Using European Fama–French Factors for German Equity Regressions

Summary

The note explains how to estimate monthly idiosyncratic volatility for German equities using a three-factor Fama–French regression. It advises using the precomputed European factors from Ken French’s data library rather than calculating a separate factor set for each stock or substituting U.S. factors. The regression relates excess returns to the market, size, and value factors, with the residual representing variation not explained by those factors.

Monthly return intervals and factor dates must align precisely; for example, a return labeled June spans the end of May to the end of June. The choice of regression window involves a tradeoff: longer histories provide more observations, while factor exposures can shift over time. The note also cautions that European factors may use a market index that overlaps substantially with the DAX, which could affect how informative the resulting idiosyncratic volatility measure is. It does not prescribe a particular estimation window or resolve that benchmark concern.

Key ideas

• Use European Fama–French factors for German equity analysis rather than the standard U.S. factor series.
• Estimate excess returns against market, size, and value factors in a monthly time-series regression.
• Align monthly return intervals with the corresponding factor observations.
• Longer regression histories add data but may obscure changing factor exposures.
• The market benchmark used for European factors may overlap strongly with the DAX and limit interpretation.

Tags

# Fama French 3 model factors for German equities


# Fama French 3 model factors for German equities












I want to calculate monthly idiosyncratic volatility for the DAX index comprising 30 stocks. I will use the Fama french 3 model. My question is whether I have to calculate these factors for each stock or there is any database which will provide me with this data. I am a beginner in this model and wish I CAN HAVE A clear answer for this. It will save my life. Thanks in advance.

## Answer by Matthew Gunn (score 3)

https://quant.stackexchange.com/a/43397

Ken French's data library has all the factors here. You should probably use their calculated factors for Europe rather than their regular factors computed from U.S. markets.

Let $R_t$ denote the return of the DAX from $t-1$ to $t$. You want to run a monthly, time-series regression of returns in excess of the risk free rate on the three factors:

$$ R_{t} - R_{ft} = \alpha + \beta_1 \mathit{SMB}_{t} + \beta_2 \mathit{HML} + \beta_3 \mathit{RMRF}_t + \epsilon_{t}$$

It's important to get the timing to match up exactly. Eg. the June return is from May 31 to June 30. There's a somewhat subjective question of how long a time-series regression to run. The upside of a longer time-series regression is utilizing more data the downside is that betas aren't stable over time.

You may want to carefully read French's data website and the Fama-French international factors paper to see exactly what they're using for the market index. Since the DAX is itself rather market indexy, idiosyncratic volatility relative to whatever market index Fama-French use for their European factors may not be that great.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.