Financial data: Yahoo Finance

License

MIT