Factor Investing: From Statistical Foundations to Fama-French Models - Factor Investing: From Statistical Foundations to Fama-French Models

A self-contained tutorial series on factor investing for anyone with an undergraduate-level background in probability and statistics. Key results are derived and demonstrated on data small enough to verify by hand, then applied to real market data.

Notebooks¶

Work through these in order — each builds on the previous one.

#	Notebook	Topics
1	Statistical Foundations	Descriptive stats, CLT, hypothesis testing, OLS by hand, HC sandwich estimator, Newey-West HAC derivation — all on hand-checkable datasets (5–8 observations)
2	Time-Series Foundations	Prices vs returns, stationarity & ADF tests, ACF/PACF, Ljung-Box, volatility clustering & ARCH, ergodicity & mixing, full assumption stack for FF regressions
3	Fama-French 3-Factor Model	FF3 theory, real data (Ken French + Yahoo Finance ETFs: SPY, IWN, VTV), OLS normal equations, statsmodels, R²/F-tests, diagnostics (JB, BP, DW, VIF), robust SEs, Newey-West, interpretation & predictions, multi-ETF comparison
4	Advanced Factor Models	FF5, Carhart 4-factor, Novy-Marx profitability, reversals, beta anomaly & BAB, meta-research (replication, out-of-sample decay, Adaptive Markets Hypothesis), the practitioner debate (DFA, Avantis, premium timing), rolling-window analysis
5	Practical Factor Investing	Real ETF data (AVWS, AVWC, AVEM, IWMO), monthly FF5+Momentum regressions with regional factor matching, portfolio construction (50/40/10), constrained optimisation, factor profiles, limitations of short samples
6	Solution Manual	Complete worked solutions to all 18 exercises from notebooks 01–04, with code, output interpretation, and explanations

Read Online¶

View the book on GitHub Pages →

Quick Start¶

git clone https://github.com/beachnoir/FamaFrench.git && cd FamaFrench
python -m venv venv && source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt
jupyter notebook

Requires Python 3.9+.

Preview the book locally¶

pip install -r requirements-book.txt
jupyter-book build --execute --html
jupyter-book start                     # opens at http://localhost:3000

Project Structure¶

FamaFrench/
├── 01_Statistical_Foundations.ipynb
├── 02_Time_Series_Foundations.ipynb
├── 03_Fama_French_3Factor.ipynb
├── 04_Advanced_Factor_Models.ipynb
├── 05_Practical_Factor_Investing.ipynb
├── 06_Solution_Manual.ipynb
├── custom.css               ← Styling for the Jupyter Book
├── myst.yml                 ← Jupyter Book v2 / MyST config + TOC
├── runtime.txt              ← Python version pin for hosted runtimes
├── .github/workflows/
│   └── deploy-book.yml      ← GitHub Actions: build + deploy to Pages
├── requirements.txt         ← Runtime deps (numpy, pandas, statsmodels, …)
├── requirements-book.txt    ← Book build deps (jupyter-book v2)
├── .gitignore
└── README.md

Data Sources¶

Fama-French factors (US and international): Downloaded at runtime from Kenneth French’s Data Library
ETF prices: Downloaded at runtime via yfinance
- Notebook 3: SPY, IWN, VTV (US large-cap, small-cap value, large-cap value)
- Notebook 5: AVWS, AVWC, AVEM, IWMO (Avantis UCITS + iShares Momentum, Xetra-listed)

No data files are committed to the repo — everything is fetched when the notebooks execute.

License¶

This educational material is provided as-is for learning purposes.