Analytical Methodology

Transparency is critical for academic and investment research. This page details the specific calculations and statistical methods used throughout the Fama-French Value Spread Analytics dashboard.

1. Value Spread Calculation

The "Value Spread" represents the difference in valuation multiples between value stocks and growth stocks. We use the Book-to-Market (B/M) ratio as the primary valuation metric, following Fama and French's standard methodology.

Where:
- B/M_Value: The Book-to-Market ratio of the "High" B/M portfolio (top 30% or decile).
- B/M_Growth: The Book-to-Market ratio of the "Low" B/M portfolio (bottom 30% or decile).
- Logarithm: We take the natural log to normalize the distribution, as ratios tend to be right-skewed.

2. Rolling Statistics (Z-Score)

To determine if the current spread is "wide" or "narrow" relative to history, we calculate a Z-Score (Standard Score).

We typically use a full historical window (starting from 1926) to capture the entire regime history, rather than a short rolling window which might suffer from "recency bias."

3. Percentile Rank

The percentile rank indicates what percentage of historical months had a lower value spread than the current month.
- 99th Percentile: Extremely wide spreads (Value is very cheap relative to Growth).
- 1st Percentile: Extremely narrow spreads (Value is expensive relative to Growth).

4. Data Adjustments

Inflation: Returns data is nominal. However, since the Value Spread is a ratio of ratios at a single point in time, it is naturally inflation-neutral.
Rebalancing: We assume annual rebalancing at the end of June, matching the Kenneth French library update schedule.