Fama-French Analytics

Using Factor Attribution in Real Portfolios

You have a portfolio of stocks or funds. How do you know if you are actually exposed to the factors (Value, Momentum, Quality) you read about? This process is called Factor Attributionor "Factor Loading" analysis.

The Regression Model

Mathematically, factor loads are calculated using a multi-variable regression. If you are analyzing a fund $R_p$, the equation looks like this:

R_p - R_f = α + β_mkt(Mkt-RF) + β_smb(SMB) + β_hml(HML) + ... + ε

The coefficients (β) tell you your exposure:

  • β > 0: Positive exposure. When the factor goes up, your portfolio tends to go up.
  • β close to 0: No meaningful exposure. You are neutral to this factor.
  • β < 0: Negative exposure. You are betting against this factor.

Example: A classic "Growth Fund" typically has a highly negative HML beta (e.g., -0.4). This means when Value stocks outperform (HML is positive), the Growth fund drags.

Interpreting the "Alpha" (α)

In the regression above, Alpha (α) is the intercept—the return left over after explaining everything else.

True Skill vs. Hidden Factors:
Decades ago, Warren Buffett's returns looked like massive Alpha. However, when researchers like AQR's Andrea Frazzini analyzed Buffett's record using modern factors (specifically Quality and Low Beta), much of that "Alpha" turned out to be "Beta" — consistent exposure to high-quality, safe companies, levered up.

This is not a criticism! It means Buffett identified these factors decades before academics did. But for an investor today, it means you shouldn't pay high fees for "Alpha" that is just disguised "Factor Beta."

Practical Utilities

Tools like Portfolio Visualizer allow retail investors to run these regressions easily. Use them to:

  1. Verify Style Purity: Does your "Small Cap Value" ETF actually have high SMB and high HML loadings? Sometmes "Value" funds hold growth stocks due to index construction rules.
  2. Check Overlaps: If you hold 5 different funds, you might find they all bet on the same factor (e.g., all have high Market Beta), offering less diversification than you think.
  3. Fee Analysis: If a manager charges 1.5% but their returns are 99% explained by a simple 3-factor model, you can likely replicate their performance with cheaper ETFs.

Conclusion

Factor attribution lifts the hood on portfolio performance. It moves the conversation from "This manager is a genius" to "This strategy harvests the Profitability premium efficiently."